tag:blogger.com,1999:blog-45032929495327606182024-03-18T19:52:13.656-07:00DSHR's BlogI'm David Rosenthal, and this is a place to discuss the work I'm doing in Digital Preservation.David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.comBlogger927125tag:blogger.com,1999:blog-4503292949532760618.post-38481451612575248722024-03-12T08:00:00.000-07:002024-03-13T09:59:28.465-07:00Petabit Optical Media?<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi3o4iVw3a7KlNcltB3flBGEvE7aybYm5G6liubjeBThYpeFQh7gAwCAy1M5MD9Tm6pECxVVJxc35y6hkknOa4YkEiWliSpcJgVMDNxdFjB7DnubN4hsnH3cwDguYF1TGajAownfksnJUvcPUknMRMVUL3mpTNkl8djjvtMBvJxbdW94PDVD3aa1AgYKXKp/s218/Sabine.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="218" data-original-width="214" height="218" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi3o4iVw3a7KlNcltB3flBGEvE7aybYm5G6liubjeBThYpeFQh7gAwCAy1M5MD9Tm6pECxVVJxc35y6hkknOa4YkEiWliSpcJgVMDNxdFjB7DnubN4hsnH3cwDguYF1TGajAownfksnJUvcPUknMRMVUL3mpTNkl8djjvtMBvJxbdW94PDVD3aa1AgYKXKp/s1600/Sabine.png" width="214" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.youtube.com/watch?v=6a_yxsJuOMY">Source</a></td></tr></tbody></table>
<a href="https://www.youtube.com/watch?v=6a_yxsJuOMY"></a><a>Sabine Hossenfelder</a> does the <a href="https://blog.dshr.org/search?q=Pangloss&max-results=20&by-date=true">good Dr. Pangloss</a> proud in her report on <a href="https://doi.org/10.1038/s41586-023-06980-y"><i>A 3D nanoscale optical disk memory with petabit capacity</i></a> by Miao Zhao <i>et al</i>. Their abstract claims that:<br />
<blockquote>
we increase the capacity of [optical data storage] to the petabit level by extending the planar recording architecture to three dimensions with hundreds of layers, meanwhile breaking the optical diffraction limit barrier of the recorded spots. We develop an optical recording medium based on a photoresist film doped with aggregation-induced emission dye, which can be optically stimulated by femtosecond laser beams. This film is highly transparent and uniform, and the aggregation-induced emission phenomenon provides the storage mechanism. It can also be inhibited by another deactivating beam, resulting in a recording spot with a super-resolution scale. This technology makes it possible to achieve exabit-level storage by stacking nanoscale disks into arrays, which is essential in big data centres with limited space.
</blockquote>
Below the fold I discuss this technology.<br />
<span><a name='more'></a></span>
What the authors mean by "petabit level" <a href="https://doi.org/10.1038/s41586-023-06980-y">is</a>:<br />
<blockquote>
The ODS has a capacity of up to 1.6 Pb for a DVD-sized disk area through the recording of 100 layers on both sides of our ultrathin single disk.
</blockquote>
1.6 petabit is 200TB per disk, which is 2,000 times the capacity of <a href="https://en.wikipedia.org/wiki/Blu-ray">triple-level Blu-ray media</a>. So this is a big increase. But weirdly, the caption to their Figure 1 <a href="https://doi.org/10.1038/s41586-023-06980-y">claims that</a>:<br />
<blockquote>
The capacity of a single 3D nanoscale disk is approximately equivalent to that of a petabit-level Blu-ray library (15.2 Pb, DA-BH7010, Hualu, China) or an HDD data array (12.64 Pb, EMC PowerVault ME5084, Dell, USA).
</blockquote>
A decade ago, <a href="https://blog.dshr.org/2014/09/more-on-facebooks-cold-storage.html">Facebook's Blu-ray library</a> put 10,000 100GB disks in a single rack for 1 Peta<i>byte</i> or 8 Peta<i>bit</i> capacity. This is 5 times as much as the authors' claim for a single disk. The caption's claim of 15.2Pb for the DA-BH7010 is 9.5 times their claim of the capacity of a single disk. Note also that they compare the volume of a single disk to the volume of complete read-write systems, which is comparing apples to oranges. I guess if your meaning of "approximately" is "within an order of magnitude" that makes sense.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiOgGwOj5l65aHujHZwqZQLXZL9Xga03LgxVCKNk-ZJo0DAjuIBqKq-xQdHRZmqmy0Gmio2x_52R9nasszMtoARyRfhsVC0OJYmQxdnywxwmzlVZqT1gYEc5-5XGWo7XBm5NKc-ZepisikV_fOOqPgZceKmRc9b0ibPGIGxZsQIx5A-vMrCUwRs_Iu_evoS/s902/TriStateTransitions.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="299" data-original-width="902" height="66" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiOgGwOj5l65aHujHZwqZQLXZL9Xga03LgxVCKNk-ZJo0DAjuIBqKq-xQdHRZmqmy0Gmio2x_52R9nasszMtoARyRfhsVC0OJYmQxdnywxwmzlVZqT1gYEc5-5XGWo7XBm5NKc-ZepisikV_fOOqPgZceKmRc9b0ibPGIGxZsQIx5A-vMrCUwRs_Iu_evoS/w200-h66/TriStateTransitions.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://doi.org/10.1038/s41586-023-06980-y">Figure 3a</a></td></tr></tbody></table>
The recording material on the disk has three states, as shown in the <a href="https://doi.org/10.1038/s41586-023-06980-y">schematic Figure 3a</a>:<br />
<blockquote>
The transition from the second to the third state is initiated by the 515-nm femtosecond Gaussian-shaped laser beam and deactivated by the 639-nm CW doughnut-shaped laser beam.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi-Pmz-P1BpAGdZJbm6t00W-evELOTYIC-Rg7hhMZ3UjnjJt5lJPNaGGpuHUjs6bvre7W1nfyiBw7gVakV94FEXZ_MGs8EBIOh3mfoZjmfIvfyWxfelj9_5qNxgLB9JM43EGZgEVJnm4NkJ9Eu72uE331UgImZirla6m6-CJ_BRevysZXLYr1Ue0yuZgd0x/s500/EmissionSpectra.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="305" data-original-width="500" height="122" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi-Pmz-P1BpAGdZJbm6t00W-evELOTYIC-Rg7hhMZ3UjnjJt5lJPNaGGpuHUjs6bvre7W1nfyiBw7gVakV94FEXZ_MGs8EBIOh3mfoZjmfIvfyWxfelj9_5qNxgLB9JM43EGZgEVJnm4NkJ9Eu72uE331UgImZirla6m6-CJ_BRevysZXLYr1Ue0yuZgd0x/w200-h122/EmissionSpectra.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://doi.org/10.1038/s41586-023-06980-y">Figure 3c</a></td></tr></tbody></table>
I assume that because this transition involves polymerization it is irreversible, making the media write-once. Comparing the dark blue line (second state) with the yellow and pink lines (third state) in Figure 3c shows that the second and third states are readily distinguishable by their <a href="https://doi.org/10.1038/s41586-023-06980-y">emissions when illuminated by >1mw 480nm</a>.<br />
<br />
There are a number of reasons to be less enthusiastic about the potential of this technology than Hossenfelder. It is true that they have demonstrated the ability to read and write petabit-scale data on a CD-sized medium. To do the reading they use two lasers, a 480nm pulsed and a 592nm continuous laser. To do the writing they used two lasers, a 515nm <a href="https://en.wikipedia.org/wiki/Mode_locking">femtosecond laser</a> and a 639nm continuous-wave laser. I haven't been able to find a price for a 515nm femtosecond laser, but <a href="https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=14348">here</a> is a 1550nm femtosecond laser for $48,880. The femtosecond laser they actually (<a href="https://www.acculasers.com/en/product/fs_lasers/">Acculasers ACL-AFS-515-CUS</a>) used is a substantial box with fans and an AC power input.<br />
<br />
The authors make claims of the density of the <i>medium</i> but not of the <i>system</i>. Clearly, current femtosecond lasers are too expensive and too large to use in equivalents of the decade-old <a href="https://blog.dshr.org/2014/09/more-on-facebooks-cold-storage.html">Facebook Blu-Ray technology</a>. Something like Microsoft Research's system that uses femtosecond lasers to write in Silica allows the cost of the lasers to be amortized over an entire data-center aisle of media. If you are going to build something like this, there is no reason to use the CD form factor.<br />
<br />
The repetition rate of the femtosecond laser was 42MHz. I believe it writes one bit per pulse, so the write bandwidth is limited to around 5MB/sec, meaning that writing an entire disk would take around <strike>10.5</strike> 10,000 hours. A system using this technology would be write-once, and have a long read latency while the robot fetched the needed disk. It would thus only be suitable for the niche archival market, and in this market the slow write rate would require many drives writing in parallel. This all makes this claim by the authors <a href="https://doi.org/10.1038/s41586-023-06980-y">somewhat hyperbolic</a>:<br />
<blockquote>
the development of next-generation industry-oriented nanoscale ODS that is much less expensive than state-of-the-art optical disk libraries and HDD data arrays will fulfil the vast data storage requirements of the big-data era.
</blockquote>
It would have similar product issues to those I outlined in <a href="https://blog.dshr.org/2024/03/microsofts-archival-storage-research.html"><i> Microsoft's Archival Storage Research</i></a>:<br />
<blockquote>
Six years ago <a href="https://blog.dshr.org/2018/02/dnas-niche-in-storage-market.html">I wrote</a>:<br />
<blockquote>
<a href="http://blog.dshr.org/2016/12/the-medium-term-prospects-for-long-term.html">time-scales in the storage industry are long</a>. Disk is a <a href="https://en.wikipedia.org/wiki/History_of_IBM_magnetic_disk_drives#Early_IBM_HDDs">60-year-old technology</a>, tape is at least <a href="https://en.wikipedia.org/wiki/IBM_7_track">65 years old</a>, CDs are <a href="https://en.wikipedia.org/wiki/Compact_disc">35 years old</a>, flash is <a href="https://www.google.com/patents/US5095344">30 years old</a> and has yet to impact bulk data storage.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj-ZpPqvkf3GXj_kN5rc__s2L5rvYJFx7FAmoP_wx6B-74cGM0d6zq6Y1HZdiuKwCRMx91LIhSAOod9lRwnBjlKQh8mwXLNSbQqy3nRhk6fYJRY4W8jniBiE4E8Bp5dgQZF_H6ghOlIIkdYl1JQK8wkgW3TKLT0Z92Pb2BNOgRiRyHlYVCeR-yk3CZ4CA/s800/BitShipments.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="331" data-original-width="800" height="83" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj-ZpPqvkf3GXj_kN5rc__s2L5rvYJFx7FAmoP_wx6B-74cGM0d6zq6Y1HZdiuKwCRMx91LIhSAOod9lRwnBjlKQh8mwXLNSbQqy3nRhk6fYJRY4W8jniBiE4E8Bp5dgQZF_H6ghOlIIkdYl1JQK8wkgW3TKLT0Z92Pb2BNOgRiRyHlYVCeR-yk3CZ4CA/w200-h83/BitShipments.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://digitalpreservation.gov/meetings/DSA2023/loc_dsa2023_website_0104_lauhoff_Storage%20Landscape%20__0326.pdf">Source</a></td></tr></tbody></table>
Six years on flash has finally impacted the bulk storage market, but it isn't predicted to ship as many bits as hard disks for another four years, when it will be a 40-year-old technology. Actual demonstrations of DNA storage are only 12 years old, and similar demonstrations of silica media are 15 years old. History suggests it will be decades before these technologies impact the storage market.<br />
</blockquote>
Hossenfelder makes several mistakes in her report:<br />
<ul>
<li>"new disk memory that could bring disk memory into the Petabyte range" - no, that is the Peta<i>bit</i> range.</li>
<li>Optical disks "were outcompeted by hard disks". - no, write-once removable media and on-line storage are two completely different markets. Optical disks lost out to the cloud and to a lesser extent by flash.</li>
<li>"the information density on compact disks or any optical storage is ultimately limited by the frequency of the laser light" - well yes, but she is talking about a paper describing a 2000-times increase in capacity <i>using laser light</i>.</li>
<li>"in modern flash drives the information is stored in little magnetizable cells that are a few atoms in size" - no, flash isn't a magnetic technology. She also misses that modern flash is a volumetric not a planar technology, just like the technology in the paper.</li>
<li>"figured out how to write data in multiple layers" - no, Blu-ray is a multi-layer technology more than a decade old. They figured out how to write a lot more layers of much smaller bits.</li>
<li>"this could work up to hundreds of layers" - well, they only demonstrated 100 layers, so hundreds plural is speculation. To get to the petabyte range needs at least 500 layers or much smaller bits. Note that modern flash has over 100 layers.</li>
</ul>
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com2tag:blogger.com,1999:blog-4503292949532760618.post-61341577734025082102024-03-05T08:00:00.000-08:002024-03-05T14:41:38.846-08:00Microsoft's Archival Storage Research<table border="2" cellpadding="5" cellspacing="0" class="tr-caption-container" rules="groups" style="float: right; margin-left: 1em; text-align: right;">
<caption align="bottom"><a href="https://doi.org/10.1063/1.5007621">2016 Media Shipments</a></caption>
<tbody>
</tbody>
<colgroup span="1">
</colgroup><colgroup span="3">
</colgroup><tbody>
<tr>
<td colspan="1"><br /></td>
<th align="center">Exabytes</th>
<th align="center">Revenue</th>
<th align="center">$/GB</th>
</tr>
</tbody><tbody>
<tr>
<th>Flash</th><td>120</td><td>$38.7B</td><td>$0.320</td>
</tr>
<tr>
<th>Hard Disk</th><td>693</td><td>$26.8B</td><td>$0.039</td>
</tr>
<tr>
<th>LTO Tape</th><td>40</td><td>$0.65B</td><td>$0.016</td>
</tr>
</tbody></table>
Six years ago I wrote <a href="https://blog.dshr.org/2018/03/archival-media-not-good-business.html"><i>Archival Media: Not a Good Business</i></a> and included this <a href="https://doi.org/10.1063/1.5007621">table</a>. The argument went as follows:<br />
<ul>
<li>The value that can be extracted from data decays rapidly with time.</li>
<li>Thus companies would rather invest in current than archival data.</li>
<li>Thus archival media and systems are a niche market.</li>
<li>Thus archival media and systems lack the manufacturing volume to drive down prices.</li>
<li>Thus although quasi-immortal media have low opex (running cost), they have high capex (purchase cost).</li>
<li>Especially now interest rates are non-zero, the high capex makes the net present value of their lifetime cost high.</li>
<li>Archival media compete with legacy generic media, which have mass-market volumes and have already amortized their R&D, so have low capex but higher opex through their shorter service lives.</li>
<li>Because they have much higher volumes and thus much more R&D, generic media have much higher Kryder rates, meaning that although they need to be replaced over time, each new unit at approximately equal cost replaces several old units, reducing opex.</li>
<li>Especially now interest rates are non-zero, the net present value of the lower capex but higher opex is likely very competitive.</li>
</ul>
Below the fold I look into why, despite this, Microsoft has been pouring money into archival system R&D for about a decade.<br />
<span><a name='more'></a></span>
<h3>Background</h3>
Eleven years ago Facebook announced they were building entire <a href="https://blog.dshr.org/2013/02/facebooks-cold-storage.html">data centers for cold storage</a>. They expected the major reason for reads accessing this data would be subpoenas. Eighteen months later I was finally able to report on Kestutis Patiejunas' talk explaining the technology and why it made sense in <a href="https://blog.dshr.org/2014/09/more-on-facebooks-cold-storage.html"><i>More on Facebook's "Cold Storage"</i></a>. They used two different technologies, the first exploiting legacy generic media in the form of mostly-powered-down hard drives. Facebook's <a href="https://blog.dshr.org/2014/09/more-on-facebooks-cold-storage.html">design for this was</a>:<br />
<blockquote>
aimed at limiting the worst-case power draw. It exploits the fact that this storage is at the bottom of the storage hierarchy and can tolerate significant access latency. Disks are assigned to groups in equal numbers. One group of disks is spun up at a time in rotation, so the worst-case access latency is the time needed to cycle through all the disk groups. But the worst-case power draw is only that for a single group of disks and enough compute to handle a single group.<br />
<br />
Why is this important? Because of the synergistic effects knowing the maximum power draw enables. The power supplies can be much smaller, and because the access time is not critical, need not be duplicated. Because Facebook builds entire data centers for cold storage, the data center needs much less power and cooling. It can be more like cheap warehouse space than expensive data center space. Aggregating these synergistic cost savings at data center scale leads to really significant savings.
</blockquote>
Patiejunas figured out that, only at cloud scale, the economics of archival storage could be made to work by reducing the non-media, non-system costs. This insight led to the second technology, <a href="https://blog.dshr.org/2014/09/more-on-facebooks-cold-storage.html">robots full of long-lived optical media</a>:<br />
<blockquote>
It has 12 Blu-ray drives for an entire rack of cartridges holding 10,000 100TB Blu-ray disks managed by a robot. When the robot loads a group of 12 fresh Blu-ray disks into the drives, the appropriate amount of data to fill them is read from the currently active hard disk group and written to them. This scheduling of the writes allows for effective use of the limited write capacity of the Blu-ray drives. If the data are ever read, a specific group has to be loaded into the drives, interrupting the flow of writes, but this is a rare occurrence. Once all 10,000 disks in a rack have been written, the disks will be loaded for reads infrequently. Most of the time the entire Petabyte rack will sit there idle.
</blockquote>
In theory Blu-ray disks have a 50-year life, but this is <a href="https://blog.dshr.org/2014/09/more-on-facebooks-cold-storage.html">irrelevant</a>:
<blockquote>
No-one expects the racks to sit in the data center for 50 years, at some point before then they will be obsoleted by some unknown new, much denser and more power-efficient cold storage medium
</blockquote>
Eight years ago in <a href="https://blog.dshr.org/2016/05/the-future-of-storage.html"><i>The Future Of Storage</i></a> I explained:<br />
<blockquote>
Every few months there is another press release announcing that some new, <a href="http://blog.dshr.org/2014/06/more-on-long-lived-media.html">quasi-immortal medium</a> such as 5D quartz or stone DVDs has solved the problem of long-term storage. But the problem stays resolutely unsolved. Why is this? Very long-lived media are inherently more expensive, and are a niche market, so they lack economies of scale. Seagate could easily make <a href="http://www.digitalpreservation.gov/news/events/other_meetings/storage09/docs/5-4_Anderson-seagate-v3_archive_study.pdf">disks with archival life</a>, but they did a study of the market for them, and discovered that no-one would pay the relatively small additional cost. The drives currently marketed for "archival" use have a shorter warranty and a shorter MTBF than the enterprise drives, so they're not expected to have long service lives.<br />
<br />
The fundamental problem is that long-lived media only make sense at very low Kryder rates. Even if the rate is only 10%/yr, after 10 years you could store the same data in 1/3 the space. Since space in the data center racks or even at Iron Mountain isn't free, this is a powerful incentive to move old media out. If you believe that Kryder rates will get back to 30%/yr, after a decade you could store 30 times as much data in the same space.
</blockquote>
The key parameter of these archival storage systems isn't the media life, it is the write bandwidth needed to keep up with the massive flow of <a href="https://blog.dshr.org/2014/09/more-on-facebooks-cold-storage.html">data to be archived</a>:<br />
<blockquote>
While a group of disks is spun up, any reads queued up for that group are performed. But almost all the I/O operations to this design are writes. Writes are erasure-coded, and the shards all written to different disks in the same group. In this way, while a group is spun up, all disks in the group are writing simultaneously providing huge write bandwidth. When the group is spun down, the disks in the next group take over, and the high write bandwidth is only briefly interrupted.
</blockquote>
The lesson Facebook taught a decade ago was that the keys to cost-effective archival storage were first, massive scale, and second, high write bandwidth. Microsoft has learned the lesson and has been working to develop cloud-scale, high write bandwidth systems using two quasi-immortal media, DNA and silica.<br />
<h3>DNA</h3>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgqldpbCFfWyenDsCcOHRkumVjQEiOMB8Vxv6nSacYWJVuyYmy4pLbk2AnmPuRRZJu9b4hyphenhyphenYElZf_U6DKOulSPK9IwY4aA5BMIXYVSJMkFDTI0k3m__IdN4IGOieXr6Y4pep7zJJo1Or96-7pbPZYKoRseeMepNGUp2IVlwX5aLJEI0VfWDPA9RfKnVvPqA/s964/DNA-Prototype.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="591" data-original-width="964" height="122" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgqldpbCFfWyenDsCcOHRkumVjQEiOMB8Vxv6nSacYWJVuyYmy4pLbk2AnmPuRRZJu9b4hyphenhyphenYElZf_U6DKOulSPK9IwY4aA5BMIXYVSJMkFDTI0k3m__IdN4IGOieXr6Y4pep7zJJo1Or96-7pbPZYKoRseeMepNGUp2IVlwX5aLJEI0VfWDPA9RfKnVvPqA/w200-h122/DNA-Prototype.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://spectrum.ieee.org/dna-data-storage">Source</a></td></tr></tbody></table>
I have been tracking the development of DNA as a storage medium since 2012, when I wrote <a href="https://blog.dshr.org/2012/10/forcing-frequent-failures.html"><i>Forcing Frequent Failures</i></a> based on a <a href="http://www.extremetech.com/extreme/134672-harvard-cracks-dna-storage-crams-700-terabytes-of-data-into-a-single-gram">Harvard team</a> writing 700KB. Much of the recent progress has been driven by the collaboration between <a href="https://www.microsoft.com/en-us/research/project/dna-storage/">Microsoft Research and the University of Washington</a>. In 2019 they published the first <a href="https://doi.org/10.1038/s41598-019-41228-8">automated write-to-read prototype system</a> (pictured), <a href="https://spectrum.ieee.org/dna-data-storage">which was slow</a>:<br />
<blockquote>
This single-channel device, which occupied a tabletop, had a throughput of 5 bytes over approximately 21 hours, with all but 40 minutes of that time consumed in <a href="https://news.microsoft.com/source/features/innovation/hello-data-dna-storage/">writing “HELLO” into the DNA</a>.
</blockquote>
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhKXQ3KU9GFfnb9O-ZpTrrytpkUxaRqpYZmL-Y-facKjxQtn019OboY0cL7hiIfq5MuiiGlN4x8cPWNQSymVftq6QYwoZCAsocqcEhr9Z7cZQ-oaTS7CqWWvLOoRqfcamVsRoPC6syBHVnNfYT-cHuQLDFoBQJWNIoMqidkJ5-r3No6nGlV1bZtemwEmruM/s857/Gartner.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="594" data-original-width="857" height="139" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhKXQ3KU9GFfnb9O-ZpTrrytpkUxaRqpYZmL-Y-facKjxQtn019OboY0cL7hiIfq5MuiiGlN4x8cPWNQSymVftq6QYwoZCAsocqcEhr9Z7cZQ-oaTS7CqWWvLOoRqfcamVsRoPC6syBHVnNfYT-cHuQLDFoBQJWNIoMqidkJ5-r3No6nGlV1bZtemwEmruM/w200-h139/Gartner.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://spectrum.ieee.org/dna-data-storage">Source</a></td></tr></tbody></table>
Rob Carlson's <a href="https://spectrum.ieee.org/dna-data-storage"><i>The Quest for a DNA Data Drive</i></a> provides a useful overview of the current state of the art. Alas he starts by using one of my pet hates, the graph showing an immense gap between the "requirements" for data storage and the production of storage media. Carlson <a href="https://spectrum.ieee.org/dna-data-storage">captions the graph</a>:<br />
<blockquote>
Prior projections for data storage requirements estimated a global need for about 12 million petabytes of capacity by 2030. The research firm Gartner recently issued new projections, raising that estimate by 20 million petabytes. The world is not on track to produce enough of today’s storage technologies to fill that gap.
</blockquote>
<b><rant></b> Carlson's point is to suggest that there is a huge market for DNA storage. But this ignores economics. There will always be a "requirement" to store more data than the production of storage media, because some data is not valuable enough to justify the cost of storing it. The "gap" could only be filled if media were free. Customers will buy the storage systems they can afford and prioritize the data in them according to the value that can be extracted from it. <b></rant></b>.<br />
<br />
The size of the market for DNA storage systems depends upon their cost. In 2018's <a href="https://blog.dshr.org/2018/02/dnas-niche-in-storage-market.html"><i>DNA's Niche in the Storage Market</i></a> I imagined myself as the marketing person for a DNA storage system and posed these challenges:<br />
<blockquote>
Engineers, your challenge is to increase the speed of synthesis by a factor of a quarter of a trillion, while reducing the cost by a factor of fifty trillion, in less than 10 years while spending no more than $24M/yr.<br />
<br />
Finance team, your challenge is to persuade the company to spend $24M a year for the next 10 years for a product that can then earn about $216M a year for 10 years.
</blockquote>
Carlson is realistic about the <a href="https://spectrum.ieee.org/dna-data-storage">engineering challenges</a>:<br />
<blockquote>
For a DNA drive to compete with today’s archival tape drives, it must be able to write about 2 gigabits per second, which at demonstrated DNA data storage densities is about 2 billion bases per second. To put that in context, I estimate that the total global market for synthetic DNA today is no more than about 10 terabases per year, which is the equivalent of about 300,000 bases per second over a year. The entire DNA synthesis industry would need to grow by approximately 4 orders of magnitude just to compete with a single tape drive.
</blockquote>
One of the speed-ups that is needed is <a href="https://spectrum.ieee.org/dna-data-storage">chip-based massive parallelism</a>:<br />
<blockquote>
One of our goals was to build a semiconductor chip to enable high-density, high-throughput DNA synthesis. <a href="https://www.science.org/doi/10.1126/sciadv.abi6714">That chip</a>, which we completed in 2021, demonstrated that it is possible to digitally control electrochemical processes in millions of 650-nanometer-diameter wells.
</blockquote>
It was faster, but it's chemistry had <a href="https://spectrum.ieee.org/dna-data-storage">problems</a>:<br />
<blockquote>
The main problem is that it employs a volatile, corrosive, and toxic organic solvent (<a href="https://pubchem.ncbi.nlm.nih.gov/compound/Acetonitrile">acetonitrile</a>), which no engineer wants anywhere near the electronics of a working data center.<br />
<br />
Moreover, based on a sustainability analysis of a theoretical DNA data center performed my colleagues at Microsoft, I conclude that the volume of acetonitrile required for just one large data center, never mind many large data centers, would become logistically and economically prohibitive.
</blockquote>
Although it is the industry standard, this isn't the only way to <a href="https://spectrum.ieee.org/dna-data-storage">write DNA</a>:<br />
<blockquote>
Fortunately, there is a different emerging technology for constructing DNA that does not require such solvents, but instead uses a benign salt solution. Companies like <a href="https://www.dnascript.com/">DNA Script</a> and <a href="https://molecularassemblies.com/">Molecular Assemblies</a> are commercializing automated systems that use enzymes to synthesize DNA. These techniques are replacing traditional chemical DNA synthesis for some applications in the biotechnology industry. The current generation of systems use either simple plumbing or light to control synthesis reactions. But it’s difficult to envision how they can be scaled to achieve a high enough throughput to enable a DNA data-storage device operating at even a fraction of 2 gigabases per second.
</blockquote>
The Microsft/UW chip is an <a href="https://spectrum.ieee.org/dna-data-storage">alternative way to control enzymatic synthesis</a>:<br />
<blockquote>
The University of Washington and Microsoft team, collaborating with the enzymatic synthesis company Ansa Biotechnologies, recently took the first step toward this device. Using our high-density chip, we successfully demonstrated <a href="https://doi.org/10.1021/acssynbio.3c00044">electrochemical control of single-base enzymatic additions</a>.
</blockquote>
The link is to <a href="https://doi.org/10.1021/acssynbio.3c00044"><i>Spatially Selective Electrochemical Cleavage of a Polymerase-Nucleotide Conjugate</i></a> by Jake A. Smith <i>et al</i>:<br />
<blockquote>
Novel enzymatic methods are poised to become the dominant processes for de novo synthesis of DNA, promising functional, economic, and environmental advantages over the longstanding approach of phosphoramidite synthesis. Before this can occur, however, enzymatic synthesis methods must be parallelized to enable production of multiple DNA sequences simultaneously. As a means to this parallelization, we report a polymerase-nucleotide conjugate that is cleaved using electrochemical oxidation on a microelectrode array. The developed conjugate maintains polymerase activity toward surface-bound substrates with single-base control and detaches from the surface at mild oxidative voltages, leaving an extendable oligonucleotide behind. Our approach readies the way for enzymatic DNA synthesis on the scale necessary for DNA-intensive applications such as DNA data storage or gene synthesis.
</blockquote>
This has, as the article points out, potential for dramatically reducing the current cost of writing DNA, but it is still many orders of magnitude away from being competitive with a tape drive. The <a href="https://blog.dshr.org/search?q=Pangloss&max-results=20&by-date=true">good Dr. Pangloss</a> can continue to enjoy the optimism for many more years.<br />
<h3>Silica</h3>
The idea of writing data into fused silica with a femtosecond laser is at least a <a href="https://www.storagenewsletter.com/2014/10/31/rd-hitachi-rw-of-digital-data-in-fused-silica-glass/">decade and a half old</a><br />
<blockquote>
In 2009, Hitachi focused on fused silica known for its excellent heat and water resistance as a recording medium for long-term digital storage. After proposing the use of computed tomography to read data recorded with a femtosecond pulse laser, fused silica glass was confirmed as an effective storage medium. Applying this technology, it is possible to achieve multi-layer recording by changing the laser’s focal point to form microscopic regions (dots) with differing refractive indices. In 2012, a method was developed with Kyoto University to read the recorded dots in 4 layers using an optical microscope (recording density equivalent to a CD), and in 2013, 26 layer recording was achieved (recording density equivalent to a DVD). In order to increase recording density for practical applications, one means is to increase the number of recording layers. At the 100-layer level of recording density equivalent to a Blu-ray disc however issues arose in dot quality degradation and read errors resulting from crosstalk of data recorded in other layers.
</blockquote>
But in the last few years Microsoft Research has taken this idea and run with it, as they report in a 68-author paper at SOSP entitled <a href="https://doi.org/10.1145/3600006.3613208"><i>Project Silica: Towards Sustainable Cloud Archival Storage in Glass</i></a>. It is a fascinating paper that should be read by anyone interest in archival storage.
Like <a href="https://blog.dshr.org/2018/02/dnas-niche-in-storage-market.html">me</a>, the authors are skeptical of the near-term prospects for <a href="https://doi.org/10.1145/3600006.3613208">DNA storage</a>:<br />
<blockquote>
Technologies like DNA storage offer the promise of an extremely dense media for long-term data storage. However, the high costs and low throughputs of oligonucleotide synthesis and sequencing continue to hamper the technology’s feasibility. The total amount of data demonstrably stored in DNA remains on the order of MBs, and building a functional storage system that can offer reasonable SLAs underpinned by DNA is a substantial challenge. Alternative DNA-based technologies like dNAM attempt to bypass costly sequencing and synthesis steps, sacrificing density down to densities comparable with magnetic tape.
</blockquote>
Hence their focus on a medium with, in theory, a somewhat lower volumetric density. From their <a href="https://doi.org/10.1145/3600006.3613208">abstract</a>:<br />
<blockquote>
This paper presents Silica: the first cloud storage system for archival data underpinned by quartz glass, an extremely resilient media that allows data to be left in situ indefinitely. The hardware and software of Silica have been co-designed and co-optimized from the media up to the service level with sustainability as a primary objective. The design follows a cloud-first, data-driven methodology underpinned by principles derived from analyzing the archival workload of a large public cloud service.
</blockquote>
Their analysis of the workload of a tape-based cloud archival storage system in Section 2 <a href="https://doi.org/10.1145/3600006.3613208">shows that</a>:<br />
<blockquote>
on average for every MB read there are 47 MBs written, and for every read operation there are 174 writes. We can see some variation across months, but writes always dominate by over an order of magnitude.<br />
...<br />
Small files dominate the workload, with 58.7% of the reads for files of 4 MiB or smaller. However, these reads only contribute 1.2% of the volume of data read. Files larger than 256 MiB comprise around 85% of bytes read but less than 2% of total read requests. Additionally, there is a long tail of request sizes: there is ∼ 10 orders of magnitude between the smallest and largest requested file sizes.<br />
...<br />
We observe a variability in the workload within data centers, with up to 7 orders of magnitude difference between the median and the tail, as well as large variability across different data centers.<br />
...<br />
At the granularity of a day, the peak daily [ingress] rate is ∼16x higher than the mean daily rate. As the aggregation time increases beyond 30 days, the peak over mean ratio decreases significantly down to only ∼2, indicating that the average write rate is similar across different 30-day windows.<br />
...<br />
To summarize, as expected for archival storage, the workload is <i>heavily write-dominated</i>. However, unexpectedly, the IO operations are dominated by <i>small file accesses</i>.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgP3MmIfoEohS5K_2OPkmJsNV7rfL-qRf-eNHdGT-PFJPzZjwoq-X6pDHC0SJvqbDk7NOWKuO9aDBNiOdKKcFu9kpsqT3Hv5HFbgir51mzUV1YAJ08OUwnK2NMJWyVbZxKJUNe_zoSL8LuN/s647/FacebookWarmFig3.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="620" data-original-width="647" height="192" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgP3MmIfoEohS5K_2OPkmJsNV7rfL-qRf-eNHdGT-PFJPzZjwoq-X6pDHC0SJvqbDk7NOWKuO9aDBNiOdKKcFu9kpsqT3Hv5HFbgir51mzUV1YAJ08OUwnK2NMJWyVbZxKJUNe_zoSL8LuN/w200-h192/FacebookWarmFig3.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.usenix.org/system/files/conference/osdi14/osdi14-paper-muralidhar.pdf">Figure 3</a></td></tr></tbody></table>
Subramanian Muralidhar and a team from Facebook, USC and Princeton had a paper at the 2014 OSDI that described Facebook's warm layer above the two cold storage layers and below <a href="https://www.usenix.org/event/osdi10/tech/full_papers/Beaver.pdf">Haystack</a>, the hot storage layer. Section 3 of <a href="https://www.usenix.org/system/files/conference/osdi14/osdi14-paper-muralidhar.pdf"><i>f4: Facebook's Warm BLOB Storage System</i></a> provides workload data for this layer, which filters out most of the IOPS before they hit the archival layers. I explained in <a href="https://blog.dshr.org/2014/10/facebooks-warm-storage.html"><i>Facebook's Warm Storage</i></a> that:<br />
<blockquote>
A BLOB is a Binary Large OBject. Each type of BLOB contains a single type of immutable binary content, such as photos, videos, documents, etc.<br />
...<br />
Figure 3 shows that the rate of I/O requests to BLOBs drops rapidly through time. The rates for different types of BLOB drop differently, but all 9 types have dropped by 2 orders of magnitude within 8 months, and all but 1 (profile photos) have dropped by an order of magnitude within the first week.<br />
<br />
The vast majority of Facebook's BLOBs are warm, as shown in Figure 5 - notice the scale goes from 80-100%. Thus the vast majority of the BLOBs generate I/O rates at least 2 orders of magnitude less than recently generated BLOBs.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhdiS9TuQFo-H4BD1VrPF_uw833Zbrc0cA2A_SuoOfp2tam3Fo7tNqEDjRPtXck9zJ9zm5af1wRHK69ipZjJKp45gHZa7PWMPeisjA6oqWnufM_2GCW7PA7sHbxjDdSvkMLnuRHeAB6HnpF/s595/FacebookFigure5.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="443" data-original-width="595" height="149" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhdiS9TuQFo-H4BD1VrPF_uw833Zbrc0cA2A_SuoOfp2tam3Fo7tNqEDjRPtXck9zJ9zm5af1wRHK69ipZjJKp45gHZa7PWMPeisjA6oqWnufM_2GCW7PA7sHbxjDdSvkMLnuRHeAB6HnpF/w200-h149/FacebookFigure5.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.usenix.org/system/files/conference/osdi14/osdi14-paper-muralidhar.pdf">Figure 5</a></td></tr></tbody></table>
Note these important differences between Microsoft's and Facebook's storage hierarchies:<br />
<ul>
<li>Microsoft stores generic data and depends upon user action to migrate it down the hierarchy to the archive layer, whereas Facebook stores 9 specific types of application data which is migrated automatically based upon detailed knowledge of the workload for each of the types.</li>
<li>Because Facebook can migrate data automatically, it can interpose a warm layer above the archive layer of the hierarchy, and because it has detailed knowledge about the behavior of each of the data types it can make good decision about when to move each type down the hierarchy.</li>
<li>Because the warm layer responds to the vast majority of the read requests and schedules the downward migrations, Facebook's archive layer's IOPS are a steady flow of large writes with very few reads, making efficient use of the hardware.</li>
</ul>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhOK0u8vnqW0yd42tTDcpfQpqklYoJx1PkPhusifIk_aYgSXNluVydDdAswnbB_WAScPTI0for7eQFstp7fGX6T6IVEElNeKHzUS0OSxQw6lXTkbnVeev7SKLmdTT-E6-sNU4QO4Hk0xInSwXDvPqLhs8-x6nrq_Vxz7MWl7E1S4Zhiu15hOjvDGYXdPeEi/s903/SilicaFig2.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="468" data-original-width="903" height="104" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhOK0u8vnqW0yd42tTDcpfQpqklYoJx1PkPhusifIk_aYgSXNluVydDdAswnbB_WAScPTI0for7eQFstp7fGX6T6IVEElNeKHzUS0OSxQw6lXTkbnVeev7SKLmdTT-E6-sNU4QO4Hk0xInSwXDvPqLhs8-x6nrq_Vxz7MWl7E1S4Zhiu15hOjvDGYXdPeEi/w200-h104/SilicaFig2.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://doi.org/10.1145/3600006.3613208">Figure 2</a></td></tr></tbody></table>
Contrast Facebook's scheduled ingest flow with the bursty ingest rate shown in Figure 2 of the Silica paper, <a href="https://doi.org/10.1145/3600006.3613208">which finds that</a>:<br />
<blockquote>
At the granularity of a day, the peak daily rate is ∼16x higher than the mean daily rate.
</blockquote>
Another interesting aspect of the Silica design is that the technologies, and thus the hardware, used for writing and reading are completely different. The authors point out the <a href="https://doi.org/10.1145/3600006.3613208">implications for the design</a>:<br />
<blockquote>
As different technologies are used to read and write, after a platter is written it must be fully read using the same technology that will be used to read it subsequently. This happens before a platter is stored in the library and any staged write data is deleted.<br />
<br />
This design has an interesting consequence: during the period when user data is being written into the library, the workload is going to become <i>read</i>-dominated. Every byte written must be read to be verified, in addition to the user reads. The read bandwidth has to be provisioned for peak user read rate, however the read workloads are very bursty, so read drive utilization is extremely low on average. Thus, the verification workload simply utilizes what would otherwise be idle read drives.
</blockquote>
Using separate write and read drives has <a href="https://doi.org/10.1145/3600006.3613208">two advantages</a>:<br />
<blockquote>
This allows independent scaling of read and write throughput. Additionally, this design allows us to create the first storage system that offers true <i>air-gap-by-design</i>: the robotics are unable to insert a glass platter into a write drive once the glass media has been written.
</blockquote>
Since the majority of reads are for verification, the design needs to make specific provision for <a href="https://doi.org/10.1145/3600006.3613208">user reads</a>:<br />
<blockquote>
To enable high drive efficiency, two platters can be mounted simultaneously in a read drive: one undergoing verification, and one servicing a customer read. Customer traffic is prioritized over verification, with the read drive switching away when a platter is mounted for a customer read. As soon as the customer platter stops being accessed, the read drive has the ability to quickly switch to the other platter and continue verification. This is similar to avoiding head-of-line blocking of mice flows by elephant flows in networked systems.
</blockquote>
Like Facebook's, the prototype Silica systems are <a href="https://doi.org/10.1145/3600006.3613208">data-center size</a>:<br />
<blockquote>
A <i>Silica library</i> is a sequence of contiguous write, read, and storage racks interconnected by a platter delivery system. Along all racks there are parallel horizontal rails that span the entire library. We refer to a side of the library (spanning all racks) as a <i>panel</i>. A set of <i>free roaming</i> robots called <i>shuttles</i> are used to move platters between locations.<br />
...<br />
A read rack contains multiple read drives. Each read drive is independent and has slots into which platters are inserted and removed. The number of shuttles active on a panel is limited to twice the number of read drives in the panel. The write drive is full-rack-sized and writes multiple platters concurrently.
</blockquote>
Their performance evaluation focuses on the ability to respond to read requests within 15 hours. Their cost evaluation, like Facebook's, focuses on the savings from using warehouse-type space to house the equipment, although is isn't clear that they have actually done so. The rest of their cost evaluation is somewhat hand-wavy, as is natural for a system that <a href="https://doi.org/10.1145/3600006.3613208">isn't yet in production</a>:<br />
<blockquote>
The Silica read drives use polarization microscopy, which is a commoditized technique widely used in many applications and is low-cost. Currently, system cost in Silica is dominated by the write drives, as they use femtosecond lasers which are currently expensive and used in niche applications. This highlights the importance of resource proportionality in the system, as write drive utilization needs to be maximized in order to minimize costs. As the Silica technology proliferates, it will drive up the demand for femtosecond lasers, commoditizing the technology.
</blockquote>
I'm skeptical of this last point. Archival systems are a niche in the IT market, and one on which companies are loath to spend money, The only customers for systems like Silica are the large cloud providers, who will be reluctant to commit their archives to technology owned by a competitor. Unless a mass-market application for femtosecond lasers emerges, the scope for cost reduction is limited.<br />
<h3>Conclusion</h3>
Six years ago <a href="https://blog.dshr.org/2018/02/dnas-niche-in-storage-market.html">I wrote</a>:<br />
<blockquote>
<a href="http://blog.dshr.org/2016/12/the-medium-term-prospects-for-long-term.html">time-scales in the storage industry are long</a>. Disk is a <a href="https://en.wikipedia.org/wiki/History_of_IBM_magnetic_disk_drives#Early_IBM_HDDs">60-year-old technology</a>, tape is at least <a href="https://en.wikipedia.org/wiki/IBM_7_track">65 years old</a>, CDs are <a href="https://en.wikipedia.org/wiki/Compact_disc">35 years old</a>, flash is <a href="https://www.google.com/patents/US5095344">30 years old</a> and has yet to impact bulk data storage.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj-ZpPqvkf3GXj_kN5rc__s2L5rvYJFx7FAmoP_wx6B-74cGM0d6zq6Y1HZdiuKwCRMx91LIhSAOod9lRwnBjlKQh8mwXLNSbQqy3nRhk6fYJRY4W8jniBiE4E8Bp5dgQZF_H6ghOlIIkdYl1JQK8wkgW3TKLT0Z92Pb2BNOgRiRyHlYVCeR-yk3CZ4CA/s800/BitShipments.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="331" data-original-width="800" height="83" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj-ZpPqvkf3GXj_kN5rc__s2L5rvYJFx7FAmoP_wx6B-74cGM0d6zq6Y1HZdiuKwCRMx91LIhSAOod9lRwnBjlKQh8mwXLNSbQqy3nRhk6fYJRY4W8jniBiE4E8Bp5dgQZF_H6ghOlIIkdYl1JQK8wkgW3TKLT0Z92Pb2BNOgRiRyHlYVCeR-yk3CZ4CA/w200-h83/BitShipments.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://digitalpreservation.gov/meetings/DSA2023/loc_dsa2023_website_0104_lauhoff_Storage%20Landscape%20__0326.pdf">Source</a></td></tr></tbody></table>
Six years on flash has finally impacted the bulk storage market, but it isn't predicted to ship as many bits as hard disks for another four years, when it will be a 40-year-old technology. Actual demonstrations of DNA storage are only 12 years old, and similar demonstrations of silica media are 15 years old. History suggests it will be decades before these technologies impact the storage market.<br />
<br />David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com1tag:blogger.com,1999:blog-4503292949532760618.post-26472506381236955682024-02-22T08:00:00.000-08:002024-02-22T08:00:00.131-08:00Competition-proofing<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjfeROHBB3VcUVaQSjLlxDb4rpKVWB9soukI9gSNqLzUNXEc_JEAhk8VgH8Va-Wc-Y6FRtQnDoBjx1N3rQuol19rdxj70OTe4lp8DwlpTsHsF-Mc4ryTHaEBeg95wDkhFONw9s5Yhhc7yrRK5t1BrMz2sG5yaqakcK9Lfpp3AKyWSV8Dk5zfnUdDkDpEBMg/s1141/MarketCaps.jpeg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="783" data-original-width="1141" height="138" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjfeROHBB3VcUVaQSjLlxDb4rpKVWB9soukI9gSNqLzUNXEc_JEAhk8VgH8Va-Wc-Y6FRtQnDoBjx1N3rQuol19rdxj70OTe4lp8DwlpTsHsF-Mc4ryTHaEBeg95wDkhFONw9s5Yhhc7yrRK5t1BrMz2sG5yaqakcK9Lfpp3AKyWSV8Dk5zfnUdDkDpEBMg/w200-h138/MarketCaps.jpeg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.ft.com/content/c43661bb-a087-4a54-a4b4-a6549d9156c3">Source</a></td></tr></tbody></table>
Apart from getting started in the midst of one of Silicon Valley's regular downturns, another great thing about the beginnings of Nvidia was that instead of insisting on the "minimum viable product" our VCs, Sutter Hill and Sequoia, gave us the time to develop a real architecture for a family of chips. It enabled us to get an amazing amount of functionality into a half-micron gate array; <a href="https://blog.dshr.org/2014/12/hardware-io-virtualization.html">I/O virtualization</a>, a DMA engine, a graphics processor that rendered curved surfaces directly, not by approximating them with triangles, a sound engine and support for game controllers. As I write, after a three decade-long history of bringing innovations to the market, Nvidia is <a href="https://www.ft.com/content/c43661bb-a087-4a54-a4b4-a6549d9156c3">America's third most valuable company</a>. <br />
<br />
I've written <a href="https://blog.dshr.org/search/label/venture%20capital">several times</a> about how in pursuit of a quicker buck, VCs have largely discarded the slow process of building an IPO-ready company like Nvidia in favor of building one that will be acquired by one of the dominant monopolists. These VCs don't support innovation. Even if their acquisition-bound companies do innovate in their short lives, their innovations are rarely tested in the market after the acuisition.<br />
<br />
Below the fold I discuss a new paper that presents a highly detailed look at the mechanisms the dominant companies use to neutralize the threats startups could pose to their dominance.<br />
<span><a name='more'></a></span>
<br />
In <a href="https://dx.doi.org/10.2139/ssrn.4713845"><i>Coopting Disruption</i></a> law professors Mark Lemley (Stanford) and Matthew Wansley (Cardozo) ask a good question:<br />
<blockquote>
Our economy is dominated by five aging tech giants—Alphabet, Amazon, Apple, Meta, and Microsoft. Each of these firms was founded more than twenty years ago: Apple and Microsoft in the 1970s, Google and Amazon in the 1990s, and Facebook in 2004. Each of them grew by successfully commercializing a disruptive technology—personal computers (Apple), operating systems (Microsoft), online shopping (Amazon), search engines (Google), and social networks (Facebook). Each of them displaced the incumbents that came before them. But in the last twenty years, no company has commercialized a new technology in a way that threatens the tech giants. Why?
</blockquote>
The TL;DR of Lemley and Wansley's answer to their question <a href="https://dx.doi.org/10.2139/ssrn.4713845">is</a>:<br />
<blockquote>
While there are many reasons for the tech giants’ continued dominance, we think an important and overlooked one is that they have learned how to coopt disruption. They identify potentially disruptive technologies, use their money to influence the startups developing them, strategically dole out access to the resources the startups need to grow, and seek regulation that will make it harder for the startups to compete. When a threat emerges, they buy it off. And after they acquire a startup, they redirect its people and assets to their own innovation needs.
</blockquote>
<a href="https://dx.doi.org/10.2139/ssrn.4713845">They observe that</a>:<br />
<blockquote>
a company that is started with the goal of being swallowed by a tech giant probably isn’t contributing much to society.
</blockquote>
<h3>Introduction</h3>
They start by identifying the advantages and disadvantages the incumbents possess in their efforts to monetize innovations. Their list of advantages is:<br />
<ul>
<li>"large incumbents can take advantage of economies of scale" not just in manufacturing, but also in marketing an distribution by exploiting their existing customer relationships.</li>
<li>"Large incumbents can also take advantage of economies of scope. Innovation creates “involuntary spillovers”—new knowledge that has economic value beyond the specific product that the firm was developing."</li>
<li>"large incumbents can access capital at a lower cost" for example from retained earnings from their cash cows.</li>
<li>"large incumbents may have another potential advantage—a longer investment time horizon" even more so now with the compression of VC time horizons.</li>
</ul>
Their list of incumbents disadvantages in innovation is more interesting:<br />
<ol>
<li>"their success will cannibalize their own market share" or "More generally, a monopolist has diminished incentives to introduce new products, improve product quality, or lower prices because any new sales generated replace its existing sales." Economists call this "Arrow's replacement effect"; more specifically: "The general lesson is, all else equal, the larger a firm’s market share and the less it is threatened by competition, the weaker its incentives to innovate. So we should expect large incumbents to not innovate much. And if they can dispense with the competitors rather than have to compete with them, they will do that."</li>
<li>"their managers prefer to deliver incremental innovations to their existing customers". Unlike Arrow's theory, "Christensen’s theory of disruptive innovation, ... focuses on the career incentives of middle managers ... Incumbent managers have an incentive to deliver sustaining innovations—incremental improvements in quality to the firm’s existing products that will please its existing customers. But they have substantial disincentives to pursue projects that upset the apple cart, even if doing so would bring new customers to the firm" The fundamental problem is that "Housing an innovation project inside a firm with diverse lines of business creates conflict with those other businesses. Some firm assets—cash, cloud computing, equipment, facilities, and engineers’ time—are rivalrous and finite, so executives must be willing to fight internal constituencies to devote those resources to innovation." Ingenuity, NASA's wildly successful Mars helicopter is a good example, as Eric Berger reports in <a href="https://arstechnica.com/space/2024/02/before-ingenuity-ever-landed-on-mars-scientists-almost-managed-to-kill-it/"><i>Before Ingenuity ever landed on Mars, scientists almost managed to kill it</i></a>. It was competing for cost, weight and risk with Perseverance's primary mission.</li>
<li>"their single veto point decision-making structure encourages risk-aversion" More specifically: "Inside a large incumbent, decisions about whether to fund an innovative project must pass through one veto point. In the venture capital market, many competing investors independently decide whether to finance an innovative idea. Inside a firm, an employee with an innovative idea must pitch an idea to managers who ultimately report to one executive gate-keeper. In the venture capital market, if a would-be startup founder pitches an idea to ten VC firms, and nine of them are not persuaded, the idea gets funded." The advantage of market-based finance over internal finance applies not just to the initiation but also the continuation of an innovation project. Inside a firm, an executive who has soured on a project can terminate it. In the venture capital market, when a startup’s initial investors grow skeptical, the company can still pitch outsiders on infusing more cash." The authors make this important point (my emphasis): "And while economists often describe markets as efficient, there is no reason to believe individual corporate executives make efficient (or even rational) decisions. Just ask Twitter. Markets work not because private executives make good decisions but because the ones who make bad decisions get driven out. But <b>that dynamic only works with competition</b>."</li>
<li>"they cannot appropriately compensate employees working on innovation projects." The reason they cannot is that: "Startups solve this problem by giving employees stock options. Every employee with significant equity knows that if the startup successfully exits, they will be rewarded. Stock in a large, diversified public company does not create similar incentives. The incentives are diluted because the value of the stock will be affected by too many variables unrelated to the success of the specific innovation project." And that: "large firms do not recognize internal “property rights” to innovations that employees develop. If they did, employees might become reluctant to share information. But not protecting internal property rights gives innovative employees incentive to leave. If employees at a large firm found their own startup and raise venture capital to fund it, they will earn a much greater share of the profits of the innovation."</li>
</ol>
The authors go on to describe five techniques incumbents use to neutralize the threat of disruption that innovative startups might pose; network effects, self-preferencing, paying for defaults, cloning, and coopting the disruptor. They claim other have described the first four, but they don't amount to an adequate explanation for why the tech giants haven't been disrupted. I will summarize each of the four in turn..
<h3>Network effects</h3>
Nearly three decades ago W. Brian Arthur, in <a href="http://www.amazon.com/Increasing-Returns-Dependence-Economics-Cognition/dp/0472064967"><i>Increasing Returns and Path Dependence in the Economy</i></a> explained how increasing returns to scale, or network effects, of technology markets typically led to them being dominated by one player. Consider a new market opened up by a technologcal development. Several startups enter, for random reasons one gets bigger then the others, network effects amplify its advantage in a feedback loop.<br />
<br />
This effect is more important now, as the the giants' business models have evolved to become <a href="https://dx.doi.org/10.2139/ssrn.4713845">platforms</a>:<br />
<blockquote>
The tech giants’ core businesses are built on platforms. A platform is an intermediary in a two-sided market. It connects users on one side of the market with users on the other side for transactions or interactions.<br />
...<br />
Platforms tend to exhibit network effects—the addition of a new user increases the value of a platform to existing users and attracts new users.
</blockquote>
This is precisely the mechanism Brian Arthur described, but applied to a business model that has since been enabled by the Internet.<br />
<h3>Self-preferencing</h3>
Self-preferencing happens when a platform isn't just a two-sided market, but one in which the platform itself is <a href="https://dx.doi.org/10.2139/ssrn.4713845">a vendor</a>:<br />
<blockquote>
Amazon, for example, both invites third party vendors to sell their products in its online marketplace and sells its own house brands that compete with those vendors. Amazon has a powerful advantage in that competition. It has access to data on all of its competitors—who their customers are, which products are selling well, and which prices work best. And it controls which ads consumers see when they search for a specific product. Assuming Amazon uses that information to prefer its own products to those of its competitors (either by pricing strategically or by promoting its own products in search results) – something alleged but not yet proven in a pending antitrust case -- the result is to bias competition. Vendors cannot realistically protest Amazon’s self-preferencing (or just go elsewhere) because Amazon has such a dominant share in the online retail market.
</blockquote>
<h3>Paying for defaults</h3>
The value of the default position is notorious <a href="https://dx.doi.org/10.2139/ssrn.4713845">because</a>:<br />
<blockquote>
Alphabet pays Apple a reported $18 billion (with a b) each year for Google to be the default search engine on iOS devices. Android and iOS together account for 99% of the U.S. mobile operating system market. Consequently, almost everyone who uses a smartphone in America is accustomed to Google search. Alphabet claims that “competition is just a click away.” But research and experience have shown that defaults can be somewhat sticky. So controlling the default position can give Alphabet (or whoever wins the Apple bid) an advantage. That said, someone has to be the default, and it might be better for consumers if the default is the search engine most users already prefer. The real problem might be the idea of paying for placement, whoever wins the bidding war.
</blockquote>
<h3>Cloning</h3>
There are many examples of a tech giant tryng to neutralize the threat from a startup by using the threat of cloning their product to force the startup to sell itself, or of actually cloning the product and using their market power to swamp the startup. Meta's addition of Reels to Instagram in response to Tik Tok is an obvious example. But the authors make two good points: <a href="https://dx.doi.org/10.2139/ssrn.4713845">First</a>:<br />
<blockquote>
Cloning is only objectionable if the tech giant wins out not by competition on the merits, but by exclusionary conduct.
</blockquote>
Second, that cloning <a href="https://dx.doi.org/10.2139/ssrn.4713845">often fails</a>:<br />
<blockquote>
Google+, Google’s effort to build a social media service that combined the best of Facebook and Twitter was an abject failure. Apple’s effort to control the music world’s move to streaming by offering its own alternative to Spotify hasn’t prevented Spotify from dominating music streaming and eclipsing the once-vibrant (and Apple-dominated) market for music downloads. Meta’s effort to copy Snap, then TikTok, by introducing Stories and Reels has not proven terribly successful, and certainly has not prevented those companies from building their markets.
</blockquote>
The fact that the giants can clone a startup's product leads the authors to <a href="https://dx.doi.org/10.2139/ssrn.4713845">ask</a>:<br />
<blockquote>
If the product is cloneable, then why would you buy the company and burn cash paying off its VCs?
</blockquote>
There are two possible answers. It may be faster and easier, though likely not cheaper, to "acquihire" the startup's talent than to recruit equivalent talent in the open market. Or it may be faster and easier, though likely not cheaper, to acquire the company and its product rather than cloning it.
<h3>Inadequate Explanation</h3>
The authors use the example of <a href="https://dx.doi.org/10.2139/ssrn.4713845">Microsoft</a>:<br />
<blockquote>
Microsoft enjoyed strong network effects in the 1990s as the dominant maker of operating system software – far more dominant than it is today. It cloned internet browser technology from upstarts like Netscape, and it engaged in anticompetitive conduct designed to ensure that it, not Netscape, became the browser of choice.82 But Microsoft’s victory over Netscape was short-lived. New startups – Mozilla and then Google – came out of nowhere and took the market away from it. Microsoft still benefits from network effects, and it still uses cloning and self-preferencing to send users to its Edge browser. But it doesn’t work. Microsoft employed all the tools of a dominant firm in a network market, but it still faced disruption.
</blockquote>
So these four techniques aren't an explanation for the recent dearth of disruption.<br />
<h3>Coopting disruption</h3>
The authors imagine themselves as a tech giant, asking what else they would do to prevent disruption, and coming up with <a href="https://dx.doi.org/10.2139/ssrn.4713845">four new techniques</a>:<br />
<ul>
<li>"First, you would learn as much as you can about which companies had the capability to develop disruptive innovations and try to steer them away from competing with you – perhaps by partnering with them, or perhaps by investing in them."</li>
<li>"Second, you would make sure that those companies could not access the critical resources they would need to transform their innovation into a disruptive product."</li>
<li>"Third, you would tell your government relations team to seek regulation that would build a competitive moat around your position and keep disruption out."</li>
<li>"Fourth, if one of the companies you were tracking nevertheless did start to develop a disruptive product, you would want extract that innovation—and choke off the potential competition—in an acquisition."</li>
</ul>
These are the techniques they call "coopting disruption", pointing out that the <a href="https://dx.doi.org/10.2139/ssrn.4713845">tech giants have</a>:<br />
<ul>
<li>"built a powerful reconnaissance network covering emerging competitive threats by investing in startups as corporate VCs and by cultivating relationships with financial VCs."</li>
<li>"accumulated massive quantities of data that are essential for many software and AI innovations, and they dole out access to this data and to their networks selectively."</li>
<li>"asked legislators to regulate the tech industry—in a way that will buttress incumbents."</li>
<li>"repeatedly bought potentially competitive startups in a way that has flown—until a few years ago—below the antitrust radar."</li>
</ul>
The authors detail many examples of each of these techniques, for example Facebook conditioning access to user data on the purchase of advertising, and Google's purchase of DoubleClick and YouTube. Interestingly, they contrast the recent purchasing of the tech giants with Cisco's famously successful purchases in the 90s:
<blockquote>
The Cisco story exemplifies how the venture capital market, as a market, is better at exploring a series of risky ideas than a firm with a single risk-averse gatekeeper. It also illustrates how the advantages of a large incumbent—in this case access to markets and existing customer relationships—can sometimes extract more market value out of a technology than a new entrant.
</blockquote>
The rapid evoluution of networking technology at the time meant that even Cisco, the largest company in the market, didn't have the R&D resources to explore all the opportunities. They depended upon VCs to fund the initial explorations, rewarding them by paying over the odds for the successes. Their market power then got the successes deployed much faster than a startup could.<br />
<h3>Why Is Cooption Bad?</h3>
The authors explain the <a href="https://dx.doi.org/10.2139/ssrn.4713845">harms of cooption</a>:<br />
<blockquote>
Our claim here is that the same dynamics that inhibit disruptive innovation by longstanding employees of large incumbents inhibit disruptive innovation by new employees from acquired startups.<br />
...<br />
The tech giants win from coopting disruption even though it destroys social value. In fact, they benefit in two ways. They make faster incremental progress on the sustaining innovations that they want. They get the new code, the valuable intellectual property, and the fresh ideas of the startup. And, critically, they also kill off a competitor. They no longer have to worry about the startup actually developing the more disruptive innovation and leapfrogging them or other tech giants acquiring the startup and using its assets to compete with them.
</blockquote>
And, by making the innovators from the startup rich, the acquirer greatly reduces their incentives for future innovation. <a href="https://blog.dshr.org/2024/02/the-stanford-digital-library-project.html">Andy Bechtolsheim</a> is an outlier.<br />
<h3>Remedies?</h3>
Lemley and Wansley, who seem to think in fours, make a set of four proposals for how these harms might be reduced:<br />
<ul>
<li><b>Unlocking Directorates</b> — under the Clayton Act "interlocking officers and directors between companies that compete, even in part, are illegal <i>per se</i> – without any inquiry into whether the companies in fact restrained competition because of their overlapping interests or whether the conduct offered procompetitive benefits." Companies with less than $4.1M in revenue are exampt, which excludes most startups; this should be revised.</li>
<li><b>Limiting Leveraging of Data and Networks</b> — "we would impose on incumbent tech monopolists a presumptive duty of nondiscrimination in access where the defendant (1) provides or sells data or network access to at least some unaffiliated companies and (2) refuses to provide or sell the same data or network access to the plaintiff company on comparable terms, but (3) the plaintiff does not operate a competing network or otherwise compete with the defendant in the market from which it collected the relevant data."</li>
<li><b>Regulating Regulation</b> — "Done right, regulation of technology can be beneficial and even necessary to the development of that technology, minimizing the risk of harm to third parties and ensuring that the world views the technology as safe and trustworthy. But all too often regulation has become a way to insulate incumbents from competition, with predictable results." The authors' suggestions exemplify this difficulty, being rather vague and aspirational.</li>
<li><b>Blocking Cooptive Acquisitions</b> — this is the most complex of the four proposals, and builds on <a href="https://www.jstor.org/stable/45467503"><i>Nascent Competitors</i></a> by C. Scott Hemphill & Tim Wu, who write:<br />
<blockquote>
We favor an enforcement policy that prohibits anticompetitive conduct that is reasonably capable of contributing significantly to the maintenance of the incumbent s market power. That approach implies enforcement even where the competitive significance of the nascent competitor is uncertain.
</blockquote>
Justifying blocking mergers because of a nascent threat that might never materialize is problematic. But it is only more so than the current way anti-trust works, by projecting likely harm to consumer welfare, which also might never materialize (although it almost always does). Lemley and Wansley explain <a href="https://dx.doi.org/10.2139/ssrn.4713845">the dilemma</a>:<br />
<blockquote>
antitrust enforcers need a strategy for blocking cooptive acquisitions that works within existing case law (or plausible improvements to that law) and is surgical enough to avoid chilling investment.
</blockquote>
Some cases are <a href="https://dx.doi.org/10.2139/ssrn.4713845">obvious</a>:<br />
<blockquote>
For cooptive acquisitions like Facebook/Instagram deal, we think Hemphill and Wu’s strategy makes sense. Zuckerberg’s email arguing for acquiring startups like Instagram because “they could be very disruptive to us” is a smoking gun of anticompetitive intent.
</blockquote>
But Lemley and Wansley go further, arguing for blocking megers based the startup's ability to <a href="https://dx.doi.org/10.2139/ssrn.4713845">innovate distruptive technology</a>:<br />
<blockquote>
Of course, an approach to policing startup acquisitions based on innovation capabilities need limits. Many startups have some innovation capabilities that could have a significant effect on competition. We can cabin enforcement in three ways—by focusing on specific technologies and specific firms and by looking at the cumulative effects of multiple acquisitions.
</blockquote>
Their examples of technologies include generative AI and virtual and augmented reality, both cases where it is already too late. The companies they identify "Alphabet, Amazon, Apple, Microsoft, and Meta" are all veterans of multiple acquisitions in these areas. But they argue that committing to
<a href="https://dx.doi.org/10.2139/ssrn.4713845">challenge fuure mergers</a>:<br />
<blockquote>
would create socially desirable incentives for startups. A startup developing one of the listed technologies would gain stronger incentives to turn its innovations into the products that its management team believed would garner the highest value on the open market—rather than the one most valuable to the tech giants. They would also gain stronger incentives to build a truly independent business and go public since an acquisition by the tech giants would be a less likely exit.
</blockquote>
</li>
</ul>
I think these would all be worthwhile steps, and I'm all in favor of updating anti-trust law and, even better, actually enforcing the laws on the books. But I am skeptical that the government can spot potentially disruptive technologies before the tech giants spot and acquire them. Especially since the government can't be embedded in the VC industry the way the tech giants are. Note that many of the harms Lemley and Wansley identify happen shortly after the acquisition. Would forcing Meta to divest Instagram at this late date restore the innovations the acquisition killed off?<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com0tag:blogger.com,1999:blog-4503292949532760618.post-82530659476818164282024-02-13T08:00:00.000-08:002024-02-20T10:28:19.024-08:00Clouds Over The MinesIn early December 2022 when I wrote skeptically about the economics of Bitcoin mining in <a href="https://blog.dshr.org/2022/12/foolish-lenders.html"><i>Foolish Lenders</i></a> the Bitcoin "price" was around $17K. It has now climbed 153% to around $43K and, below the fold, I am still posting skeptically about the economics of mining.<br />
<span><a name='more'></a></span>
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgYIjjIeeerm-Zzz0z4uKi7-O4gmkaveLT2Wgg0hpDpm6yK_sxoG7fdLWniOb84aINDI2WDOegrlfrI8MB_YE56oCM5FqLWRU6aoaaGuHLuaqZBABOiNXW0Q4S3VmiT8vePMIfwIpAFiE1Fk4unMFV8ANOnw3N9cuxA94DxXVyJLsg1pNKexP88G5IhvZi7/s932/MinersHODLings.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="537" data-original-width="932" height="115" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgYIjjIeeerm-Zzz0z4uKi7-O4gmkaveLT2Wgg0hpDpm6yK_sxoG7fdLWniOb84aINDI2WDOegrlfrI8MB_YE56oCM5FqLWRU6aoaaGuHLuaqZBABOiNXW0Q4S3VmiT8vePMIfwIpAFiE1Fk4unMFV8ANOnw3N9cuxA94DxXVyJLsg1pNKexP88G5IhvZi7/w200-h115/MinersHODLings.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2024-02-07/bitcoin-btc-price-outlook-clouded-by-falling-miner-reserves-before-halving">Source</a></td></tr></tbody></table>
The first clue that the future for miners is clouded comes in <a href="https://www.bloomberg.com/news/articles/2024-02-07/bitcoin-btc-price-outlook-clouded-by-falling-miner-reserves-before-halving"><i>Bitcoin Outlook Clouded by Falling Miner Reserves Ahead of April’s Halving</i></a> by Sidhartha Shukla and David Pan:<br />
<blockquote>
Bitcoin miners are getting a jump on an anticipated decline in revenue from the so-called halving in April, when the blockchain’s network protocol will reduce rewards for verifying transactions by half.<br />
<br />
Miner reserves — unsold Bitcoin held in digital wallets associated with the companies — have dropped by 8,400 tokens since the start of 2024 to 1.8 million, a level last seen in June 2021, according to data compiled by CryptoQuant. Analysts said the decrease indicates miners are selling tokens.
</blockquote>
The somewhat misleading graph of the miners HODL-ings actually shows only a 2% drop in the number of BTC from the peak in August 2022. But the "value" of those HODL-ings has risen 75% from $44.8B to $78.5B.
One way of looking at it is that the mining industry started August 2022 with 1.865M BTC and, in the 18 months since mined 821,250 BTC for a total of 2,686,250 but they now have only 1.825M BTC so they must have sold 861,250, or about 5% more than they mined.
<a href="https://www.bloomberg.com/news/articles/2024-02-07/bitcoin-btc-price-outlook-clouded-by-falling-miner-reserves-before-halving">Nevertheless</a>:<br />
<blockquote>
“Miners have begun to sell more of their coins to bolster balance sheets and fund growth capex ahead of tougher times for margins when block rewards are halved in April,” said Matthew Sigel, head of digital-asset research at VanEck. “After the halving, scale will matter even more.”
</blockquote>
Before the great EFT pump, it was generally believed that the frantic efforts to pump BTC over $30K suggested that the mining industry's break-even point was around $30K. On 6<sup>th</sup> October BTC was $26K and the hash rate was around 412M TH/s. Lets assume that the industry was breaking even at BTC=$30K and a hash rate of 412MTH/s, so the industry's costs were covered by 45K BTC/month or $1.35B/month. Assuming the increase in efficiency is roughly cancelled by less efficient operators entering, the hash rate is now 527 TH/s and so costs are around $1.73B. But income is around $1.94B/month so margins are around 12%.<br />
<br />
After the halvening, income is 22.5K BTC/month. At BTC=$43K, this is $968M/month or 56% of current costs. To maintain a 12% margin, costs need to be cut to $852M/month, or 49% of current costs. Alternatively, if costs stay the same, the Bitcoin "price" needs to increase to $86K in April.<br />
<br />
The industry isn't going to halve its costs in the next three months, and even massive printing of unbacked Tether isn't going to double the BTC "price", so "tougher times for margins" are certainly among the clouds overhead.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiMQw8ldiqYcm2zhKUDU2Bv3vYJSD9ml5KKG5R4RzSdtzRcoFsDMk-NpVq8GR9hz0K6lumcY3KwoIIwf5MwfWMpEVFhB7G1AH9C8enXxZHtynNj74CYlzk2LYsRgx_TG3i0D3ve2PLD-CeZok5qQWFKvULWFZVEddIxGc_s6edjCaYpIKfh8JlidgTt9UsS/s957/Hut8.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="524" data-original-width="957" height="109" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiMQw8ldiqYcm2zhKUDU2Bv3vYJSD9ml5KKG5R4RzSdtzRcoFsDMk-NpVq8GR9hz0K6lumcY3KwoIIwf5MwfWMpEVFhB7G1AH9C8enXxZHtynNj74CYlzk2LYsRgx_TG3i0D3ve2PLD-CeZok5qQWFKvULWFZVEddIxGc_s6edjCaYpIKfh8JlidgTt9UsS/w200-h109/Hut8.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;">Source</td></tr></tbody></table>
I am not the only skeptic. Short-seller J Capital Research investigated Hut 8, a large public miner, and on January 18<sup>th</sup> published <a href="https://www.jcapitalresearch.com/uploads/2/0/0/3/20032477/2024_01_18_hut.pdf"><i>The Coming HUT Pump and Dump</i></a>. As the chart shows, it was effective. The report featured a list of 22 different red flags, focused mainly on the recent merger between Hut 8 (HUT) and U.S. Bitcoin Corp. (USBTC) and asking:<br />
<blockquote>
Why then did HUT pay $745 mln to acquire this company and its planned payments?
</blockquote>
This is a good question, given <a href="https://www.jcapitalresearch.com/uploads/2/0/0/3/20032477/2024_01_18_hut.pdf">that</a>:<br />
<blockquote>
One person highly familiar with USBTC told us, “without the merger, [USBTC] would have done a structured bankruptcy.”
</blockquote>
It wasn't as if Hut 8 didn't have <a href="https://www.jcapitalresearch.com/uploads/2/0/0/3/20032477/2024_01_18_hut.pdf">problems of its own</a>:<br />
<blockquote>
Hut 8’s North Bay mining facility has been non-operational for an extended period of time, and problems at its Drumheller facility “have been causing miners to fail.”
</blockquote>
And the result of the merger is a <a href="https://www.jcapitalresearch.com/uploads/2/0/0/3/20032477/2024_01_18_hut.pdf">company that</a>:<br />
<blockquote>
has an industry-low efficiency rate and, post halving, will produce Bitcoin at a loss of close to $20,000 per Bitcoin at current spot prices.
</blockquote>
In other words, the merged company can barely make money now and cannot survive when the industry's income is halved in less than 3 months. This all looks like rats leaving the sinking ship with whatever they can carry. An impression reinforced by David Pan in <a href="https://www.bloomberg.com/news/articles/2024-02-07/bitcoin-miner-hut-8-hut-ceo-exits-three-weeks-after-short-seller-allegations"><i>Bitcoin Miner Hut 8 CEO Exits Three Weeks After Short-Seller Allegations </i></a>:<br />
<blockquote>
Hut 8 Corp., one of the largest publicly traded Bitcoin mining companies, named Asher Genoot to succeed Jaime Leverton as chief executive officer, three weeks after a short-seller released a report critical of its recent merger.<br />
<br />
The transition is effective immediately. Genoot served as the chief operating officer and the president of US Bitcoin Corp. Miami-based US Bitcoin, which has large-scale mining facilities across the US including Texas, completed its merger with then-Canadian miner Hut 8 in late 2023.<br />
<br />
The leadership transition comes amid increasing competition among the miners, a Bitcoin code update set to drastically reduce mining revenue in two months as well as the Jan. 18 report from short-seller J Capital Research alleging the merged company was a “pump and dump” waiting to happen. Hut 8 has disputed the claim.
</blockquote>
Genoot has <a href="https://www.jcapitalresearch.com/uploads/2/0/0/3/20032477/2024_01_18_hut.pdf">allegedly</a>:<br />
<blockquote>
abandoned several failed start-ups.
</blockquote>
His co-founder was USBTC's CEO and is now HUT's CSO/director <a href="https://www.jcapitalresearch.com/uploads/2/0/0/3/20032477/2024_01_18_hut.pdf">and</a>:<br />
<blockquote>
is a 30-year-old used-car salesman from Vancouver whose history is littered with involvement in SEC-defined pump-and-dumps, sporting share-price declines of 83%,
</blockquote>
The key inputs for profitable mining are low-cost power and state-of-the-art chips to use it as effficently as possible. Clouds are gathering over both of them.<br />
<br />
As regards access to cheap electricity, utilities are increasingly reluctant to provide it. Take, for example,
<a href="https://www.msn.com/en-ca/news/canada/crypto-mining-company-loses-bid-to-force-bc-hydro-to-provide-power/ar-BB1hPaZE"><i>Crypto mining company loses bid to force BC Hydro to provide power
</i></a> by Darryl Greer of The Canadian Press:<br />
<blockquote>
A cryptocurrency firm has lost a bid to force BC Hydro to provide the vast amounts of power needed for its operations, upholding the provincial government's right to pause power connections for new crypto miners. <br />
<br />
Conifex Timber Inc., a forestry firm that branched out into cryptocurrency mining, had gone to the B.C. Supreme Court to have the policy declared invalid.<br />
<br />
But Justice Michael Tammen ruled Friday that the government's move in December 2022 to pause new connections for cryptocurrency mining for 18 months was "reasonable" and not "unduly discriminatory."<br />
<br />
BC Hydro CEO Christopher O'Riley had told the court in an affidavit that the data centres proposed by Conifex would have consumed 2.5 million megawatt-hours of electricity each year.
</blockquote>
In the US, David Pan's <a href="https://www.bloomberg.com/news/articles/2024-02-01/bitcoin-miners-in-us-consume-up-to-2-3-of-nation-s-electricity"><i>US Bitcoin Miners Use as Much Electricity as Everyone in Utah </i></a> shows that regulators are starting to worry:<br />
<blockquote>
Bitcoin miners in the US are consuming the same amount of electricity as the entire state of Utah, among others, according to a new analysis by the US Energy Information Administration. And that’s considered the low end of the range of use.<br />
<br />
Electricity usage from mining operations represents 0.6% to 2.3% of all the country’s demand in 2023, according to the report released Thursday. It is the first time EIA has shared an estimate. The mining activity has generated mounting concerns from policymakers and electric grid planners about straining the grid during periods of peak demand, energy costs and energy-related carbon dioxide emissions.<br />
<br />
“This estimate of U.S. electricity demand supporting cryptocurrency mining would equal annual demand ranging from more than three million to more than six million homes,” the report said.
</blockquote>
Globally, this increase in demand is part of a bigger picture that is concerning, as Eamon Farhat reports in <a href="https://www.bloomberg.com/news/articles/2024-01-24/cryptocurrency-ai-electricity-demand-seen-doubling-in-three-years"><i>Electricity Demand at Data Centers Seen Doubling in Three Years</i></a>:<br />
<blockquote>
Global electricity demand from data centers, cryptocurrencies and artificial intelligence could more than double over the next three years, adding the equivalent of Germany’s entire power needs, the International Energy Agency forecasts in its latest report.<br />
<br />
There are more than 8,000 data centers globally, with about 33% in the US, 16% in Europe and close to 10% in China, with more planned. In Ireland, where data centers are developing rapidly, the IEA expects the sector to consume 32% of the country’s total electricity by 2026 compared to 17% in 2022. Ireland currently has 82 centers; 14 are under construction and 40 more are approved.<br />
<br />
Overall global electricity demand is expected to see a 3.4% increase until 2026, the report found. The increase, however, will be more than covered by renewables, such as wind, solar and hydro, and all-time high nuclear power.
</blockquote>
The first step towards stricter regulation is information collection which, as Kristoffer Tigue reports in <a href="https://arstechnica.com/tech-policy/2024/02/large-cryptocurrency-miners-in-us-now-have-to-report-energy-use-to-government/"><i>Large cryptocurrency miners in US now have to report energy use to government</i></a> has started:<br />
<blockquote>
The Biden administration is now requiring some cryptocurrency producers to report their energy use following rising concerns that the growing industry could pose a threat to the nation’s electricity grids and exacerbate climate change.<br />
<br />
The Energy Information Administration announced last week that it would start collecting energy use data from more than 130 “identified commercial cryptocurrency miners” operating in the US. The survey, which started this week, aims to get a sense of how the industry’s energy demand is evolving and where in the country cryptocurrency operations are growing fastest.<br />
<br />
“As cryptocurrency mining has increased in the United States, concerns have grown about the energy-intensive nature of the business and its effects on the US electric power industry,” the EIA said in a new report, following the announcement. “Concerns expressed to EIA include strains to the electricity grid during periods of peak demand, the potential for higher electricity prices, as well as effects on energy-related carbon dioxide emissions.”
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj14LmRDj5Qv_KdDA7OamCTjicibUyIkDrn8j5PcuhsVd6QEKnqomD_78uV1bvx_fnLtE9iBQzr2MRLvHqUugPXsxhwkM-n13jURWC9c9Na_6c6Aq0mlnYt0IAJjO_Uo34GO5aOpN7vq7KklNB85wr-5JenwvUxepMcR6IfW94LjmdIpVJSisgsy8X1UFr4/s1200/hash-rate.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="600" data-original-width="1200" height="100" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj14LmRDj5Qv_KdDA7OamCTjicibUyIkDrn8j5PcuhsVd6QEKnqomD_78uV1bvx_fnLtE9iBQzr2MRLvHqUugPXsxhwkM-n13jURWC9c9Na_6c6Aq0mlnYt0IAJjO_Uo34GO5aOpN7vq7KklNB85wr-5JenwvUxepMcR6IfW94LjmdIpVJSisgsy8X1UFr4/w200-h100/hash-rate.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.blockchain.com/explorer/charts/hash-rate">Source</a></td></tr></tbody></table>
Bitcoin is not nearly as decentralized as advocates claim. It appears a third of the industry is in Texas and thus subject to both US regulators and concerns about <a href="https://www.nytimes.com/2023/12/25/technology/bitrush-bitcoin-cryptocurrency-china.html">Chinese</a> <a href="https://www.nytimes.com/2023/10/13/us/bitcoin-mines-china-united-states.html">ownership</a>. Turner Wright reports that <a href="https://cointelegraph.com/news/bitcoin-hash-rate-drops-freezing-texas"><i>Bitcoin hash rate drops by 34% amid freezing temperatures in Texas</i></a>:
<blockquote>
A sudden freeze in Texas may have contributed to a 34% drop in the Bitcoin hash rate, as some miners were forced to curtail operations amid demand on the state’s energy grid.<br />
<br />
Beginning on Jan. 14, temperatures in many parts of Texas dropped below freezing for one of the first times since a massive ice storm in February 2023. According to data from YCharts, the total Bitcoin network hash rate fell from more than 629 exahashes per second (EH/s) on Jan. 11 to roughly 415 EH/s on Jan. 15 — a 34% drop. The analytics site reported the hash rate increased to more than 454 EH/s on Jan. 16 as temperatures in Austin briefly rose above freezing during the day.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiphQCQ1Q-bLryzRusVK3OLU3VGNCs_uzgYqgRgNnTdV_hcntuILWhBoVGCjpDVHx8ies7Zr_HSw1p2OYkyZb6eDA4N9bEY7m8Y52aELEjyoiT6tEe_VG1fJOWtBPYkHe40a-an0DJ2-mQ8oDUlE4VKZ2QlCmjqI8_iFoSFRIaNyxUDKLzj73P3ACVYs1Yn/s969/RigShipments.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="610" data-original-width="969" height="126" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiphQCQ1Q-bLryzRusVK3OLU3VGNCs_uzgYqgRgNnTdV_hcntuILWhBoVGCjpDVHx8ies7Zr_HSw1p2OYkyZb6eDA4N9bEY7m8Y52aELEjyoiT6tEe_VG1fJOWtBPYkHe40a-an0DJ2-mQ8oDUlE4VKZ2QlCmjqI8_iFoSFRIaNyxUDKLzj73P3ACVYs1Yn/w200-h126/RigShipments.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2024-02-08/bitcoin-mining-why-ethiopia-is-attracting-chinese-crypto-miners">Source</a></td></tr></tbody></table>
It isn't just that Texas has maybe a third of the industry, it is also that the US industry is growing much faster than anywhere else. Figures from Luxor Logistics show that in 2023 the US consumed <a href="https://www.bloomberg.com/news/articles/2024-02-08/bitcoin-mining-why-ethiopia-is-attracting-chinese-crypto-miners">nearly two-thirds of all new mining rigs</a>. The figures come from <a href="https://www.bloomberg.com/news/articles/2024-02-08/bitcoin-mining-why-ethiopia-is-attracting-chinese-crypto-miners"><i>Chinese Bitcoin Miners Find a New Crypto Haven in Ethiopia</i></a> by David Pan and Fasika Tadesse:<br />
<blockquote>
Ethiopia has emerged as a rare opportunity for all firms that mine the original cryptocurrency, as climate change and power scarcity fuel a backlash against the $16 billion-a-year industry (at Bitcoin’s current price) elsewhere. But it holds special appeal for Chinese companies, which once dominated Bitcoin mining but have struggled to compete with local rivals in Texas, the current hub.<br />
</blockquote>
The haven isn't without its own <a href="https://www.bloomberg.com/news/articles/2024-02-08/bitcoin-mining-why-ethiopia-is-attracting-chinese-crypto-miners">clouds</a>:<br />
<blockquote>
It is also a risky gamble, for the companies and Ethiopia alike. A succession of developing countries like Kazakhstan and Iran initially embraced Bitcoin mining, only to turn on the sector when its energy use threatened to fuel domestic discontent. China’s reign as the epicenter of Bitcoin mining came to an abrupt end in 2021, when the government banned it. Dozens of companies were forced to leave.<br />
<br />
Ethiopian officials are wary of the controversy that accompanies Bitcoin mining, according to industry executives who spoke on condition of anonymity to avoid jeopardizing government relations. Even after new generation capacity came online, almost half the population live without access to electricity, making mining a delicate topic. At the same time, it represents a potentially lucrative source of foreign-exchange earnings.<br />
...<br />
The reliance on abundant power is also a major vulnerability because it can put miners in competition for electricity with factories and households, exposing them to political backlash.<br />
<br />
When Kazakzstan imposed fresh curbs and taxes on miners, “it basically killed the industry,” said Hashlabs co-founder Alen Makhmetov. Two years after the clampdown, his 10-megawatt facility there is still sitting idle.<br />
<br />
And in an era when rising temperatures wreak havoc around the world, Bitcoin mining is increasingly seen as a contributor to global warming that doesn’t serve any productive purpose — even though miners have claimed they’re increasingly tapping clean energy. A study by United Nations University published in October estimated that two-thirds of the electricity used for Bitcoin mining in 2020 and 2021 was generated using fossil fuels.
</blockquote>
Developing countries aren't the only ones where miners face "domestic discontent". <a href="https://www.nytimes.com/2024/02/03/us/bitcoin-arkansas-noise-pollution.html"><i>Anxiety, Mood Swings and Sleepless Nights: Life Near a Bitcoin Mine</i></a> by Gabriel J.X. Dance reports on an example in Arkansas:<br />
<blockquote>
The Arkansas Data Centers Act, popularly called the Right to Mine law, offers Bitcoin miners legal protections from communities that may not want them operating nearby. Passed just eight days after it was introduced, the law was written in part by the Satoshi Action Fund, a nonprofit advocacy group based in Mississippi whose co-founder worked in the Trump administration rolling back Obama-era climate policies.
</blockquote>
The law <a href="https://www.nytimes.com/2024/02/03/us/bitcoin-arkansas-noise-pollution.html">ins't popular</a>:<br />
<blockquote>
A furious backlash has some lawmakers considering a statewide ban.
</blockquote>
The Satoshi Action Fund <a href="https://www.nytimes.com/2024/02/03/us/bitcoin-arkansas-noise-pollution.html">over-reached</a>:<br />
<blockquote>
Despite efforts to build bipartisan support, the Satoshi fund has succeeded predominantly in red states. But in Arkansas, where the state legislature is dominated by Republicans, it is conservatives who have led calls to repeal the law, including Senator Bryan King, a poultry farmer whose district includes a property purchased by one of the companies tied to the Chinese government. He said it was not fair that the Bitcoin operators received special protections under the law, which shields them from “<a href="https://www.arkleg.state.ar.us/Home/FTPDocument?path=%2FACTS%2F2023R%2FPublic%2FACT851.pdf">discriminatory industry specific regulations and taxes</a>,” including noise ordinances and zoning restrictions.
</blockquote>
At least the Ethiopian mines don't emit much CO2, they run on <a href="https://www.bloomberg.com/news/articles/2024-02-08/bitcoin-mining-why-ethiopia-is-attracting-chinese-crypto-miners">hydropower</a>:<br />
<blockquote>
The opening of the GERD project increased Ethiopia’s installed generation capacity to 5.3 gigawatt, 92% of which comes from hydropower, a renewable energy source.<br />
<br />
Once GERD is fully completed, Ethiopia’s generation capacity will double, according to Ethiopian Electric Power. It charges Bitcoin miners a fixed rate of 3.14 US cents per kilowatt hour for electricity drawn from substations, Marketing and Business Development Director Hiwot Eshetu said in an interview.<br />
<br />
While that’s similar to the average in Texas, rates in the Lone Star State can swing wildly, Luxor’s Vera said, making profits there less predictable. In Ethiopia, the price will fall once miners connect directly to power plants, according to Hiwot.
</blockquote>
But if the utility can make money selling power to the mines right by the dam they have little incentive to build out the grid than could get the power to the unserved half of the population.<br />
<br />
As regards the longer-term issue of access to state-of-the-art chips, it is important to note that the best mining chips are sold by Bitmain, a Chinese company, but manufactured at TSMC in Taiwan. There are two major risks here. The first is that the US appears determined to prevent China <a href="https://www.cnn.com/2023/10/18/tech/us-china-chip-export-curbs-intl-hnk/index.html">importing leading-edge chips</a> and the <a href="https://www.theguardian.com/technology/2024/jan/02/asml-halts-hi-tech-chip-making-exports-to-china-reportedly-after-us-request">equipment to make them</a>. China remains at least a <a href="https://arstechnica.com/gadgets/2024/02/china-close-to-shipping-5nm-chips-despite-western-curbs/">generation and a half</a> behind TSMC and Samsung, and reportedly has poor yields on its leading-edge process. These restrictions could well prevent Chinese mining companies acquiring leading-edge rigs, and might cause TSMC problems in fab-ing Bitmain products.<br />
<br />
Second, there is the looming threat of a Chinese blockade or even invasion of Taiwan. Of course, difficulties for Bitcoin miners are hardly the major impact if these threats are made good. One might think that, even if supplies of new mining chips were cut off, existing rigs would continue working. In the short term they would, but there is a long history of <a href="https://blog.dshr.org/2022/05/generally-accepted-accounting-principles.html">rigs being obsolete</a> after <a href="https://doi.org/10.1016/j.resconrec.2021.105901">about 18 months</a>. So they aren't designed or operated for longevity.<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com5tag:blogger.com,1999:blog-4503292949532760618.post-73612101844794119352024-02-08T08:00:00.000-08:002024-02-08T08:00:00.147-08:00Tracing The Pig Butchers <table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjZuJMbwKu8ueXvgRZ0zbwb2FIiebX-fzffvdFFG7GvL4q0ILSTy_skuzBXdSv0g0JNKBr2J30ZNPKxe8xXYbkFbbhAi1hyJvBY4JJzxnAKmfFOpx_GWpgxhaOi4tTMCkfTnMuvY4vCWYZIczr67FbEXXydkuyKLe4CJk4AbkXeamwuTIAod9X_0kpwiFiY/s2200/Vicky.jpg" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="2200" data-original-width="1467" height="200" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjZuJMbwKu8ueXvgRZ0zbwb2FIiebX-fzffvdFFG7GvL4q0ILSTy_skuzBXdSv0g0JNKBr2J30ZNPKxe8xXYbkFbbhAi1hyJvBY4JJzxnAKmfFOpx_GWpgxhaOi4tTMCkfTnMuvY4vCWYZIczr67FbEXXydkuyKLe4CJk4AbkXeamwuTIAod9X_0kpwiFiY/w133-h200/Vicky.jpg" width="133" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/newsletters/2023-08-17/crypto-text-message-scams-lead-back-to-cambodia-slave-labor">"Vicky"</a></td></tr></tbody></table>
Chapter 18 of Zeke Faux's <a href="https://www.amazon.com/Number-Go-Up-Cryptos-Staggering-ebook/dp/B0BYV2Y3XT/ref=sr_1_1"><i>Number Go Up: Inside Crypto's Wild Rise and Staggering Fall</i></a> is entitled "Pig Butchering". It starts when he receives a supposed wrong-number text from a "Vicky":<br />
<blockquote>
I showed my phone to my friend and explained that I was stringing Vicky along because I’d heard about a new kind of investment fraud that often started with a random text message. I had a hunch that this was why “Vicky” was texting me. The scam was called “pig butchering” because the scammers liked to build up the victim’s confidence with a pretend romantic relationship and made-up investment gains before stealing all their money in one fell swoop—like how hogs are fattened up before their slaughter.
</blockquote>
This is a romance- and cryptocurrency-enabled version of the "Wee Forest Folk" scam we described in our 2003 <a href="http://dx.doi.org/10.1145/945445.945451">SOSP paper</a>.<br />
<br />
Below the fold, I look into the details of pig-butchering scams, and how the tracing techniques I discussed in <a href="https://blog.dshr.org/2024/01/criming-on-blockchain.html"><i>Criming On The Blockchain</i></a> are being applied to identify the cryptocurrency companies facilitating it.<br />
<span><a name='more'></a></span>
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj93HUM1LUCu7NySHdWO7TYSQPSyk7XsiUQ6WjzJAiM8XcQPyFPea_cQhAYNlN9O3Cj6P_JphOpi9XFPsbYBcKnZozw56OYbnN0bQrRsE0OKLWYvIRilEXMwgof59kxKjOroVJ5koKOFVPgSxSBRTNB7ku3d_3oZTbe64z6jqrMQrraKxSbflG_9aNWbo27/s1500/ZekeFauxBook.jpg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="1500" data-original-width="987" height="200" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj93HUM1LUCu7NySHdWO7TYSQPSyk7XsiUQ6WjzJAiM8XcQPyFPea_cQhAYNlN9O3Cj6P_JphOpi9XFPsbYBcKnZozw56OYbnN0bQrRsE0OKLWYvIRilEXMwgof59kxKjOroVJ5koKOFVPgSxSBRTNB7ku3d_3oZTbe64z6jqrMQrraKxSbflG_9aNWbo27/w132-h200/ZekeFauxBook.jpg" width="132" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.amazon.com/Number-Go-Up-Cryptos-Staggering-ebook/dp/B0BYV2Y3XT/ref=sr_1_1">Source</a></td></tr></tbody></table>
Faux strung "Vicky" along <a href="https://www.amazon.com/Number-Go-Up-Cryptos-Staggering-ebook/dp/B0BYV2Y3XT/ref=sr_1_1">and</a>:<br />
<blockquote>
My short-lived texting flirtation with Vicky confirmed something that I had suspected: The scammers were using Tether to move the money. I spoke with several other pig-butchering victims, and they said they’d been asked to send Tether too.<br />
...<br />
From the scammer’s perspective, Tether was a clear improvement on bribing bankers or money mules. It was instant, there was no recourse for refunds, and it didn’t ask for anyone’s name or address. And unlike other cryptocurrencies, its value didn’t fluctuate from hour to hour, making it less scary for potential victims and easier to manage for criminals.
</blockquote>
<a href="https://www.amazon.com/Number-Go-Up-Cryptos-Staggering-ebook/dp/B0BYV2Y3XT/ref=sr_1_1">How much Tether are we talking?</a>:<br />
<blockquote>
News stories revealed that people were losing huge amounts of money. A project finance lawyer in Boston with terminal cancer handed over $2.5 million. A divorced mother of three in St. Louis was defrauded of $5 million. A twenty-four-year-old social media producer in Tennessee lost $300,000 she inherited from the sale of her childhood home.<br />
...<br />
I came across a group that claimed to be raising money dedicated to helping victims. It was called the Global Anti-Scam Organization. The group said it had helped a huge number of pig-butchering victims: 1,483 worldwide, who’d lost more than $250 million combined.
</blockquote>
Despite Tether's ability to <a href="https://medium.com/chainargos/what-happens-when-tether-freezes-your-tether-5a8ece2bd508">freeze wallets</a>, <a href="https://www.amazon.com/Number-Go-Up-Cryptos-Staggering-ebook/dp/B0BYV2Y3XT/ref=sr_1_1">Faux writes</a>:<br />
<blockquote>
Icetoad and other volunteers from the Global Anti-Scam group told me that Tether refused to help them by freezing accounts or seizing stolen money, even when presented with evidence that an account held the proceeds of fraud.<br />
...<br />
In another exchange provided by Global Anti-Scam, Tether told a Hong Kong police official that it did have the ability to intervene. But the company still refused, saying the case was too small. “We must sometimes decline fraud cases involving relatively small amounts of stolen USDT,” the Tether representative wrote, using the ticker symbol for the cryptocurrency. It said they would only intervene if the case was “directly connected to acts of violence.”
</blockquote>
Faux got Rich Sanders of CipherBlade to trace his <a href="https://www.amazon.com/Number-Go-Up-Cryptos-Staggering-ebook/dp/B0BYV2Y3XT/ref=sr_1_1">payment to "Vicky"</a>:<br />
<blockquote>
a small dark circle represented Vicky Ho’s wallet. This was where I’d sent the eighty-one Tethers. Sanders explained that the wallet was active for about two months and received many deposits from addresses that were known to be associated with U.S. and Canadian Canadian crypto exchanges. There were transfers of $3,600, $180, $400, $500, and $9,774, always in Tether. These were other people who’d been scammed by the same operation, Sanders said. In other words, I wasn’t Vicky’s only love interest. Once Vicky Ho had collected the money, they sent it to another numbered address, represented on the screen by a larger white circle. This one was likely controlled by the scammers too, Sanders said, and it held $9.4 million in Tethers. From there, many of the Tethers were sent to addresses associated with the crypto exchange Binance, and to others that belonged to Sam Bankman-Fried’s FTX.
</blockquote>
<a href="https://www.amazon.com/Number-Go-Up-Cryptos-Staggering-ebook/dp/B0BYV2Y3XT/ref=sr_1_1">Global Anti-Scam told Faux that</a>:<br />
<blockquote>
most butchering scams were orchestrated by gangsters based in Cambodia or Myanmar. These bosses would lure young men and women from across Southeast Asia to move abroad with the promise of well-paying jobs in customer service or online gambling. Then, when the workers arrived, they’d be held captive and forced to work on online scams. Icetoad and his colleagues told me there were thousands of people who’d been tricked this way. Entire office towers were filled with floor after floor of people forced to send spam messages around the clock, under threat of torture or death. These were the people using Tether to move their money.
</blockquote>
His next chapter recounts his trip to Cambodia to see the pig-butchering campuses. Like the rest of the book, it is well worth reading. Conditions in these campuses were <a href="https://www.bloomberg.com/news/newsletters/2023-08-17/crypto-text-message-scams-lead-back-to-cambodia-slave-labor">horrific</a>:<br />
<blockquote>
Workers who missed quotas were beaten, starved, made to hit one another. One said he’d seen people forcibly injected with methamphetamine to increase productivity. Two others said they’d seen workers murdered, with the deaths passed off as suicides. They said the bosses would buy and sell captive laborers like livestock.<br />
...<br />
Local news reports described a string of suspicious deaths near Chinatown—one body was found hanging at a construction site, and another corpse was dug up handcuffed from a shallow grave in a field nearby. A local vendor told a Cambodian outlet that there had been many suicides at the complex. “If an ambulance doesn’t go inside at least twice a week, it is a wonder,” he said.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgxigDJLqPqz8XzZlIJcor1eoAlk73uofS2zjkQ_L0GyYtMbzE1jlFL0zgqCpFROXfJwNo76Fk-8WT3xBrAP_GKBApJcZtpI6P1Mc3gaMw0nyTDGdI3x5NQ7gpNFrMMMIhyNf-nTOrej9okin9n0dQBZprNYFesCGaHXVXKUvOAJ0z4mx6-ZlkQTszbi3FM/s1920/Wang.jpg" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="1351" data-original-width="1920" height="141" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgxigDJLqPqz8XzZlIJcor1eoAlk73uofS2zjkQ_L0GyYtMbzE1jlFL0zgqCpFROXfJwNo76Fk-8WT3xBrAP_GKBApJcZtpI6P1Mc3gaMw0nyTDGdI3x5NQ7gpNFrMMMIhyNf-nTOrej9okin9n0dQBZprNYFesCGaHXVXKUvOAJ0z4mx6-ZlkQTszbi3FM/w200-h141/Wang.jpg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.reuters.com/investigates/special-report/fintech-crypto-fraud-thailand/">Wang and cybercrime cop</a>></td></tr></tbody></table>
Faux wasn't alone in reporting on the pig-butchers. Poppy McPherson and Tom Wilson of Reuters published a special report entitled <a href="https://www.reuters.com/investigates/special-report/fintech-crypto-fraud-thailand/"><i>Crypto scam: Inside the billion-dollar ‘‘pig-butchering‘‘ industry</i></a>. They traced funds flowing to Wang Yicheng in Bangkok:<br />
<blockquote>
At a Thai police headquarters in October 2022, Chinese businessman Wang Yicheng congratulated one of Bangkok’s most senior cybercrime investigators on his recent promotion, presenting the official with a large bouquet of flowers wrapped in red paper and a bow.<br />
<br />
Wang, the vice president of a local Chinese trade group, wished the new cybercrime investigator “smooth work and new achievements,” according to the group’s website, which displays photographs of the event.<br />
<br />
Over the past two years, Wang has forged relationships with members of Thailand’s law-enforcement and political elite, the trade group’s online posts show. During that time, a cryptocurrency account registered in Wang’s name was receiving millions of dollars linked to a type of cryptocurrency investment scam known as pig butchering, a Reuters investigation has found.<br />
<br />
In total, crypto worth more than $90 million flowed into the account between January 2021 and November 2022, according to registration documents and transaction logs reviewed by Reuters. Of that, at least $9.1 million came from a crypto wallet that U.S. blockchain analysis firm TRM Labs said was linked to pig-butchering scams. Two other major crypto-tracking firms also said the account received funds linked to such scams.<br />
...<br />
Lisa Wolk, a blockchain intelligence analyst at TRM, said the crypto account in Wang’s name “is a node in a money laundering network and not necessarily the ultimate recipient of funds.” Reuters was unable to determine whether anyone else benefited from the account or used Wang’s identity to open it.<br />
</blockquote>
Where was <a href="https://www.reuters.com/investigates/special-report/fintech-crypto-fraud-thailand/">Wang's account?</a>:<br />
<blockquote>
The crypto account registered to Wang was held at Binance, the world’s largest crypto exchange, according to three blockchain analysis firms. Asked about the account, Binance spokesperson Jessica Jung declined to comment on individual users or Reuters’ findings. In an August post on its website, the company said the number of reports of pig-butchering scams it had received this year was double that of 2022, an increase it attributed to an influx of inexperienced crypto investors and scammers looking to exploit them.<br />
...<br />
In April, the U.S. Department of Justice said it seized about $112 million worth of crypto linked to pig-butchering scams, without identifying suspects. A warrant that resulted in the seizure of more than half that amount specified a Binance account registered in Thailand.<br />
</blockquote>
Wang's account moved a <a href="https://www.reuters.com/investigates/special-report/fintech-crypto-fraud-thailand/">lot of Tether</a>:<br />
<blockquote>
The crypto account in Wang’s name was registered in November 2020, according to the financial records Reuters reviewed. The three blockchain-analysis firms determined the account was with Binance. The documents, which relate to the Binance account, span two years to November 2022. They show that the account was accessed mostly from Bangkok.<br />
<br />
Deposits started arriving in the account in early 2021 and quickly increased in size to include sums of more than $100,000, the financial records show.<br />
...<br />
In 2022, the volume of crypto reaching the account in Wang’s name mushroomed to almost $79 million, a near six-fold increase on the previous year, the account records seen by Reuters show.<br />
<br />
Coinfirm, the blockchain analysis firm that reviewed the transaction records for Reuters, said deposits from funds linked to pig-butchering scams it previously investigated began entering the account as early as February 2022. Stolen crypto was moved to the account via a “complex layering scheme,” involving multiple different wallets, Coinfirm’s then-head of fraud investigations Roman Bieda told Reuters in May. The crypto moved between “dozens” of wallets and was mixed with funds from other sources, he said.<br />
<br />
More than $102,000 originating from fraudsters’ wallets that the California man identified were deposited into the account in Wang’s name between June 26 and Oct. 23, 2022, according to Coinfirm.<br />
<br />
Nearly all of the crypto deposited in the Wang-registered account was moved to other wallets, the transaction records reviewed by Reuters show. Between January 2021 and October 2022, about $87.5 million was transferred to almost 50 other crypto wallets, including at least five registered at regionally based crypto exchanges, Coinfirm said.
</blockquote>
The pigs weren't <a href="https://www.reuters.com/investigates/special-report/fintech-crypto-fraud-thailand/">butchered in Bangkok</a>:<br />
<blockquote>
Several scams connected to deposits made into the account in Wang’s name were run from an industrial park on the Myanmar-Thai border, one of the blockchain analysis firms told Reuters. Workers are trafficked to the area, known as KK Park, by gangs that force them to con people online, according to two former workers and groups that support workers or scam victims.<br />
</blockquote>
This was clearly a different pig abattoir from the Cambodia-based one Faux investigated.<br />
<br />
A more detailed account of tracing the money flows of the pig butchers can be found in <a href="https://dx.doi.org/10.2139/ssrn.4687064"><i>Connecting Chinese and American Scam Victims</i></a>, a collaboration betwween Jonathan Reiter's company ChainArgos and BitTrace, a Chinese blockchain tracing company. The abstract reads:
<blockquote>
The world is experiencing an epidemic of online scams with at least tens of billions of dollars lost across dozens of countries. One particular class of scam known as “pig butchering” has grown dramatically and often involves the use of cryptocurrency both to collect funds from victims and to launder the proceeds. Here we are going to explore the use of cryptocurrency in pig butchering scams beginning with victims in both the People’s Republic of China and United States of America, and demonstrating the remarkable degree of similarity for cases that have no reason <i>a priori</i> to be similar at all.<br />
<br />
Specifically we will show that scammers with victims in both these countries share cryptocur- rency addresses, use overlapping sets of money-laundering services, and are therefore likely parts of the same group or syndicate. Our analysis begins with reports from victims in both countries where we find source victim and scam wallet addresses. From there we are able to trace scam proceeds through the cryptocurrency ecosystem to prove these connections. Finally, a range of exchanges and other service providers are identified through a combination of on-chain and documentary research, to demonstrate the cross-border nature of modern scams leveraging cryptocurrencies, and highlight the challenges facing law enforcement agencies globally when it comes to such scams.
</blockquote>
The two teams use press accounts of two cases with victims in China and legal filings from two cases with victims in Florida and California. They provide detailed traces, including wallet addresses, of flows from the victims to exchanges including Binance, Bitkub, FTX, OKX, Coinbase, Huboi, Maskex, Paribu and Peatio amounting to $635M. In particular, one of the Binance wallets was one of those identified in Reuters' reporting.
The paper <a href="https://dx.doi.org/10.2139/ssrn.4687064">concludes</a>:
<blockquote>
<ol>
<li>Similar simultaneously-run scams have victims in both China and the United States.</li>
<li>Funds taken from both jurisdictions flow through common service providers.</li>
<li>The sums involved range from the tens to possibly hundreds of millions of dollars.</li>
</ol>
</blockquote>
The ChainArgos team followed up with <a href="https://medium.com/chainargos/laundering-the-proceeds-of-crime-cryptos-killer-app-02686113ff66"><i>Laundering the Proceeds of Crime: Crypto’s Killer App?</i></a>:<br />
<blockquote>
Now we are going to connect a few more dots to argue that scamming and the laundering of scam proceeds was not just a use case for cryptocurrencies in Asia, but likely a major driver of flows through several well-known industry players.<br />
<br />
This may well even be, empirically, crypto’s killer app.
</blockquote>
They note that <a href="https://protos.com/genesis-block-ftx-had-more-than-prostitutes-in-thailand/">Genesis Block</a>, whose <a href="https://www.youtube.com/watch?v=pKeMfxLV8K8">Charles Yang admitted</a> maintaining bank accounts in South-East Asia under false pretences, appears to have been the <a href="https://www.dirtybubblemedia.com/p/blockchain-evidence-suggests-genesis">on- and off-ramp for FTX</a> in the region. The <a href="https://dx.doi.org/10.2139/ssrn.4687064">earlier paper</a> showed <a href="https://medium.com/chainargos/laundering-the-proceeds-of-crime-cryptos-killer-app-02686113ff66">that</a>:<br />
<blockquote>
FTX was involved in processing over US$250 million in flows downstream from scams and Bitkub, Thailand’s largest crypto exchange, was involved in over US$100 million.
</blockquote>
Could this be "<a href="https://medium.com/chainargos/laundering-the-proceeds-of-crime-cryptos-killer-app-02686113ff66">banking Cambodia's unbanked</a>"?:<br />
<blockquote>
Cambodia’s population is only 16 million, with a per capita GDP of about US$1,600 — it’s hard to imagine Cambodia driving hundreds of millions of dollars worth of stablecoin flows.
</blockquote>
They refer to the <a href="https://www.youtube.com/watch?v=pKeMfxLV8K8">interview with Charles Yang</a> and <a href="https://medium.com/chainargos/laundering-the-proceeds-of-crime-cryptos-killer-app-02686113ff66">note that</a>:<br />
<blockquote>
Genesis Block’s head trader explicitly talks about scam proceeds leading to bank account freezes and how they actively manage this problem.<br />
<br />
So Genesis Block were clearly a downstream service provider and knew that at least some of their clients were, or at least were dealing with, criminals.<br />
<br />
With regards to Genesis Block’s Chinese client base, in a <a href="https://www.youtube.com/watch?v=l86eSpiUbKw">different interview</a> the same person says:<br />
<blockquote><i>
We’ve been seeing a lot of our partners pretty much disappear…a lot of our friends over there are kind of in big trouble…it’s not really about crypto it’s more about the source of the RMB… [which was] a lot of pyramid schemes, a lot of scams.
</i></blockquote>
Genesis Block were clearly dealing with people who were either criminals themselves, or service providers to criminals, and Genesis Block appeared comfortable talking about such activities on camera.
</blockquote>
Genesis Block was <a href="https://medium.com/chainargos/laundering-the-proceeds-of-crime-cryptos-killer-app-02686113ff66">plugged in to the top level of cryptocurrency infrastructure</a>:<br />
<blockquote>
Genesis Block OTC was on the <a href="https://www.dirtybubblemedia.com/p/a-walled-garden-full-of-snakes-the">Signet client list</a>, as were many of the other exchanges referenced in our earlier work.<br />
<br />
A service provider that talked openly about dealing with criminals was active on a platform another commentator described as “<a href="https://www.dirtybubblemedia.com/p/a-walled-garden-full-of-snakes-the">a walled garden filled with snakes</a>.”<br />
<br />
That service provider (Genesis Block) was integrated with FTX/Alameda Research), which was both the <a href="https://medium.com/chainargos/usdt-on-tron-ftx-wtf-is-really-happening-ef0cb807019a">largest recipient of new USDT</a> and a major exchange operator with <a href="https://www.nytimes.com/2023/11/02/technology/sam-bankman-fried-fraud-trial-ftx.html">myriad, sometimes-remarkably-creative, schemes to access fiat currency</a>.
</blockquote>
This leads ChainArgos to <a href="https://medium.com/chainargos/laundering-the-proceeds-of-crime-cryptos-killer-app-02686113ff66">argue that</a>:<br />
<blockquote>
it is entirely plausible that:<br />
<ul>
<li>Genesis Block’s client base included many of these scammers and/or crimionals;</li>
<li>the outsized involvement of FTX was this flow; and/or</li>
<li>so was the Bitkub flow, via Genesis Block or otherwise;</li>
</ul>
and this isn’t just plausible, it’s provable, because the illicit flows documented in our prior work were laundered by someone.<br />
<br />
If it wasn’t this route it means there is at least one more multi-billion-dollar laundering pipeline connected to this activity we know nothing about.<br />
<br />
That is possible of course, but seems unlikely.<br />
<br />
Certainly the simplest explanation is that no second as-yet-undiscovered group of scammers or criminals exists and it’s all the same organization.<br />
<br />
Now remember these actors specifically mentioned that the traditional financial system was blocking them and freezing their accounts.
</blockquote>
Thus ChainArgos <a href="https://medium.com/chainargos/laundering-the-proceeds-of-crime-cryptos-killer-app-02686113ff66">concludes</a>:<br />
<blockquote>
We are looking at a gigantic amount of bad flow that was caught by banks and how cryptocurrencies were used to circumvent those controls.<br />
<br />
The contents of those frozen accounts could have been, and at least partially would have been, sent back to the victims instead of being irreversibly exfiltrated.<br />
<br />
So this may well be crypto’s real killer app — irreversible cross-border transfers for criminals.
</blockquote>
Taken together, what these documents show is lives being wrecked, both by the scams themselves leading to suicides, and by the human trafficing that provides the scammers workforce, leading to beatings, imprisonment and suicides. Pillars of the cryptocurrency community are enabling this:<br />
<ul>
<li>Tether, which refused to freeze wallets despite detailed blockchain evidence of pig-butchering even from law enforcement. Contrast this with Tether's behavior in the case of <a href="https://medium.com/chainargos/what-happens-when-tether-freezes-your-tether-5a8ece2bd508">The Victim</a> documented by Patrick Tan, where a simple request from law enforcement with only suspicion froze a wallet.</li>
<li>Major exchanges, including Binance, FTX, Huboi and many others, were on notice that they were handling funds from the pig butchers. Take Binance <a href="https://www.reuters.com/investigates/special-report/fintech-crypto-fraud-thailand/">for example</a>:<br />
<blockquote>
Binance publicly said it assisted Thai police with a probe into “a significant pig butchering scam,” and that about $277 million of assets were confiscated.
</blockquote>
and they <a href="https://www.reuters.com/investigates/special-report/fintech-crypto-fraud-thailand/">blamed the victims</a>:<br />
<blockquote>
the number of reports of pig-butchering scams it had received this year was double that of 2022, an increase it attributed to an influx of inexperienced crypto investors and scammers looking to exploit them.
</blockquote>
and they responded to a <a href="https://www.reuters.com/investigates/special-report/fintech-crypto-fraud-thailand/">Dept. of Justice warrant</a> about these funds.</li>
</ul>
Thus these businesses cannnot claim ignorance of the flow of pig-butchering funds, or of their scale. But they are clearly motivated to turn a blind eye to them whenever possible.<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com6tag:blogger.com,1999:blog-4503292949532760618.post-12832702034827622782024-02-01T08:00:00.000-08:002024-02-01T08:00:00.135-08:00The Stanford Digital Library Project<a href="https://doi.org/10.1145/205323.205338"><i>The Stanford Digital Library Project</i></a> stated its goal thus:<br />
<blockquote>
The Stanford Integrated Digital Library Project will develop enabling technologies for an integrated “virtual” library to provide an array of new services and uniform access to networked information collections. The Integrated Digital Library will create a shared environment linking everything from personal information collections, to collections of conventional libraries, to large data collections shared by scientists.
</blockquote>
Stanford librarians <a href="http://diglib.stanford.edu:8091/diglib/pub/people.shtml">Vicky Reich and Rebecca Wesley</a> provided the "library" input for the research.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhVUMd9QzEAJGFvn6ExiorNPalduqn-l7tac26cnhh30NqeqJpdVOwRUuaVgZ25fVZe9dgCdpIHnKG74xVG-aaEvGEgjJb9IdOR8x8JBVQFprIWh_wHe_MfjtOy4SwKeNjIN_v3HL04EmYTfYwO1BUXQQ5lPI9cz9DBTa2grkiEQaVL4hoI2dUlMfxMyPPa/s482/Google.png" style="clear: right; display: block; margin-left: auto; margin-right: auto; padding: 1em 0px; text-align: center;"><img alt="" border="0" data-original-height="130" data-original-width="482" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhVUMd9QzEAJGFvn6ExiorNPalduqn-l7tac26cnhh30NqeqJpdVOwRUuaVgZ25fVZe9dgCdpIHnKG74xVG-aaEvGEgjJb9IdOR8x8JBVQFprIWh_wHe_MfjtOy4SwKeNjIN_v3HL04EmYTfYwO1BUXQQ5lPI9cz9DBTa2grkiEQaVL4hoI2dUlMfxMyPPa/s200/Google.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://web.archive.org/web/19981111184551/http://google.com/">Wayback Machine, 11/11/98</a></td></tr></tbody></table>
In particular Vicky explained <a href="https://en.wikipedia.org/wiki/Citation_index">citation indices</a>, the concept behind <a href="https://en.wikipedia.org/wiki/PageRank">Page Rank</a>, to <a href="https://en.wikipedia.org/wiki/Larry_Page">Larry Page</a> and <a href="https://en.wikipedia.org/wiki/Sergey_Brin">Sergey Brin</a>. <a href="https://en.wikipedia.org/wiki/Andy_Bechtolsheim">Andy Bechtolsheim</a> was famously instrumental in persuading them to turn their demo of a Page Rank search engine into Google, the company. In his fascinating interview in the Computer History Museum's <a href="https://www.computerhistory.org/collections/oralhistories/?s=date">oral history</a> collection, Andy explains why the idea of ranking pages by their inbound links was so important.<br />
<br />
Below the fold I have taken the liberty of transcribing and cleaning up the relevant section of Andy's stream of conciousness, both because it is important history and because it exactly reflects the Andy I was privileged to know in the early days of Sun Microsystems.<br />
<span><a name='more'></a></span>
<br />
<div class="separator" style="clear: both; text-align: center;"><iframe allowfullscreen="" class="BLOG_video_class" height="294" src="https://www.youtube.com/embed/GjR7sRASjdo" width="440" youtube-src-id="GjR7sRASjdo"></iframe></div>
<br />
This rough transcript runs from <a href="https://youtu.be/GjR7sRASjdo?feature=shared&t=2867">[47:47]</a> to <a href="https://youtu.be/GjR7sRASjdo?feature=shared&t=3215">[53:35]</a>. Andy speaks:<br />
<br />
The true story is that I met the founders of Google before the company existed because they were, like I was, a student at Stanford.
I'm not sure Larry was really ready to abandon the PhD program and jump into this.
So I was in this same position once myself and I used the same line "you'll always finish your PhD later".
Now the concern I had is that if they didn't get going this great idea may be not happening.
<br /><br />
Of course I was involved in the company itself but it was really really a very good idea, one of the best ideas I've ever seen.
This notion of relevant search and relevant ads and this business model that solved it
Put it this way, I was very familiar with scientific publishing where what matters is not how many papers you write but how many people cite your papers.
So if you apply the same thing to the Web clearly what is relevant is what other people link to and notice.
You could automatically build a graph, a structure that said what's more important than others.
<br /><br />
At the time people didn't think automatic search was actually possible because Alta Vista which was popular at the time just looked at keywords.
What people would do is they would add the whole dictionary as a dark page behind the document.
Since every word you are looking for is in the dictionary you couldn't find anything any more because every document had every word.
That wasn't a path to success; you couldn't actually look at the document because it could just be spam.
<br /><br />
Yahoo believed fairly strongly, they actually had at one point talked to Larry and Sergey and maybe even Larry was offering them to sell them the idea but Yahoo passed on the belief that you just couldn't do it.
They really thought you could hire people, like newspaper editors, they would make the sports section and the garden section like a newspaper; select content from the Web that then they would present to the front page.
Clearly that was not going to scale if there were millions of Web pages you just couldn't keep it up.
<br /><br />
So Larry and Sergey believed very strongly that it had to be automatic and that if that wasn't possible — it was the only way to do it basically.
That first demo which they had on a laptop was actually quite compelling it looked the same as before — here's the lucky button.
The only worry was they wanted to sort of demo they could scale the search engine to like a couple of racks of computers before they would raise the real venture and there was some question whether it was scalable.
For that they just wanted to raise a small amount of money to build the first couple of racks with motherboards.
<br /><br />
The money that I put in there on day one was actually before the company started helped to demonstrate that.
So I bought them this check.
They had the name of the company but the check was to the name of the company that didn't exist.
At the time some of the law firms were so busy they didn't want to take on new clients that didn't have some funding behind them.
I figured if I write them a check it would help them to get the right firm.
But I can't claim any credit for what they've done its all due to the insight they had and the team they built.
<br /><br />
Let me back up here.
I couldn't find stuff on Alta Vista so I was desperate for better search.
A lot of my time was looking for data sheets and information I was looking for and if I couldn't find anything on the Web the Web wouldn't be very useful.
Like, how do you find stuff?
Its the most important thing.
So for me it was a personal goal to have something that would actually work.
<br /><br />
But in any new company the first question is "what's your business model?"
Even at the time they had this model of sponsored links that would take your search query and link it to this ad inventory.
I asked them "how much is it per click?" and they said 5 cents per click which is still their bottom price today - this is before they got into the competitive bidding - and I did this math in the back of my mind "a million clicks a day and 5 cents a click is $50K a day" — they can't go broke.
<br /><br />
I had no idea how this would scale, in effect I don't think anybody understood this but it was clear that there was enough interest in finding the people who were looking for stuff.
Lets say you search for a tennis racket.
That means you're probably a tennis player and most likely there will be an ad that shows you tennis rackets or tennis balls or something that relates to your interest.
And the key was you have an unlimited ad inventory, instead of having these banner ads that I have never clicked on in my life, maybe once.
<br /><br />
Banner ads are a waste of bandwidth essentially.
These ads were highly more relevant and even today except for spam mail its the most cost-effective way of advertising or finding customers.
It took advantage of the fact that the Internet is a full-duplex communication path whereas banner ads were more like TV — here's your ad break and here's what you have to consume before you get to the next page.
<br /><br />
Google was an absolutely brilliant idea but funnily enough their business really took off after the dot-com implosion in 2001.
If you look at their historical revenue what happened at that point I believe was that people realized that they spent on banner ads was basically wasted and they got back to an ROI calculation of where do we apply money more cost-effectively and, yeah, so Google.
Once people started bidding for the keywords of course the price per click went up but it still provides very very good value for advertisers.
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com2tag:blogger.com,1999:blog-4503292949532760618.post-83664617484725858012024-01-30T08:00:00.000-08:002024-01-30T08:00:00.282-08:00Criming On The BlockchainI apologize for the delay in posting but, as you will see, the post I was working on grew rather long.<br />
<br />
It seems obvious that doing crimes and writing the receipts to an immutable public ledger is risky, but many criminals have been convinced that there is no risk because cryptocurrencies such as Bitcoin are anonymous. Although there are cryptocurrencies with anonymous transactions, such as Monero and zCash, they are much more difficult to use and much less liquid than pseudonymous cryptocurrencies like Bitcoin. As many criminals have discovered, without an unrealistically intense focus on <a href="https://blog.dshr.org/2022/02/inadequate-opsec.html">operational security (opsec)</a>, the identity behind the pseudonym can be revealed. An entire industry has evolved to do these revelations, tracing the flow of coins through their blockchains.<br />
<br />
Below the fold I discuss the techniques and results of blockchain tracing, based on four main sources:<br />
<ul>
<li>Andy Greenberg's new book entitled <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/"><i> Tracers in the Dark: The Global Hunt for the Crime Lords of Cryptocurrency</i></a>.</li>
<li>Chainalysis' report <a href="https://www.chainalysis.com/blog/2024-crypto-crime-report-introduction/"><i>2024 Crypto Crime Trends: Illicit Activity Down as Scamming and Stolen Funds Fall, But Ransomware and Darknet Markets See Growth</i></a>
and his <a href="https://www.wired.com/story/stablecoin-sanctions-violations-crypto-crime/"><i>‘Stablecoins’ Enabled $40 Billion in Crypto Crime Since 2022</i></a>.</li>
<li>The UN Office for Drugs and Crime (UNODC) report <a href="https://www.unodc.org/roseap/uploads/documents/Publications/2024/Casino_Underground_Banking_Report_2024.pdf"><i>Casinos, Money Laundering, Underground Banking, and Transnational Organized Crime in East and Southeast Asia: A Hidden, Accelerating Threat</i></a>.</li>
<li>Recent posts on <a href="https://medium.com/chainargos/">ChainArgos blog</a>.</li>
</ul>
<span><a name='more'></a></span>
There are two main use cases for cryptocurrencies, speculation and crime. Although speculation is likely behind the majority of transactions its externalities, such as people losing their life savings, have cause it to be downgraded from "harmless" to "<a href="https://en.wikipedia.org/wiki/The_Hitchhiker%27s_Guide_to_the_Galaxy_(radio_series)">mostly harmless</a>", the minority of criminal transactions are definitely harmful. I've written about these harms in, among others, my <a href="https://blog.dshr.org/2022/02/ee380-talk.html">EE380 talk</a>, <a href="https://blog.dshr.org/2023/05/the-cryptocurrency-use-case.html"><i>The Cryptocurrency Use Case</i></a> and <a href="https://blog.dshr.org/2022/09/cryptocurrency-enabled-crime.html"><i>Cryptocurrency-enabled Crime</i></a>.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi9TbM75u2Xff7AXUBO5u3BOajEyJqin8yL9uZ3OnNB-iQW_hI1vU0LMzL6w6LVsUFbb6H2Z6HEU1MezOmWIdFmH6t8HX93eNKzauopYST8hvipW5GDWyntHAG1x1lHbxYuyIG0ATSRHvGfE21h19l8jeEeSdgXTmpXy47XKeekKredC074cc_AwgvzzOHP/s1500/Tracers.jpg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="1500" data-original-width="970" height="200" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi9TbM75u2Xff7AXUBO5u3BOajEyJqin8yL9uZ3OnNB-iQW_hI1vU0LMzL6w6LVsUFbb6H2Z6HEU1MezOmWIdFmH6t8HX93eNKzauopYST8hvipW5GDWyntHAG1x1lHbxYuyIG0ATSRHvGfE21h19l8jeEeSdgXTmpXy47XKeekKredC074cc_AwgvzzOHP/w129-h200/Tracers.jpg" width="129" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/ref=sr_1_1">Source</a></td></tr></tbody></table>
Greenberg's book lays out the history of blockchain tracing technology, starting from Sarah Meiklejohn <i>et al</i>'s 2013 paper entitled <a href="https://doi.org/10.1145/2504730.2504747"><i>A Fistful of Bitcoins: Characterizing Payments Among Men with No Names</i></a>. Their abstract reads:<br />
<blockquote>
Bitcoin is a purely online virtual currency, unbacked by either physical commodities or sovereign obligation; instead, it relies on a combination of cryptographic protection and a peer-to-peer protocol for witnessing settlements. Consequently, Bitcoin has the unintuitive property that while the ownership of money is implicitly anonymous, its flow is globally visible. In this paper we explore this unique characteristic further, using heuristic clustering to group Bitcoin wallets based on evidence of shared authority, and then using re-identification attacks (i.e., empirical purchasing of goods and services) to classify the operators of those clusters. From this analysis, we characterize longitudinal changes in the Bitcoin market, the stresses these changes are placing on the system, and the challenges for those seeking to use Bitcoin for criminal or fraudulent purposes at scale.
</blockquote>
Meiklejohn started from an observation by Satoshi Nakamoto. <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">Greenberg quotes Nakamoto</a>:<br />
<blockquote>
"Some linking is still unavoidable with multi-input transactions, which necessarily reveal that their inputs were owned by the same owner,” Satoshi wrote. “The risk is that if the owner of a key is revealed, linking could reveal other transactions that belonged to the same owner.”
</blockquote>
Linking the inputs of multi-input transactions roughly halved the then number of Bitcoin users. Meiklejohn then developed the <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">"change address" technique</a>:<br />
<blockquote>
When you pay someone 6 bitcoins from a 10-coin address, 6 coins go to their address. Your change, 4 coins, is stored at a new address, which your wallet software creates for you. The challenge, when looking at that transaction on the blockchain as a sleuthing observer, is that the recipient’s address and the change address are both simply listed as outputs, with no label to tell them apart.<br />
<br />
But sometimes, Meiklejohn realized, spotting the difference between the change address and the recipient address was easy: If one address had been used before and the other hadn’t, the second, totally fresh address could only be the change address
</blockquote>
Meiklejohn's first criminal case started when "Flycracker" raised funds to mail Brian Krebs a baker's dozen bags of <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">heroin from Silk Road</a>:<br />
<blockquote>
Flycracker had made it easy. By posting a Bitcoin address to the cybercriminal forum, he’d given Meiklejohn a starting point. She simply copied the thirty-four-character string into her blockchain software and looked at the transactions at that address. After collecting 2 bitcoins in donations at the address he’d posted, worth around $200 at the time, a little over three-quarters of the money had been sent to another address, with a third collecting the change. At a glance, Meiklejohn immediately identified the change address and checked the money’s destination against her database. Sure enough, the address was one of the nearly 300,000 she had already tagged as belonging to the Silk Road. Meiklejohn had just connected Flycracker’s address directly to the source of the heroin he’d tried to use to frame Krebs.
</blockquote>
The first major cryptocurrency bust Greenberg recounts was the arrest of Silk Road's Dread Pirate Roberts in a San Francisco library. It <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">did not depend upon these tracing techniques</a>:<br />
<blockquote>
The FBI has described that cybersurveillance coup as the result of a misconfiguration in the site’s use of the Tor anonymity software but has been reluctant to ever officially explain that error in a courtroom.
</blockquote>
<a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">In fact</a>:<br />
<blockquote>
it had been the IRS’s Gary Alford, sitting in his New Jersey home four months earlier, who’d done the meticulous, unglamorous work that had led to the case’s first real breakthrough. Alford had been using Google to dig up the earliest online posts about the Silk Road on drug forums when he’d found a curious artifact: Someone going by the name “altoid” had posted to a site called the Shroomery in January 2011 recommending the Silk Road’s just-launched dark web market as a source for drugs. Around the same time, a user with the same handle had also asked for programming help on a coding forum. On that page, altoid had listed his email address: rossulbricht@gmail.com.
</blockquote>
Another IRS agent, Tigran Gambaryan, received a tip that Carl Force, one of the DEA agents working on Silk Road, had used a fake ID to set up an account at Bitstamp, a cryptocurrency exchange, and deposited a lot of BTC He had cashed out $200K and, as Gambaryan <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">examined his financial records</a>:<br />
<blockquote>
He found that Force had, in late 2013, paid off his home’s entire mortgage, an outstanding loan of $130,000. He’d repaid, too, a $22,000 loan he’d taken out against his federal retirement account. He’d even made a gift of tens of thousands of dollars to his local church, the sort of largesse that, Gambaryan knew all too well, was tough to afford on a federal agent’s salary. The numbers only got shadier from there: Gambaryan found records of real estate investments in which Force had listed his net worth as more than 1 million. That wealth was almost entirely due, it became clear, to a massive influx of liquidated bitcoins from cryptocurrency exchanges like Bitstamp and CampBX that had flowed into Force’s bank accounts. The payments totaled $776,000 beyond his $150,000 annual DEA salary over the two prior years that he’d worked on the Silk Road case. With that ample financial padding, Force had then retired from the DEA, just days before Gambaryan began to look into his records.
</blockquote>
Gambaryan could get Force's wallet addresses from the exchanges he used, and he found an unencrypted message from DPR referencing a 525 BTC payment to Force's investigative alias, but he needed proof, So, <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">Greenberg writes</a>:<br />
<blockquote>
Despite having read Meiklejohn’s paper, he possessed none of the data that she’d assembled over months of clustering Bitcoin addresses and identifying them with test transactions. So he simply started copying Bitcoin addresses from Carl Force’s account records—the ones he’d gotten from exchanges such as CampBX and Bitstamp—and pasting them into the search field on Blockchain.info, which displayed the entire blockchain on the web. At first, the collections of garbled character strings seemed meaningless to Gambaryan. But almost immediately, he could see he was onto something. On September 27, 2013, just a few days before Ross Ulbricht’s arrest, Gambaryan saw with a jolt of recognition that one of Force’s CampBX addresses had received a 525-bitcoin payment—the magic number that DPR had mentioned in his conveniently unencrypted message.
</blockquote>
Gambaryan manually followed the chains backward, from their inputs to the outputs that caused them, <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">until finally</a>:<br />
<blockquote>
Following the money at each of the remaining addresses back one more step, he now saw the coins had originally come from just four sources. Each of those addresses had received their funds on the same day: August 4, 2013—the exact date when the Dread Pirate Roberts had told Nob he’d paid him. Gambaryan mentally recorded the payments: They were for 127, 61, 134, and 203 bitcoins. He added the numbers in his head. They summed up to 525 bitcoins.<br />
...<br />
The next morning, after a few hours’ sleep, Gambaryan began texting his DHS contact Jared Der-Yeghiayan, the Armenian American agent in Chicago whom he’d befriended. He needed to check the four addresses he’d found with someone who had access to the Dread Pirate Roberts’s Bitcoin wallet. As a member of the Silk Road investigation team, Der-Yeghiayan still had access to all the site’s server data, including its Bitcoin addresses. Der-Yeghiayan called Gambaryan back a few hours later and confirmed what Gambaryan already knew: Each of the four addresses belonged to DPR.
</blockquote>
He thus became apparently the first law enforcer to use blockchain tracing as evidence in an investigation. Its first use in a trial appears to be when, with help from Nick Weaver, the prosecution of Ross Ulbricht introduced a trace of his payment for <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/"> a murder-for-hire attempt</a>:<br />
<blockquote>
But the day when the prosecution found the incontrovertible, public, and unerasable proof of Ulbricht’s Silk Road millions, argues Nick Weaver, remains a milestone in the history of cryptocurrency and crime. “That is the date,” Weaver says, “that you can state unequivocally that law enforcement learned that the blockchain is forever.”
</blockquote>
The blockchain tracing industry's pioneer, Chainalysis, spun out of the Kraken exchange as a result of the next big crime Greenberg covers, the <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">collapse of the Mt. Gox exchange</a>:<br />
<blockquote>
Kraken’s management, in a pro bono attempt to help rescue the cryptocurrency ecosystem from the rippling shock of Mt. Gox’s failure—and the collapse in Bitcoin’s price that followed—had agreed to help distribute any remaining bitcoins that could be found to Mt. Gox’s thousands upon thousands of angry creditors.<br />
<br />
Michael Gronager, for his part, had taken on a far more uncertain task. He’d agreed to find the missing coins. By all appearances, this was not a rational decision. The Danish entrepreneur had left his relatively comfortable position as the COO of Kraken to found a new start-up whose sole client, for the moment, was this roomful of Japanese bankruptcy lawyers asking him to track down Mt. Gox’s gigantic, wayward fortune. Even calling them a client would be a stretch: He would receive no fee, and no portion of the recovered funds, if he could manage to find any.
</blockquote>
The co-evolution of Bitcoin's and tracing technology started with the revelation that Chainalysis, by running a node in the Bitcoin network, could discover the IP address associated with many wallets, which garnered both hostility and customers. With a head-start, Chainalysis rapidly became the leader in their emerging market, as <a href="http://www.amazon.com/Increasing-Returns-Dependence-Economics-Cognition/dp/0472064967">Brian Arthur</a> would have predicted.<br />
<br />
The next investigation Greenberg covers was into the BTC-e exchange, <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">whose</a>:<br />
<blockquote>
computers where the exchange was hosted weren’t on the dark web, protected by Tor. They ought to be discoverable with a simple “traceroute” command, an operation that anyone with a computer and an internet connection can run to find a site’s IP address—no harder than looking up a commercial service’s number in a phone book. Gambaryan checked, and it turned out the only layer of misdirection that had prevented curious observers from learning the location of BTC-e’s servers in the first place was a company called Cloudflare, a web infrastructure provider and security service that shielded the exchange’s IPs from prying eyes like Gambaryan’s.
</blockquote>
Subpoenas to Cloudflare revealed they were hosted in the US, which allowed them to be <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">imaged</a>:<br />
<blockquote>
Gambaryan dug into the data his team had copied from the BTC-e server. What he found was a revelation: The IP address for the account trading in stolen Mt. Gox coins on BTC-e matched one of the few IP addresses on the BTC-e server’s allow list for the administrators’ connections. In other words, the person who had siphoned hundreds of thousands of bitcoins from Mt. Gox into BTC-e wasn’t just any BTC-e user. They were a
BTC-e administrator. Specifically, an admin with the username WME. “The gears started turning in my head,” Gambaryan remembers. “What better way to launder hundreds of thousands of bitcoins than to launch your own Bitcoin exchange?”
</blockquote>
WME was Alexander Vinnik but, alas, he was in Russia.<br />
<br />
Greenberg goes on to describe the takedown of Alexandre Cazes, who ran the Alphabay dark-web market and the related takeover of the Hansa dark-web market, and then of the Welcome to Video child sexual abuse site.
This led to the arrest and indictment of a Texas-based <a href="https://www.amazon.com/Tracers-Dark-Global-Crime-Cryptocurrency-ebook/dp/B09SKW8WRV/">Border Patrol agent</a>:<br />
<blockquote>
The Texas man had taken a rare approach to his legal defense: He’d pleaded guilty to possession of child sexual abuse materials, but he also appealed his conviction. He argued that his case should be thrown out because IRS agents had identified him by tracking his Bitcoin payments—without a warrant—which he claimed violated his Fourth Amendment right to privacy and represented an unconstitutional “search.”<br />
<br />
A panel of appellate judges considered the argument—and rejected it. In a nine-page opinion, they explained their ruling, setting down a precedent that spelled out in glaring terms exactly how far from private they determined Bitcoin’s transactions to be.<br />
<br />
“Every Bitcoin user has access to the public Bitcoin blockchain and can see every Bitcoin address and its respective transfers. Due to this publicity, it is possible to determine the identities of Bitcoin address owners by analyzing the blockchain,” the ruling read. “There is no intrusion into a constitutionally protected area because there is no constitutional privacy interest in the information on the blockchain.”<br />
<br />
A search requires a warrant, the American judicial system has long held, only if that search enters into a domain where the defendant has a “reasonable expectation of privacy.” The judges’ ruling argued that no such expectation should have existed here: The HSI agent wasn’t caught in the Welcome to Video dragnet because IRS agents had violated his privacy. He was caught, the judges concluded, because he had mistakenly believed his Bitcoin transactions to have ever been private in the first place.
</blockquote>
This firmly established blockchain tracing as a legitimate form of evidence.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjbbjMzX65f3S0ZT5tNMBXgCgMURaRDBwo6S5FCdgATslp7MUp6vokxzYwa5tVjt19QqGK_GDPGujJC1CihsEbGFIG4qqTZ26mMqfCq2r87xKWHlyTFDRcL1pwogYJ-ZD6RuZwKZ9-oKiXuB_eLBPPI8vdv7ccHKHtcL1dbuhfDFOorY8fbG4x74Qn6JspD/s1000/VCE4-Transfers.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="826" data-original-width="1000" height="165" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjbbjMzX65f3S0ZT5tNMBXgCgMURaRDBwo6S5FCdgATslp7MUp6vokxzYwa5tVjt19QqGK_GDPGujJC1CihsEbGFIG4qqTZ26mMqfCq2r87xKWHlyTFDRcL1pwogYJ-ZD6RuZwKZ9-oKiXuB_eLBPPI8vdv7ccHKHtcL1dbuhfDFOorY8fbG4x74Qn6JspD/w200-h165/VCE4-Transfers.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.justice.gov/opa/press-release/file/1470186/download">Source</a><br /></td></tr></tbody></table>
One case Greenberg mentions only in passing is the theft of nearly 120K BTC from Bitfinex, to which <a href="https://www.bbc.com/news/technology-66390639">Heather Morgan and Ilya Lichtenstein pled guilty</a>. They were intitially flagged as suspects during the takeover of AlphaBay that Greenberg describes in detail. I discussed the tracing steps revealed by the <a href="https://www.justice.gov/opa/press-release/file/1470186/download"><i>Statement of Facts</i></a> from their indictment in <a href="https://blog.dshr.org/2022/02/inadequate-opsec.html"><i>Inadequate OpSec</i></a>. The image shows a small part of the tracing evidence in this case. The two VCE4 accounts used Russian e-mail addresses, but the VCE7 and VCE8 accounts were in the name of companies controlled by Lichtenstein and Morgan. Tracing the chains back connected the VCE4 accounts to the suspects, who had taken the precaution of funding VCE4 with Monero.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhqlcEYPthnP28Q-HN8kmweEuIL5TWrGgK6HXJVAxoYy2_mTcUnb5n5KFMaldB0TOQm8eePa2y2OkJzlyslViPsnIQeYUhVFHaJBaNOdUv3fZhmQH7PhJfvaml_3fEIPIcx-zxKXAf5Zia615xIea5uG4dlQzedpa8uPo4ZVBGCf_TbEONPhJXIvE2JtKCx/s800/TotalIllicitValue.png" style="clear: right; display: block; margin-left: auto; margin-right: auto; padding: 1em 0px; text-align: center;"><img alt="" border="0" data-original-height="509" data-original-width="800" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhqlcEYPthnP28Q-HN8kmweEuIL5TWrGgK6HXJVAxoYy2_mTcUnb5n5KFMaldB0TOQm8eePa2y2OkJzlyslViPsnIQeYUhVFHaJBaNOdUv3fZhmQH7PhJfvaml_3fEIPIcx-zxKXAf5Zia615xIea5uG4dlQzedpa8uPo4ZVBGCf_TbEONPhJXIvE2JtKCx/s200/TotalIllicitValue.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.chainalysis.com/blog/2024-crypto-crime-report-introduction/">Source</a></td></tr></tbody></table>
Deanonymizing individual wallets and flows is valuable to Chainalysis' clients; law enforcement for evidence and financial institutions for risk-assessment. But their vast collection of tagged wallet addresses and transactions is also valuable in aggregate. It enables statistical analysis of the cryptosphere, such as Chainalysis' annual report on cryptocurrency crime. This year's is introduced in <a href="https://www.chainalysis.com/blog/2024-crypto-crime-report-introduction/"><i>2024 Crypto Crime Trends: Illicit Activity Down as Scamming and Stolen Funds Fall, But Ransomware and Darknet Markets See Growth</i></a>:<br />
<blockquote>
2023 saw a significant drop in value received by illicit cryptocurrency addresses, to a total of $24.2 billion. As always, we have to caveat by saying that these figures are lower bound estimates based on inflows to the illicit addresses we’ve identified today. One year from now, these totals will almost certainly be higher, as we identify more illicit addresses and incorporate their historic activity into our estimates. For instance, when we published our Crypto Crime Report last year, we estimated $20.6 billion worth of illicit transaction volume for 2022. One year later, our updated estimate for 2022 is $39.6 billion. Much of that growth came from the identification of previously unknown, highly active addresses hosted by sanctioned services, as well as our addition of transaction volume associated with services in sanctioned jurisdictions to our illicit totals.<br />
<br />
Another key reason the new total is so much higher, besides the identification of new illicit addresses: We’re now counting the $8.7 billion in creditor claims against FTX in our 2022 figures. In last year’s report,
</blockquote>
Although these arae large sums, Chainalysys estimate they represent a fairly small proportion of the total cryptocurrency volume, falling from 0.42% in 2022 to 0.34% in 2023. Of course, it is unlikely that they have identified <i>all</i> the illicit transactions.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg4uLYTQ9ZsDZNFU81xIFkplLGDiFnnWGY-G1H-iGLnqdZEoYEAniBjX-931DTLyAFEACQ30EYDC_tMDHxzBceF2TwoKLjkn4R3qG0Tk4etXvfU1w6ik2i2tSCoYfRioAgf5bBXCwPUOn2Z2gRBhVkgEqNO9A27GFDJqbqqb6qN2-wDWkEuxoBUrP-ZQf_a/s800/AssetTypeVolume.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="449" data-original-width="800" height="113" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg4uLYTQ9ZsDZNFU81xIFkplLGDiFnnWGY-G1H-iGLnqdZEoYEAniBjX-931DTLyAFEACQ30EYDC_tMDHxzBceF2TwoKLjkn4R3qG0Tk4etXvfU1w6ik2i2tSCoYfRioAgf5bBXCwPUOn2Z2gRBhVkgEqNO9A27GFDJqbqqb6qN2-wDWkEuxoBUrP-ZQf_a/w200-h113/AssetTypeVolume.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.chainalysis.com/blog/2024-crypto-crime-report-introduction/">Source</a></td></tr></tbody></table>
The report has a <a href="https://www.chainalysis.com/blog/2024-crypto-crime-report-introduction/">big surprise</a>:<br />
<blockquote>
Through 2021, Bitcoin reigned supreme as the cryptocurrency of choice among cybercriminals, likely due to its high liquidity. But that’s changed over the last two years, with stablecoins now accounting for the majority of all illicit transaction volume. This change also comes alongside recent growth in stablecoins’ share of all crypto activity overall, including legitimate activity.
</blockquote>
Bitcoin's volatility is great for speculation, but when it fails to proceed moonwards it is a big problem for criminals, and <a href="https://www.chainalysis.com/blog/2024-crypto-crime-report-introduction/">especially for sanctions-busters</a>:<br />
<blockquote>
Some forms of illicit cryptocurrency activity, such as darknet market sales and ransomware extortion, still take place predominantly in Bitcoin. Others, like scamming and transactions associated with sanctioned entities, have shifted to stablecoins. Those also happen to be the biggest forms of crypto crime by transaction volume, thereby driving the larger trend. Sanctioned entities, as well as those operating in sanctioned jurisdictions or involved with terrorism financing, also have a greater incentive to use stablecoins, as they may face more challenges accessing the U.S. dollar through traditional means, but still want to benefit from the stability it provides.
</blockquote>
The report notes that stablecoin users,criminal or not, run the risk of having their wallets and thus their funds "frozen", as Tether has been doing recently. Patrick Tan covered the case of an Indian user (The Victim) in detail in <a href="https://medium.com/chainargos/what-happens-when-tether-freezes-your-tether-5a8ece2bd508"><i>What happens when Tether “freezes” your Tether?</i></a>. On 7<sup>th</sup> December 2023 Tether changed its <a href="https://tether.to/en/legal/">Terms of Service</a> and, in <a href="https://medium.com/chainargos/3-things-you-must-know-about-tethers-terms-of-service-c46c0172b892"><i>3 Things You Must Know About Tether’s Terms of Service</i></a>, Tan delves into the deliberately confusing details and ends up agreeing with <a href="https://datafinnovation.medium.com/3249639e677a">Jonathan Reiter</a> about the The Victim's problem:<br />
<blockquote>
On a basic level this user was relying on an unlicensed money transmitter where they have 0 access to any authority that feels accountable to them.<br />
<br />
Tether isn’t an Indian money services business. Nor is it regulated in the victim’s country. Or anywhere with a real process.<br />
...<br />
This — precisely this — is the cost of living outside the law. You may end up with no recourse. Or not.<br />
<br />
But you don’t even have someone to complain to that feels accountable for your problems (i.e. your local police or elected representative, or an employee of a business accountable to a regulator you can contact).
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhV92Ly0XyArXKwagxLzoBznW2n1X_MYePLr7nlyGJjPkAdsy9uP_42kiCiC_xt1rjGUZUHDOzmj-FLKEUS6t2lVWrKwNWyM-nBobG_vWCI_04ftKeMKN8gUlGPuiZe-t_lvP7pMqTOpS1U1zG-WmOPHW7aO7tX_seXbux0v1qsfgNrcWMGBepVa9-il9c5/s800/AssetTypeUsage.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="476" data-original-width="800" height="119" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhV92Ly0XyArXKwagxLzoBznW2n1X_MYePLr7nlyGJjPkAdsy9uP_42kiCiC_xt1rjGUZUHDOzmj-FLKEUS6t2lVWrKwNWyM-nBobG_vWCI_04ftKeMKN8gUlGPuiZe-t_lvP7pMqTOpS1U1zG-WmOPHW7aO7tX_seXbux0v1qsfgNrcWMGBepVa9-il9c5/w200-h119/AssetTypeUsage.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.chainalysis.com/blog/2024-crypto-crime-report-introduction/">Source</a></td></tr></tbody></table>
It turns out that evading sanctions is the <a href="https://www.chainalysis.com/blog/2024-crypto-crime-report-introduction/">major criminal use case</a>:<br />
<blockquote>
Perhaps the most obvious trend that emerges when looking at illicit transaction volume is the prominence of sanctions-related transactions. Sanctioned entities and jurisdictions together accounted for a combined $14.9 billion worth of transaction volume in 2023, which represents 61.5% of all illicit transaction volume we measured on the year. Most of this total is driven by cryptocurrency services that were sanctioned by the U.S. Department of the Treasury’s Office of Foreign Assets Control (OFAC), or are located in sanctioned jurisdictions, and can continue to operate because they’re in jurisdictions where U.S. sanctions are not enforced. <br />
<br />
While those services can and have been used for nefarious purposes, it also means that some of that $14.9 billion in sanctions-related transaction volume includes activity from average crypto users who happen to reside in those jurisdictions. For example, Russia-based exchange Garantex, which was sanctioned by OFAC and OFSI in the U.K. for its facilitation of money laundering on behalf of ransomware attackers and other cybercriminals, was one of the biggest drivers of transaction volume associated with sanctioned entities in 2023. Garantex continues to operate because Russia does not enforce U.S. sanctions. So, does that mean all of Garantex’s transaction volume is associated with ransomware and money laundering? No. Nevertheless, exposure to Garantex introduces serious sanctions risk for crypto platforms subject to U.S. or U.K. jurisdiction, which means those platforms must remain ever-more vigilant and screen for exposure to Garantex in order to be compliant.
</blockquote>
Translation: platforms need to subscribe to Chainalysis to be safe. Andy Greenberg's <a href="https://www.wired.com/story/stablecoin-sanctions-violations-crypto-crime/"><i>‘Stablecoins’ Enabled $40 Billion in Crypto Crime Since 2022</i></a> quotes Chainalysis' Andrew Fierman:<br />
<blockquote>
As examples, Fierman points to Nobitex, the largest cryptocurrency exchange operating in the sanctioned country of Iran, as well as Garantex, a notorious exchange based in Russia that has been specifically sanctioned for its widespread criminal use. Stablecoin usage on Nobitex outstrips bitcoin by a 9:1 ratio, and on Garantex by a 5:1 ratio, Chainalysis found. That's a stark difference from the roughly 1:1 ratio between stablecoins and bitcoins on a few nonsanctioned mainstream exchanges that Chainalysis checked for comparison.
</blockquote>
Of course, when Chainalysis says "stablecoin" they essentially mean Tether. Three years ago, this <a href="https://www.youtube.com/watch?v=pKeMfxLV8K8">interview of Charles Yang</a>, head trader of Genesis Block based in Hong Kong, by John Riggins descibed how Tether was the basis for trade flows in South-East Asia because it evaded governments' currency controls. <a href="https://www.youtube.com/watch?v=pKeMfxLV8K8">Yang noted</a>:<br />
<blockquote>
bank acccounts are the absolute most valuable thing — you have to set up a bunch of different companies, a lot of different bank accounts just to facilitate trades that aren't that big, maybe $50K. The moment you tell them this is for a USDT trade, you're basically asking them to shut your bank account down.
</blockquote>
Last September DataFinnovation posted <a href="https://datafinnovation.medium.com/usdt-on-tron-ftx-wtf-is-really-happening-ef0cb807019a"><i>USDT-on-TRON, FTX & WTF Is Really Happening</i></a>. In summary:<br />
<blockquote>
FTX/Alameda minted nearly all the USDT-on-TRON and operate as something like a central bank or reserve manager for a shadow East Asian USD payment system. We provide convincing evidence from novel on-chain analysis that shows how a real, albeit mostly-not-kosher, crypto use case works. This data also makes plain that Binance/Cumberland runs the Ethereum part of the same ecosystem and that these two groups of parties probably coordinate their actions in some way.<br />
...<br />
<b>we are going to show that this entire complex looks an awful lot like a funnel to establish backing for a USD payment network aimed at people who cannot (easily or legally, depending) hold USD or transfer them</b>. This also exposes how USDT is split into a China-and-surroundings slice and a rest-of-world slice with a different major crypto entity handling each part.
</blockquote>
The UN Office for Drugs and Crime (UNODC) just published a report entitled <a href="https://www.unodc.org/roseap/uploads/documents/Publications/2024/Casino_Underground_Banking_Report_2024.pdf"><i>Casinos, Money Laundering, Underground Banking, and Transnational Organized Crime in East and Southeast Asia: A Hidden, Accelerating Threat</i></a>:<br />
<blockquote>
Online gambling platforms, and especially those that are operating illegally, have emerged as among the most popular vehicles for cryptocurrencybased money launderers, particularly for those using Tether or USDT on the TRON blockchain,<br />
...<br />
USDT on the TRON blockchain has become a preferred choice for crypto money launderers in East and Southeast Asia due to its stability and the ease, anonymity, and low fees of its transactions. Law enforcement and financial intelligence authorities in the region have reported USDT among the most popular cryptocurrencies used by organized crime groups in the region, particularly those involved in the regional cyberfraud industry, demonstrated by a surging volume of cases and unauthorized online gambling and cryptocurrency exchange platforms offering undergroud [sic] USDT-based services.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhe7xw9GWfydAqZ2bIiimuIWd8BunWhfRn5Vf24Mq8qWS7gtPuJmbEtLvZaKZGJBKr4pIql0EilsXQKEy-yZfhgLIK-j0ZmYnF0SwKo-ywebf0mgm_XZ0KNZYGMBmqgGfg5RHQATlkuD-9z2JFR5B593o4T1NcjfiOtoM-5HMLTdh4RMAQCeNbP2qnBkeNt/s708/RunningPoints.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="708" data-original-width="534" height="200" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhe7xw9GWfydAqZ2bIiimuIWd8BunWhfRn5Vf24Mq8qWS7gtPuJmbEtLvZaKZGJBKr4pIql0EilsXQKEy-yZfhgLIK-j0ZmYnF0SwKo-ywebf0mgm_XZ0KNZYGMBmqgGfg5RHQATlkuD-9z2JFR5B593o4T1NcjfiOtoM-5HMLTdh4RMAQCeNbP2qnBkeNt/w151-h200/RunningPoints.png" width="151" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.unodc.org/roseap/uploads/documents/Publications/2024/Casino_Underground_Banking_Report_2024.pdf">Source</a></td></tr></tbody></table>
The report details techniques such as <a href="https://www.unodc.org/roseap/uploads/documents/Publications/2024/Casino_Underground_Banking_Report_2024.pdf">points-running and motorcades</a>:<br />
<blockquote>
As third- and fourth-party payments have become better understood by authorities and more widely reported following ‘Operation Chain Break’ and other measures in China, organized crime groups have responded by accelerating the integration of cryptocurrencies into their illegal betting operations, creating significant challenges for investigators. In recent years, law enforcement and financial intelligence authorities have reported the growing use of sophisticated, high-speed money laundering ‘motorcade’ teams specializing in underground USDT – fiat currency exchanges (卡接回U) across East and Southeast Asia. This has also included the mass recruitment of mule bank accounts across virtually all jurisdictions in the Asia Pacific region which can be purchased for as little as US $30.<br />
<br />
Due to the rise of cryptocurrency-integrated motorcades, points running syndicates, and other challenges, in 2021 the Government of China banned cryptocurrency transactions, trading, and mining. The industry subsequently migrated to various jurisdictions, particularly driving up already rising cryptocurrency adoption in several countries in Southeast Asia, together with the establishment of high-risk and underground cryptocurrency exchanges. At the same time, it is worth noting that cryptocurrency flows connected to organized crime have been cited as being vastly underestimated by industry experts as well as law enforcement and regulatory authorities in the region. Experts have pointed to a number of shortcomings related to existing analyses including massive gaps in crime attribution on the blockchain, fabricated reporting by crypto exchanges, and the prevalence of wash trading which inflates crypto transaction volumes, thereby shrinking the portion of illicit transactions identified.
</blockquote>
The US is rightly concerned that Tether is undermining their sanctions system, but countries like China with strict controls on cross-boarder currency flows are also worried about similar undermining. Fortunately, the flows of Tether are observable on the Ethereum and Tron blockchains, so tracing techniques can be and, as I discussed in <a href="https://blog.dshr.org/2022/11/the-stablecoin-saga.html"><i>The Stablecoin Saga</i></a>, <a href="https://blog.dshr.org/2022/11/the-stablecoin-saga-continued.html"><i>The Stablecoin Saga Continued</i></a> and <a href="https://blog.dshr.org/2023/11/alamedas-on-ramp.html"><i>Alameda's On-Ramp</i></a> are being, applied.<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com11tag:blogger.com,1999:blog-4503292949532760618.post-12979209616514522592024-01-18T08:00:00.000-08:002024-01-18T08:00:00.313-08:00A Lesson LearnedYou know how backups work great until you really need them? Below the fold, a lesson learned from my recent example of this phenomenon.<br />
<span><a name='more'></a></span>
<br />
A long time ago I was feeling rich, so I spent nearly $2000 on one of the first-generation Mac Airs. This led to me travelling with a <a href="https://blog.dshr.org/2014/12/a-note-of-thanks.html">laptop I could afford to lose</a> instead. I loved the Air for all sorts of reasons but one of its important tasks was to back up my iPhone every day. And, of course, being something of an expert on preserving digital information, then back up the Air to an external USB hard disk using Time Machine. <a href="https://www.lockss.org">Lots Of Copies Keep Stuff Safe</a>.<br />
<br />
In the end I had to upgrade to a more recent M1 Air, not because the old one stopped working, but simply because the USB ports stoppped working, so it could no longer backup the iPhone or update the Time Machine disk. The new Air was impressively fast and had two USB-C ports, which it used both for data and for power. I continued to back up the iPhone and the Air daily<br />
<br />
Fast forward to last month. The new Air's USB-C ports had been a bit flaky for a while, but now one failed. It wouldn't transfer data or power. A few days later the other port stopped transferring power, leaving me with no way to charge it. It was about 85% charged, so I powered it off.<br />
<br />
I thought "I'm OK, I have yesterday's Time Machine so I can get a new M2 Air and initialize it from the disk". So I ordered a new one and, as soon as it was delivered, plugged in the Time Machine disk and fired up the Migration Assistant. It asked me what disk it should use. I wasn't able to select the Time Machine disk because, to my surprise, the Air couldn't see it. Plugging the disk into my Linux desktop revealed that in the couple of weeks since the day before the M1 Air failed, the Time Machine disk had died.<br />
<br />
At this point I remembered that when I powered it off, the M1 Air still had a USB-C port that transferred data, even if it didn't work for power. There were two ways to initialize the M2 Air using the remaining battery, either create a new Time Machine disk using the one USB-C port that transferred data, or via Bluetooth. I found a USB-C external SSD, which I figured would use less power, powered up the M1 Air, plugged in the drive, started a Time Machine backup, and crossed my fingers.<br />
<br />
After a long, agonizing wait the backup finished with still quite a lot of battery left, so I powered off the M1 Air and used the disk to initialize the M2 Air. Apple's Migration Assistant worked flawlessly, and I relaxed. The best thing about the M2 Air is that Apple went back to powering it with MagSafe! Their fragile USB hardware could no longer prevent the machine charging.<br />
<br />
So the lesson was that I needed to have <i>two</i> Time Machine disks and use them alternately, so even if the most recent one failed I would have a backup only one day older.<br />
<br />
The next problem came when I plugged my iPhone into the M2 Air to back it up. I got "unknown error 7014". Google had no trace of "unknown error 7014" so, baffled, I made an appointment with the "Genius" at Apple's Palo Alto store. In the faint hope that the M1 Air could be fixed, I took both Airs and the iPhone with me.<br />
<br />
The Genius was very good. He rapidly figured out that:<br />
<ul>
<li>"unknown error 7014" was caused by the fact that, despite its having been delivered a few days earlier, the M2 Air's operating system was out-of-date. A software update overnight fixed the problem.</li>
<li>He confirmed that the M1 Air's USB-C ports had failed and, to my surprise, explained that the ports were on a daughter-board that could be replaced. It was a $15 part and $90 in labor, and completed overnight.</li>
</ul>
So now I can travel with a Mac Air.<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com2tag:blogger.com,1999:blog-4503292949532760618.post-67432954894432506562024-01-09T08:00:00.000-08:002024-02-12T13:21:38.831-08:00Autonomous Vehicles: Trough of Disillusionment<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgkxH2f42x14QdnWdgQSi20Vh-SoxgiLmcSJmU6O34BlO-q9QulJkQ-mh2DMADXSqNs7lwTMwaPZp4-5cpE-1ZoV_E9EZ6h0BpfuPuntiKmrtz60tx1zXgR83W6mxonXoP1YnjtugaCtUnnvfyfRV2v60Hc9MruOZsBO25icow6Qp7b7qfUvI5sC4gaJBBI/s719/GartnerHypeCycle.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="471" data-original-width="719" height="131" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgkxH2f42x14QdnWdgQSi20Vh-SoxgiLmcSJmU6O34BlO-q9QulJkQ-mh2DMADXSqNs7lwTMwaPZp4-5cpE-1ZoV_E9EZ6h0BpfuPuntiKmrtz60tx1zXgR83W6mxonXoP1YnjtugaCtUnnvfyfRV2v60Hc9MruOZsBO25icow6Qp7b7qfUvI5sC4gaJBBI/w200-h131/GartnerHypeCycle.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://en.wikipedia.org/wiki/User:Jeremykemp">Jeremykemp</a> <a href="https://creativecommons.org/licenses/by-sa/3.0">CC BY-SA 3.0</a>, <a href="https://commons.wikimedia.org/w/index.php?curid=10547051">Link</a></td></tr></tbody></table>
This is the famous <a href="https://en.wikipedia.org/wiki/Gartner_hype_cycle"><i>Gartner hype cycle</i></a> and it certianly appears that autonomous vehicles are currently in the <i>Trough of Disillusionment</i>. Whether they will eventually soar up to the <i>Plateau of Productivity</i> is unknown, but for now it is clear even to practitioners that the hype bubble they have been riding for years has burst.<br />
<br />
Below the fold I try to catch up with the flood of reporting on the autonomous vehicle winter triggered by the bursting of the bubble.<br />
<span><a name='more'></a></span>
<h3>Cruise</h3>
The most obvious symptom is the implosion of General Motors' Cruise robotaxi division.
Brad Templeton worked on autonomy at Waymo and is an enthusiast for the technology. But in <a href="https://www.forbes.com/sites/bradtempleton/2023/12/28/robocar-2023-in-review-the-fall-of-cruise/"><i>Robocar 2023 In Review: The Fall Of Cruise</i></a> he acknowledges that things aren't going well:<br />
<blockquote>
Looking at 2023 in review for self-driving cars, one story stands out as the clear top story of the year, namely the fall of General Motors’ “Cruise” robotaxi division, which led to a pause of operations for Cruise over the whole USA, a ban on operations in California, the resignation of the founder/CEO and much more. That was actually the most prominent component of the other big story of the year, namely the battle between San Francisco and both Cruise and Waymo.<br />
<br />
Even this serious mistake should not have and perhaps would not have driven Cruise to its fate. The bigger mistake was they did not want to talk about it. They were eager to show the car’s recorded video of events to the press, and to officials, but only over Zoom. They did not mention the dragging, even when asked about it. I did not ask to see the part of the video after the crash, but others who did see it say it ended before the dragging. The DMV states they were not shown the dragging, and Cruise did not tell them about it, including in a letter they sent shortly after the events. Cruise insists they did show the full video to the DMV, and the DMV insists otherwise, but there is little doubt that they were not open about what was obviously the most important part of the chain of events when it came to understanding the role of the robotaxi in the calamity. Cruise was very eager to show that the initial crash was not their fault, but didn’t want to talk at all about their own serious mistake.
</blockquote>
Laura Dobberstein reports on part of the aftermath in <a href="https://www.theregister.com/2023/12/14/gms_cruise_sacrifices_9_execs/"><i>GM's Cruise sheds nine execs in the name of safety and integrity</i></a>:<br />
<blockquote>
GM’s self-driving taxi outfit, Cruise, has dismissed nine execs – including its chief operating officer – after staff withheld information regarding an incident in which a woman was injured by one of the firm's robotaxis.<br />
<br />
"Today, following an initial analysis of the October 2 incident and Cruise's response to it, nine individuals departed Cruise. These include key leaders from Legal, Government Affairs, and Commercial Operations, as well as Safety and Systems," a Cruise spokesperson told <i>The Register</i>.<br />
<br />
"As a company, we are committed to full transparency and are focused on rebuilding trust and operating with the highest standards when it comes to safety, integrity, and accountability and believe that new leadership is necessary to achieve these goals," the spokesperson added.
</blockquote>
It isn't just the executives who are feeling the pain, as Hayden Field and Michael Wayland report in <a href="https://www.cnbc.com/2023/12/14/gms-cruise-laying-off-900-or-24percent-of-its-workforce.html"><i>GM's Cruise laying off 900 employees, or 24% of its workforce: Read the memo here</i></a>:<br />
<blockquote>
Cruise on Thursday announced internally that it will lay off 900 employees, or 24% of its workforce, the company confirmed to CNBC.<br />
<br />
The layoffs, which primarily affected commercial operations and related corporate functions, are the latest turmoil for the robotaxi startup and come one day after Cruise <a href="https://www.cnbc.com/2023/12/13/gms-cruise-dismisses-nine-key-leaders-amid-safety-probe.html">dismissed nine “key leaders”</a> for the company’s response to an Oct. 2 accident in which a pedestrian was dragged 20 feet by a Cruise self-driving car after being struck by another vehicle.<br />
<br />
The company had 3,800 employees before Thursday’s cuts, which also follow a round of contractor layoffs at Cruise last month.
</blockquote>
Templeton's main argument is that Cruise mishandled the PR problem resulting from their vehicle dragging an injured pedestrian. They certainly did, in particular by apparently concealing information from the Dept. of Motor Vehicles. But the PR problem they faced is generic to autonomous vehicles. They are marketed as being safer than human drivers and, statistically, this may even be true. Waymo certainly makes a <a href="https://arstechnica.com/cars/2023/12/human-drivers-crash-a-lot-more-than-waymos-software-data-shows/">plausible claim</a>. But although "safer" implies it does not mean "safe", so serious crashes will inevitably occur.<br />
<br />
When they do the company's PR department is in a no-win situation. The reporters who call asking for the company's reaction know that the company has a vast amount of detailed information about what happened, logs and video. But the PR people don't, and even if they did they don't have the skills to interpret it. Given the nature of the AI driving the vehicle, it will take considerable time for the engineers to find the bug that was the root cause.<br />
<br />
The honest thing for the PR people to say is "we don't have the details, we'll get back to you when we do". Reporters are likely to hear this as buying time for a cover-up.<br />
<br />
The wrong thing for them to do is to give the reporters what little they know, and spin it in ways that minimize the company's fault. Later, when the full details emerge, they will have been shown to have been covering up the worst. Next time, even if they are honest, they won't be believed.<br />
<br />
The PR problem is even worse because it is fundamentally asymmetric. The marketing pitch is that, statistically, autonomous vehicles cause less accidents than humans. But the public will never know about accidents that would have happened were a human driving, but were averted by the AI driver. They will know about accidents that did happen because the AI driver did something that a human would not have, as with the Uber and Cruise incidents. When set against the expectation of airline-level safety, this is an insuperable problem.<br />
<br />
In <a href="https://www.theguardian.com/commentisfree/2023/dec/06/driverless-cars-future-vehicles-public-transport"><i>Driverless cars were the future but now the truth is out: they’re on the road to nowhere</i></a>, Christian Wolmar who has a book about the autonomous winter entitled <a href="https://www.amazon.com/Driverless-Cars-Road-Nowhere-Perspectives/dp/1913019217/"><i>Driverless Cars: On a Road to Nowhere?</i></a> points out that Cruise isn't the first robotaxi company felled by an accident. That would be Uber:<br />
<blockquote>
Right from the start, the hype far outpaced the technological advances. In 2010, at the Shanghai Expo, General Motors had produced a video showing a driverless car taking a pregnant woman to hospital at breakneck speed and, as the commentary assured the viewers, safely. It was precisely the promise of greater safety, cutting the terrible worldwide annual roads death toll of 1.25m, that the sponsors of driverless vehicles dangled in front of the public.<br />
<br />
And that is now proving their undoing. First to go was Uber after an accident in which one of its self-driving cars killed Elaine Herzberg in Phoenix, Arizona. The car was in autonomous mode, and its “operator” was accused of watching a TV show, meaning they did not notice when the car hit Herzberg, who had confused its computers by stepping on to the highway pushing a bike carrying bags on its handlebars. Fatally, the computer could not interpret this confusing array of objects.<br />
<br />
Until then, Uber’s business model had been predicated on the idea that within a few years it would dispense with drivers and provide a fleet of robotaxis. That plan died with Herzberg, and Uber soon pulled out of all its driverless taxi trials.
</blockquote>
In her <a href="https://www.nakedcapitalism.com/2023/12/self-driving-cars-hit-the-wall.html"><i>review of Wolmar's book</i></a> Yves Smith makes two good points:<br />
<blockquote>
Wolmar describes how this Brave New World has stalled out. The big reason is that the world is too complicated. or to put it in Taleb-like terms, there are way too many tail events to get them into training sets for AI in cars to learn about them. The other issue, which Wolmar does not make explicit, is that the public does not appear willing to accept the sort of slip-shod tech standards of buggy consumer software. The airline industry, which is very heavy regulated, has an impeccable safety record, and citizens appear to expect something closer to that…particularly citizens who don’t own or have investments in self-driving cars and never consented to their risks.
</blockquote>
It isn't just Cruise that is figuring out that <a href="https://blog.dshr.org/2023/11/robotaxi-economics.html"><i>Robotaxi Economics</i></a> don't work. Rita Liao's <a href="https://techcrunch.com/2023/12/29/china-robotaxi-apply-the-brakes/"><i>Facing roadblocks, China’s robotaxi darlings apply the brakes</i></a> reports that they don't work in China either:<br />
<blockquote>
Despite years of hype and progress in self-driving technologies, the widespread availability of robotaxis remains a distant reality. That’s due to a confluence of challenges, including safety, regulations and costs.<br />
<br />
The last factor, in particular, is what has pushed China’s robotaxi pioneers toward more opportunistic endeavors. To become profitable, robotaxis need to eventually remove human operators. Though China recently <a href="https://techcrunch.com/2023/12/13/china-autonomous-vehicle-driving-regulation/">clarified rules around the need for human supervision</a>, taxis without a driver behind the wheel are allowed only in restricted areas at present. To attract customers, robotaxi services offer deep discounts on their paid rides.<br />
<br />
Once the subsidies are gone and initial user curiosity wanes, who’s willing to pay the same amount as taxi fares for a few fixed routes?<br />
<br />
Struggling to address that question, China’s robotaxi startups have woken up to the money-burning reality of their business.
</blockquote>
So they are pivoting to a <a href="https://techcrunch.com/2023/12/29/china-robotaxi-apply-the-brakes/">viable product</a>:<br />
<blockquote>
One logical path to monetize self-driving technology is to sell a less robust version of the technology, namely, advanced driver assistance systems (ADAS) that still require human intervention.<br />
<br />
Deeproute, which is backed by Alibaba, significantly scaled back its robotaxi operations this year and plunged right into supplying ADAS to automakers. Its production-ready solution, which includes its smart driving software and lidar-powered hardware, is <a href="https://techcrunch.com/2023/09/05/autonomous-driving-darling-deeproute-ai-sets-sight-on-europe-carmakers/">sold competitively at $2,000</a>. Similarly, Baidu is “downgrading the tech stacks” to find paying customers on its way up what it calls the “Mount Everest of self-driving.”<br />
<br />
“The experience and insight gleaned from deploying our solutions in [mass-produced] vehicles is being fed into our self-driving technology, giving us a unique moat around security and data,” a Baidu spokesperson said.
</blockquote>
Not a good year for the robotaxi concept, the thing that was supposed to distinguish Tesla's cars from everyone else's because they would earn their owners money while they slept.<br />
<h3>Tesla</h3>
As usual, when it comes to self-driving Tesla's story is worse than almost everyone else's. Elon Musk famously claimed that <a href="https://www.autoevolution.com/news/elon-musk-says-tesla-is-worth-zero-without-fsd-why-he-is-completely-wrong-191416.html">Tesla is worth zero without Full Self Driving</a>. But although this is typical Musk BS, but unlike some other utterances it contains a kernel of truth. Tesla is valued as a <i>technology</i> company not a <i>car</i> company. Thus it is critical for Telsa that its technology be viewed as better than those of other car companies; anything that suggests it is limited or inadequate is a big problem not just for the company but also for Musk's personal wealth.<br />
<br />
I believe this is why Tesla hasn't implemented effective driver monitoring nor geo-fenced their systems. And why members of the Musk cult are so keen on defeating the driver monitoring system with weights on the steering wheel and smiley-face stickers on the camera. Depending upon what you believe, the technology is either groundbreaking autonomy or a below average Level-2 driver assistance system (Mercedes has fielded a <a href="https://arstechnica.com/cars/2023/09/mercedes-benzs-level-3-autonomous-driving-system-takes-over-in-heavy-traffic/">Level 3 system</a>). The congnitive dissonance between the pronouncements of the cult leader and the reality of continually being "nagged" about the limits of the technology is too hard for the cult members to take.<br />
<br />
The <i>Washington Post</i> team of Trisha Thadani, Rachel Lerman, Imogen Piper, Faiz Siddiqui and Irfan Uraizee has been delivering outstanding reporting on Tesla's technology. They started on Octoober 6<sup>th</sup> with <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/"><i>The final 11 seconds of a fatal Tesla Autopilot crash</i></a>, in which a Tesla driver enabled Autopilot in conditions for which it was not designed, and set the speed to 14mph above the limit. Then on December 10<sup>th</sup> they followed with <a href="https://www.washingtonpost.com/technology/2023/12/10/tesla-autopilot-crash/"><i>Tesla drivers run Autopilot where it’s not intended — with deadly consequences</i></a> and <a href="https://www.washingtonpost.com/technology/2023/12/10/tesla-autopilot-cross-traffic/"><i>Why Tesla Autopilot shouldn’t be used in as many places as you think</i></a>.<br />
<br />
They recount a 2019 <a href="https://www.washingtonpost.com/technology/2023/12/10/tesla-autopilot-crash/">crash in Florida</a>:<br />
<blockquote>
A Tesla driving on Autopilot crashed through a T intersection at about 70 mph and flung the young couple into the air, killing Benavides Leon and gravely injuring Angulo. In police body-camera footage obtained by The Washington Post, the shaken driver says he was “driving on cruise” and took his eyes off the road when he dropped his phone.<br />
<br />
But the 2019 crash reveals a problem deeper than driver inattention. It occurred on a rural road where Tesla’s Autopilot technology was <a href="https://www.washingtonpost.com/technology/2023/12/10/tesla-autopilot-cross-traffic/">not designed to be used</a>. Dash-cam footage captured by the Tesla and obtained exclusively by The Post shows the car blowing through a stop sign, a blinking light and five yellow signs warning that the road ends and drivers must turn left or right.
</blockquote>
Note that, just like the repeated <a href="https://www.huffpost.com/entry/us-opens-formal-probe-tesla-autopilot_n_611a651de4b0f7bc26a7eef4">crashes into emergency vehicles</a>, the victims did not volunteer to debug Tesla's software. And also that the Autopilot system was driving 15mph above the speed limit on a road it wasn't designed for, just like the <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">2019 Banner crash</a> into a semi-trailer. As I wrote about <a href="https://blog.dshr.org/2023/10/elon-musk-threat-or-menace-part-3.html">that crash</a>:<br />
<blockquote>
It is typical of Tesla's disdain for the law that, although their cars have GPS and can therefore know the speed limit, they didn't bother to program Autopilot to obey the law.<br />
...<br />
Again, Tesla's disdain for the safety of their customers, not to mention other road users, meant that despite the car knowing which road it was on and thus whether it was a road that Autopilot should not be activated on, it allowed Banner to enable it.
</blockquote>
Federal regulators have known there was a problem for more than seven years, but they <a href="https://www.washingtonpost.com/technology/2023/12/10/tesla-autopilot-crash/">haven't taken effective action</a>:<br />
<blockquote>
Nor have federal regulators taken action. After the 2016 crash, which killed Tesla driver Joshua Brown, the National Transportation Safety Board (NTSB) called for limits on where driver-assistance technology could be activated. But as a purely investigative agency, the NTSB has no regulatory power over Tesla. Its peer agency, the National Highway Traffic Safety Administration (NHTSA), which is part of the Department of Transportation, has the authority to establish enforceable auto safety standards — but its failure to act has given rise to an unusual and increasingly tense rift between the two agencies.
</blockquote>
The reason may be that car manufacturers "self-certify" <a href="https://www.washingtonpost.com/technology/2023/12/10/tesla-autopilot-crash/">conformance with safety standards</a>:<br />
<blockquote>
The <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">string of Autopilot crashes</a> reveals the consequences of allowing a rapidly evolving technology to operate on the nation’s roadways without significant government oversight, experts say. While NHTSA has several ongoing investigations into the company and specific crashes, critics argue the agency’s approach is too reactive and has allowed a flawed technology to put Tesla drivers — and those around them — at risk.<br />
<br />
The approach contrasts with federal regulation of planes and railroads, where crashes involving new technology or equipment — such as recurring issues with Boeing’s 737 Max — have resulted in sweeping action by agencies or Congress to ground planes or mandate new safety systems. Unlike planes, which are certified for airworthiness through a process called “type certification,” passenger car models are not prescreened, but are subject to a set of regulations called Federal Motor Vehicle Safety Standards, which manufacturers face the burden to meet.
</blockquote>
And Tesla's self-certification is self-serving.<br />
<br />
Self-certification would work well if the penalty for false certification was severe, but the NTHSA has declined to impose <i>any</i> penalty for Tesla's manifestly inadequate system. It seems that the team's reporting finally drove the NHTSA to do <i>something</i> about the long-standing problems of Autopilot. Reuters reported that <a href="https://www.theguardian.com/technology/2023/dec/13/tesla-recall-us-autopilot"><i>Tesla recalls more than 2m vehicles in US over Autopilot system</i></a>:<br />
<blockquote>
Tesla is recalling just over 2m vehicles in the United States fitted with its Autopilot advanced driver-assistance system to install new safeguards, after a safety regulator said the system was open to “foreseeable misuse”.<br />
<br />
The National Highway Traffic Safety Administration (NHTSA) has been investigating the electric automaker led by the billionaire Elon Musk for more than two years over whether Tesla vehicles adequately ensure that drivers pay attention when using the driver assistance system.<br />
<br />
Tesla said in the recall filing that Autopilot’s software system controls “may not be sufficient to prevent driver misuse” and could increase the risk of a crash.<br />
...<br />
Separately, since 2016, NHTSA has opened more than three dozen Tesla special crash investigations in cases where driver systems such as Autopilot were suspected of being used, with 23 crash deaths reported to date.<br />
<br />
NHTSA said there might be an increased risk of a crash in situations when the system is engaged but the driver does not maintain responsibility for vehicle operation and is unprepared to intervene or fails to recognize when it is canceled or not.
</blockquote>
The obfuscation is extraordinary — "forseeable misuse", "may not be sufficient" and "could increase". There is no "foreseeable", "may" nor "could"; multiple people have already <i>died</i> because the system <i>was</i> abused. When Tesla is sued about these deaths, their <i>defense is that the system was abused</i>! I believe the system is specifically <i>designed</i> to allow abuse, because preventing abuse would puncture the hype bubble.<br />
<br />
Fortunately from the NTHSA's point of view the recall is pure kabuki, posing no risk from Musk's attack-dog lawyers and <a href="https://techcrunch.com/2021/10/20/buttigieg-defends-safety-agency-appointment-after-musk-claims-bias/">cult</a> <a href="https://slate.com/technology/2021/10/tesla-missy-cummings-nhtsa-elon-musk.html">members</a> because the over-the-air update is cheap and doesn't actually fix the problem. The <i>Washington Post</i>'s headline writers didn't understand this whan they captioned the team's timeline <a href="https://www.washingtonpost.com/business/2023/12/16/tesla-autopilot-recall-timeline/"><i>How Tesla Autopilot got grounded</i></a>:<br />
<blockquote>
Now, more than 2 million Tesla vehicles are <a href="https://www.washingtonpost.com/technology/2023/12/16/tesla-autopilot-recall/">receiving a software update</a> to address “insufficient” controls to combat driver inattention while in Autopilot mode. Here’s how the recall unfolded, according to documents from Tesla, safety officials and reporting by The Washington Post.
</blockquote>
The team did understand that Autopilot hadn't been "grounded". In <a href="https://www.washingtonpost.com/technology/2023/12/16/tesla-autopilot-recall/"><i>Recalling almost every Tesla in America won’t fix safety issues, experts say</i></a> they lay it out:<br />
<blockquote>
Tesla this week agreed to issue a remote update to 2 million cars aimed at improving driver attention while Autopilot is engaged, especially on surface roads with cross traffic and other hazards the driver-assistance technology is not designed to detect.<br />
<br />
But the recall — the largest in Tesla’s 20-year history — quickly drew condemnation from experts and lawmakers, who said new warnings and alerts are unlikely to solve Autopilot’s fundamental flaw: that Tesla fails to limit where drivers can turn it on in the first place.<br />
...<br />
Tesla has repeatedly acknowledged in user manuals, legal documents and communications with federal regulators that Autosteer is “intended for use on controlled-access highways” with “a center divider, clear lane markings, and no cross traffic.”
</blockquote>
The <i>Washington Post</i>'s Geoffrey A. Fowler checked on the result of the kabuki, and wrote <a href="https://www.washingtonpost.com/technology/2023/12/31/tesla-autopilot-recall-test/"><i>Testing Tesla’s Autopilot recall, I don’t feel much safer — and neither should you</i></a>:<br />
<blockquote>
Last weekend, my Tesla Model Y received an over-the-air update to make its driver-assistance software safer. In my first test drive of the updated Tesla, it blew through two stop signs without even slowing down.<br />
<br />
In December, Tesla issued its largest-ever recall, affecting almost all of its 2 million cars. It is like the software updates you get on your phone, except this was supposed to prevent drivers from misusing Tesla’s Autopilot software.<br />
<br />
After testing my Tesla update, I don’t feel much safer — and neither should you, knowing that this technology is on the same roads you use.<br />
<br />
During my drive, the updated Tesla steered itself on urban San Francisco streets Autopilot wasn’t designed for. (I was careful to let the tech do its thing only when my hands were hovering by the wheel and I was paying attention.) The recall was supposed to force drivers to pay more attention while using Autopilot by sensing hands on the steering wheel and checking for eyes on the road. Yet my car drove through the city with my hands off the wheel for stretches of a minute or more. I could even activate Autopilot after I placed a sticker over the car’s interior camera used to track my attention.
</blockquote>
Fowler concludes "I found we have every reason to be skeptical this recall does much of anything". Good job, Tesla!<br />
<br />
As a final note Rishi Sunak, the UK's tech-bro Prime Minister is naturally determined to make the UK a leader in autonomous vehicles with his new legislation on the subject. But, being a tech-bro, he has no idea of the fundamental problem they pose. Wolmar does understand it, <a href="https://www.theguardian.com/commentisfree/2023/dec/06/driverless-cars-future-vehicles-public-transport">writing</a>:<br />
<blockquote>
In the UK, Tesla will fall foul of the legislation introduced into parliament last month, which prevents companies from misleading the public about the capability of their vehicles. Tesla’s troubles have been compounded by the revelations from ex-employee Lukasz Krupski who claims the self-drive capabilities of Teslas <a href="https://www.bbc.co.uk/news/technology-67591311">pose a risk to the public</a>. Manufacturers will be forced to specify precisely which functions of the car – steering, brakes, acceleration – have been automated. Tesla will have to change its marketing approach in order to comply. So, while the bill has been promoted as enabling the more rapid introduction of driverless cars, meeting its restrictive terms may prove to be an insuperable obstacle for their developers.
</blockquote>
Tesla's stock market valuation <i>depends</i> upon "misleading the public about the capability of their vehicles".<br />
<h3>Update 12<sup>th</sup> February 2024</h3>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEido-2dOBwEG44uGn0o17mqMRGEh9IbxbheSU8nEm2grsdB5J0ZDpAT5_EZbbvOFksgfJmhOu2S0VeBMslCjYIZrOBbbReP5kYqJYcW8SEMVzWdTVIdTYPnzxsqy7_-mBVU2ko2cat7VMpaUKc8uhJ0xOo6lsVtnCXPZD72Gj9Qwupgdf9dL2_oh0p6OsG2/s1200/Magnificent7.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="675" data-original-width="1200" height="113" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEido-2dOBwEG44uGn0o17mqMRGEh9IbxbheSU8nEm2grsdB5J0ZDpAT5_EZbbvOFksgfJmhOu2S0VeBMslCjYIZrOBbbReP5kYqJYcW8SEMVzWdTVIdTYPnzxsqy7_-mBVU2ko2cat7VMpaUKc8uhJ0xOo6lsVtnCXPZD72Gj9Qwupgdf9dL2_oh0p6OsG2/w200-h113/Magnificent7.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2024-02-11/tesla-collapse-has-investors-wondering-if-it-s-still-magnificent">Source</a></td></tr></tbody></table>
Reinforcement for the last point above comes from Esha Dey's <a href="https://www.bloomberg.com/news/articles/2024-02-11/tesla-collapse-has-investors-wondering-if-it-s-still-magnificent"><i>Tesla’s Slide Has Investors Wondering If It’s Still Magnificent</i></a> and in particular from this chart comparing the history of the price-earnings ratio of the "Magnificent Seven" stocks, Alphabet, Amazon, Apple, Meta, Microsoft,Nvidia, Tesla. Dey writes:<br />
<blockquote>
After doubling last year, Tesla’s stock price is down 22% to start 2024. Compare that to Nvidia Corp.’s 46% surge or Meta Platforms Inc.’s 32% gain since the beginning of the year and it’s easy to see where the questions are coming from. Indeed, it’s by far the worst performer in the Magnificent Seven Index this year.<br />
<br />
The problem for the EV maker is six of those seven companies are benefiting from the enthusiasm surrounding burgeoning artificial intelligence technology. The group hit a record 29.5% weighting in the S&P 500 last week even with Tesla’s decline, according to data compiled by Bloomberg. But despite Musk’s efforts to position his company as an AI investment, the reality is Tesla faces a unique set of challenges.<br />
<br />
“Although Elon Musk would probably disagree, investors don’t see Tesla as an AI play like most of the other Magnificent Seven stocks,” said Matthew Maley, chief market strategist at Miller Tabak + Co. “We have a much different backdrop for Tesla and the others in the Mag Seven — the demand trend for Tesla products is fading, while it’s exploding higher for those companies that are more associated with AI.”
</blockquote>
The problem for Tesla the car company is that increasing investment in launching the Cybertruck and a lower-cost sedan is <a href="https://www.bloomberg.com/news/articles/2024-01-24/tesla-says-growth-rate-to-slow-in-2024-reports-earnings-miss">meeting slowing demand for EVs</a> in general:<br />
<blockquote>
“During the year, others in the Mag Seven were able to show how AI was driving real, profitable business growth,” Brian Johnson, former auto analyst with Barclays and founder of Metonic Advisors, said in an interview. “Tesla investors just got some random Optimus videos, Musk’s admission Dojo was a moon shot and yet another full-self-driving release that may be an improvement but still a long ways from robotaxi capability.”
</blockquote>
Even if it actually were an AI company not a car manufacturer, Tesla's PE is out of line with other, real AI companies. Hence the need for Musk's relentless hype about autonomy and, not incidentally his demands that <a href="https://www.bloomberg.com/news/articles/2024-01-15/elon-musk-wants-greater-control-of-tesla-before-building-its-ai">Tesla double his stake</a> by diluting the sharholders and re-instate the $55B pay package <a href="https://www.bloomberg.com/opinion/articles/2024-01-31/elon-musk-is-overpaid">Delaware court invalidated</a>. He needs these decisions made while Tesla's PE is nearly double that of the next highest AI company, not once it is valued like a car company.David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com7tag:blogger.com,1999:blog-4503292949532760618.post-21854263142075022002024-01-02T08:00:00.000-08:002024-03-11T14:36:47.681-07:00Good News For Tether<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjESzRWPZCJ03RH3zQeeDLVrIil0qmNBGMvjVvfRS-BCPyzGpcBuVxxJ4uUHt-V-ptRTw7QahDOsLzoqwEpj24vDghxJ_yDvoIAyB-NuQgzEn4NOcgmiLmmCxMYI2MrnT6TGTiG7KMj51a96AeYAL7a7quvT_S1uvLfVZjYn4BFXpA8J-B1mLU8ucODgvsl/s630/USDT_1Y_graph_coinmarketcap.jpeg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="380" data-original-width="630" height="121" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjESzRWPZCJ03RH3zQeeDLVrIil0qmNBGMvjVvfRS-BCPyzGpcBuVxxJ4uUHt-V-ptRTw7QahDOsLzoqwEpj24vDghxJ_yDvoIAyB-NuQgzEn4NOcgmiLmmCxMYI2MrnT6TGTiG7KMj51a96AeYAL7a7quvT_S1uvLfVZjYn4BFXpA8J-B1mLU8ucODgvsl/w200-h121/USDT_1Y_graph_coinmarketcap.jpeg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://coinmarketcap.com/currencies/tether/">USDT "market cap"</a></td></tr></tbody></table>
The good news for Tether is shown in this graph, with two huge surges in "market cap" this year. One of about $15B early in the year, and another of about $6B recently. It looks like the euphoria over the prospect of <a href="https://blog.dshr.org/2023/11/desperately-seeking-retail.html">spot Bitcoin ETFs</a> has solved the <a href="https://blog.dshr.org/2022/11/greater-fool-supply-chain-crisis.html"><i>Greater Fool Supply-Chain Crisis</i></a> with the cryptosphere experiencing a massive inflow of around $20B actual dollars. As one might expect from injecting $20B whose only uses are to HODL or to buy cryptocurrency into the market, the result has been a massive bubble in cryptocurrency "prices".<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiUYsQqauaUBYn_uS9gUXrGenOZbltiSOf2DFwDyeBuflKklG5Tr9Jp9NU4I1w4liMFDiAyl9Uogft5I0XQvHoHU0Y6wccRcTttJaOb9_l9xXz6beSGfZcq6GkMT4mo4YN1WOASTQYawqFeC_XaxXxXpO0TNVgagJlLXaaxVuu2vpIaMoPtkyBFzLoKWZT8/s630/BTC_1Y_graph_coinmarketcap.jpeg" style="clear: right; display: block; margin-left: auto; margin-right: auto; padding: 1em 0px; text-align: center;"><img alt="" border="0" data-original-height="380" data-original-width="630" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiUYsQqauaUBYn_uS9gUXrGenOZbltiSOf2DFwDyeBuflKklG5Tr9Jp9NU4I1w4liMFDiAyl9Uogft5I0XQvHoHU0Y6wccRcTttJaOb9_l9xXz6beSGfZcq6GkMT4mo4YN1WOASTQYawqFeC_XaxXxXpO0TNVgagJlLXaaxVuu2vpIaMoPtkyBFzLoKWZT8/s200/BTC_1Y_graph_coinmarketcap.jpeg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://coinmarketcap.com/currencies/bitcoin/">BTC "price"</a></td></tr></tbody></table>
Bitcoin has gone from about $16K at the start of the year to around $42K recently. Ethereum has merely doubled, from about $1.2K to about $2.4K.<br />
<br />
So all is well with the world; Tether gets to keep the interest on another $20B, which at say 4% is an extra $800M/year on their bottom line, and the Bitcoin HODL-ers see their <strike>investment</strike> gamble return a 160% gain. Is all really well with the world? Follow me below the fold.<br />
<span><a name='more'></a></span>
<br />
Did hordes of retail <strike>investors</strike> gamblers really send $20B in cash to Tether? No, because Tether only deals with authorized institutions, and only in amounts over $100K, Did institutions send Tether $20B in cash? It seems unlikely. At various times over its history Tether has been caught minting USDT in return for things other than cash, such as loans to mysterious Chinese companies, or even thin air.
It has never been audited, and has been described as being "<a href="https://www.bloomberg.com/news/features/2021-10-07/crypto-mystery-where-s-the-69-billion-backing-the-stablecoin-tether">practically quilted out of red flags</a>". Matt Levine <a href="https://www.bloomberg.com/opinion/articles/2023-10-31/bad-passwords-are-securities-fraud">says</a> "I feel like eventually Tether is going to be an incredibly interesting story, but I still don’t know what it is."<br />
<br />
Part of the story was revealed by Dirty Bubble Media in <a href="https://www.dirtybubblemedia.com/p/tethers-secret-agent"><i>Tether’s Secret Agent</i></a>:<br />
<blockquote>
When people talk about the leaders of Tether, a few names typically occupy the conversation. Paolo Ardoino, recently promoted from CTO to CEO of the company, is the public face of Tether on Twitter and in the media. Giancarlo Devasini, the failed plastic surgeon, then failed software pirate, now billionaire CFO (how’s that for a career trajectory?), is widely regarded as the de facto leader of the company. And people often joke about Tether’s absentee former CEO, JL van der Velde, asking whether he even exists (he does, and he also was a serial failure before joining Tether).<br />
...<br />
We have been puzzled by Tether’s leadership for some time. The massive success of the company and the complexity of its operations seem beyond the abilities of a small-time scam operator and repeat failure (Giancarlo), an inexperienced and sweaty front man (Paolo), and another failed businessman and absentee executive (JL). Notably, none of these guys had any significant prior experience in finance, and it is unlikely that any of them had the sorts of business and political ties that would be essential for keeping a controversial company like Tether afloat.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgvboHXiL5MuriIDGEQjmhXK9igJZnhEo3RM_GN_rbN9QLfZy3a1RunR-UWHPha-g03jzyKdKHOEXJCDNFIpGFMJfoNv3bWc9L3ikiQjrhKWBY-WjOTN6dSLoaNaJA6DnJoMoGEvGc-WwqaET6csgMQOJrapOejTXngdnx8AQxdw9wnz5pArn5jEvtsVAxZ/s639/Harborne.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="627" data-original-width="639" height="196" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgvboHXiL5MuriIDGEQjmhXK9igJZnhEo3RM_GN_rbN9QLfZy3a1RunR-UWHPha-g03jzyKdKHOEXJCDNFIpGFMJfoNv3bWc9L3ikiQjrhKWBY-WjOTN6dSLoaNaJA6DnJoMoGEvGc-WwqaET6csgMQOJrapOejTXngdnx8AQxdw9wnz5pArn5jEvtsVAxZ/w200-h196/Harborne.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.dirtybubblemedia.com/p/tethers-secret-agent">Harborne?</a></td></tr></tbody></table>
But Dirty Bubble Media identifies <a href="https://www.dirtybubblemedia.com/p/tethers-secret-agent">someone else involved</a>:<br />
<blockquote>
His name is Christopher Charles Sherriff Harborne, AKA Chakrit Sakunkrit. Styling himself a “digital nomad,” Mr. Harborne’s busy hands reach across continents, industries, and political movements. The scope of Mr. Harborne’s activities and the apparent wealth backing his activities is staggering. Among those many diverse interests, it appears that Tether has become one of his most important interests. And Mr. Harborne has been far more than a passive investor in the company. Indeed, the available evidence suggests that Mr. Harborne’s involvement in Tether is far more significant than generally recognized…
</blockquote>
The details of <a href="https://en.wikipedia.org/wiki/Christopher_Harborne">Harborne</a>'s career are <a href="https://www.dirtybubblemedia.com/p/tethers-secret-agent">fascinating</a>:<br />
<blockquote>
Mr. Harborne has major ownership interests in other notable companies. Through a Delaware corporation he is the largest minority shareholder in QinetiQ, a major British defense contractor; this stake is worth around $200 million at present. He is also the sole owner of IFX Payments, a British fintech company specializing in moving large sums of money around the globe (hm). In total, over a dozen corporate entities spread across the globe have been linked to Mr. Harborne, with many more likely still hidden.
</blockquote>
He was a major funder of the <a href="https://www.theguardian.com/politics/2023/dec/30/britons-brexit-bad-uk-poll-eu-finances-nhs">disastrous Brexit campaign</a> and the UK's second <a href="https://www.theguardian.com/politics/2022/oct/20/iceberg-lettuce-in-blonde-wig-outlasts-liz-truss">worst recent Prime Minister</a>, Boris Johnson. He owns "<a href="https://www.dirtybubblemedia.com/p/tethers-secret-agent">around 12% of Tether and a similar percentage of equity in Bitfinex</a>". After the <a href="https://ag.ny.gov/press-release/2021/attorney-general-james-ends-virtual-currency-trading-platform-bitfinexs-illegal">Crypto Capital Corp seizure</a> they lost access to the US banking system, and Harborne was apparently <a href="https://www.dirtybubblemedia.com/p/tethers-secret-agent">critical to rescuing them both</a>. The details of this are the subject of a lawsuit <a href="https://amycastor.com/2024/03/09/tether-ftx-and-deltec-bank-money-time/">Harborne filed against the <i>WSJ</i></a>. <a href="https://www.dirtybubblemedia.com/p/tethers-secret-agent">Dirty Bubble Media</a> updated the post:<br />
<blockquote>
<i>Note: On February 28, 2024, Mr. Harborne sued the Wall Street Journal regarding alleged defamation. Mr Harborne alleges that the Journal misrepresented his opening of a bank account at Signature Bank as attempting to assist Bitfinex during a time of crisis. The Journal removed this portion of the article several days prior to the suit being filed. Mr. Harborne’s attorneys subsequently contacted Dirty Bubble Media requesting parts of the article referring to the Journal’s story be removed; while the case is pending we have acquiesced to their request. Additionally, we have added clarification regarding Mr. Harborne’s role as a “principal” at Bitfinex.</i>
</blockquote>
With a cast of characters this sketchy, skepticism is warranted, and in <a href="https://davidgerard.co.uk/blockchain/2023/12/06/bitcoin-goes-up-can-5-billion-unbacked-tethers-kickstart-a-fresh-crypto-bubble/"><i>Bitcoin goes up! Can 5 billion unbacked tethers kickstart a fresh crypto bubble?</i></a> Amy Castor and David Gerard supply some:<br />
<blockquote>
Bitcoin is over $44,000! In just the last week, the <a href="https://davidgerard.co.uk/blockchain/2022/04/10/the-national-posts-invisible-hand-society-nft-collection-the-facepalm-manifesto/">invisible hand of the market</a> suddenly decided that bitcoins are really good now!<br />
<br />
By complete coincidence, Tether has printed five billion USDT stablecoins in the past month out of thin air as “loans” — backed in the Tether reserve only by the “loans” themselves.<br />
<br />
How high can you pump a number with five billion fake dollars to deploy?<br />
...<br />
In just one month, from November 5 to December 5, Tether’s issuance climbed from 85 billion to 90 billion.<br />
</blockquote>
And the printer is still going. Molly White reports that <a href="https://web3isgoinggreat.com/?id=tether-christmas-2023-mint"><i>Tether mints itself a $1 billion Christmas present</i></a>:<br />
<blockquote>
On December 25, Tether minted 1 billion of its USDT dollar-pegged stablecoin. CEO Paolo Ardoino announced on Twitter that the mint was an "authorized but not issued transaction, meaning that this amount will be used as inventory for next period issuance requests and chain swaps". This seems to be a recent trend for Tether, as similar language was used for a $1 billion mint in September.<br />
<br />
The activity has raised more questions around where the real money backing Tether is coming from, and if it even exists at all. Some have argued that these recent Tether mints are being used to artificially inflate the price of Bitcoin, which has been on an upward trend since mid-October.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhJbZz4zTM1ZScRjtjTrv7Fec7vVLS1naQjx_u1sJgv9eHU4ghbCClZ1yCIUQEK_N8BHPsA3P-p3dclgjk6_EgrqVRAWMBVyojcbBjMe9iBBnd6WaTh8n8gcZd9hWPfgp9dP_0AGwwylNiazsvkSmm1-Yx9As52rKaWzfi1aJoKrYoRAu5C_AFpljluoHMw/s975/USDC_1Y_graph_coinmarketcap.jpeg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="396" data-original-width="975" height="81" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhJbZz4zTM1ZScRjtjTrv7Fec7vVLS1naQjx_u1sJgv9eHU4ghbCClZ1yCIUQEK_N8BHPsA3P-p3dclgjk6_EgrqVRAWMBVyojcbBjMe9iBBnd6WaTh8n8gcZd9hWPfgp9dP_0AGwwylNiazsvkSmm1-Yx9As52rKaWzfi1aJoKrYoRAu5C_AFpljluoHMw/w200-h81/USDC_1Y_graph_coinmarketcap.jpeg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://coinmarketcap.com/currencies/usd-coin/">USDC "market cap"</a></td></tr></tbody></table>
Amy Castor and David Gerard <a href="https://davidgerard.co.uk/blockchain/2023/12/06/bitcoin-goes-up-can-5-billion-unbacked-tethers-kickstart-a-fresh-crypto-bubble/">ask the right question</a>:<br />
<blockquote>
You would think, with that kind of totally genuine and organic market demand for stablecoins, USDC’s issuance would also be going up — but no. USDC’s issuance is 24.4 billion, having seen a steady decrease from 44 billion in March 2023.<br />
<br />
So where is Tether getting all the dollars to back these tethers?<br />
<br />
It isn’t. Tether’s printing press is not fueled by demand. This is Tether issuing loans to some of its biggest customers — printing pseudo-dollars out of thin air, with the only “backing” being the loan itself, counted as an asset. The loans are secured by cryptos held as collateral — not as reserves. No actual dollars flow into the system this way.<br />
</blockquote>
Why would crypto-bros prefer USDT to USDC if it is doubtful that it is fully backed? It might have something to do with USDT lacking "regulatory clarity". How do we know Tether is printing USDT out of thin air? <a href="https://davidgerard.co.uk/blockchain/2023/12/06/bitcoin-goes-up-can-5-billion-unbacked-tethers-kickstart-a-fresh-crypto-bubble/">Because</a>:<br />
<blockquote>
Tether spent years denying that they issued tethers from thin air as loans — then Alex Mashinsky of Celsius Network confirmed in October 2021 that Celsius had been taking out such loans from Tether. It came out in the <a href="https://davidgerard.co.uk/blockchain/2021/10/29/the-cftc-settlement-with-tether-and-bitfinex-42-5-million-dollars-in-fines/">CFTC settlement</a> later that month that they had been doing this precise thing for a while.<br />
<br />
Tether admitted in September that it was making “secured” loans again — after saying in December 2022 that it would reduce its secured loans to zero. [<a href="https://www.wsj.com/finance/currencies/tether-is-lending-its-stablecoins-again-b11705f2">WSJ</a>]
</blockquote>
Presumably, during the "crypto winter", there would have been a need for customers to exchange USDT for USD. But <a href="https://davidgerard.co.uk/blockchain/2023/12/06/bitcoin-goes-up-can-5-billion-unbacked-tethers-kickstart-a-fresh-crypto-bubble/">did they?</a>:<br />
<blockquote>
In mid-2022, after <a href="https://davidgerard.co.uk/blockchain/2022/05/10/terras-stablecoin-does-a-2008-crisis-ust-crashes-and-takes-bitcoin-with-it/">the Terra-Luna collapse</a>, Tether <a href="https://davidgerard.co.uk/blockchain/2022/06/13/celsius-goes-fahrenheit-451-and-number-goes-down/">bragged</a> that it had <a href="https://davidgerard.co.uk/blockchain/2022/06/22/crypto-collapse-latest-the-contagion-spreads/">“redeemed” 16 billion USDT</a>. We would assume most or all of that was loans being canceled and the tethers burned. We certainly don’t know of any independently verifiable evidence that a single actual dollar was transferred in return.<br />
<br />
For comparison, USDC reserves are held in short-term treasuries and cash in US bank accounts. A USDC appears to have an actual dollar backing it — and now that interest rates are up, Circle has been making a ton of money.<br />
<br />
If Tether had billions of real dollars backing its tethers — as it claims — then the folks running Tether could also make a ton of money simply by putting the reserve into Treasury bills. They do not need to be making loans.<br />
<br />
In late 2022, CZ from Binance was deeply upset that Sam Bankman-Fried from FTX might destabilize tethers by <a href="https://amycastor.com/2022/12/18/crypto-collapse-binance-is-not-so-fine-ftx-delaware-vs-ftx-bahamas-celsius-voyager-gemini-tether/">trying to cash out … $250,000 worth</a>. That’s out of a supposed reserve in the <i>billions</i>. This brings into serious question how many actual dollars are anywhere near Tether — clearly not enough.
</blockquote>
What is going on has similarities with Celsius' <a href="https://celsius.network/cel-token-explained">"flywheel"</a>. <a href="https://davidgerard.co.uk/blockchain/2023/12/06/bitcoin-goes-up-can-5-billion-unbacked-tethers-kickstart-a-fresh-crypto-bubble/">Castor and Gerard explain</a>:<br />
<blockquote>
Crypto institutions — exchanges, hedge funds — use the tethers to buy leverage and pump the price. They post their inflated crypto as collateral to borrow more USDT and keep pumping. [<a href="https://www.dirtybubblemedia.com/p/examining-tethers-secret-loan-portfolio">Dirty Bubble</a>]
</blockquote>
These crypto-backed loans are the fuel for the pump inflating the cryptocurrency bubble. <a href="https://www.dirtybubblemedia.com/p/examining-tethers-secret-loan-portfolio">Dirty Bubble Media concludes</a>:<br />
<blockquote>
Based on the current dataset, we can estimate that Tether issued many billions of USDT backed by crypto collateral. The impact is far larger than one might assume just from looking at the loan balances. For example, Tether lent Celsius Network just over $4 billion in total. Our data indicates that other parties like Amber and 3AC similarly received billions in loans, which round-tripped their way through the crypto-conomy without ever touching the real financial system or ever being backed by real money….<br />
<br />
Many questions remain unanswered:<br />
<ul>
<li>What percentage of Tether’s “redemptions” are actually loan repayments?</li>
<li>What is the impact of cycling billions of crypto-backed USDT through the crypto markets?</li>
<li>And, why did Tether’s reported secured loans massively diverge from this data starting in May/June 2022, around the same time as Terra, Celsius Network, and Three Arrows Capital collapsed?</li>
</ul>
</blockquote>
US regulators have been suspicious of Tether for a long time, and they <a href="https://amycastor.com/2023/12/28/crypto-collapse-mt-gox-payouts-tether-hooks-up-the-feds-sec-says-no-to-coinbase-crypto-media-mergers/">seem to have lost patience</a>:<br />
<blockquote>
The US government isn’t entirely happy with Tether’s financial shenanigans. But they’re really unhappy about sanctions violations, especially with what’s going on now in the Middle East. <br />
<br />
So Tether has announced that it will now be freezing OFAC-sanctioned blockchain addresses — and it’s onboarded the US Secret Service and FBI onto Tether! [<a href="https://tether.to/en/tether-introduces-new-policy-to-strengthen-ecosystem-security/">Tether</a>, <a href="https://archive.is/MZtxn">archive</a>; <a href="https://assets.ctfassets.net/vyse88cgwfbl/6KDtp7U4IcH03zPWnpG11n/1b052835c72f2c7be0bb5ec5bd5a89fc/Tether_Lummis_Hill_Follow_up_Letter.pdf">letter</a>, PDF, <a href="https://web.archive.org/web/20231216090755/https://assets.ctfassets.net/vyse88cgwfbl/6KDtp7U4IcH03zPWnpG11n/1b052835c72f2c7be0bb5ec5bd5a89fc/Tether_Lummis_Hill_Follow_up_Letter.pdf">archive</a>]<br />
<br />
Tether doesn’t do anything voluntarily. We expect they were told that they would allow this or an extremely large hammer would come down upon them.
</blockquote>
It looks like Tether has achieved some "regulatory clarity". It isn't just US authorities who can get Tether to freeze wallets. Patrick Tan asks <a href="https://medium.com/chainargos/what-happens-when-tether-freezes-your-tether-5a8ece2bd508"><i>What happens when Tether “freezes” your Tether?</i></a>, and recounts the tale of The Victim, whose wallet was frozen at the request of Indian law enforcement. Tan <a href="https://medium.com/chainargos/what-happens-when-tether-freezes-your-tether-5a8ece2bd508">concludes</a>:<br />
<blockquote>
The Victim’s transactions are at very most 3 hops away from known bad actors, so it’s not entirely unreasonable for Indian authorities to require more information and detailed documents, not to mention the backdrop of ongoing scams abusing already stretched Indian law enforcement agencies.<br />
<br />
For Tether’s part, it looks as though they received a request from Indian law enforcement and followed it.<br />
<br />
But perhaps, and somewhat more significantly, there is also the risk that Tether blacklists the USDT in your wallet, in response to government requests, regardless if those requests are lawful or not.<br />
<br />
It’s entirely possible for government officials or authorities with a personal vendetta, to target causes, or political opponents who receive donations or are known to transact in USDT, and for Tether to err on the side of caution and comply.<br />
<br />
It’s entirely possible that many “innocent” addresses are blacklisted in such opaque processes, collateral damage in purges.
</blockquote>
The Victim might well be collateral damage, but it was essentially impossible to supply the "more information and detailed documents" that law enforcement required to lift the freeze. Such are the risks of operating without "regulatory clarity". But the kind of "regulatory clarity" the US has imposed on Tether isn't likely to help The Victim or others who become collateral damage, it is likely to increase their numbers. An increasing number of users unable to access their funds is a double-edged sword for Tether; the good news is that Tether gets to keep the interest on frozen funds, the <i>good</i> news is that more and more people figure out how risky USDT is.<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com4tag:blogger.com,1999:blog-4503292949532760618.post-32822521695994320272023-12-27T08:00:00.000-08:002023-12-27T08:00:00.150-08:00Make Up Your Mind<table align="center" cellpadding="0" cellspacing="0" class="tr-caption-container" style="margin-left: auto; margin-right: auto;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiV2HiLFRj_8F0-Vbx3TQYxuOYrEtPkBkjqmKLpo6_CGbE-sGYARbhfdRJc-JmPDGq3uVLQmo-Zr294pYRnTY5tSknxHdFPcMBHQ-2g3nD_W2uFxbRe9N9QQ8Oq0I9t5AFYjtJFZUt0hY2b1ycLwdoj9H_cDBbKXG8NLb-wzn4XnR5sV_VueU91XwBb6afO/s923/ShotChaser.png" style="display: block; margin-left: auto; margin-right: auto; padding: 1em 0px; text-align: center;"><img alt="" border="0" data-original-height="542" data-original-width="923" height="235" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiV2HiLFRj_8F0-Vbx3TQYxuOYrEtPkBkjqmKLpo6_CGbE-sGYARbhfdRJc-JmPDGq3uVLQmo-Zr294pYRnTY5tSknxHdFPcMBHQ-2g3nD_W2uFxbRe9N9QQ8Oq0I9t5AFYjtJFZUt0hY2b1ycLwdoj9H_cDBbKXG8NLb-wzn4XnR5sV_VueU91XwBb6afO/w400-h235/ShotChaser.png" width="400" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/crypto">Source</a></td></tr></tbody></table>This isn't <a href="https://www.theonion.com/"><i>America's Finest News Source</i></a>, it is this morning's <a href="https://www.bloomberg.com/crypto">Crypto page</a> from Bloomberg. I couldn't resist posting it.
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com0tag:blogger.com,1999:blog-4503292949532760618.post-65670905632514287122023-12-26T08:00:00.001-08:002023-12-26T08:00:00.131-08:00There Is No Planet B: Part 2<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgkY4YXGd1kdelthAcg8Uzpn3peaOLN2kMrNOvIFUP3QlPfEC_ihhXJWI2TbRhwvSbXPLoxBcvk2YXa4Qp50Q3uqRL9tb595C0RepkjyUmQCKEZqA4jRytGERH6FFA_FYhDpYayg0D1jAgVZirk3xsQGqW5DmxVAGy2OVJMxtkrsJSqxipC7eTKbchQHUYV/s1500/CityOnMars.jpg" style="clear: right; display: block; margin-left: auto; margin-right: auto; padding: 1em 0px; text-align: center;"><img alt="" border="0" data-original-height="1500" data-original-width="987" height="200" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgkY4YXGd1kdelthAcg8Uzpn3peaOLN2kMrNOvIFUP3QlPfEC_ihhXJWI2TbRhwvSbXPLoxBcvk2YXa4Qp50Q3uqRL9tb595C0RepkjyUmQCKEZqA4jRytGERH6FFA_FYhDpYayg0D1jAgVZirk3xsQGqW5DmxVAGy2OVJMxtkrsJSqxipC7eTKbchQHUYV/s200/CityOnMars.jpg" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.amazon.com/gp/product/B0BWH5YYTR/">Source</a></td></tr></tbody></table>
In <a href="https://blog.dshr.org/2023/11/there-is-no-planet-b-part-1.html">Part 1</a> I applied basic arithmetic to the logistics of Elon Musk's claimed plans for colonizing Mars in 2050 to show they were implausible. Below the fold I continue, first by discussing Maciej Cegłowski's equally basic dissection of NASA's economically implausible plans for Mars in <a href="https://idlewords.com/2023/1/why_not_mars.htm"><i>Why Not Mars?</i></a>. Second, by reviewing the host of non-logistical and non-economic problems facing humans attempting to survive on Mars based on:<br />
<ul>
<li>Kelley and Zach Weinersmith's <a href="https://www.amazon.com/gp/product/B0BWH5YYTR/"><i>A City on Mars</i>.</a></li>
<li><a href="https://dx.doi.org/10.48550/arxiv.2309.16806"><i>Towards Sustainable Horizons: A Comprehensive Blueprint for Mars Colonization</i></a> by Florian Neukart.</li>
</ul>
<br />
<span><a name='more'></a></span>
<br />
There are two known programs planning to land humans on Mars, Musk's and NASA's. Cegłowski's massively footnoted (which I omit) post starts by asking the obvious question of each "<a href="https://idlewords.com/2023/1/why_not_mars.htm">What are they trying to achieve?"</a>:<br />
<ul>
<li>Musk's is the establishment of a backup planet for humans in case the primary fails. <a href="https://idlewords.com/2023/1/why_not_mars.htm">Cegłowski writes</a>:<br />
<blockquote>
Elon Musk, the Martian spiritual leader, has talked about the need to “preserve the light of consciousness” by making us a multiplanetary species. As he sees it, Mars is our only way off of a planet crawling with existential risk. And it's not just enough to explore mars; we have make it a backup for all civilization. Failing to stock it with subsistence farming incels would be tantamount to humanity lying down in its open grave.
</blockquote>
The best that can be said for this is that it is "aspirational", many orders of magnitude more aspirational than, for example, the idea of <a href="https://blog.dshr.org/2023/11/robotaxi-economics.html">autonomous Tesla robotaxis</a> earning while their owners sleep.
</li>
<li>NASA's primary goal is much less aspirational; it appears to be to use the idea of "one small step for a man" on Mars to ensure the continuation of the manned space-flight program. Cegłowski is scathing about NASA's inability to articulate a rational justification, <a href="https://idlewords.com/2023/1/why_not_mars.htm">quoting NASA administrator Bill Nelson</a>:<br />
<blockquote>
First of all, we are explorers and adventurers as a species. That basically is the fulfilment of our destiny. But, in that exploration, we’re going to learn new things and develop new things that is going to improve, just as it’s been under our space program, our lives here on Earth.
</blockquote>
The destiny that is being fulfilled here is NASA's manned space-flight program, which has conspicuously failed to deliver "new things ... to improve ... our lives here on earth". The things that <i>have</i> improved lives, from weather forecasts, satellite communications, earth observation and so on have all come from the unmanned side of NASA. As have all the scientific discoveries from astronomical missions such as Hubble and James Webb, and of greater relevance here, planetary missions such as Opportunity and Perseverance. Note how Nelson elides the difference.
</li>
</ul>
If NASA's justification for landing on Mars is science, helping to answer the <a href="https://hitchhikers.fandom.com/wiki/Earth">Ultimate Question of Life, the Universe and Everything</a> by discovering life there, there are three big problems:<br />
<ul>
<li>Economics: using humans increases the cost by at least a couple of orders of magnitude.</li>
<li>Schedule: using humans delays results by at least a couple of decades.</li>
<li>Science: using humans inevitably contaminates Mars with Earth lifeforms, rendering a definite answer to the research questiuon impossible.<br />
</li></ul>
Cegłowski points out that the last couple of decades have seen a vast expansion of understanding of <a href="https://idlewords.com/2023/1/why_not_mars.htm">life on Earth</a>:<br />
<blockquote>
Microbiologists had long suspected that the 12,000 or so known species of microbes were just a fraction of the total, with perhaps another hundred thousand “unculturable” species left to discover. But when new sequencing technology became available at the turn of the century, it showed the number of species might be as high as one trillion. In the genomic gold rush that followed, researchers discovered not just dozens of unsuspected microbial phyla, but two entire new branches of life.<br />
<br />
These new techniques confirmed that earth’s crust is inhabited to a depth of kilometers by a ‘deep biosphere’ of slow-living microbes nourished by geochemical processes and radioactive decay. One group of microbes was discovered still living their best lives[24] 100 million years after being sealed in sedimentary rock. Another was found enjoying a rewarding, long-term relationship with fungal partners deep beneath the seafloor[25].
</blockquote>
And that what this means for the search for <a href="https://idlewords.com/2023/1/why_not_mars.htm">life on Mars</a> is critical:<br />
<blockquote>
The fact that we failed to notice 99.999% of life on Earth until a few years ago is unsettling and has implications for Mars. The existence of a deep biosphere in particular narrows the habitability gap between our planets to the point where it probably doesn’t exist—there is likely at least one corner of Mars that an Earth organism could call home. It also adds support to the theory that life may have started as an interplanetary infection, a literal Venereal disease that spread across the early solar system by meteorite[36]. If that is the case, and if our distant relatives are still alive in some deep Martian cave, then just about the worst way to go looking for them would be to land in a septic spacecraft.
</blockquote>
Cegłowski points out that NASA has to care about <a href="https://idlewords.com/2023/1/why_not_mars.htm">contaminating Mars</a>:<br />
<blockquote>
The requirement to avoid contamination is a clause in the <a href="https://en.wikipedia.org/wiki/Outer_Space_Treaty">1967 Outer Space Treaty</a>. The detailed guidelines for what this means are formulated by an international body called COSPAR.
</blockquote>
Musk's likely attitude to this treaty can be inferred from the title of Douglas Rushkoff's <a href="https://www.theguardian.com/books/2023/nov/25/we-will-coup-whoever-we-want-the-unbearable-hubris-of-musk-and-the-billionaire-tech-bros"><i>‘We will coup whoever we want!’: the unbearable hubris of Musk and the billionaire tech bros</i></a>.<br />
<br />
Cegłowski has noticed <a href="https://idlewords.com/2023/1/why_not_mars.htm">NASA manned spaceflight's track record</a>:<br />
<blockquote>
Sticking a flag in the Martian dust would cost something north of half a trillion dollars, with no realistic prospect of landing before 2050. To borrow a quote from John Young, keeping such a program funded through fifteen consecutive Congresses would require a series “of continuous miracles, interspersed with acts of God”. Like the Space Shuttle and Space Station before it, the Mars program would exist in a state of permanent redesign by budget committee until any logic or sense in the original proposal had been wrung out of it.<br />
...<br />
How long such a program could last is anyone’s guess. But if landing on the Moon taught us anything, it’s that taxpayer enthusiasm for rock collecting has hard limits. At ~$100B per mission, and with launch windows to Mars one election cycle apart, NASA would be playing a form of programmatic Russian roulette. It’s hard to imagine landings going past the single digits before cost or an accident shut the program down.
</blockquote>
Members of the cult of Musk will ignore the elephant in the room, which is life support, <a href="https://idlewords.com/2023/1/why_not_mars.htm">but NASA cannot</a>:<br />
<blockquote>
The things that make going to Mars hard are not fun space things, like needing a bigger rocket, but tedious limits of human physiology. Understanding these limits well enough to get to Mars will require years of human experiments beyond low Earth orbit.<br />
<br />
In particular, we need preliminary data on the physiological effects of partial gravity, and a better estimate of the risk from heavy ion radiation. Since core tradeoffs around crew safety depend on the outcome, these experiments have to be done before NASA can finalize a mission design.
</blockquote>
But partial gravity and radiation aren't the <a href="https://idlewords.com/2023/1/why_not_mars.htm">big life support problem</a>:<br />
<blockquote>
The chief technical obstacle to a Mars landing is not propulsion, but a lack of reliable closed-loop life support. With our current capability, NASA would struggle to keep a crew alive for six months on the White House lawn, let alone for years in a Martian yurt.<br />
<br />
The technology program required to close this gap would be remarkably circular, with no benefits outside the field of applied zero gravity zookeeping. The web of Rube Goldberg devices that recycles floating animal waste on the space station has already cost twice its weight in gold and there is little appetite for it here on Earth, where plants do a better job for free.<br />
...<br />
What makes life support so vexing is that all the subcomponents interact with each other and with the crew. There’s no such thing as a life support unit test; you have to run the whole system in space under conditions that mimic the target mission. Reliability engineering for life support involves solving mysteries like why gunk formed on a certain washer on Day 732, then praying on the next run that your fix doesn’t break on Day 733. The process repeats until the first crew makes it home alive (figuratively speaking), at which point you declare the technology reliable and chill the champagne.
</blockquote>
NASA's requirement is for a completely closed-loop life support system capable of keeping say 6 people healthy for at least 30 months. This is way beyond the state of the art. Musk's is for a similar system capable of keeping a million people healthy indefinitely.<br />
<br />
There isn't a good justification for sending humans to Mars, and the technology for doing it doesn't exist yet. Even assuming the bill was only $500B in 2023 dollars, the idea that Congress would consistently fund $20B every year for a program that wouldn't produce its headline result for two-and-a-half decades fails the laugh test. Implausible is a polite way to describe the economics of NASA's plan.<br />
<br />
Lets suspend disbelief and assume some humans do land on Mars.
Two popular works of hard science fiction about Mars that are fairly plausible about logistics both ignore some of the known problems:<br />
<ul>
<li>Andy Weir's <i>The Martian</i> ignores two big problems. First, Mark Watney spends 549 sols (564 days) on the Martian surface plus the outbound and return trips, for a total of say 1000 days radiation exposure with no shielding. Second, he survives by eating potatoes grown in a mixture of Martian soil and human excrement.</li>
<li>Kim Stanley Robinson's <i>Martian Trilogy</i> describes how Mars is terraformed, including providing it with a breathable atmosphere. He doesn't describe how the atmosphere is prevented from leaking back into space.</li>
</ul>
Problem 1: The collapse of Mars' magnetic field caused it to lose its primordial atmosphere. Without a magnetic field, the same processes would require the terraformed atmosphere to be continually resupplied from elsewhere. Providing Mars with a magnetic field is an interesting engineering problem.<br />
<br />
Problem 2: <a href="https://link.springer.com/article/10.1007/s00159-021-00136-5"><i>Radiation environment for future human exploration on the surface of Mars: the current understanding based on MSL/RAD dose measurements</i></a> by Jingnan Guo <i>et al</i> states:<br />
<blockquote>
The accumulated GCR dose equivalent, via a Hohmann transfer, is about 0.65±0.24 sievert and 1.59±0.12 sievert during solar maximum and minimum periods, respectively.<br />
...<br />
A typical round-trip mission scenario to Mars using a Hohmann transfer trajectory both ways thus includes about 400–500 days of transit time and 500 days of surface stay.
</blockquote>
<a href="https://research.csu.edu.au/integrity-ethics-compliance/radiation/forms-templates-proformas/radiation-life/ionising/how-much">A 1 sievert dose</a>:<br />
<blockquote>
In a short term dose is about the threshold for causing immediate radiation sickness in a person of average physical attributes, but would be unlikely to cause death. Above 1000 mSv, severity of illness increases with dose.
If doses greater than 1000 mSv occur over a long period they are less likely to have early health effects but they create a definite risk that cancer will develop many years later.
</blockquote>
Lets say the radiation dose for the emigrants is around 0.75 sieverts/year. <a href="https://research.csu.edu.au/integrity-ethics-compliance/radiation/forms-templates-proformas/radiation-life/ionising/how-much">0.02 sieverts/year</a>:<br />
<blockquote>
Averaged over 5 years is the limit for radiological personnel such as employees in the nuclear industry, uranium or mineral sands miners and hospital workers (who are all closely monitored).
</blockquote>
So the emigrants are subject to 36 times the radiation level considered safe for persons subjected to radiation in their jobs. Watney would likely get cancer. Robinson acknowledges but underplays the radiation issue.<br />
<br />
Problem 3: Plants grown in Martian soil will take up <a href="https://www.space.com/21554-mars-toxic-perchlorate-chemicals.html">perchlorates</a>:<br />
<blockquote>
"Anybody who is saying they want to go live on the surface of Mars better think about the interaction of perchlorate with the human body," he warned. "At one-half percent, that's a huge amount. Very small amounts are considered toxic. So you'd better have a plan to deal with the poisons on the surface."<br />
<br />
Any humans exploring Mars, Smith said, will find it hard to avoid the finest of dust particles. "It'll get into everything…certainly into your habitat."<br />
...<br />
The research emphasizes that perchlorate is widespread in Martian soils at concentrations of between 0.5 to 1 percent. There are dual implications of calcium perchlorate on Mars. On one hand, at such concentrations, perchlorate could be an important source of oxygen. But it could also become a critical chemical hazard to astronauts.
</blockquote>
Perchlorates aren't just toxic, <a href="https://19january2017snapshot.epa.gov/sites/production/files/2014-03/documents/ffrrofactsheet_contaminant_perchlorate_january2014_final.pdf">but also</a>:<br />
<blockquote>
Perchlorate is highly soluble in water, and relatively stable and mobile in surface and subsurface aqueous systems.
</blockquote>
Watney would have been poisoned both by his potatoes and by breathing the Martian dust. So would Robinson's Martians, who would also be poisoned by the Martian water.<br />
<br />
The implications of the toxicity of the <a href="https://en.wikipedia.org/wiki/Perchlorate#On_Mars">Martian dust</a> for human exploration are so awkward that the response seems to be "ignore it and hope it goes away". For example <a href="https://www.nytimes.com/2023/11/27/science/mars-needs-insects.html"><i>Mars Needs Insects</i></a> by Sarah Scoles in the <i>NYT</i> is about the virtues of soldier fly larvae for growing plants in Martian regolith. It is based on research using simulated Martian regolith from a company called Martian Garden that <a href="https://www.themartiangarden.com/tech-specs"><i>doesn't contain perchlorates</i></a>. Presumably this is because its toxicity would severely limit the potential market.<br />
<br />
But toxicity isn't the only problem posed by Martian dust. It is a real problem for <a href="https://www.nasa.gov/solar-system/the-fact-and-fiction-of-martian-dust-storms/">solar power</a>:<br />
<blockquote>
Mars’ dust storms aren’t totally innocuous, however. Individual dust particles on Mars are very small and slightly electrostatic, so they stick to the surfaces they contact like Styrofoam packing peanuts.<br />
<br />
“If you’ve seen pictures of Curiosity after driving, it’s just filthy,” Smith said. “The dust coats everything and it’s gritty; it gets into mechanical things that move, like gears.”<br />
<br />
The possibility of dust settling on and in machinery is a challenge for engineers designing equipment for Mars. <br />
<br />
This dust is an especially big problem for solar panels. Even dust devils of only a few feet across — which are much smaller than traditional storms — can move enough dust to cover the equipment and decrease the amount of sunlight hitting the panels. Less sunlight means less energy created.<br />
<br />
In “The Martian,” Watney spends part of every day sweeping dust off his solar panels to ensure maximum efficiency, which could represent a real challenge faced by future astronauts on Mars.<br />
</blockquote>
This alone likely makes solar power impractical for powering human habitation on Mars. But there's an even worse problem; every few years Mars is completely covered in a <a href="https://www.nasa.gov/solar-system/the-fact-and-fiction-of-martian-dust-storms/">dust storm that lasts months</a>:<br />
<blockquote>
“Once every three Mars years (about 5 ½ Earth years), on average, normal storms grow into planet-encircling dust storms, and we usually call those ‘global dust storms’ to distinguish them,” Smith said.<br />
...<br />
Global storms can also present a secondary issue, throwing enough dust into the atmosphere to reduce sunlight reaching the surface of Mars.<br />
...<br />
Large global dust storms put enough dust in the air to completely cover the planet and block out the sun, but doing so ultimately dooms the storm itself. The radiative heat of sunlight reaching the surface of the planet is what drives these dust storms.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhwuiwZEvdFaNBDVxBkca-AY4T0r_F-3Hg2_XCEeoTztUBQTH95gKFNeYBrgcO_pJYU3TaHjYRR2YkDSNzcEI1Et_KRX-QKTPcnk5D3tpRZ_DTLfAwxxgG5e-d0HzT4KYEfd8uMeuS03-uiJA4dPxD0dDk2_W-9zyRQBBnDIll_hkgVlEf-9BRrOLoF2QnZ/s800/Mars_dust_opacities_MER-B_Sol_1205_to_1235.jpg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="533" data-original-width="800" height="133" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhwuiwZEvdFaNBDVxBkca-AY4T0r_F-3Hg2_XCEeoTztUBQTH95gKFNeYBrgcO_pJYU3TaHjYRR2YkDSNzcEI1Et_KRX-QKTPcnk5D3tpRZ_DTLfAwxxgG5e-d0HzT4KYEfd8uMeuS03-uiJA4dPxD0dDk2_W-9zyRQBBnDIll_hkgVlEf-9BRrOLoF2QnZ/w200-h133/Mars_dust_opacities_MER-B_Sol_1205_to_1235.jpg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://en.wikipedia.org/wiki/Climate_of_Mars#Atmospheric_properties_and_processes">Source</a></td></tr></tbody></table>
This <a href="https://en.wikipedia.org/wiki/Climate_of_Mars#Atmospheric_properties_and_processes">image is a</a>:<br />
<blockquote>
Time-lapse composite of the Martian horizon as seen by the Opportunity rover over 30 Martian days; it shows how much sunlight the July 2007 dust storms blocked; Tau of 4.7 indicates 99% sunlight was blocked.
</blockquote>
It is clearly uneconomic to over-provision solar panels by a factor of 100, even if they could be swept clear of dust. Nor is it practical to store enough power to last an unpredictable number of months while waiting out the storm. Nuclear reactors are the only feasible way to power human habitation on Mars, but putting tons of nuclear fuel on rockets that can fail and dump their contents in Earth's atmosphere isn't going to be popular. Let alone the chances of contaminating the Martian atmosphere when a reentry is misjudged.<br />
<br />
These are far from the only known problems facing humans on Mars; there are bound to be many others, as yet unknown. Mitigating each of them will involve both difficult engineering and massive expense. All this for no clearly expressed benefit, and very clearly expressed negative scientific benefit. A far more practical and afforable policy would be to divert a small proportion of the funding from manned space-flight to a massive increase in automated exploration of the solar system and its plaents. And divert the rest to dealing with actual pressing problems on Earth, such as climate change.<br />
<br />David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com2tag:blogger.com,1999:blog-4503292949532760618.post-77543778611144861292023-12-12T08:00:00.000-08:002023-12-12T08:00:00.140-08:00Why Worry About Resources?The attitude of the crypto-bros and tech more generally is that they are going to make so much money that paying for whatever resource they need to make it will be a drop in the ocean. Amd that externalities such as carbom emissions are someone else's problem.<br />
<br />
I discussed Proof-of-Work's scandalous waste of energy in my EE380 talk, <a href="https://blog.dshr.org/2022/02/ee380-talk.html"><i>Can We Mitigate Cryptocurrencies' Externalities?</i></a> and elsewhere, <a href="https://blog.dshr.org/2017/09/web-drm-enables-innovative-business.html">since 2017</a> often citing the work of Alex de Vries. Two years ago de Vries and Christian Stoll's <a href="https://doi.org/10.1016/j.resconrec.2021.105901"><i>Bitcoin's growing e-waste problem</i></a> pointed out that in addition to mining rig's direct waste of power, their short economic life drove a massive e-waste problem, adding the embedded energy of the hardware to the problem.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg5-UmqeNGj1ixEna29-6orGK_frJU_RoxZIj-vSvyK-KKkDIlD8YuUHpSKJwyNeEMFTKdjQhzDEEazn3N4LkR4FVAbTRhpevKtp-RjzQH3Aab_7WQpYMcA2DCcNP80HMce40cXPBmyDIVY4KNkytcj4WghDF5SUvut2QSKEHuhrMm4w27JJPCLvbeczHpN/s2367/deVriesFig1.jpg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="2367" data-original-width="2233" height="200" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg5-UmqeNGj1ixEna29-6orGK_frJU_RoxZIj-vSvyK-KKkDIlD8YuUHpSKJwyNeEMFTKdjQhzDEEazn3N4LkR4FVAbTRhpevKtp-RjzQH3Aab_7WQpYMcA2DCcNP80HMce40cXPBmyDIVY4KNkytcj4WghDF5SUvut2QSKEHuhrMm4w27JJPCLvbeczHpN/w189-h200/deVriesFig1.jpg" width="189" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://doi.org/10.1016/j.crsus.2023.100004">de Vries Fig. 1</a></td></tr></tbody></table>
Now, de Vries' <a href="https://doi.org/10.1016/j.crsus.2023.100004"><i>Bitcoin’s growing water footprint</i></a> reveals that supporting <a href="https://blog.dshr.org/2023/05/the-cryptocurrency-use-case.html">gambling, money laundering and crime</a> causes yet another massive waste of resources.<br />
<br />
But that's not all. de Vries has joined a growing chorus of researchers showing that the VC's <i>pivot to AI</i> wastes similar massive amounts of power. Can analysis of AI's e-waste and water consumption be far behind?<br />
<br />
Below the fold I discuss papers by de Vries and others on this issue.<br />
<span><a name='more'></a></span>
<br />
First, de Vries examines the problem highlighted by Jacky Sawicky of the <a href="https://amycastor.com/2023/10/23/texans-versus-bitcoin-jackie-sawicky-and-the-texas-coalition-against-cryptomining/">Texas Coalition Against Cryptomining</a>, water usage by Riot Blockchain's massive Texas mining operation:<br />
<blockquote>
Corsicana will sell Riot the water. Jackie only discovered how much water Riot needed via records requests. “They asked for 1.6 million gallons a day in the height of the summer, and we are all being told to conserve,” Jackie said.<br />
<br />
“When Riot tells people they are going to be wasting 1.6 million gallons a day, more than the iron smelt, more than the candy factory, they are going to be the number one user of water,” said Jackie.
</blockquote>
In <a href="https://doi.org/10.1016/j.crsus.2023.100004"><i>Bitcoin’s growing water footprint</i></a> Alex de Vries discovers some very large water losses:<br />
<blockquote>
Amid growing concerns over the impacts of climate change on worldwide water security, Bitcoin’s water footprint has rapidly escalated in recent years. The water footprint of Bitcoin in 2021 significantly increased by 166% compared with 2020, from 591.2 to 1,573.7 GL. The water footprint per transaction processed on the Bitcoin blockchain for those years amounted to 5,231 and 16,279 L, respectively. As of 2023, Bitcoin’s annual water footprint may equal 2,237 GL.
</blockquote>
That is, each Bitcoin transaction consumes a <a href="https://www.msn.com/en-us/money/technology/a-single-bitcoin-transaction-uses-enough-water-to-fill-a-swimming-pool/ar-AA1kJBzC">small swimming pool worth</a> of water.<br />
<br />
de Vries <a href="https://doi.org/10.1016/j.crsus.2023.100004">computes two kinds of losses</a>:<br />
<blockquote>
The first involves onsite (direct) water use for cooling systems and air humidification. Water usage depends on cooling system types and local climate conditions. It is important to differentiate between water withdrawal and water consumption in terms of this usage. Water withdrawal pertains to the water taken from surface water or groundwater sources, while water consumption refers to the portion of water that becomes unavailable for reuse after withdrawal, primarily due to evaporation in cooling systems. Water consumption is not extensively studied in Bitcoin mining or generic data center research, as reliable data on water consumption factors are challenging to obtain.
</blockquote>
<a href="https://doi.org/10.1016/j.crsus.2023.100004">And</a>:<br />
<blockquote>
The second way in which miners use water relates to the (indirect) water consumption associated with generating the electricity necessary to power their devices. Thermoelectric power generation plays a major role in water consumption, as a portion of the withdrawn water for cooling purposes evaporates (unless dry cooling technologies utilizing air are employed).
</blockquote>
Note some <a href="https://doi.org/10.1016/j.crsus.2023.100004">caveats</a> (my emphasis):<br />
<blockquote>
These systems can utilize both freshwater and non-freshwater sources. This commentary, however, exclusively focuses on freshwater <i>consumption</i>. ...
The total water footprint of Bitcoin examined in this commentary encompasses the freshwater <i>consumed</i> due to both direct and indirect water <i>consumption</i> during the operational stage of Bitcoin mining devices.
</blockquote>
It is important to note that water consumed is not the only concern, withdrawn water can be too. <a href="https://amycastor.com/2023/10/23/texans-versus-bitcoin-jackie-sawicky-and-the-texas-coalition-against-cryptomining/">Sawicky points out that</a>:<br />
<blockquote>
The water gets quite hot — over 100 degrees Fahrenheit — and that hot water needs to go somewhere. Where will it go?<br />
<br />
Jackie hasn’t been able to get a clear answer, but she has concerns. “The property Riot purchased here in Navarro Country feeds into Richland Creek and Richland Creek feeds into Richland-Chambers Reservoir, and that’s the tap water for Arlington-Fort Worth.”
</blockquote>
Hot water kills fish, and any pollution from the Riot facility will end up in the water supply.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi0XL69evW-Cerb1uI4NtUiQKejGiq6P6ddYNTnMe1rBliqmPKioKh7NTiY1IGA90Kan-2YuB6FZQoOy3FIRU2JniWSZnS1UhU4e7BgGFqCBA1cYG0u1U-PSeJJXcoO89JuZypaPQ-2dE_5xTbtY2PD_9uv7E426aE9oSL455Tj3SyNwqmb9jsFcqYK9lPO/s2233/deVriesFig2.jpg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="1857" data-original-width="2233" height="166" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi0XL69evW-Cerb1uI4NtUiQKejGiq6P6ddYNTnMe1rBliqmPKioKh7NTiY1IGA90Kan-2YuB6FZQoOy3FIRU2JniWSZnS1UhU4e7BgGFqCBA1cYG0u1U-PSeJJXcoO89JuZypaPQ-2dE_5xTbtY2PD_9uv7E426aE9oSL455Tj3SyNwqmb9jsFcqYK9lPO/w200-h166/deVriesFig2.jpg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://doi.org/10.1016/j.crsus.2023.100004">de Vries Fig. 2</a></td></tr></tbody></table>
The US hosts a significant part of the total Bitcoin mining fleet. The <a href="https://www.nytimes.com/2023/04/09/business/bitcoin-mining-electricity-pollution.html"><i>New York Times</i> list</a> of <a href="https://doi.org/10.1016/j.crsus.2023.100004">large US miners</a>:<br />
<blockquote>
includes 34 Bitcoin mines, with the power requirement for each mine ranging from 38 to 450 megawatts (MW) as of March 2023. Together, these 34 mines are responsible for 3.91 GW of power demand, representing roughly a quarter of Bitcoin’s total estimated power demand in the same month (i.e., 16.2 GW) and a majority of the share that can be attributed to the US according to the CCAF (i.e., 37.84% as of January 2022).
</blockquote>
de Vries <a href="https://doi.org/10.1016/j.crsus.2023.100004">estimates that</a>:<br />
<blockquote>
This means the total water footprint of US Bitcoin miners, after adding their direct water consumption, could be an annual 93–120 GL, which is 10%–41% more than the estimated indirect water consumption of 84.9 GL per year. It also means the total water footprint of US Bitcoin miners could be equivalent to the average annual water consumption of around 300,000 US households, comparable with a city such as Washington, DC.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgVqfyBDVLXth7VciNxVTEDtNRSQMLfD9x8hfv2hi9eQZ0xmezaGu8mLkdlmdvdCQgIY-__f7tHAoJqMbuskHNloAwVANh-wilhcmAKoh85ly3QJFSDoPZQYy6Rqn6XRjO1xGXcbr4Pm79Ww_YqsmlbSxRiBBL2shNm4m81S8Mj8AOP6r7yjCJtwuCTSJ8W/s850/CBECI.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="530" data-original-width="850" height="125" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgVqfyBDVLXth7VciNxVTEDtNRSQMLfD9x8hfv2hi9eQZ0xmezaGu8mLkdlmdvdCQgIY-__f7tHAoJqMbuskHNloAwVANh-wilhcmAKoh85ly3QJFSDoPZQYy6Rqn6XRjO1xGXcbr4Pm79Ww_YqsmlbSxRiBBL2shNm4m81S8Mj8AOP6r7yjCJtwuCTSJ8W/w200-h125/CBECI.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://ccaf.io/cbnsi/cbeci">Source</a></td></tr></tbody></table>
As I write, the Bitcoin blockchain is processing <a href="https://www.blockchain.com/explorer/charts">498,790 transactions/day</a> using <a href="https://ccaf.io/cbnsi/cbeci">16.84GW</a>.
<a href="https://www.nber.org/papers/w29396">Igor Makarov and Antoinette Schoar write</a>:<br />
<blockquote>
90% of transaction volume on the Bitcoin blockchain is not tied to economically meaningful activities but is the byproduct of the Bitcoin protocol design as well as the preference of many participants for anonymity.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhnIHFUIkCy-7Ry0xH6_RFR-ipfkkIa7rkdvjFQsHVuEqoz-HGnc3k3NyfabFF_3R878LDBecf_oYYQgqX_qRCM-x9famPwddZyf6VHFqWFaL3EgoGQ1ln5_-JjVFURpUD2DNbawKhIu05ACepZhH6GpNbqwIKCmTLNgi1-2J10v-Eg-w0W9xhGw4lmThGP/s1440/n-transactions.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="810" data-original-width="1440" height="113" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhnIHFUIkCy-7Ry0xH6_RFR-ipfkkIa7rkdvjFQsHVuEqoz-HGnc3k3NyfabFF_3R878LDBecf_oYYQgqX_qRCM-x9famPwddZyf6VHFqWFaL3EgoGQ1ln5_-JjVFURpUD2DNbawKhIu05ACepZhH6GpNbqwIKCmTLNgi1-2J10v-Eg-w0W9xhGw4lmThGP/w200-h113/n-transactions.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.blockchain.com/explorer/charts/n-transactions">Source</a></td></tr></tbody></table>
Thus only about <a href="https://www.nber.org/papers/w29396">10% are "economically meaningful"</a> transactions between individuals and exchanges (about 2080/hour), so the average <i>economically meaningful</i> transaction consumes 4MWh. It generates <a href="https://blog.dshr.org/2022/02/ee380-talk.html">an average of a quarter of a MacBook Air</a> of e-waste, and uses 160K liters of water.<br />
<br />
For comparison, our house uses 4.3MWh/year, generates perhaps 2 MacBook Airs of e-waste per year, and consumes 223K liters of water per year. So each economically meaningful Bitcoin transaction consumes almost our annual electricity usage, about 6 weeks of our e-waste generation, and about 70% of our annual water usage.<br />
<br />
de Vries then follows the current hype cycle and pivots to AI. In <a href="https://doi.org/10.1016/j.joule.2023.09.004"><i>The growing energy footprint of artificial intelligence</i></a> he starts with the notoriously compute-intensive training phase:<br />
<blockquote>
Hugging Face reported that its BigScience Large Open-Science Open-Access Multilingual (BLOOM) model consumed 433 MWh of electricity during training.4 Other LLMs, including GPT-3, Gopher and Open Pre-trained Transformer (OPT), reportedly used 1,287, 1,066, and 324 MWh, respectively, for training. Each of these LLMs, was trained on terabytes of data and has 175 billion or more parameters.
</blockquote>
de Vries then turns to the less-studied <a href="https://doi.org/10.1016/j.joule.2023.09.004">inference phase</a>:<br />
<blockquote>
Research firm SemiAnalysis suggested that OpenAI required 3,617 of NVIDIA’s HGX A100 servers, with a total of 28,936 graphics processing units (GPUs), to support ChatGPT, implying an energy demand of 564 MWh per day. Compared to the estimated 1,287 MWh used in GPT-3’s training phase, the inference phase’s energy demand appears considerably higher. Furthermore, Google reported that 60% of AI-related energy consumption from 2019 to 2021 stemmed from inference.
</blockquote>
He estimates for Google the <a href="https://doi.org/10.1016/j.joule.2023.09.004">energy needed per inference in LLMs</a>:<br />
<blockquote>
SemiAnalysis estimated that implementing AI similar to ChatGPT in each Google search would require 512,821 of NVIDIA’s A100 HGX servers, totaling 4,102,568 GPUs. At a power demand of 6.5 kW per server, this would translate into a daily electricity consumption of 80 GWh and an annual consumption of 29.2 TWh. New Street Research independently arrived at similar estimates, suggesting that Google would need approximately 400,000 servers, which would lead to a daily consumption of 62.4 GWh and an annual consumption of 22.8 TWh. With Google currently processing up to 9 billion searches daily, these scenarios would average to an energy consumption of 6.9–8.9 Wh per request. This estimate aligns with Hugging Face’s BLOOM model, which consumed 914 kWh of electricity for 230,768 requests,4 averaging to 3.96 Wh per request.<br />
...<br />
Alphabet’s chairman indicated in February 2023 that interacting with an LLM could ‘‘likely cost 10 times more than a standard keyword search.6" As a standard Google search reportedly uses 0.3 Wh of electricity,9 this suggests an electricity consumption of approximately 3 Wh per LLM interaction.
</blockquote>
And similarly for <a href="https://doi.org/10.1016/j.joule.2023.09.004">ChatGPT</a>:<br />
<blockquote>
SemiAnalysis’ assessment of ChatGPT’s operating costs in early 2023, which estimated that ChatGPT responds to 195 million requests per day, requiring an estimated average electricity consumption of 564 MWh per day, or, at most, 2.9 Wh per request.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgcn7ww8koW7xd3bINrqFogxoq-BYZ7GKUNaNkYJ-nNiDDgyDTMGFU7aMiiu5hYorN1SLBaKo8O4iScS3unDlJDE1v4O74qTbDZUAZM-4Ja4V8ur9fMbuAhKFeyja9gSKdZLv1c1N5zRMdRXbK3AR4IcivmRpF2hsnm-5hTrku-d_yIF0wFLeGtH0tmzkg_/s1564/deVriesFig1.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="892" data-original-width="1564" height="114" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgcn7ww8koW7xd3bINrqFogxoq-BYZ7GKUNaNkYJ-nNiDDgyDTMGFU7aMiiu5hYorN1SLBaKo8O4iScS3unDlJDE1v4O74qTbDZUAZM-4Ja4V8ur9fMbuAhKFeyja9gSKdZLv1c1N5zRMdRXbK3AR4IcivmRpF2hsnm-5hTrku-d_yIF0wFLeGtH0tmzkg_/w200-h114/deVriesFig1.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://doi.org/10.1016/j.joule.2023.09.004">de Vries Fig. 1</a></td></tr></tbody></table>
de Vries considers a worst-case scenario in which <i>every</i> search involved interacting with an LLM:<br />
<blockquote>
In 2021, Google’s total electricity consumption was 18.3 TWh, with AI accounting for 10%–15% of this total. The worst-case scenario suggests Google’s AI alone could consume as much electricity as a country such as Ireland (29.3 TWh per year),
</blockquote>
Fortunately, the hardware supply chain can't deliver the chips needed to do this, and even if it could Google would not want to make the roughly $100B investment. A more realistic scenario is based on Nvidia's projected sales of 100,000 AI servers in 2023:<br />
<blockquote>
these servers would have a combined power demand of 650–1,020 MW. On an annual basis, these servers could consume up to 5.7–8.9 TWh of electricity. Compared to the historical estimated annual electricity consumption of data centers, which was 205 TWh, this is almost negligible.
</blockquote>
Note that this estimate assumes 100% duty cycle but ignores the power needed to cool the servers, two effects which tend to cancel each other. It also ignores the embedded energy in the servers, and their eventual contribution to the e-waste problem. de Vries concludes:<br />
<blockquote>
While the exact future of AI-related electricity consumption remains difficult to predict, the scenarios discussed in this commentary underscore the importance of tempering both overly optimistic and overly pessimistic expectations. Integrating AI into applications such as Google Search can significantly boost the electricity consumption of these applications. However, various resource factors are likely to restrain the growth of global AIrelated electricity consumption in the near term. Simultaneously, it is probably too optimistic to expect that improvements in hardware and software efficiencies will fully offset any longterm changes in AI-related electricity consumption. These advancements can trigger a rebound effect whereby increasing efficiency leads to increased demand for AI, escalating rather than reducing total resource use.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjKPZhAos7JVO96Yhde6Bm8aN2qu0jAH4diezaueejfEekKbh7enWfmRUQsftZIkRN9xOHdc3IymqYiCcBfzKZO0XWm547Ib55kp-t8JMzUAdx0rKsPEFZPhK3cbtGbxHO4Gde79hjDmTI0o0gVQckqZ6ZSU0B5YqMk0MDhGQKm2HrjNctyNWyinaNZPQcR/s1339/LuccioniFig1.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="687" data-original-width="1339" height="103" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjKPZhAos7JVO96Yhde6Bm8aN2qu0jAH4diezaueejfEekKbh7enWfmRUQsftZIkRN9xOHdc3IymqYiCcBfzKZO0XWm547Ib55kp-t8JMzUAdx0rKsPEFZPhK3cbtGbxHO4Gde79hjDmTI0o0gVQckqZ6ZSU0B5YqMk0MDhGQKm2HrjNctyNWyinaNZPQcR/w200-h103/LuccioniFig1.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://arxiv.org/pdf/2311.16863.pdf">Luccioni Fig. 1</a></td></tr></tbody></table>
de Vries' pivot to the cost of AI inference is joined by Alexandra Sasha Luccioni, Yacine Jernite and Emma Strubell with <a href="https://arxiv.org/pdf/2311.16863.pdf"><i>Power Hungry Processing: Watts Driving the Cost of AI Deployment?</i></a> who:<br />
<blockquote>
propose the first systematic comparison of the ongoing inference cost of various categories of ML systems, covering both task-specific (i.e. finetuned models that carry out a single task) and ‘general-purpose’ models, (i.e. those trained for multiple tasks). We measure deployment cost as the amount of energy and carbon required to perform 1,000 inferences on representative benchmark dataset using these models. We find that multi-purpose, generative architectures are orders of magnitude more expensive than task-specific systems for a variety of tasks, even when controlling for the number of model parameters.
</blockquote>
As with de Vries, they justify their <a href="https://arxiv.org/pdf/2311.16863.pdf">focus on inference</a>:<br />
<blockquote>
According to AWS, the largest global cloud provider, inference is estimated to make up 80 to 90% of total ML cloud computing demand [2, 28], whereas a 2021 publication by Meta attributed approximately one-third of their internal end-to-end ML carbon footprint to model inference, with the remainder produced by data management, storage, and training [56]; similarly, a 2022 study from Google attributed 60% of its ML energy use to inference, compared to 40% for training [40].
</blockquote>
Luccioni <i>et al</i>'s Figure 1 (note Y axis logarithmic) shows that the range of carbon emissions for the various tasks they tested spans 3 orders of magnitude. Generative tasks are by far more expensive, with image generation particularly costly.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiKmeU0L_guZ3xVOLbae3X6iOxIvfvpiLlliGaVeCmfw_phjizrsKJluM4expzSApPQbk1sOBq4nU_MfvmcLuaqfXgLcAssh60PsaLp-gHG_aE-mfrFtyuUoURGA7AJnM8IUtR3Xvch1QGTn3VoNv75UOT_b7UDz6FhLWv-rGEVGPafKu_LIRznf2lZbU_W/s1350/LuccioniFig2.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="504" data-original-width="1350" height="74" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiKmeU0L_guZ3xVOLbae3X6iOxIvfvpiLlliGaVeCmfw_phjizrsKJluM4expzSApPQbk1sOBq4nU_MfvmcLuaqfXgLcAssh60PsaLp-gHG_aE-mfrFtyuUoURGA7AJnM8IUtR3Xvch1QGTn3VoNv75UOT_b7UDz6FhLWv-rGEVGPafKu_LIRznf2lZbU_W/w200-h74/LuccioniFig2.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://arxiv.org/pdf/2311.16863.pdf">Luccioni Fig. 2</a></td></tr></tbody></table>
They report that although increasing model size increases emissions, the effect is <a href="https://arxiv.org/pdf/2311.16863.pdf">less than that of the task</a>:<br />
<blockquote>
We do observe a relationship between model size and quantity of emissions produced during inference, with differing progressions for each modality – however, the task structure accounts for more of the variation than the model size does. We can observe once again that text-to-image is by far the most carbon- and energy-intensive task, with smaller image generation models such as <tt>segmind/tiny-sd</tt> that have around 500M parameters producing magnitudes more carbon than text-to-category models (100g vs. 0.6 g of CO 2 per 1,000 inferences).<br />
...<br />
For context, the most carbon-intensive image generation model <tt>(stable-diffusion-xl-base-1.0)</tt> generates 1,594 grams of CO 2 for 1,000 inferences, which is roughly the equivalent to 4.1 miles driven by an average gasoline-powered passenger vehicle.
</blockquote>
They <a href="https://arxiv.org/pdf/2311.16863.pdf">conclude</a>:<br />
<blockquote>
<i>Using multi-purpose models for discriminative tasks is more energy-intensive compared to task-specific models for these same tasks.</i> This is especially the case for text classification (on IMDB, SST 2 and Rotten Tomatoes) and question answering (on SciQ, SQuAD v1 and v2), where the gap between task-specific and zero-shot models is particularly large, and less so for summarization (for CNN-Daily Mail, SamSUM and XSum). As can be seen in Table 4, the difference between multi-purpose models and task-specific models is amplified as the length of output gets longer.<br />
...<br />
While we see the benefit of deploying generative zero-shot models given their ability to carry out multiple tasks, we do not see convincing evidence for the necessity of their deployment in contexts where tasks are well-defined, for instance web search and navigation, given these models’ energy requirements.
</blockquote>
One of the most-hyped applications of AI is <a href="https://blog.dshr.org/2023/11/robotaxi-economics.html">autonomous vehicles</a>, and Soumya Sudhakar, Vivienne Sze and Sertac Karaman examine their externalities in <a href="https://doi.org/10.1109/MM.2022.3219803"><i>Data Centers on Wheels: Emissions From Computing Onboard Autonomous Vehicles</i></a> which is summarized by Brandon Vigliarolo in <a href="https://www.theregister.com/2023/01/13/selfdriving_car_computers_could_equal/"><i>Self-driving car computers may be 'as bad' for emissions as datacenters</i></a>. Sudhakar <i>et al</i> use probabilistic models to estimate the power needed for autonomous navigation:<br />
<blockquote>
Based on current trends, a widespread AV adoption scenario where approximately 95% of all vehicles are autonomous requires computer power to be less than 1.2 kW for emissions from computing on AVs to be less than emissions from all data centers in 2018 in 90% of modeled scenarios. Anticipating a future scenario with high adoption of AVs, business-as-usual decarbonization, and workloads doubling every three years, hardware efficiency must double every 1.1 years for emissions in 2050 to equal 2018 data center emissions. The rate of increase in hardware efficiency needed in many scenarios to contain emissions is faster than the current rate.
</blockquote>
Again, note that in aggregate autonomous vehicles will consume far more power in inference than in training.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhl1u20zOZRInFckjq7WDscGFCXMV3V1a3ckVfmXogcXNqAAXTmsVPr6qRM4WSETbWYupqOt2iI67tS5w2SbHbsDIjRKLVWTICGHb3nNu17VU3HCcCqKfOm8eIVeCeEwcmkOzkCi8N7g3yL5hCA5FnKBaYQBS_O2nFZCaEvl27I9y5_8gMI6p4ob17OHWd8/s762/SudhakarFig1.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="645" data-original-width="762" height="169" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhl1u20zOZRInFckjq7WDscGFCXMV3V1a3ckVfmXogcXNqAAXTmsVPr6qRM4WSETbWYupqOt2iI67tS5w2SbHbsDIjRKLVWTICGHb3nNu17VU3HCcCqKfOm8eIVeCeEwcmkOzkCi8N7g3yL5hCA5FnKBaYQBS_O2nFZCaEvl27I9y5_8gMI6p4ob17OHWd8/w200-h169/SudhakarFig1.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://doi.org/10.1109/MM.2022.3219803">Sudhakar <i>et al</i> Fig. 1</a><br /></td></tr></tbody></table>
History shows that the efficiency of the on-board computers in autonomous vehicles has doubled every 2.8 years. Their models, using plausible assumptions as to parameters such as the adoption rate of the technology, and the growth of the navigation workload, show that unless the hardware efficiency increases much faster emissions from autonomous vehicles will rapidly eclipse the total from all data centers in 2018. For example, their Figure 1 shows:<br />
<blockquote>
Emissions from computing onboard AVs driving 1 h/day. With one billion AVs, an average computer power of 0.84 kW yields emissions equal to emissions of all data centers.
</blockquote>
In all these cases we see the industry's total lack of concern for externalities such as carbon emissions, grid stability, e-waste and water consumption. In cryptocurrency's case we can add <a href="https://blog.dshr.org/2023/05/the-cryptocurrency-use-case.html">crime</a> and <a href="https://blog.dshr.org/2022/09/cryptocurrency-enabled-crime.html">money laundering</a>. In the case of autonomous vehicles we can add <a href="https://blog.dshr.org/2023/10/elon-musk-threat-or-menace-part-3.html">death</a>, <a href="https://blog.dshr.org/2023/11/robotaxi-economics.html">injury</a> and <a href="https://blog.dshr.org/2023/11/robotaxi-economics.html">traffic congestion</a>. All these papers propose ways to reduce the externalities they document, but in almost all cases my reaction is "good luck with that!".<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com5tag:blogger.com,1999:blog-4503292949532760618.post-88673825054138196112023-11-30T08:00:00.000-08:002023-11-30T08:00:00.141-08:00There Is No Planet B: Part 1<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://upload.wikimedia.org/wikipedia/commons/d/d7/Starship-IFT2-ascent_(cropped).jpg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="800" data-original-width="421" height="200" src="https://upload.wikimedia.org/wikipedia/commons/d/d7/Starship-IFT2-ascent_(cropped).jpg" width="105" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://commons.wikimedia.org/w/index.php?title=User:Osunpokeh">Source<br /></a><a href="https://creativecommons.org/licenses/by-sa/4.0/">CC-BY-SA-4.0</a></td></tr></tbody></table>
Anything Elon Musk says must be treated skeptically. This is particularly true of anything involving timescales (see <a href="https://blog.dshr.org/2021/08/autonowashing.html">Tesla robotaxis</a>). And it is even more true of Musk's plans for <a href="https://www.spacex.com/human-spaceflight/mars/">visiting</a> and eventually colonizing Mars.<br />
<br />
Below the fold in part 1 of this two-part post, I apply some arithmetic just to the logistics of Musk's plans for Mars. Part 2 isn't specific to Musk's plans; I discuss two attempts to list the set of "knowns" about Mars exploration, for which the science is fairly clear but the engineering and the economics don't exist, and the much larger set of "known unknowns", critical aspects requiring robust solutions for which the science, let alone the engineering, doesn't exist:<br />
<ul>
<li>Kelley and Zach Weinersmith's <a href="https://www.amazon.com/gp/product/B0BWH5YYTR/"><i>A City on Mars</i>.</a>
</li><li>Maciej Cegłowski's <a href="https://idlewords.com/2023/1/why_not_mars.htm"><i>Why Not Mars?</i></a>.</li>
</ul>
<br />
<span><a name='more'></a></span>
<blockquote>
<i>Amateurs talk strategy, professionals talk logistics</i><br />
<blockquote>
Attributed to <a href="https://www.quora.com/Where-does-the-quote-of-Amateurs-talk-strategy-professionals-talk-logistics-come-from-Is-it-true-or-not">General Omar Bradley</a>
</blockquote>
</blockquote>
Chris Young reported that <a href="https://interestingengineering.com/innovation/elon-musk-spacex-1000-starships-humans-mars"><i>Elon Musk: SpaceX will build over 1,000 Starships to move 1 million humans to Mars</i></a>:<br />
<blockquote>
The plan is to "build 1000+ Starships to transport life to Mars. Basically, (very) modern Noah's Arks," Musk wrote, reiterating a statement he had made during a recent interview with TED curator Chris Anderson. In that interview, he stated that SpaceX would achieve this goal by 2050.
</blockquote>
<a href="https://en.wikipedia.org/wiki/SpaceX_Starship#Space_colonization">Wikipedia reports that</a>:<br />
<blockquote>
SpaceX and Musk have stated their goal of colonizing Mars to ensure the long-term survival of humanity, with an ambition of sending a thousand Starship spacecraft to Mars during a Mars launch window in a very far future.
</blockquote>
SpaceX's Starship is claimed to be able to "<a href="https://www.science.org/content/article/starship-spacex-mega-rocket-makes-it-to-space">carry more than 100 tons of payload to the lunar surface in a single flight</a>" and, using on-orbit refueling, "<a href="https://www.spacex.com/human-spaceflight/mars/">up to 100 tons all the way to Mars</a>". The Mars mission depends upon refueling the Starship on-orbit from fuel tankers launched beforehand on the Super Heavy booster.<br />
<br />
"<a href="https://en.wikipedia.org/wiki/SpaceX_Starship#Potential_uses">Musk has predicted that a Starship orbital launch will eventually cost $1 million</a>" but a more realistic estimate is about $1.5M, or $10/Kg. Each launch requires a <i>lot</i> of fuel.
"<a href="https://en.wikipedia.org/wiki/SpaceX_Starship#Planned_mission_profile">Roughly four hundred truck deliveries are needed for one launch</a>"<br />
<br />
The problem with refueling on-orbit is that getting the fuel up for one mission requires a <i>lot</i> of launches. For a manned Moon landing <a href="https://en.wikipedia.org/wiki/SpaceX_Starship#Starship_spacecraft">the Government Accountability Office said that SpaceX would "require 16 launches overall"</a>. So delivering 100,000 tonnes to Mars would require 16,000 Super Heavy launches costing around $24B and involving 6.4M truck deliveries. One might think that the launches could take the whole of the 780-day interval between <a href="https://en.wikipedia.org/wiki/Exploration_of_Mars#Launch_windows">Mars launch windows</a>, or a rate of just over 20/day. But because the fuel in the orbital tanks boils off over time, the launches have to happen much faster than that, perhaps around 50/day. in 2022 the <a href="https://en.wikipedia.org/wiki/Falcon_9#Launch_history">Falcon 9 launched 60 times</a>, or 0.16/day. Getting to 50/day requires scaling up over 300 times.<br />
<br />
SpaceX's Falcon 9 has an astonishing <a href="https://en.wikipedia.org/wiki/Falcon_9#Launch_history">99.3% success rate</a>. If the Mars vehicle had the same rate, an extra 113 launches would be needed to cover the failures. Of the 1,000 manned launches, 7 would be failures, carrying 7,000 people.<br />
<br />
1000 Starships carrying 1,000,000 people is 1,000 per Starship, which can deliver 100 tons to Mars. That is 220lb/person, so apart from the person and the spacesuit they will need to disembark, they won't have many carry-on bags in the overheads, let alone the air and food needed for the journey. Everything they need to survive on Mars must already have been delivered by earlier missions. Lets guess that each person needs 10 times their weight in life support and other equipment. So the 1M person launch window must be preceded by 10 similar launch windows delivering freight. Now we are talking $264B in launch costs alone, or $264K per person. Even for Elon Musk, over a quarter-trillion dollars is real money. Even that may not be enough. The history of Mars missions shows that Mars landing is a high-risk endeavour. It is <i>very</i> unlikely that all 10,000 freight missions would be successful.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgolhDFB3zfoi15DzUSF3cbuTJtoGjmiy8fDKTeXSE63pvEsZRXh3aoJVlHo87XPFGHqI6SUOVubSLlnGFlUNir3rdRGnYBNGXejPNq5dlLzIOz3qsygWcYecaPN9LqcTFIhti0tEG4UWcCEqNffgEDCwaIRh7qr2tH1KStenvGPdPGhkv5eNwXBQq5EAif/s1150/LaunchWindows.png" imageanchor="1" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="731" data-original-width="1150" height="127" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgolhDFB3zfoi15DzUSF3cbuTJtoGjmiy8fDKTeXSE63pvEsZRXh3aoJVlHo87XPFGHqI6SUOVubSLlnGFlUNir3rdRGnYBNGXejPNq5dlLzIOz3qsygWcYecaPN9LqcTFIhti0tEG4UWcCEqNffgEDCwaIRh7qr2tH1KStenvGPdPGhkv5eNwXBQq5EAif/w200-h127/LaunchWindows.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.reddit.com/r/SpaceXLounge/comments/dtm5bc/mars_launch_windows_20202030/?rdt=55849">Source</a></td></tr></tbody></table>
10 launch windows is 7,800 days or more than 21 years. So if Musk wants to land a million humans by 2050 he needs to start launching 50 Starships a day in 2029, a bit more than 5 years from now. Fortunately, he has time to experiment. There are <a href="https://www.reddit.com/r/SpaceXLounge/comments/dtm5bc/mars_launch_windows_20202030/?rdt=55849">two launch windows</a> before the one when he needs to start in earnest.<br />
<br />
If the Starships are to be reused, they have to be re-fueled on Mars using fuel made locally from Martian resources. Each Starship requires <a href="https://en.wikipedia.org/wiki/SpaceX_Starship#Starship_spacecraft">1,200 tons</a> of fuel. Thus unless the Starships are to be expended, during the 2024 and 2026 launch windows Musk needs to deliver a factory to Mars capable of producing 1.2M tons of fuel every 780 days, or nearly 1600 tons/day. Clearly, the factory must weigh several times its daily output. Lets guess it weighs 20 times, or 32,000 tons. So in each of those two launch windows Musk needs to send another 160 Starships to Mars, requiring another 2,560 launches.<br />
<br />
But there's another huge problem with Musk's fantasy. The 1,000 Martian emigrants will spend something like 180 days cooped up in each Starship's payload bay, which has a volume of <a href="https://en.wikipedia.org/wiki/SpaceX_Starship#Starship_spacecraft">1,000m<sup>3</sup></a>. The average human's volume is around <a href="https://www.syfy.com/syfy-wire/the-human-cube-the-volume-of-humanity">0.06m<sup>3</sup></a>. The <a href="https://en.wikipedia.org/wiki/Boeing_737#Specifications">cabin volume per passenger of a 1-class 737</a> is around 0.22m<sup>3</sup>, so each emigrant will spend 180 days in the equivalent of 4 seats in coach in a 737. I think people paying more than a quarter-million dollars in launch costs alone would be imagining something more like business class! The migrants will end up fighting each other long before they arrive.<br />
<br />
And, of course, once the million people land on Mars the story isn't over. They will still be dependent upon supplies from Earth until they can build a completely self-sustaining ecology. They may not need 100,000 tons of supplies every 780 days, but there will still need to be a lot of launches to get fuel into orbit to get a smaller number of Starships with supplies to Mars.<br />
<br />
Look, I totally understand that what people like Elon Musk and Jeff Bezos really want is to call up the <a href="https://www.bbc.co.uk/cult/hitchhikers/guide/magrathea.shtml">Magratheans</a> and order one of their custom-made luxury planets lovingly made to their exacting specifications, so they don't have to deal with taxes, government, competitors or people who disagree with them. Let alone having to survive in a 2.5C warmer world racked with war, migration and starvation that is depleting their support staff. It must be really frustrating that niggling little issues like the speed of light mean that the only planet they can afford isn't just a long, slow, expensive commute, but also needs a <i>lot</i> of work to make it a suitable home for a multi-billionaire. That work is the subject of part 2.<br />
<br />
All it took for me to get this level of understanding of just the logistics of Musk's fantasy was his own numbers, an Internet connection, a couple of hours, and basic arithmetic. Given Musk's notorious lack of credibility when it comes to schedules, it is disappointing that as far as I can tell no journalist made the effort to inform the public that Musk was BS-ing. Chris Young expressed mild skepticism, "just not very realistic" and "risks putting him in similar territory as he was with Tesla's progress on Level 5 autonomy". But that is a long way from explaining Musk's specific implausibilities.<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com1tag:blogger.com,1999:blog-4503292949532760618.post-36633408495268810292023-11-28T08:00:00.000-08:002023-11-28T08:00:00.136-08:00Decentralized Finance Isn'tA major theme of this blog since 2014's <a href="https://blog.dshr.org/2014/10/economies-of-scale-in-peer-to-peer.html"><i>Economies of Scale in Peer-to-Peer Networks</i></a> has been that decentralized systems aren't, because economic forces overwhelm the technologies of decentralization. Last year I noted that this rule applied to Decentralized Finance (DeFi) in <a href="https://blog.dshr.org/2022/03/shadow-banking-20.html"><i>Shadow Banking 2.0</i></a> based on Prof. Hilary Allen's <a href="https://ssrn.com/abstract=4038788"><i>DeFi: Shadow Banking 2.0?</i></a> which she summarizes thus:<br />
<blockquote>
TL;DR: DeFi is neither decentralized, nor very good finance, so regulators should have no qualms about clamping down on it to protect the stability of our financial system and broader economy.
</blockquote>
And also <a href="https://www.bis.org/publ/qtrpdf/r_qt2112b.htm"><i> DeFi risks and the decentralisation illusion </i></a> by Sirio Aramonte, Wenqian Huang and Andreas Schrimpf of the Bank for International Settlements who write:<br />
<blockquote>
While the main vision of DeFi’s proponents is intermediation without centralised entities, we argue that some form of centralisation is inevitable. As such, there is a “decentralisation illusion”. First and foremost, centralised governance is needed to take strategic and operational decisions. In addition, some features in DeFi, notably the consensus mechanism, favour a concentration of power.
</blockquote>
Below the fold I look at new evidence that the process of centralizing DeFi is essentially complete.<br />
<span><a name='more'></a></span>
<br />
In <a href="https://www.bloomberg.com/news/articles/2023-11-16/defi-is-becoming-less-competitive-a-year-after-ftx-s-collapse-battered-crypto"><i>DeFi Is Becoming Less Competitive a Year After FTX’s Collapse Battered Crypto</i></a> Muyao Shen highlights the situation:<br />
<blockquote>
A small number of participants are dominating the world of decentralized finance as the crypto sector, which seeks to replicate financial markets without middlemen, still hasn’t recovered from FTX’s collapse a year ago.<br />
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgHHD8mf564mXHnAdezpbg7qrMsZnK0HeaYaGgr_iSm87BQQ3gOqllPvS7fhnBtNamb3kpWmsfCOi68JHidBwcoc4e1f3uO7vN31YJAKBT-FZTwd-hjv30oDGpWy6I8ooVT-CoT-_OHceYD08yQrnsBLnLoU9v1wlDk9cefVR82vY1pQKWjkmzdOrX1CoTr/s938/H-H-Index.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="796" data-original-width="938" height="170" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgHHD8mf564mXHnAdezpbg7qrMsZnK0HeaYaGgr_iSm87BQQ3gOqllPvS7fhnBtNamb3kpWmsfCOi68JHidBwcoc4e1f3uO7vN31YJAKBT-FZTwd-hjv30oDGpWy6I8ooVT-CoT-_OHceYD08yQrnsBLnLoU9v1wlDk9cefVR82vY1pQKWjkmzdOrX1CoTr/w200-h170/H-H-Index.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2023-11-16/defi-is-becoming-less-competitive-a-year-after-ftx-s-collapse-battered-crypto">Source</a></td></tr></tbody></table>
Shen uses <a href="https://www.bloomberg.com/news/articles/2023-11-16/defi-is-becoming-less-competitive-a-year-after-ftx-s-collapse-battered-crypto">research from a "crypto-risk modeling company"</a>:<br />
<blockquote>
Most categories in DeFi — from peer-to-peer lending to decentralized exchanges — are seeing capital largely held in a few major projects, according to data compiled by crypto-risk modeling company Gauntlet. The firm used a popular measure of market concentration and competition called the Herfindahl-Hirschman Index.
</blockquote>
<a href="https://en.wikipedia.org/wiki/Herfindahl%E2%80%93Hirschman_index">Wikipedia</a> explains the HHI thus:<br />
<blockquote>
HHI is calculated by squaring the market share of each competing firm in the industry—expressed as either fractions, decimals, or whole numbers—and then summing the resulting numbers ... The result is proportional to the average market share, weighted by market share. As such, it can range from 0 to 1.0, moving from a huge number of very small firms to a single monopolistic producer.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjrCgX3F1JTuA_5YlT3D5QFg9CbgFF112-Vgra8JJFgIHlWDDhPt0DRWXzZSTSSBo1gUYzJ1i4lSFXDM6aMBHxLxTmDFh-ImLIEgTKi-DKnEU1gIIeP-K6Nj7Uu-e7-UVtdwiYHy1ETO56RWk6Zux7nUm-sBqH0zI6VKYJt-D81EkVdmpPfXHH91opvZZU-/s940/DeFiRevenue.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="752" data-original-width="940" height="160" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjrCgX3F1JTuA_5YlT3D5QFg9CbgFF112-Vgra8JJFgIHlWDDhPt0DRWXzZSTSSBo1gUYzJ1i4lSFXDM6aMBHxLxTmDFh-ImLIEgTKi-DKnEU1gIIeP-K6Nj7Uu-e7-UVtdwiYHy1ETO56RWk6Zux7nUm-sBqH0zI6VKYJt-D81EkVdmpPfXHH91opvZZU-/w200-h160/DeFiRevenue.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2023-11-16/defi-is-becoming-less-competitive-a-year-after-ftx-s-collapse-battered-crypto">Source</a></td></tr></tbody></table>
In Shen's graph above the HHI is expressed in "points" from 0 to 10,000. Shen <a href="https://www.bloomberg.com/news/articles/2023-11-16/defi-is-becoming-less-competitive-a-year-after-ftx-s-collapse-battered-crypto">comments</a>:<br />
<blockquote>
Based on the metric, the most competition exists between decentralized finance exchanges, with the top four venues holding about 54% of total market share. Other categories including decentralized derivatives exchanges, DeFi lenders, and liquid staking, are much less competitive. For example, the top four liquid staking projects hold about 90% of total market share in that category, according to Gauntlet.
</blockquote>
<font size="-2">
<table align="right" border="3" cols="3">
<thead>
<tr bgcolor="lightgray"><th>Protocol</th><th>Revenue</th><th>Market</th></tr>
<tr bgcolor="lightgray"><th> </th><th>$M</th><th>Share %</th></tr>
</thead><tbody>
<tr><td>Lido</td><td>304</td><td>55.2</td></tr>
<tr><td>Uniswap V3</td><td>55</td><td>10.0</td></tr>
<tr><td>Maker DAO</td><td>48</td><td>8.7</td></tr>
<tr><td>AAVE V3</td><td>24</td><td>4.4</td></tr>
<tr bgcolor="lightgray"><td><b>Top 4</b></td><td> </td><td>78.2</td></tr>
<tr><td>Venus</td><td>18</td><td>3.3</td></tr>
<tr><td>GMX</td><td>14</td><td>2.5</td></tr>
<tr><td>Rari Fuse</td><td>14</td><td>2.5</td></tr>
<tr><td>Rocket Pool</td><td>14</td><td>2.5</td></tr>
<tr><td>Pancake Swap AMM V3</td><td>13</td><td>2.4</td></tr>
<tr><td>Compound V2</td><td>13</td><td>2.4</td></tr>
<tr><td>Morpho Aave V2</td><td>10</td><td>1.8</td></tr>
<tr><td>Goldfinch</td><td>9</td><td>1.6</td></tr>
<tr><td>Aura Finance</td><td>8</td><td>1.5</td></tr>
<tr><td>Yearn Finance</td><td>7</td><td>1.3</td></tr>
<tr><td>Stargate</td><td>5</td><td>0.9 </td></tr>
<tr bgcolor="lightgray"><td><b>Total</b></td><td>551</td><td> </td></tr>
</tbody>
</table>
</font>
Shen provides a breakdown of the last 180 days revenue of DeFi projects, which I reformatted into this table. It shows that the top project, Lido, has 55% of the revenue, the top two have 2/3, and the top four projects have 78%. This is clearly a highly concentrated market, typical of cryptocurrency markets in general.<br />
<br />
Why does the centralization of Defi matter? First, for all the bad things concentration causes in general markets — see for example Matt Stoller's <a href="https://www.thebignewsletter.com/"><i>Big</i></a>. But more importantly, because the technologies of decentralization impose massive costs over and above those of equivalent centralized systems. The incentive to pay these additional costs is to reap the profits centralized systems sacrifice to regulation. If you are paying the costs but not evading regulation, what is the point? I discussed this in <a href="https://blog.dshr.org/2023/02/economic-incentives.html"><i>Economic Incentives</i></a>.<br />
<br />
I've argued many times, most recently to the <a href="https://blog.dshr.org/2023/09/brief-remarks-to-iosco-defi-wg.html">IOSCO DeFi Working Group</a>, that regulators should ignore the alleged "decentralization" of cryptocurrency systems and <a href="https://blog.dshr.org/2022/12/dominoes.html">go after the major players</a> in each area. With lawsuits against <a href="https://web3isgoinggreat.com/?id=sam-bankman-fried-convicted">FTX' management</a>, <a href="https://davidgerard.co.uk/blockchain/2023/06/07/the-sec-sues-coinbase-its-on/">Coinbase</a>, <a href="https://davidgerard.co.uk/blockchain/2023/10/30/new-york-vs-dcg-genesis-and-gemini-kids-kids-youre-both-ugly/">DCG/Genesis and Gemini</a>, <a href="https://web3isgoinggreat.com/?id=kraken-sued-by-us-sec">Kraken</a>, and <a href="https://davidgerard.co.uk/blockchain/2023/07/15/crypto-collapse-alex-mashinsky-of-celsius-arrested-ripples-bizarre-xrp-win-gemini-sues-genesis-gisele-bundchen-knew-nothing/">Celsius and Alex Mashinsky</a>, and now their <a href="https://www.justice.gov/opa/pr/binance-and-ceo-plead-guilty-federal-charges-4b-resolution">$4.3B criminal conviction of Binance and CZ</a>, the US regulators no longer seem to be going after the small fry. Presumably, they used winning against the small fry to build precedents they are now using against bigger fish.<br />
<br />
If despite claims of "decentralization" regulators can shut down the few big players in a market, the decentralized systems will be paying the massive extra costs but not reaping the unregulated profits. Of course, there is a <a href="https://en.wikipedia.org/wiki/Whac-A-Mole">Whac-A-Mole</a> aspect; kill a big player and one or more smaller players will get bigger enough to be worth taking out. But it won't take many of these cycles for the market to get the message.<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com3tag:blogger.com,1999:blog-4503292949532760618.post-82794511442344303802023-11-21T08:00:00.000-08:002023-11-21T08:00:00.154-08:00Desperately Seeking Retail<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiBgFzjRaYYqik8Gg39yutgFI2TVwfs7tyobeAAKWSF_CxSLlwCYRDFS5pbWf6aGD0Mv_0vUE_srt-qfyfZAfBlvl6TV83XyhrSPEkCEV4QjFbZxB0x6STQzmy0ZYDX96I-QD6NB5sj58hFUOUr4fk3WTCIAco6Inte4yV48c35Wo7Gr9VhXuIk0kFn3sYL/s1200/BTC-rally.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="675" data-original-width="1200" height="113" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiBgFzjRaYYqik8Gg39yutgFI2TVwfs7tyobeAAKWSF_CxSLlwCYRDFS5pbWf6aGD0Mv_0vUE_srt-qfyfZAfBlvl6TV83XyhrSPEkCEV4QjFbZxB0x6STQzmy0ZYDX96I-QD6NB5sj58hFUOUr4fk3WTCIAco6Inte4yV48c35Wo7Gr9VhXuIk0kFn3sYL/w200-h113/BTC-rally.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2023-11-10/bitcoin-etf-exuberance-drives-four-week-nothing-for-sale-rally">Source</a></td></tr></tbody></table>
The SEC has a long history of refusing to approve spot Bitcoin ETFs, on the reasonable basis that the Bitcoin market was heavily manipulated. Crypto-skeptics like <a href="https://hackernoon.com/meet-spoofy-how-a-single-entity-dominates-the-price-of-bitcoin-39c711d28eb4">Bitfinex'ed</a> and <a href="https://davidgerard.co.uk/blockchain/2018/05/24/explaining-the-justice-department-cftc-criminal-investigation-into-bitcoin-trading-and-likely-bitfinex-tether/">Davd Gerard</a> have been pointing out obvious instances of manipulation for many years, and there is a considerable academic literature demonstrating manipulation, such as <a href="https://www.nber.org/papers/w30783"><i>Crypto Wash Trading</i></a> by Lin William Cong <i>et al</i>, which demonstrates:<br />
<blockquote>
abnormal first-significant-digit distributions, size rounding, and transaction tail distributions on unregulated exchanges reveal rampant manipulations unlikely driven by strategy or exchange heterogeneity. We quantify the wash trading on each unregulated exchange, which averaged over 70% of the reported volume.
</blockquote>
Unfortunately, despite knowing about the manipulation, the <a href="https://www.coindesk.com/policy/2022/09/29/bitcoin-could-double-in-price-under-cftc-regulation-chairman-behnam-says/">CFTC approved Bitcoin futures ETFs</a>. Among the requests that the <a href="https://blog.dshr.org/2022/07/pump-and-dump-schemes.html">SEC refused</a> was one from <a href="https://blog.dshr.org/2022/04/grayscale-bitcoin-trust.html">Grayscale Bitcoin Trust</a>. Grayscale sued the SEC and, back in August, the panel of judges <a href="https://www.reuters.com/business/finance/whats-stake-grayscales-spot-bitcoin-etf-case-against-sec-2023-08-29/">ruled in their favor</a>:<br />
<blockquote>
The court's panel of judges said Grayscale showed that its proposed bitcoin ETF is "materially similar" to the approved bitcoin futures ETFs. That's because the underlying assets— bitcoin and bitcoin futures - are "closely correlated," and because the surveillance sharing agreements with the CME are "identical and should have the same likelihood of detecting fraudulent or manipulative conduct in the market for bitcoin."<br />
<br />
With that in mind, the court ruled that the SEC was "arbitrary and capricious" to reject the filing because it "never explained why Grayscale owning bitcoins rather than bitcoin futures affects the CME’s ability to detect fraud."
</blockquote>
The prospect of the SEC approving this ETF and others from, for example, BlackRock and Fidelity led to something of a buying frenzy in Bitcoin as shown in the "price" chart, and in headlines such as <a href="https://www.bloomberg.com/news/articles/2023-11-10/bitcoin-etf-exuberance-drives-four-week-nothing-for-sale-rally"><i>Bitcoin ETF Exuberance Drives Four-Week ‘Nothing for Sale’ Rally</i></a>:<br />
<blockquote>
Bitcoin is climbing for a fourth consecutive week, with the digital token’s price lingering just below an 18-month high of $38,000, as more investors bet that US exchange-traded funds that hold the largest cryptocurrency are on the verge of winning regulatory approval.
</blockquote>
Below the fold I look into where this euphoria came from, and why it might be misplaced<br />
<span><a name='more'></a></span>
<br />
There already are spot Bitcoin ETFs, just not in the USA. I believe the <a href="https://amycastor.com/2022/04/19/welcome-to-grayscales-hotel-california/">first launched</a> in:<br />
<blockquote>
February 2021, when the <a href="https://www.globenewswire.com/news-release/2021/02/18/2177751/0/en/Purpose-Investments-Launches-World-s-First-Bitcoin-ETF-Invested-Directly-in-the-Digital-Asset.html">Purpose bitcoin ETF</a> launched in Canada. Unlike GBTC, which trades over-the-counter, Purpose trades on the Toronto Stock Exchange, close to NAV. At 1%, its management fees are half that of GBTC. Within a month of trading, Purpose quickly absorbed more than $1 billion worth of assets.
</blockquote>
So why would a US-based Bitcoin ETF be a big deal? The thesis is that by making it easier for US <strike>investors</strike>gamblers to put money into BTC it would increase demand and thus the price. This anticipated flood of retail money would solve the long-standing problem for cryptocurrency exchanges and whales, the <a href="https://blog.dshr.org/2022/11/greater-fool-supply-chain-crisis.html"><i>Greater Fool Supply-Chain Crisis</i></a>. There is a <a href="https://www.bloomberg.com/news/articles/2023-11-11/ftx-bankruptcy-one-year-later-here-s-how-crypto-has-changed?srnd=cryptocurrencies-v2">shortage of suckers</a>:<br />
<blockquote>
The number of over-the-counter desks has declined, with mainly the more conservative ones remaining, according to Tegan Kline, co-founder of Edge & Node, which developed a crypto project called The Graph. That, combined with the erosion of leverage, has sapped liquidity.<br />
<br />
“Leverage is gone,” Kline said. “A lot of people have pulled money out of the system or they have money stuck at FTX.”
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgxNdf6HWfgyFgRJKIgFmK4_pdS8Gw6SQW2OOrXyBw0OiJ2Zyj1QHX5ocDC67b52I-PF8m3sbb8zeByxY-43tUx3DLYLIGXzWpyFVdepJs0tZN0W0EcyP4KIriuZxjTlWZHEsN-83Wk2dvozfoEyxRjypNO29T_5zv-rf4jH_ejT_8Wrgp3xreWRltdrCgI/s958/CoinbaseTradingVolume.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="538" data-original-width="958" height="113" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgxNdf6HWfgyFgRJKIgFmK4_pdS8Gw6SQW2OOrXyBw0OiJ2Zyj1QHX5ocDC67b52I-PF8m3sbb8zeByxY-43tUx3DLYLIGXzWpyFVdepJs0tZN0W0EcyP4KIriuZxjTlWZHEsN-83Wk2dvozfoEyxRjypNO29T_5zv-rf4jH_ejT_8Wrgp3xreWRltdrCgI/w200-h113/CoinbaseTradingVolume.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2023-10-11/coinbase-coin-trading-volume-is-likely-lowest-since-before-public-debut">Source</a></td></tr></tbody></table>
The "price" has been soaring, even though the anticipated flood of retail dollars hasn't yet, and may never, show up. That is the real problem, as shown in the graph of Coinbase's trading volume. Volume, and thus liquidity, has collapsed since the heady days of 2021. In <a href="https://amycastor.com/2023/11/07/coinbase-q3-earnings-regulatory-clarity-is-all-we-need-and-a-miracle-or-two/"><i>Coinbase Q3 earnings: Regulatory clarity is all we need. And a miracle or two</i></a>, Amy Castor and David Gerard lay out the recent history of Coinbase's income from trading fees, which I have re-formatted into a table.<br />
<br />
<table align="right" border="5" cols="4">
<thead>
<tr><th>Quarter</th><th>Retail<br />$M</th><th>Institutions<br />$M</th><th>Institutions<br />%</th>
</tr></thead>
<tbody>
<tr><td>Q4 2021</td><td align="right">2,185.8</td><td align="right">90.8</td><td align="right">4.2</td></tr>
<tr><td>Q1 2022</td><td align="right">965.8</td><td align="right">47.2</td><td align="right">4.9</td></tr>
<tr><td>Q2 2022</td><td align="right">616.2</td><td align="right">39.0</td><td align="right">6.3</td></tr>
<tr><td>Q3 2022</td><td align="right">346.1</td><td align="right">19.8</td><td align="right">5.7</td></tr>
<tr><td>Q4 2022</td><td align="right">308.8</td><td align="right">13.4</td><td align="right">4.3</td></tr>
<tr><td>Q1 2023</td><td align="right">352.4</td><td align="right">22.3</td><td align="right">6.3</td></tr>
<tr><td>Q2 2023</td><td align="right">310.0</td><td align="right">17.1</td><td align="right">5.5</td></tr>
<tr><td>Q3 2023</td><td align="right">274.5</td><td align="right">14.1</td><td align="right">5.1</td></tr>
</tbody>
</table>
The numbers show three things clearly:<br />
<ul>
<li>Retail trading fee income has collapsed, down 87% from its peak.</li>
<li>Institutional trading fee income has collapsed 84% from its peak.</li>
<li>Coinbase is wholly dependent upon retail traders, who account for around 95% of its fee income.</li>
</ul>
Castor and Gerard <a href="https://amycastor.com/2023/11/07/coinbase-q3-earnings-regulatory-clarity-is-all-we-need-and-a-miracle-or-two/">write</a>:<br />
<blockquote>
Trading volume is the lifeblood of a crypto exchange — and Coinbase’s is through the floor. The exchange saw $76 billion in total trading volume in Q3, down from $92 billion in Q2. (They did $547 billion in trading volume in Q4 2021, their last profitable quarter.)
</blockquote>
So volume is down 84% from its peak. No wonder Coinbase laid off staff in <a href="https://nypost.com/2022/06/18/coinbases-layoffs-show-growing-callous-side-of-cryptos-plummet/">June 2022</a> and <a href="https://www.nytimes.com/2023/01/10/business/coinbase-layoffs-crypto.html">January 2023</a>. And why <a href="https://www.bloomberg.com/news/articles/2023-11-14/crypto-exchange-blockchain-com-raises-110-million-at-fraction-of-prior-value?srnd=cryptocurrencies-v2">Blockchain.com just raised money</a> at less than half its valuation in 2022.<br />
<br />
The problems are exacerbated by the fact that last June the <a href="https://davidgerard.co.uk/blockchain/2023/06/07/the-sec-sues-coinbase-its-on/">SEC sued Coinbase</a> for:<br />
<blockquote>
selling unregistered securities and running an exchange, a broker-dealer, and a clearinghouse as part of the same operation — and without registering any of these.
</blockquote>
Castor and Gerard <a href="https://amycastor.com/2023/11/07/coinbase-q3-earnings-regulatory-clarity-is-all-we-need-and-a-miracle-or-two/">point out that</a>:<br />
<blockquote>
If the SEC wins, Coinbase may have to stop trading in any cryptos other than CFTC-regulated commodity coins such as bitcoin. There isn’t enough volume in those for Coinbase to live on.<br />
<br />
This is why Coinbase is so insistent on trading blatant unregistered securities — it’s all they have left for a business model.
</blockquote>
Lynn Parramore transcribes an interview with Jim Chanos in <a href="https://www.nakedcapitalism.com/2023/11/jim-chanos-the-crypto-ecosystem-is-well-suited-for-the-dark-side-of-finance.html"><i>Jim Chanos: “The Crypto Ecosystem Is Well-Suited for the Dark Side of Finance.”</i></a> that expresses the famed short-seller's skepticism:<br />
<blockquote>
JC: ... The last iteration that you just touched upon is the hope by aspects of Wall Street that crypto gets institutionalized because the one thing that Bitcoin really has — and we’ve pointed this out to our clients — is ridiculous levels of transaction fees associated with it. So Coinbase, for example, which is one of our shorts, charges customers over 4% per round trip trade to transact in a so-called currency.<br />
<br />
<b>LP: Why would people be willing to pay that?</b><br />
<br />
JC: Well, exactly. They think the asset price is going to go up. It’s like a Nasdaq stock, not a currency. So they’ll pay, and Wall Street wants the fees. The cost structure in crypto is quite high and so the fees are really high. You need retail investors because institutions aren’t going to be paying 4% per round trip to buy and sell Bitcoin. Mom and Pop are, so Wall Street needs to keep the public interested in the crypto space. The paradox, of course, is that if BlackRock and Vanguard and whoever do get ETFs, it’s actually going to force fees down, not up, because ETF fees will be a fraction of the cost of what Coinbase or Binance charge. It’s like mutual fund fees and stock trading fees: they’ve all converged to zero.
</blockquote>
So, even if there <i>is</i> a flood of retail money into Bitcoin ETFs, the exchanges' fee income isn't going to recover. Chanos' students share his <a href="https://www.nakedcapitalism.com/2023/11/jim-chanos-the-crypto-ecosystem-is-well-suited-for-the-dark-side-of-finance.html">skepticism about the industry</a>:<br />
<blockquote>
<b>LP: Are your Wisconsin students still excited about crypto?</b><br />
<br />
JC: They were much more enthusiastic in 2021, 2020, 2019. There’s a lot less enthusiasm now in terms of students going to work for a crypto startup or whatever. I had a number of them in 2020 and 2021 who said they were going to work in the digital currency or blockchain space. I don’t hear that as much anymore.<br />
<br />
<b>LP: How about your Yale students?</b><br />
<br />
JC: The last two springs, they’ve been skeptical. During the spring semester of 2022, Bankman-Fried’s now-infamous interview with Bloomberg dropped about an hour before my class started. I ran to the audiovisual department to see if I could get the interview up on our screen because to me it was such a bombshell. This was where Bankman-Fried basically called the crypto ecosystem a giant Ponzi scheme. I told the class it was very rare to get an industry leader telling you that his industry is a Ponzi scheme. That was seven months before the collapse.
</blockquote>
Will the flood happen?
Teresa Xie's <a href="https://www.bloomberg.com/news/articles/2023-11-04/former-crypto-day-traders-say-no-thanks-even-as-bitcoin-roars-back"><i>Former Crypto Day Traders Say No Thanks Even as Bitcoin Roars Back</i></a> presents some negative evidence:<br />
<blockquote>
Craig Murray, who estimates he made almost $200,000 in the market, says he escaped losing everything to FTX by a hair after friends in the industry heard whispers about its imminent collapse. By that point, the 23-year-old — who lives in New York and recently dropped out of Vanderbilt University — had made up his mind.<br />
<br />
“That kind of put me over the edge,” Murray said. “I just decided it wasn’t worth it. Why would I have my money in this space when there’s a chance that one day it could just all go away?”<br />
<br />
Another sign that retail investing in crypto isn’t returning to previous levels can be seen in weekday versus weekend volumes, with the presumption that the typical person trading on a weekend is a day trader.<br />
<br />
“It’s not unusual nowadays to see weekday trading volume average 50% higher values than weekend trading volume, whereas in the past this ratio was almost 1:1,” said Fredrick Collins, chief executive and founder of crypto data platform Velo Data.
</blockquote>
So it seems unlikely that there will be a huge flood of retail dollars into the market. Even if there were, they would go into Bitcoin ETFs, which would present the exchanges with two big problems:<br />
<ul>
<li>much less retail trading, because it would migrate to ETFs for low fees and the security of dealing with institutions like BlackRock,</li>
<li>and a much higher proportion of very low fee institutional trading.</li>
</ul>
It seems likely that even if the cryptosphere's hopes for ETFs materialize the exchanges will end up in an even worse position than they now are.<br />
<br />David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com18tag:blogger.com,1999:blog-4503292949532760618.post-29008337528700726692023-11-16T08:00:00.119-08:002023-11-16T11:15:06.828-08:00NDSA Sustainability Excellence Award
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjsW7x_4i-TUwFG6xNVH2q_nk5jXabNhRGxUbnEeDV6vKapZoAmu9eDP1JY4Oavw508eDSbFy734Gxb2crvCI-eAQVjL-3LjYp0olWhqIR4OM6AhehgFJTKFxrKEdrLJHUOdsrPCeODqpu4dVDUT2sggb73rBj_EomjFI1YirWH2THzPUgP4_KwTIDX1Q/s151/LOCKSS.logo.png" style="clear: right; float: right; margin-bottom: 1em; margin-left: 1em;"><img border="0" data-original-height="151" data-original-width="151" height="151" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjsW7x_4i-TUwFG6xNVH2q_nk5jXabNhRGxUbnEeDV6vKapZoAmu9eDP1JY4Oavw508eDSbFy734Gxb2crvCI-eAQVjL-3LjYp0olWhqIR4OM6AhehgFJTKFxrKEdrLJHUOdsrPCeODqpu4dVDUT2sggb73rBj_EomjFI1YirWH2THzPUgP4_KwTIDX1Q/s1600/LOCKSS.logo.png" width="151" /></a>Yesterday, at the DigiPres conference, Vicky Reich and I were awarded a "<a href="https://ndsa.org/awards/">Sustainability Excellence Award</a>" by the <a href="https://ndsa.org/">National Digital Stewardship Alliance</a>. This is a tribute to the sustained hard work of the entire LOCKSS team over more than a <a href="https://blog.dshr.org/2023/10/lockss-program-turns-25.html">quarter-century</a>.<br />
<br />
Below the fold are the <a href="https://www.diglib.org/announcing-the-2023-ndsa-excellence-award-winners/">citation</a> and our response.<br />
<span><a name='more'></a></span>
<h3>Citation</h3>
2023 marks a significant date for the LOCKSS Program: It is the twenty-fifth anniversary of the friendly hikes in Joseph Grant State Park and Big Basin where Victoria Reich and Dr. David S.R.[sic] Rosenthal first conceived of LOCKSS, “lots of copies keeps stuff safe,” as a guiding principle for long-term access and preservation to digital library resources. It is also the twentieth anniversary of the transition of the LOCKSS project from beta testing in 2002 to full production release. In the words of the duo’s nominator, “Even if you don’t know David Rosenthal and Vicky Reich by name, you’ve almost certainly heard their rallying cry for digital preservation: Lots Of Copies Keep Stuff Safe. This ethos has informed digital preservation best practices since its introduction, and has shaped the design and implementation of the LOCKSS open-source software.”<br />
<br />
David Rosenthal and Vicky Reich’s brainchild—and the enduring preservation networks that it has made possible—was at the leading edge of a global wave of digital preservation initiatives in the early 20th century. The effectiveness and reliability of LOCKSS software has been validated through rigorous third-party evaluation, including ongoing certification of the CLOCKSS archive as a trustworthy digital repository under the TRAC standard since 2014. The LOCKSS project has provided enduring proof of the concept that large-scale digital preservation work can be accomplished cost-effectively and with community benefits (not vendor profits) as the primary driver.<br />
<br />
Rosenthal and Reich’s work on LOCKSS stands as a benchmark against which other approaches to digital preservation and persistent access to digital resources are measured.
<br />
<h3>Response</h3>
The LOCKSS Program was <a href="https://blog.dshr.org/2023/10/lockss-program-turns-25.html">25 years old</a> last month. Vicky and David would like to thank:<br />
<ul>
<li>The international user community and Stanford staff who are using, evolving and improving the software.</li>
<li>Our far sighted funders who got the program up and running — <a href="https://en.wikipedia.org/wiki/Mike_Lesk">Michael Lesk</a> at the NSF and then <a href="https://blog.dshr.org/2019/08/a-tribute-to-don-waters.html">Don Waters</a> at the Mellon Foundation.</li>
<li>And three brilliant Stanford Computer Science PhD students who provided the <a href="http://dx.doi.org/10.1145/945445.945451">award winning science</a> behind the technology — Mema Roussopoulos, Petros Maniatis and TJ Giuli.</li>
</ul>
We learned many lessons over the two decades we worked on the program, none of them a surprise to anyone in the preservation business. Two that are key. For a program to survive the long term it is critical to:<br />
<ul>
<li>Build and nurture relationships, among the staff, among the user community, and with the funders.</li>
<li>Document, document, document. Write it down, whatever it is. Over a quarter-century it is easy to forget things such as "why was it done this way?".</li>
</ul>
We are honored by and grateful for this award, and happy that the program we left for our new career as full-time grandparents is flourishing in good hands.<br />
<br />David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com3tag:blogger.com,1999:blog-4503292949532760618.post-65353911931718821442023-11-14T08:00:00.004-08:002023-11-14T08:00:00.136-08:00Alameda's On-RampTether has been one of the major mysteries of the cryptosphere for a long time. It has never been audited, and has been described as being "<a href="https://www.bloomberg.com/news/features/2021-10-07/crypto-mystery-where-s-the-69-billion-backing-the-stablecoin-tether">practically quilted out of red flags</a>". Matt Levine <a href="https://www.bloomberg.com/opinion/articles/2023-10-31/bad-passwords-are-securities-fraud">says</a> "I feel like eventually Tether is going to be an incredibly interesting story, but I still don’t know what it is." He was commenting on Emily Nicolle's <a href="https://www.bloomberg.com/news/articles/2023-10-30/bankman-fried-trial-renews-speculation-about-alameda-s-tether-usdt-pile"><i>Bankman-Fried Trial Renews Conjecture About Alameda’s $40 Billion Tether Stablecoin Pile</i></a> by Emily Nicolle. It includes a lot of interesting information, starting with this:<br />
<blockquote>
Alameda was Tether’s largest non-exchange customer between 2020 and 2022, with blockchain data showing it received almost $40 billion in transfers of its stablecoin USDT directly from the company — equal to roughly 20% of all USDT tokens ever issued.<br />
</blockquote>
Below the fold, I discuss the questions Nicolle raises, and go on to ask one she doesn't<br />
<span><a name='more'></a></span>
<br />
The <a href="https://www.bloomberg.com/news/articles/2023-10-30/bankman-fried-trial-renews-speculation-about-alameda-s-tether-usdt-pile">obvious question is</a>:<br />
<blockquote>
The vast amount has raised questions about where Alameda got the money to fund the issuance of the stablecoin, which is typically used as a dollar proxy and is the world’s most traded cryptocurrency. That even led to speculation Alameda might have used other means to help fund the purchases, similar to its bets on startups or dealings with lenders.
</blockquote>
By "other means" Nicolle means using FTX's FTT token. But, if you can believe Tether, that <a href="https://www.bloomberg.com/news/articles/2023-10-30/bankman-fried-trial-renews-speculation-about-alameda-s-tether-usdt-pile">isn't what happened</a>:<br />
<blockquote>
Tether’s incoming chief executive Paolo Ardoino appeared to clear up the mystery in a social media <a href="https://x.com/paoloardoino/status/1715083160109683181?s=20">post</a> this month, reiterating an earlier statement that the hedge fund wired dollars to Tether in exchange for the cryptocurrency.
</blockquote>
An important part of the story is that most people and institutions can't buy USDT from, or sell USDT to, Tether directly. They have to go via one of the small number of institutions that are Tether customers. And those customers can only buy or sell in large amounts. Alameda was one of these customers. Nicolle quotes <a href="https://www.bloomberg.com/news/articles/2023-10-30/bankman-fried-trial-renews-speculation-about-alameda-s-tether-usdt-pile">Jonathan Reiter</a>:<br />
<blockquote>
“This was one major on-ramp, maybe the major on-ramp” for traders seeking access to USDT, said Jonathan Reiter, CEO of blockchain analytics firm ChainArgos. “It is certainly peculiar that FTX, which is now alleged to have run an essentially fraudulent exchange, still somehow managed to intermediate tens of billions in one-way flow from USD into USDT.”
</blockquote>
My oversimplified mental model of stablecoins like USDT is that they are the chips in the casino. If you want to play the tables, you need to hand dollars to the cashier and get chips in return, because the tables only accept chips. The cashiers in the cryptocurrency casino need accounts at banks, so they can receive dollars or other fiat, and accounts at Tether, so they can receive USDT. Since the early days of Tether, it has been increasingly difficult for cryptocurrency traders and exchanges to maintain accounts at banks. For example, in this interview of <a href="https://www.youtube.com/watch?v=pKeMfxLV8K8">Charles Yang by John Riggins</a>, Yang (head trader of Genesis Block based in Hong Kong) says:<br />
<blockquote>
bank acccounts are the absolute most valuable thing — you have to set up a bunch of different companies, a lot of different bank accounts just to facilitate trades that aren't that big, maybe $50K. The moment you tell them this is for a USDT trade, you're basically asking them to shut your bank account down.
</blockquote>
Alameda, run by ex-Wall Street whiz-kid US citizens, had the savvy and credentials to have bank accounts through which <a href="https://www.bloomberg.com/news/articles/2023-10-30/bankman-fried-trial-renews-speculation-about-alameda-s-tether-usdt-pile">billions of dollars could be moved</a>:<br />
<blockquote>
Alameda received around $39.6 billion in USDT tokens from Tether’s treasury wallet, according to on-chain data tracking its wallets, as labeled by the Arkham Intelligence platform.<br />
<br />
Meanwhile wallets connected to FTX sent $3.9 billion in USDT tokens back to Tether, all during the crash of TerraUSD and Luna in May 2022 — meaning that Alameda largely transferred the tokens it received to other entities, rather than cashing it out for dollars with Tether directly. It’s worth noting that this may not record all of FTX or Alameda’s redeemed tokens due to the way Tether swaps its coins between different blockchains, according to Coinbase’s Conor Grogan, who updated Alameda’s USDT total after most of its transactions were <a href="https://protos.com/tether-papers-crypto-stablecoin-usdt-investigation-analysis/">first reported</a> by Protos in 2021.
</blockquote>
Presumably, Alameda received dollars via their accounts at US banks such as the late lamented Signature and Silvergate, aggregated them into orders big enough for Tether, <a href="https://www.bloomberg.com/news/articles/2023-10-30/bankman-fried-trial-renews-speculation-about-alameda-s-tether-usdt-pile">then</a>:<br />
<blockquote>
To mint or redeem USDT, FTX would transfer funds via its own accounts at Tether’s Bahamas-based bank Deltec Bank or a US-based correspondent bank, Bankman-Fried said in a series of social media posts in 2021. A popular solution utilized by Tether and other crypto firms were banking networks established by Signature Bank and Silvergate Bank, both of which collapsed earlier this year.
</blockquote>
The mismatch between the minting and redeeming is why Reiter talks about a "one-way flow from USD into USDT". It seems clear that Alameda was exploiting its access to both the banking system and Tether to facilitate this flow. How big a player were they <a href="https://www.bloomberg.com/news/articles/2023-10-30/bankman-fried-trial-renews-speculation-about-alameda-s-tether-usdt-pile">in this flow?</a>:<br />
<blockquote>
A rough estimate using transactional data for USDT’s main two networks — Ethereum and Tron, which support about 98% of USDT’s current circulation — shows Tether has issued about $190 billion in USDT between its inception and Oct. 20, according to analysis by ChainArgos.<br />
<br />
Conversely, Tether has redeemed about $108 billion in its lifetime. The difference is about $82 billion, compared with around $84 billion in circulation. The discrepancy can be attributed to the amount of USDT trading on other blockchains, according to ChainArgos.
</blockquote>
Alameda bought $40B of USDT out of a total mint of $190B and sold $4B out of total burn of $108B. They were ~21% of the on-ramp but only 4% of the off-ramp.<br />
<br />
Nicolle again <a href="https://www.bloomberg.com/news/articles/2023-10-30/bankman-fried-trial-renews-speculation-about-alameda-s-tether-usdt-pile">quotes Reiter</a>:<br />
<blockquote>
“This does not really look like a market-maker or trading firm as those terms are normally used. It looks more like a money changer,” said Reiter.<br />
<br />
Inflows and outflows from Alameda’s main wallet each month were almost equal, Reiter said, indicating that the main purpose for the mints was most likely to facilitate USDT purchases on behalf of FTX and its customers. This is supported by Bankman-Fried’s statements that FTX would transfer funds via bank accounts to Tether, in addition to evidence submitted to his trial which alleged that FTX often relied on Alameda’s bank accounts for exchange purposes.<br />
<br />
“These guys were running a one-way USD to USDT bureau de change,” said Reiter, adding that any USDT that Alameda actually held on to was never in large size relative to its minting activity.
</blockquote>
This explains how a significant part of the $190B Tether minted was funded. But it begs the question of where the $104B redeemed went that didn't go to Alameda? Who were the Tether customers who made the redemptions and what did they do with the dollars?<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg7HGis5DNHrBvK6W-LNHaIL-bf6CPvPD-b51Phk0rNOxHClJToM33ZPuQ50dilmVu1s_CMalSxxElS1EaksSb769fd9ThOBKF1BPw_pRpbsqxCrATJejhjIgwaxbf8jVrW73Rw2ykO2ZI3zwlkiQMXhsm2y0hx1hyphenhyphenAFeMnNadmLDM3ceLS3cJCxArLkxbM/s824/WhaleActivity.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="447" data-original-width="824" height="109" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg7HGis5DNHrBvK6W-LNHaIL-bf6CPvPD-b51Phk0rNOxHClJToM33ZPuQ50dilmVu1s_CMalSxxElS1EaksSb769fd9ThOBKF1BPw_pRpbsqxCrATJejhjIgwaxbf8jVrW73Rw2ykO2ZI3zwlkiQMXhsm2y0hx1hyphenhyphenAFeMnNadmLDM3ceLS3cJCxArLkxbM/w200-h109/WhaleActivity.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://medium.com/chainargos/a-tale-of-tether-on-tron-before-and-after-the-russian-invasion-of-ukraine-b9d4c63196b4">Source</a></td></tr></tbody></table>
On the Chain Argos blog, Patrick Tan's <a href="https://medium.com/chainargos/a-tale-of-tether-on-tron-before-and-after-the-russian-invasion-of-ukraine-b9d4c63196b4"><i>A Tale of Tether on TRON Before and After the Russian Invasion of Ukraine</i></a> suggests one answer. Tan notes that:<br />
<blockquote>
Although most attention is focused on the use of Tether on the Ethereum blockchain, the reality is the bulk of transactions occur on TRON for a simple reason — low or zero transaction fees.<br />
...<br />
Using data from ChainArgos, 50 so-called “whale wallets” for USDT on the TRON blockchain were analyzed in the weeks leading up to, and following, the Russian invasion of Ukraine.<br />
<br />
From the chart above, it’s clear that the 50 biggest wallets that transact in Tether on TRON saw a larger cluster of activity just before and after the Russian invasion of Ukraine in February 2022.<br />
<br />
None of the wallets (as far as has been identified) are wallets belonging to cryptocurrency exchanges.
</blockquote>
<a href="https://medium.com/chainargos/a-tale-of-tether-on-tron-before-and-after-the-russian-invasion-of-ukraine-b9d4c63196b4">But</a>:<br />
<blockquote>
one of those wallet addresses:<br />
<br />
TCTrwC2rxGFRS74XihgdMkp5mLYGezdrDL<br />
<br />
(the “Alameda Wallet”) also happens to be the wallet used by Alameda Research (yes that one) to deposit funds into cryptocurrency exchange FTX.com.<br />
<br />
Even as the trial of FTX.com founder Sam Bankman-Fried proceeds in earnest, the Alameda Wallet saw a massive $300 mm in Tether flow into it, occurring just days before the Russian invasion.
</blockquote>
<a href="https://medium.com/chainargos/a-tale-of-tether-on-tron-before-and-after-the-russian-invasion-of-ukraine-b9d4c63196b4">And then</a>:<br />
<blockquote>
Of particular note was this wallet address:<br />
<br />
TYZseM53iSPB75aN1V5N762GgzAjBp7PcX<br />
<br />
which sent a whopping $194 mm USDT in just two weeks around the time of the invasion.<br />
<br />
And a quick Google search of the address reveals Telegram (the messaging service) posts by “The Ghost of New Russia” containing this wallet address that appeared to be raising funds for Russia’s invasion.<br />
<br />
On further analysis, this wallet, and others with large USDT transactions in the run-up to the Russian invasion are part of a massive Russian-Kazakh pyramid scheme that had already been blacklisted by the Russian central bank in 2022.<br />
<br />
After the USDT is collected into:<br />
<br />
TYZseM53iSPB75aN1V5N762GgzAjBp7PcX<br />
<br />
it’s then withdrawn into a collection of wallets, but with two in particularly large sizes:<br />
<ol>
<li>TXxU8LJ9wwQxUdg1YNP7LPz4YSXB5guhc5</li>
<li>TCtVdT6XSoWcQifTHbBzFcuaUDX9oo8Yr5</li>
</ol>
Then there’s a few withdrawals for less than $2 mm, almost as if someone is taking their “commission” or “fees” for facilitating the transactions.<br />
<br />
What happens next is relatively predictable, from these withdrawal wallets, the USDT was sent to cryptocurrency exchange Binance, probably for the final cashing out into fiat currency.
</blockquote>
That <a href="https://blog.dshr.org/2023/01/binances-time-in-barrel.html">Binance</a> provided an off-ramp isn't unexpected. What was unexpected was that the spike in off-ramping started two weeks before the invasion. The Russians maintained good enough operational secrecy that Ukraine was suprised and the troops thought they were on exercise. Knowing two weeks ahead that Ukraine would be invaded means the off-rampers were very close to Putin. Tan speculates that the funds were destined for the <a href="https://medium.com/chainargos/a-tale-of-tether-on-tron-before-and-after-the-russian-invasion-of-ukraine-b9d4c63196b4">Wagner Group</a>:<br />
<blockquote>
One possible explanation is Russia’s heavy use of the Wagner Group — mercenaries closely affiliated with the Kremlin, but disavowed when convenient.<br />
<br />
And as is known, mercenaries, whether at the Battle of Waterloo or the Donbas, need to be paid, whether in gold, or a hard currency like the dollar, failing which Tether will do.<br />
<br />
This explanation becomes all the more plausible because the shadowy Wagner Group has been operating in the Donbas since 2014.<br />
<br />
So if Russia wanted to wage an all-out invasion of Ukraine, and the Wagner Group needed to recruit mercenaries in a hurry without an audit trail linked to the Kremlin, what better way than using stablecoins such as Tether?<br />
<br />
And if a Wagner Group mercenary wanted to ensure he’d get paid no matter what, which would he prefer, the ruble against a backdrop of soon-to-be sanctioned Russian banks, or Tether on TRON?<br />
</blockquote>
It is very plausible that the uptick in USDT activity was related to the invasion, but I am more skeptical about the far end of the off-ramp being the Wagner Group. As I understand it, the Russians did not plan an "all-out invasion of Ukraine"; they expected their armored thrusts and airborne landings to swiftly capture Kyiv and its government, which would be replaced by a puppet administration supported by the mass of the Ukrainian population. <a href="https://en.wikipedia.org/wiki/Wagner_Group#2022_invasion_of_Ukraine">Wikipedia suggests</a> that Wagner's initial involvement was limited to two of three attempts to assassinate Zelenskyy.<br />
<br />
Doubtless, <a href="https://en.wikipedia.org/wiki/Yevgeny_Prigozhin">Yevgeny Prigozhin</a> and Wagner's leadership would have known of the invasion plans, probably much more than two weeks ahead, and seen it as prudent to accumulate fiat. But I doubt that this fiat was intended for recruiting additional mercenaries, which wouldn't have been needed had the invasion plans worked out.<br />
<br />
Nevertheless, I agree with Tan when <a href="https://medium.com/chainargos/a-tale-of-tether-on-tron-before-and-after-the-russian-invasion-of-ukraine-b9d4c63196b4">he argues</a>:<br />
<blockquote>
that’s why it’s so important to monitor stablecoin flows — because they can provide an orthogonal data point that most fund and risk managers may miss.<br />
<br />
Whether as an indicator of leverage or risk appetite, or as a precursor to a potentially economically destabilizing conflict, stablecoin flows are becoming an increasingly important metric to monitor.
</blockquote>
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com2tag:blogger.com,1999:blog-4503292949532760618.post-85033971794751257812023-11-09T08:00:00.052-08:002023-11-09T08:00:00.133-08:00Robotaxi Economics<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi5fXjGAnTr3KC6wlwGhQeT8Rpwdm0psh1VdgtSMaowZxE71uhF3oBdpWblfdJ1Y9rjf0VhwbFijLaLL0MpF-NTk3t3qy5wR-UkZ0smP0C-npjjXJqfKsLw_f4acRLFlXog3s_-AxE5v-hFhx1e6nDWymWI3PMnZHBVOnKdyLIQCJdeMLZAO162Sl7pc6CG/s1145/CruiseInSF.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="710" data-original-width="1145" height="124" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi5fXjGAnTr3KC6wlwGhQeT8Rpwdm0psh1VdgtSMaowZxE71uhF3oBdpWblfdJ1Y9rjf0VhwbFijLaLL0MpF-NTk3t3qy5wR-UkZ0smP0C-npjjXJqfKsLw_f4acRLFlXog3s_-AxE5v-hFhx1e6nDWymWI3PMnZHBVOnKdyLIQCJdeMLZAO162Sl7pc6CG/w200-h124/CruiseInSF.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.nytimes.com/2022/10/04/us/san-francisco-driverless-cars.html">Source</a></td></tr></tbody></table>
The <i>New York Times</i> team of Tripp Mickle, Cade Metz and Yiwen Lu have been covering San Francisco's experiment with robotaxis from <a href="https://www.nytimes.com/2023/08/21/technology/waymo-driverless-cars-san-francisco.html">Waymo</a> and <a href="https://www.nytimes.com/2022/09/28/technology/driverless-cars-san-francisco.html">Cruise</a> for some time. Their latest report is <a href="https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-driving-cars.html"><i>G.M.’s Cruise Moved Fast in the Driverless Race. It Got Ugly.</i></a>, and it is a doozy. Leaving aside the serious safety issues that finally forced the DMV to <a href="https://www.nytimes.com/2023/10/24/technology/cruise-driverless-san-francisco-suspended.html">stop Cruise operating</a>, their history of interference with <a href="https://www.cnbc.com/2023/08/18/cruise-self-driving-car-in-san-francisco-fire-truck-crash-one-injured.html">first responders</a>, and the <a href="https://www.kron4.com/news/bay-area/video-cruise-cars-halt-traffic-in-sf-cite-connectivity-issues-linked-to-outside-lands-festival/">traffic congestion</a> they exacerbate, the really devastating part of the article is the economics. Below the fold, I lay out the numbers.<br />
<span><a name='more'></a></span>
<br />
The reason robotaxis look like a good investment is that the up-front investment in equipping cars with high-tech sensors and powerful computers would earn returns in two ways, by removing the need to pay drivers, and by operating 24/7. In the real world, it isn't quite that simple.<br />
<br />
First, the investment is very significant.
The NYT team reports on <a href="https://www.nytimes.com/2023/08/21/technology/waymo-driverless-cars-san-francisco.html">Waymo</a>:<br />
<blockquote>
The <a href="https://www.nytimes.com/2021/05/24/technology/self-driving-cars-wait.html">cars are staggeringly expensive</a>, outfitted with high-tech sensors and cameras, and are worth as much as $200,000.
</blockquote>
And on <a href="https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-driving-cars.html">Cruise</a>:<br />
<blockquote>
Each Chevrolet Bolt that Cruise operates costs $150,000 to $200,000, according to a person familiar with its operations.
</blockquote>
So the upfront investment is around $150K/car. Assuming a 5-year life and 7% interest, the annual cost would be around $32K, or about $90/day. After paying about $12/hr for fuel, insurance, repairs, depreciation and so on an <a href="https://therideshareguy.com/how-much-do-uber-drivers-make/">Uber driver makes $18/hr</a>, or assuming an 8hr day, $144/day. The robotaxi looks like it is about $54/day ahead.<br />
<br />
Proponents will claim that the robotaxi works 24hrs/day, not just 8. But the demand for rides isn't constant through the day. The robotaxi fleet needs enough cars to satisfy peak demand, so most will be idle for most of the time, still consuming interest and depreciation. Uber doesn't pay for the cars, and Uber drivers only work when there is enough demand for rides. I think this is roughly a wash, so the robotaxi stays about $54/day ahead.<br />
<br />
So far I have assumed that the robotaxi fleet needs no humans. But, as the NYT team reports, that <a href="https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-driving-cars.html">isn't the case</a>:<br />
<blockquote>
Half of Cruise’s 400 cars were in San Francisco when the driverless operations were stopped. Those vehicles were supported by a vast operations staff, with 1.5 workers per vehicle.
</blockquote>
With 1.5 workers at Uber's pay rate the robotaxi's staff and equipment is costing $306/day, so the robotaxi is $162/day behind, and Cruise is unlikely to be able to pay only $18/hr. What were the 1.5 people per car <a href="https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-driving-cars.html">actually doing</a>?<br />
<blockquote>
The workers intervened to assist the company’s vehicles every 2.5 to five miles, according to two people familiar with is operations.
</blockquote>
Lora Kolodny followed up the NYT team's reporting in <a href="https://www.cnbc.com/2023/11/06/cruise-confirms-robotaxis-rely-on-human-assistance-every-4-to-5-miles.html"><i>Cruise confirms robotaxis rely on human assistance every four to five miles</i></a>:<br />
<blockquote>
Vogt confirmed that the General Motors-owned company does have a remote assistance team, in response to a discussion under the header, "GM's Cruise alleged to rely on human operators to achieve 'autonomous' driving."<br />
<br />
The CEO wrote, "Cruise AVs are being remotely assisted (RA) 2-4% of the time on average, in complex urban environments. This is low enough already that there isn't a huge cost benefit to optimizing much further, especially given how useful it is to have humans review things in certain situations."
</blockquote>
Clearly, the robotaxi fleet needs some staff. The cars need refueling, say every 200 miles. They need cleaning. Lets guess the average ride is 5 miles and they need cleaning every 10 rides. That would be staff required every 40 miles. So the vast majority of interventions are because Cruise's self-driving technology just can't handle San Francisco traffic. They may be called robotaxis, just pay no attention to the driver behind the curtain. These remote drivers will be <i>more expensive</i> than the in-car drivers they are replacing!<br />
<br />
Overall, their robotaxi operation is costing GM a <a href="https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-driving-cars.html">boatload of money</a>:<br />
<blockquote>
G.M. has spent an average of $588 million a quarter on Cruise over the past year, a 42 percent increase from a year ago.
</blockquote>
$2,232B/year for 400 cars is $5.6M per car per year! How could each car <a href="https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-driving-cars.html">earn this much</a>?<br />
<blockquote>
The company charged an average of $10.50 per ride in the city.
</blockquote>
To generate $5.6M each car would have to service 510K rides/year or 1,400 rides/day, or one ride per minute. Cruise's goal, before they got shut down, <a href="https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-driving-cars.html">was less ambitious</a>:<br />
<blockquote>
The shutdown complicates Cruise’s ambition of hitting its goal of $1 billion of revenue in 2025.
</blockquote>
Covering less than half their costs in 2025 is about 91M rides, or about 250K/day. Lets be generous, and assume 10 rides/day/car. They would need 25K Chevrolet Bolts. Scaling the operation from 400 to 25K in under 2 years would be hard. It would be made harder because <a href="https://gmauthority.com/blog/2023/11/gm-confirms-2023-chevy-bolt-ev-bolt-euv-production-will-end-in-december/">Bolt production stops next month</a>. But even if the Bolts were available, this would be an additional investment of between $3.7B and $5B in order to lose only $1.2B/year.<br />
<br />
I'm not the only one pointing out that the numbers don't add up. Here, for example, is <a href="https://www.eschatonblog.com/2023/11/imagine-theres-no-people.html">Atrios</a>:<br />
<blockquote>
Focusing on the notoriously high margin taxi business has always been odd, in that even if they "work" it's not entirely clear where the big profit opportunity is! Replacing low wage workers with expensive machines that require a lot of attention from people who have more specialized skills than "driving" doesn't seem to be such a great plan!<br />
<br />
I get the appeal for the carbrained. What "we" all want is a 24 hour driver at an affordable rate. It's just not quite clear what the business model is, even if the technology works!<br />
<br />
People always point to niche cases in which robot car would be particularly useful. But they're niche cases and you need scale to make anything in this industry profitable! Scale is necessary for a lot of reasons, but one is maintenance and spare parts production. Your self-driving campus vans won't last long if there's only one guy who knows to maintain them and the parts manufacturer just went bankrupt.
</blockquote>
Last August, Hubert Horan continued his long-running series with <a href="https://www.nakedcapitalism.com/2023/08/hubert-horan-can-uber-ever-deliver-part-thirty-three-uber-isnt-really-profitable-yet-but-is-getting-closer-the-antitrust-case-against-uber.html"><i>Part Thirty-Three: Uber Isn’t Really Profitable Yet But is Getting Closer; The Antitrust Case Against Uber</i></a>, which starts:<br />
<blockquote>
<b>After $33 Billion in Losses Over 14 Years, Uber is Finally Approaching GAAP Breakeven</b><br />
<br />
Uber claimed its first ever quarterly GAAP profit when it released its second quarter and first half financial results on August 1. The claim was a bit of stretch as the reported $394 million second quarter profit ($237 million for the first half) was entirely explained by an alleged $386 million second quarter gain ($707 million in the first half) in the value of untradable securities they hold in companies like Didi, Grab, and Aurora that have nothing to do with their ongoing operations. Readers of this series will know that Uber has aggressively used claims like this to justify misleading claims about its corporate financial performance ever since 2018 when it inflated published net income numbers by $5.8 billion just prior to its IPO.
</blockquote>
Uber has always viewed drivers as a temporary necessity to tide them over until robotaxis take over and boost them into actual instead of manipulated profits. Last month they started offering <a href="https://www.cnbc.com/2023/10/26/uber-begins-offering-rides-in-self-driving-waymo-cars.html">robotaxi service in Phoenix</a> using Waymo's cars. But unless Waymo is showing them radically different numbers than the NYT team found for Cruise, Horan's series is going to run and run.<br />
<br />
These numbers look even worse for Tesla. Last year Matthew Loh reported that <a href="https://www.businessinsider.com/elon-musk-tesla-worth-basically-zero-without-self-driving-2022-6"><i>Elon Musk says the difference between Tesla being 'worth a lot of money or worth basically zero' all comes down to solving self-driving technology</i></a>, and the reason was that owners would rent out their Teslas as robotaxis when they weren't using them. This was always obviously a stupid idea; who wants drunkards home-bound from the pub throwing up on their Tesla's seats? But the fact that the numbers don't add up for robotaxis in general, and the fact that <a href="https://www.theverge.com/2023/10/27/23934691/hertz-tesla-uber-ev-plans-damage-repair-price-cuts"><i>Hertz is scaling back its EV ambitions because its Teslas keep getting damaged</i></a> because half of them are being used by Uber drivers as taxis, make the idea even more laughable.<br />
<br />
It is really surprising that the DMV took so long to shut Cruise down. In August <a href="https://www.nytimes.com/2023/08/21/technology/waymo-driverless-cars-san-francisco.html">the NYT team wrote</a>:<br />
<blockquote>
Last week, a Cruise driverless car collided with a fire truck responding to an emergency. <a href="https://www.nytimes.com/2023/08/17/us/driverless-car-accident-sf.html">Another Cruise vehicle</a> got stuck in wet concrete. The week before, <a href="https://www.kron4.com/news/bay-area/video-cruise-cars-halt-traffic-in-sf-cite-connectivity-issues-linked-to-outside-lands-festival/">several Cruise cars blocked traffic</a> in the city’s North Beach neighborhood.
</blockquote>
More recently, <a href="https://theintercept.com/2023/11/06/cruise-self-driving-cars-children/">Sam Biddle reported that</a>:<br />
<blockquote>
previously unreported internal materials such as chat logs show Cruise has known internally about two pressing safety issues: Driverless Cruise cars struggled to detect large holes in the road and have so much trouble recognizing children in certain scenarios that they risked hitting them. Yet, until it came under fire this month, Cruise kept its fleet of driverless taxis active, maintaining its <a href="https://getcruise.com/news/blog/2023/human-ridehail-crash-rate-benchmark/">regular reassurances of superhuman safety</a>.
</blockquote>
In contrast, all the NYT team could find to report <a href="https://www.nytimes.com/2023/08/21/technology/waymo-driverless-cars-san-francisco.html">about Waymo was</a>:<br />
<blockquote>
In May, one of its <a href="https://www.dmv.ca.gov/portal/file/waymo_052123-pdf/">cars struck and killed a small dog</a>. A few years ago, <a href="https://techcrunch.com/2021/12/16/waymo-says-self-driving-car-that-struck-pedestrian-in-san-francisco-was-in-manual-mode/">a driverless Waymo car with a human safety driver operating the wheel</a> hit a pedestrian who needed to be taken to the hospital. The company has been collecting fares in the Phoenix area for several years and now has a fleet navigating some 200 miles across that region, including to and from the airport.
</blockquote>
Why is Cruise so much less safe? The NYT team <a href="https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-driving-cars.html">write</a>:<br />
<blockquote>
Company insiders are putting the blame for what went wrong on a tech industry culture — led by the 38-year-old Mr. Vogt — that put a priority on the speed of the program over safety. In the competition between Cruise and its top driverless car rival, Waymo, Mr. Vogt wanted to dominate in the same way Uber dominated its smaller ride-hailing competitor, Lyft.<br />
<br />
“Kyle is a guy who is willing to take risks, and he is willing to move quickly. He is very Silicon Valley,” said Matthew Wansley, a professor at the Cardozo School of Law in New York who specializes in emerging automotive technologies. “That both explains the success of Cruise and its mistakes.”
</blockquote>
As someone who has been sharing the roads with Waymo cars for many years, I have always believed that Waymo, unlike Cruise and much worse <a href="https://blog.dshr.org/2023/10/elon-musk-threat-or-menace-part-3.html">Tesla</a>, understood that it was necessary to <i>prove</i> that self-driving technology was safe, not just gaslight the public as <a href="https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-driving-cars.html">Mr. Vogt did</a>:<br />
<blockquote>
To make streets safer, he said in an interview, cities should embrace self-driving cars like those designed by Cruise, a subsidiary of General Motors. They do not get distracted, drowsy or drunk, he said, and being programmed to put safety first meant they could substantially reduce car-related fatalities.
</blockquote>
Two months later Cruise's safety record got them shut down. <a href="https://theintercept.com/2023/11/06/cruise-self-driving-cars-children/">The very same day</a>:<br />
<blockquote>
CEO Mary Barra told financial analysts that Cruise “is safer than a human driver and is constantly improving and getting better.”
</blockquote>
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com9tag:blogger.com,1999:blog-4503292949532760618.post-81908237015009497802023-11-07T08:00:00.005-08:002023-11-07T08:00:00.135-08:00My Old Car<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiT_y21n_Wr_mKZzMU380nmMxofC_AyifmjZJcQQQdc3KN1gAWwZbyWb36Xj7g3EDq1wnfJYVdMyGLtZ-ATyXNS2rWcas3tESBb-BGjvAeS137QNGg5sQASLWDB7-W3BlD33KWYpmwWY9Tw41Mjs0L152HUTnMnDcE1mq78w75Auf9vhoKJpkODZZGRsJiJ/s3704/RX-7-1.jpeg" style="clear: right; float: right; margin-bottom: 1em; margin-left: 1em;"><img border="0" data-original-height="2148" data-original-width="3704" height="116" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiT_y21n_Wr_mKZzMU380nmMxofC_AyifmjZJcQQQdc3KN1gAWwZbyWb36Xj7g3EDq1wnfJYVdMyGLtZ-ATyXNS2rWcas3tESBb-BGjvAeS137QNGg5sQASLWDB7-W3BlD33KWYpmwWY9Tw41Mjs0L152HUTnMnDcE1mq78w75Auf9vhoKJpkODZZGRsJiJ/w200-h116/RX-7-1.jpeg" width="200" /></a>
This post celebrates my weird old car's 30<sup>th</sup>birthday. It is a <a href="https://en.wikipedia.org/wiki/Mazda_RX-7#Third_generation_(FD3S)">Mazda RX-7</a> dated November 1993, carrying the California license RX7 DSHR. I bought it new and in the almost 30 years since have driven it for nearly 140K miles. Unusually for an RX-7 this old, it is almost completely stock and has never been on a track.<br />
<br />
Below the fold, I recount an RX-7 saga spanning thirty-eight years and well over a quarter-million miles.<br />
<span><a name='more'></a></span>
<br />
I moved to California to work at Sun Microsystems in September 1985, so I needed a car. For the first few weeks, Andy Bechtolsheim lent me his <a href="https://en.wikipedia.org/wiki/Porsche_944">Porsche 944</a>, which was really fun to drive, but too expensive. On an earlier visit to California a friend had driven me on part of Highway One in her first-generation <a href="https://en.wikipedia.org/wiki/Mazda_RX-7#First_generation_(SA22C,_FB)">(FB) RX-7</a>. I'd been impressed by the handling and the smooth power delivery of the rotary engine, so I test drove a second-generation <a href="https://en.wikipedia.org/wiki/Mazda_RX-7#Second_generation_(FC3S_and_FC3C)">(FC) RX-7</a>. It was immediately apparent that this was, in effect, a Porsche 944 at half the price, so I bought one in white with the "Sport" package (rear spoiler, stiffer suspension).<br />
<br />
The car had more performance than I really needed and excellent handling although the stiff suspension meant you really felt the "imperfections" in the road surface. Driving to and from Lake Tahoe in ski season with chains on was a bit of a trial. The FC served me well as my daily driver for nearly a decade and around 125,000 miles. These included several long road trips up the West Coast. The longest was from the Bay Area to Port Angeles, WA to collect my mother off the ferry from Victoria, BC and drive her back along the coast.<br />
<br />
Eventually the FC got expensive to maintain and I needed a new car. By then Mazda had introduced the third-generation <a href="https://en.wikipedia.org/wiki/Mazda_RX-7#Third_generation_(FD3S)">(FD) RX-7</a>. I test-drove one and it was immediately apparent that it was a huge step up from the FC; it was a Porsche 911 for half the price. Of course, half the price of a 911 was still a lot more than half the price of a 944, so I used a broker to find a base model in white. It was hard to find; I subsequently discovered from Shawn Fanning that it was one of only 13 1994 base models imported. The FD was my daily driver until grandkids arrived; you can't put a child seat in it and even now the eldest is nearly my size their parents aren't OK with them riding in it.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiAnZ0V3KFPX8Oh65A29SXfIi3wxSalFBqllpvoCkeXodW1Y8MFIUgh6pIHdKLMXTkFKFEt2Gkx0WTlvqF-WLaGUHXefMdciz1g1wAQD4_kGAb5KrcQQ0EhJHlnHapHeSO1H5mh3AnHOuqpd8ivGyX80jXmVoUEM4pPDRmHETJfrq52QsqmiME7gHpGZvBU/s4032/RX-7-gauges.jpeg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="3024" data-original-width="4032" height="150" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiAnZ0V3KFPX8Oh65A29SXfIi3wxSalFBqllpvoCkeXodW1Y8MFIUgh6pIHdKLMXTkFKFEt2Gkx0WTlvqF-WLaGUHXefMdciz1g1wAQD4_kGAb5KrcQQ0EhJHlnHapHeSO1H5mh3AnHOuqpd8ivGyX80jXmVoUEM4pPDRmHETJfrq52QsqmiME7gHpGZvBU/w200-h150/RX-7-gauges.jpeg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;">Old school<br /></td></tr></tbody></table>
Some of the things that make the FD a wonderful driving experience are:<br />
<ul>
<li>The rotary engine, which is turbine-smooth as it races up to the 8,000 RPM red-line.</li>
<li>The light weight and stiff chassis.</li>
<li>The handling, thanks to its 49/51 weight distribution, wide track, short wheelbase, firm suspension, low polar moment of inertia, and positioning of the center of rotation right where the driver sits.</li>
<li>The gear-change which, while not quite as good as that on the MX-5, is crisp, short and in the right place.</li>
<li>The driver-focused interior with seats that really hug you, a simple set of classic analog dials, and no distractions — not even cup-holders!</li>
</ul>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhpliLN4kftQSrraPIUq6qLrCjlpmE62cGdFNs3FGxAcrC-bz0-hUWO1uqD74phFyZ_57tNruJx7R1It3KGSG-k_E5QmrNLe13cA7qyc34m_XrMbIP6gRxWTmeWX1hmDT99rW-oXnQ5kbY-lEPWhC5MT-K8onzZqsy7kLcUt2n-WhrMwlQ7IOLL3DXt15m0/s1238/GranTurismoSport.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="695" data-original-width="1238" height="113" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhpliLN4kftQSrraPIUq6qLrCjlpmE62cGdFNs3FGxAcrC-bz0-hUWO1uqD74phFyZ_57tNruJx7R1It3KGSG-k_E5QmrNLe13cA7qyc34m_XrMbIP6gRxWTmeWX1hmDT99rW-oXnQ5kbY-lEPWhC5MT-K8onzZqsy7kLcUt2n-WhrMwlQ7IOLL3DXt15m0/w200-h113/GranTurismoSport.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.youtube.com/watch?v=iX-NgifJO28">Source</a></td></tr></tbody></table>
These days I fairly often get appreciative waves and comments from other drivers. But my favorite was when, driving home from Mendocino, Vicky & I stopped for a short hike at <a href="https://www.parks.ca.gov/?page_id=438">Hendy Woods State Park</a>. A 7-year old boy and his 5-year old brother watched us park, and the elder excitedly announced "that's a really cool car!". I said "I agree, but how did you know?" He replied "its in my video game." My guess is that it was <i>Gran Turismo</i>, but it could have been any of <a href="https://www.youtube.com/watch?v=iX-NgifJO28">27 games</a>.<br />
<br />
You can get an impression of the FD from the many reviews on YouTube, for example:<br />
<ul>
<li><a href="https://www.youtube.com/watch?v=quLmDySyO7k">Hagerty</a></li>
<li><a href="https://www.youtube.com/watch?v=lZgOhP_RMSM">Doug DeMuro</a></li>
<li><a href="https://www.youtube.com/watch?v=uLaPPrXzpO0">The Straight Pipes</a></li>
<li><a href="https://www.youtube.com/watch?v=VRzgp_XFkDs">Everyday Driver</a> (300zx vs. Supra vs. RX-7)</li>
</ul>
The three things every RX-7 review agrees on are that they are unreliable, that you have to watch out for the huge surge of power when the second turbo kicks in, and that the FD is among the best-handling sports cars ever.<br />
<br />
The late, great <a href="https://en.wikipedia.org/wiki/Colin_Chapman">Colin Chapman CBE</a> famously advised designers to "<a href="https://www.classicdriver.com/en/article/genius-colin-chapman-simplify-then-add-lightness%E2%80%9D">simplify, then add lightness</a>". Some things about the RX-7 are far from simple, such as the dual sequential turbochargers, but Mazda definitely added lightness. My RX-7 weighs 2789lb (GVWR 3210lb); it was around 1000lb lighter than the competition at the time, such as the Toyota Supra. For comparison, the recent lightened, super-sporty Porsche Carrera T <a href="https://arstechnica.com/cars/2022/11/porsche-911-carrera-t-first-drive-simplify-then-add-the-right-options/">weighs 3254lb</a>, 17% more. Of course, the Porsche has a 3.0 liter twin-turbo flat-6 giving it 357HP for 220HP/ton whereas three decades earlier the RX-7 only had a 1.3 liter twin-turbo dual-rotor giving 255HP for 183HP/ton.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEia8dFl0oFvollXMPBCYQBhTFsdZEDJ_zwxb4kvS5DP4I9B4DkASptD1-8QDjngZQ_MUV1LMPGFUky0oXphX4UpSO62hYF0EfpfL3079ozLnJYRy3sSm-VF5Vl4HvcXRCpVZgfkaq5pka9xGpaP06exC8Hk1e4a_az1CdN7peuxMyf9HInHcJNXmLbRLe7u/s4032/RX-7-jack.jpeg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="3024" data-original-width="4032" height="150" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEia8dFl0oFvollXMPBCYQBhTFsdZEDJ_zwxb4kvS5DP4I9B4DkASptD1-8QDjngZQ_MUV1LMPGFUky0oXphX4UpSO62hYF0EfpfL3079ozLnJYRy3sSm-VF5Vl4HvcXRCpVZgfkaq5pka9xGpaP06exC8Hk1e4a_az1CdN7peuxMyf9HInHcJNXmLbRLe7u/w200-h150/RX-7-jack.jpeg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;">3lb jack<br /></td></tr></tbody></table>
<a href="https://worldcat.org/title/34998165"><i>The Mazda RX-7: Mazda's Legendary Sports Car</i></a>, Jack Yamaguchi and John Dinkel's incredibly detailed book, recounts how far they went. For example, this table compares some components of the FC (not a heavy car) with the FD:<br />
<table align="center" border="3" cellpadding="5" cellspacing="0" cols="4" rules="all">
<thead>
<tr><th align="center" colspan="4">Examples of Weight Reduction (lbs)</th></tr>
<tr><th align="left"> </th><th align="center">1992</th><th align="center">1993</th><th align="center">Reduction</th></tr>
</thead><tbody>
<tr><td align="left">Radiator</td><td align="center">9.5</td><td align="center">6.8</td><td align="center">2.7</td></tr>
<tr><td align="left">Intercooler</td><td align="center">6.0</td><td align="center">5.1</td><td align="center">0.9</td></tr>
<tr><td align="left">Oil Cooler</td><td align="center">6.4</td><td align="center">3.3</td><td align="center">3.1</td></tr>
<tr><td align="left">Ignition Coil</td><td align="center">7.3</td><td align="center">3.3</td><td align="center">4.0</td></tr>
<tr><td align="left">Door</td><td align="center">74.8</td><td align="center">61.6</td><td align="center">13.2</td></tr>
<tr><td align="left">Exhaust System</td><td align="center">106.7</td><td align="center">89.1</td><td align="center">17.6</td></tr>
<tr><td align="left">Wheel</td><td align="center">24.2</td><td align="center">16.3</td><td align="center">7.9</td></tr>
</tbody></table>
My favorite weight-saving example is that the jack is made out of aluminum! It weighs only 3lb. As <a href="https://www.youtube.com/watch?v=pdKLZtS2QVY">Jason Cammisa explains</a>, an obession with light weight was key to the design of the MX-5 Miata, by far the most successful sports car in history. In three and a half decades the MX-5 has gained only about 100lbs.<br />
<br />
As regards unreliability, well yes, depending on how you treat an RX-7. All that weight-saving and power increase doesn't come for free, and things under the hood get really hot. The real reason the RX-7 is still on the road after nearly 140K miles and running well is the wonderful Ivan Koinok, who has taken care of it almost the whole time. The major items he has replaced include:
<ul>
<li>Factory re-manufactured engine (water seals failed at 110K miles)</li>
<li>OEM stock turbo (oil leak)</li>
<li>OEM stock flywheel and clutch (I cooked it)</li>
<li>Competition radiator</li>
<li>OEM front struts</li>
<li>secondary air injection pump</li>
<li>central processing unit</li>
</ul>
Except in stop-and-go traffic, driving the FD is a joy. Fortunately, California is well-equipped with RX-7-friendly roads, including CA1 South from Carmel through Big Sur, and North from Mill Valley to Fort Bragg, CA20 from Fort Bragg to Willits, and CA128 from Cloverdale to the coast.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjTINjbQWoGXcNAJICLUkF1MaYq-JseOIiaTdG6xD_O4oytNlrWm8RbgZGExhqreBVWKpHNSN3s52X8l1dbgFsvDFw1w18hzpUiQPCUeo_nHCTATeN_B90A5AXwUlcMDGKVQsqLD06r7VXbHAN6SSJZeL4ZlApeZESmC2awySAK4_qTGyjXGIJEFQ52lvTA/s4032/BMW-growth.jpeg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="3024" data-original-width="4032" height="150" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjTINjbQWoGXcNAJICLUkF1MaYq-JseOIiaTdG6xD_O4oytNlrWm8RbgZGExhqreBVWKpHNSN3s52X8l1dbgFsvDFw1w18hzpUiQPCUeo_nHCTATeN_B90A5AXwUlcMDGKVQsqLD06r7VXbHAN6SSJZeL4ZlApeZESmC2awySAK4_qTGyjXGIJEFQ52lvTA/w200-h150/BMW-growth.jpeg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;">Expansion of the universe<br /></td></tr></tbody></table>
PS: Our other car is parent-approved for grandkid transportation. It is a base model 2012 <a href="https://en.wikipedia.org/wiki/BMW_1_Series_(E87)">BMW 128i coupe (E82)</a> in white with no options. It was also a really difficult car to find; it took the dealership many weeks to find one on a ship heading for Eureka via Long Beach.<br />
<br />
It is also a very enjoyable car to drive, being the last year the 128i coupe had the <a href="https://en.wikipedia.org/wiki/BMW_N52">3.0 liter naturally aspirated straight six</a>. This generation of the 1-series was the basis for the legendary skunk-works project that developed the <a href="https://www.caranddriver.com/reviews/a15125348/2011-bmw-1-series-m-coupe-test-review/">1 series M coupe</a> (<a href="https://www.youtube.com/watch?v=-xkv4PiD0To">Chris Harris loves his</a>). Our old 128i is the way non-M BMWs used to be before they got fat and soft and turned into SUVs — a relatively (but compare with the 1970 BMW 2002) small 2-door with a big straight six, rear-wheel drive, good handling, and none of the modern driver distraction technology.<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com2tag:blogger.com,1999:blog-4503292949532760618.post-7773430947398147512023-11-02T08:00:00.002-07:002023-11-02T08:00:00.131-07:00Limited LiabilityRegulation works by assigning liability for actions to specific actors. The whole idea of decentralization is that by diffusing responsibility among a large number of participants the system could evade regulation. Each participant would bear such a small part of the responsibility for the system's actions that enforcing liability for the actions to each of them would be infeasible. The problem with this is that, for <a href="https://blog.dshr.org/2014/10/economies-of-scale-in-peer-to-peer.html">economic reasons</a>, the <a href="https://blog.dshr.org/2018/10/gini-coefficients-of-cryptocurrencies.html">Gini coefficients of cryptocurrencies</a> are extremely large. It is true that <i>most</i> of the participants have only a small part of the responsibility, but there are always <i>some</i> who have a large share and are thus worth enforcing liability against.<br />
<br />
In <a href="https://blog.dshr.org/2022/10/piercing-veil.html"><i>Piercing The Veil</i></a> I discussed a <a href="https://www.cftc.gov/PressRoom/PressReleases/8590-22">Commodity Futures Trading Commission enforcement action</a> in which the bZeroX exchange converted itself into a DAO:<br />
<blockquote>
By transferring control to a DAO, bZeroX’s founders touted to bZeroX community members the operations would be enforcement-proof—allowing the Ooki DAO to violate the CEA and CFTC regulations with impunity, as alleged in the federal court action. The order finds the DAO was an unincorporated association of which Bean and Kistner were actively participating members and liable for the Ooki DAO’s violations of the CEA and CFTC regulations.
</blockquote>
Below the fold I discuss a post in which Matt Levine reports on a more <a href="https://www.bloomberg.com/opinion/articles/2023-10-24/big-bonuses-are-back-at-uk-banks">recent lawsuit making the same argument</a>.<br />
<span><a name='more'></a></span>
<br />
In <a href="https://blog.dshr.org/2022/02/list-and-dump-schemes.html"><i>List And Dump Schemes</i></a>, based on Fais Khan's <a href="https://startupsandecon.substack.com/p/you-dont-own-web3-a-coinbase-curse"><i>"You Don't Own Web3": A Coinbase Curse and How VCs Sell Crypto to Retail</i></a>, I described how A16Z and other VCs exploited crypto tokens. <a href="https://www.bloomberg.com/opinion/articles/2023-10-24/big-bonuses-are-back-at-uk-banks">Levine recaps this concisely</a>:<br />
<blockquote>
In the crypto boom, VC firms discovered a different way to invest in startups. The startup would be a crypto project, and it would issue tokens, and the VC firm would buy the tokens. The tokens might have some governance rights: The crypto project might be a DAO, a decentralized autonomous organization, whose token holders get to vote on what it should do.<br />
<br />
The advantages of this structure, for VCs, were:<br />
<ol>
<li>Crypto tokens, DAOs, web3, etc. got a lot of hype, so in 2021 you could buy these tokens and resell them for more of a profit than you might get by investing in normal stock in normal startups.</li>
<li>You could sell <i>faster</i>: Crypto tokens lived, or seemed to live, in a weird regulatory gray area, so you could buy the crypto project’s tokens, say to yourself “ahhh these tokens are not <i>securities</i>, <a href="https://www.bloomberg.com/opinion/articles/2023-07-17/branches-make-bank-runs-harder">they are</a> ‘little more than alphanumeric cryptographic sequences,’” and sell them to retail investors on some crypto exchange. Instead of buying stock and waiting for years for a startup to be successful enough to do an initial public offering, you could buy tokens and flip them quickly while the hype was strong.</li>
</ol>
</blockquote>
Then Levine turns to the <a href="https://www.bloomberg.com/opinion/articles/2023-10-24/big-bonuses-are-back-at-uk-banks">liability issue</a>:<br />
<blockquote>
But there is another problem with the token approach, which is that, while these tokens might be functionally stock, and while the SEC thinks ... that they are therefore securities, they are not literally stock, and these crypto DAOs are not literally corporations. Specifically, they are not <i>incorporated</i>. Nobody filed a form with the Delaware secretary of state to incorporate the crypto DAO as a corporation.<br />
<br />
This seems like a small technicality, but one important benefit that you get, from filling out a form and paying a fee to incorporate a corporation, is <i>limited liability</i>. If you have a business entity that is not incorporated, then it is often, by default, a general partnership. Which means that the investors who fund and participate in running it are, arguably, general partners. Shareholders of a corporation are not generally liable for the corporation’s actions, but general partners of a partnership generally <i>are</i>. For instance, they might be liable for its unregistered sales of securities.
</blockquote>
The recent lawsuit is about the <a href="https://www.bloomberg.com/opinion/articles/2023-10-24/big-bonuses-are-back-at-uk-banks">governance token for Compound</a>:<br />
<blockquote>
Late last year, some holders of Compound DAO’s COMP tokens <a href="https://assets.bwbx.io/documents/users/iqjWHBFdfxIU/r1.jaahEw6Y4/v0">sued the venture capitalists backing Compound</a> for unregistered securities sales. The idea is:<br />
<ol>
<li>“Compound is a business that allows users to borrow and lend crypto assets, in much the same way that a traditional bank allows customers to borrow and lend traditional currencies.” ...</li>
<li>Compound is governed by a DAO, and “Compound DAO is governed by the holders of a security called COMP.”</li>
<li>Compound issued some COMP tokens to its early investors and founders. It issues others to people who use the Compound smart-contract protocol: If you deposit or borrow on Compound, you can earn COMP tokens ....</li>
<li>COMP trades publicly on crypto exchanges: If you use Compound and get some tokens, you can sell them for cash, and people do.</li>
<li>“Nine people control at least 51.56% of the COMP currently issued,” including the co-founders of Compound and venture capital firms including Bain Capital Ventures, Polychain Alchemy, Andreessen Horowitz and Paradigm.</li>
<li>Those people and firms are <i>general partners</i> of Compound, liable for any bad stuff that it does.</li>
<li>One arguably bad thing that Compound does is list the COMP token on crypto exchanges, where people can buy it.</li>
<li>The plaintiffs in this case bought some COMP on the exchanges — <i>not</i> from Compound or its venture investors — and are now suing, arguing that this was an unregistered sale of securities and that the <i>venture capitalists are responsible for it</i>.</li>
</ol>
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhYK-s9Bc5mzjFszGTgNufwjhD3Y0NFpyKM3n2JXDb8Iy1aFnKl5SKANiAwlueWUSTxiJUzl7JpD8wATeumtqKjEB3hL-g86wn8nvlc-uwunh1sN1LTjE9NhRNX7LSj0ZMsXTkKogwCKaQiPDH0ARabKpWqbL6xIbXpLrx2Pws0Fv50YEYbq4E4R8N8bueR/s975/COMP_All_graph_coinmarketcap.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="396" data-original-width="975" height="81" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhYK-s9Bc5mzjFszGTgNufwjhD3Y0NFpyKM3n2JXDb8Iy1aFnKl5SKANiAwlueWUSTxiJUzl7JpD8wATeumtqKjEB3hL-g86wn8nvlc-uwunh1sN1LTjE9NhRNX7LSj0ZMsXTkKogwCKaQiPDH0ARabKpWqbL6xIbXpLrx2Pws0Fv50YEYbq4E4R8N8bueR/w200-h81/COMP_All_graph_coinmarketcap.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://coinmarketcap.com/currencies/compound/">Source</a></td></tr></tbody></table>
The reason the plaintiffs are unhappy with the VCs is obvious from the "price" chart. COMP peaked on 3<sup>rd</sup> May 2021 at $831; it currently trades around $45, down 94% from the peak. It wasn't a typical <a href="https://blog.dshr.org/2022/02/list-and-dump-schemes.html">list-and-dump scheme</a>, because it actually went up from its launch for about a year before peaking.<br />
<br />
So far, things <a href="https://www.bloomberg.com/opinion/articles/2023-10-24/big-bonuses-are-back-at-uk-banks">don't look good for the VCs</a>:<br />
<blockquote>
The VCs moved to dismiss the complaint, and last month a federal judge <a href="https://assets.bwbx.io/documents/users/iqjWHBFdfxIU/rUwpDCc4_Ni4/v0">denied their request</a>, finding “that there are sufficient allegations against each of the Partner Defendants to allow the Securities Act claim to go forward at this juncture.”
</blockquote>
Levine <a href="https://www.bloomberg.com/opinion/articles/2023-10-24/big-bonuses-are-back-at-uk-banks">eloquently summarizes the problem</a>:<br />
<blockquote>
One thing that crypto enthusiasts and venture capitalists liked to say was that crypto was a new way of organizing human economic behavior, that crypto would enable a new “Web3” in which technology was organized collectively and belonged to its users instead of being owned by big corporations. This could seem pretty cynical, as VC firms often owned big stakes in (the tokens of) these web3 projects.<br />
<br />
But also it was just a big factual mistake! Crypto and DAOs and web3 were not a new way of organizing human economic behavior; <a href="https://www.bloomberg.com/opinion/articles/2016-05-17/blockchain-company-wants-to-reinvent-companies">they were an old way</a>, the general partnership. They were a technological step <i>backward</i>: Ages ago, lawyers and financiers and governments figured out a new way to organize human behavior, the corporation, which allowed people to pool their capital in a new venture in ways that limited their risk and thus encouraged more and safer capital formation. And then crypto came along and people <i>forgot</i>.
</blockquote>
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com0tag:blogger.com,1999:blog-4503292949532760618.post-322922826181666092023-10-31T08:00:00.201-07:002023-10-31T08:00:00.157-07:00Shitcoins<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgxNdf6HWfgyFgRJKIgFmK4_pdS8Gw6SQW2OOrXyBw0OiJ2Zyj1QHX5ocDC67b52I-PF8m3sbb8zeByxY-43tUx3DLYLIGXzWpyFVdepJs0tZN0W0EcyP4KIriuZxjTlWZHEsN-83Wk2dvozfoEyxRjypNO29T_5zv-rf4jH_ejT_8Wrgp3xreWRltdrCgI/s958/CoinbaseTradingVolume.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="538" data-original-width="958" height="113" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgxNdf6HWfgyFgRJKIgFmK4_pdS8Gw6SQW2OOrXyBw0OiJ2Zyj1QHX5ocDC67b52I-PF8m3sbb8zeByxY-43tUx3DLYLIGXzWpyFVdepJs0tZN0W0EcyP4KIriuZxjTlWZHEsN-83Wk2dvozfoEyxRjypNO29T_5zv-rf4jH_ejT_8Wrgp3xreWRltdrCgI/w200-h113/CoinbaseTradingVolume.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2023-10-11/coinbase-coin-trading-volume-is-likely-lowest-since-before-public-debut">Source</a></td></tr></tbody></table>What caught my eye in Olga Kharif's <a href="https://www.bloomberg.com/news/articles/2023-10-24/cryptocurrency-delistings-from-exchanges-are-running-at-a-record-pace-this-year"><i>Crypto Delistings From Exchanges Are Already Running at a Record Pace This Year</i></a> was this:<br />
<blockquote>
Trading volume on most exchanges has plummeted in the past year, even as the number of coins has continued to multiply, with more than 1.8 million tokens listed on centralized and decentralized exchanges.
</blockquote>
Kharif reported on the collapse of trading volume earlier in <a href="https://www.bloomberg.com/news/articles/2023-10-11/coinbase-coin-trading-volume-is-likely-lowest-since-before-public-debut"><i>Coinbase’s Quarterly Crypto Trading Volume Likely Lowest Since Before Public Debut</i></a>:<br />
<blockquote>
The largest US digital-asset platform registered about $76 billion in spot trading volume, a drop of 52% from the year-ago period, according to data compiled by researcher CCData. The tally is also likely the least since before the company’s much ballyhooed direct listing on the Nasdaq Stock Market in April 2021, or just months before prices of cryptocurrencies peaked.
</blockquote>
Thus what caught my eye wasn't the drop in trading volume, it was "<i>more than 1.8 million tokens</i>"! Below the fold I ask how this could possibly make sense.<br />
<span><a name='more'></a></span>
<br />
The thing is that, like the <a href="https://blog.dshr.org/2018/10/gini-coefficients-of-cryptocurrencies.html">Gini coefficients of cryptocurrency holdings</a>, the "more than 1.8 million tokens" have an incredibly <a href="https://en.wikipedia.org/wiki/Long_tail">long-tailed distribution</a>. Almost all the 1.8M tokens are completely irrelevant, and are very likely to be scams.<br />
<br />
According to coinmarketcap.com, the top 10 tokens have a combined "market cap" of about $1.1T, with #10 TRON at $8.2B and a daily volume of $386M. They represent about 88% of the total "market cap". The next 10 have a combined "market cap" of $52B, with #20 Avalanche at $3.7B and a daily volume of $365M. The top 20 represent about 92% of the total.<br />
<br />
For contrast, tokens from 101 to 110 have a combined "market cap" of $2.7B with #110 NEM at $254M and a daily volume of $7M. Past #100 we are deep into shitcoin territory, but there is a long way yet to go.<br />
<br />
The 1000<sup>th</sup> biggest token is currently <a href="https://coinmarketcap.com/?page=10">Vesper</a>. Its "market cap" is $3.7M with a daily volume of $67K. This means that there are more than 1.799M tokens with "market caps" less than $3.7M, or that at least 99.9% of all tokens have "market caps" of less than $3.7M, and derisory daily volume.<br />
<br />
Nearly 0.5% of the way to the 1.8M mark we find things like #8767 XPRO, whose "market cap" is $5K. So it is likely that 99.5% of all tokens have "market caps" of less than $5K. Imagine how easily a token with a "market cap" of less than $5K could be <a href="https://blog.dshr.org/2023/10/the-invisible-hand-of-market.html">manipulated!</a><br />
<br />
Another way to look at this is to compare "more than 1.8 million tokens" with 4 million, the number of listed companies in the US. Do you think that 99.5% of all listed companies in the US have "market caps" of less than $5K?<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhdeFIkn5aOf55P4iVsDUcsybGZuPZo0q0xgp9xCQB0hqpmG4yeyVmhzlnZ55oa1lH8dqWipaor9JXDK12HsctU14yFwavEMIXBRZSrIIZOCz8egIX6GxdvN6QKQHhnelTG7lfRA7cwiRyjyV_VM4MFlwl_2tTw_0m2Nz7_vxwt2jNrFyWRI10JyzkX97a3/s1698/MarketDepth.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="756" data-original-width="1698" height="89" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhdeFIkn5aOf55P4iVsDUcsybGZuPZo0q0xgp9xCQB0hqpmG4yeyVmhzlnZ55oa1lH8dqWipaor9JXDK12HsctU14yFwavEMIXBRZSrIIZOCz8egIX6GxdvN6QKQHhnelTG7lfRA7cwiRyjyV_VM4MFlwl_2tTw_0m2Nz7_vxwt2jNrFyWRI10JyzkX97a3/w200-h89/MarketDepth.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2023-09-08/just-eight-exchanges-handle-90-of-all-crypto-btc-trading-volume-says-kaiko">Source</a></td></tr></tbody></table>
This incredibly skewed distribution is typical of most aspects of cryptocurrencies. For example, Vildana Hajric writes in <a href="https://www.bloomberg.com/news/articles/2023-09-08/just-eight-exchanges-handle-90-of-all-crypto-btc-trading-volume-says-kaiko"><i>Just Eight Exchanges Handle 90% of All Crypto Trading Volume</i></a> that:<br />
<blockquote>
The top eight platforms account for nearly 92% of depth — a measure of all bids and asks within 10% of the mid price — and 90% of volume, according to Kaiko. Binance, the largest crypto exchange, has this year accounted for more than 30% of global market depth and more than 60% of worldwide trade volumes. Besides Binance, the list also includes Coinbase, OKX and Huobi, among others. For context, there are hundreds of crypto exchanges, according to data from CoinGecko, though many of them see negligible or no trading at all.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEh5oudU8M657Lq3mTweZVA56R33dMDP8v2-QQQFDGJZDvGHvjeeER2-26WkE_J_I4K-ZFHZ_ZgdgJVaCgPwpwI6WNUJCsmFfjksT3Mj-Qn0_hoyw3cpyyYR0RYXfLtwJg36Cc675P7WF6FV4aP5WiBU6lBSXQcCJ2iOZndjGNK8-poqihyphenhyphen0BAgopXya0vkd/s1148/DecentralizedVolume.jpg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="609" data-original-width="1148" height="106" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEh5oudU8M657Lq3mTweZVA56R33dMDP8v2-QQQFDGJZDvGHvjeeER2-26WkE_J_I4K-ZFHZ_ZgdgJVaCgPwpwI6WNUJCsmFfjksT3Mj-Qn0_hoyw3cpyyYR0RYXfLtwJg36Cc675P7WF6FV4aP5WiBU6lBSXQcCJ2iOZndjGNK8-poqihyphenhyphen0BAgopXya0vkd/w200-h106/DecentralizedVolume.jpg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2023-10-30/bitcoin-s-btc-big-moves-are-being-magnified-by-protracted-dearth-of-liquidity">Source</a></td></tr></tbody></table>
The many tiny exchanges trading many tiny tokens are ripe for manipulation.<br />
<br />
Not that, as I discussed in <a href="https://blog.dshr.org/2023/10/the-invisible-hand-of-market.html"><i>The Invisible Hand Of The Market</i></a>, the major exchanges are immune from it. Muyao Shen reports that <a href="https://www.bloomberg.com/news/articles/2023-10-30/bitcoin-s-btc-big-moves-are-being-magnified-by-protracted-dearth-of-liquidity"><i>Bitcoin’s Big Moves Are Being Magnified by a Protracted Lack of Liquidity</i></a>:<br />
<blockquote>
A protracted dearth of liquidity in cryptocurrency markets is playing a key role in the more than 10% swings in the price of Bitcoin seen in recent weeks.<br />
<br />
Market depth remains at its lowest point this year even with the recent resurgence in trading activity being spurred on in part by expectations for a Bitcoin exchange-traded fund, FalconX’s research team found. They measure market depth by looking at the average volume of Bitcoin trading activities within 1% of its current price, on a 24-hour basis.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhtthiK9CxhvQpTkzPamNIvLy55FZWNPGknzvZyhbB0KGVCoegtR4zqPbovTZdmj6T0eNQpoKYN_FeZY4MYxZhaNU2XG75uO2u-v57yPiBrMzPJkAudBy7lWhg8V2wYNSsSzNgUcjVlMwPn3-N3yRd92qa_FFiBV2bqLdHjVR9ZSFFJLf-oSo1Onuy8Zegt/s1148/CentralizedVolume.jpg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="609" data-original-width="1148" height="106" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhtthiK9CxhvQpTkzPamNIvLy55FZWNPGknzvZyhbB0KGVCoegtR4zqPbovTZdmj6T0eNQpoKYN_FeZY4MYxZhaNU2XG75uO2u-v57yPiBrMzPJkAudBy7lWhg8V2wYNSsSzNgUcjVlMwPn3-N3yRd92qa_FFiBV2bqLdHjVR9ZSFFJLf-oSo1Onuy8Zegt/w200-h106/CentralizedVolume.jpg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2023-10-30/bitcoin-s-btc-big-moves-are-being-magnified-by-protracted-dearth-of-liquidity">Source</a></td></tr></tbody></table>
At first glance the graphs of volume on centralized and decentralized exchanges look similar, but check the Y axes. Decentralized market volume is something like $15B/month, whereas centralized exchange volume is around $200B/month. Once again, the distribution is highly skewed.<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com1tag:blogger.com,1999:blog-4503292949532760618.post-35002243637267354692023-10-24T08:00:00.033-07:002023-10-24T08:00:00.153-07:00Elon Musk: Threat or Menace Part 3<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEixbjrjdjYbxrL2KM8PPw9ur4WapKOn8m4vBTHZvMtQug7lUNKbQsrU2GkD9fSUxzE84nWysClb9NVIS8ibHV8JU_JMIhH3oSgiiVEmjpQplEoFH2CpRb-R-YVuAeJKTkO11D7szpDDXvpHfXI4KrqnGMsgoGh_Xb0Cpz6yvUVY2epCd4oBX461QG8j5irt/s634/Trailer.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="464" data-original-width="634" height="146" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEixbjrjdjYbxrL2KM8PPw9ur4WapKOn8m4vBTHZvMtQug7lUNKbQsrU2GkD9fSUxzE84nWysClb9NVIS8ibHV8JU_JMIhH3oSgiiVEmjpQplEoFH2CpRb-R-YVuAeJKTkO11D7szpDDXvpHfXI4KrqnGMsgoGh_Xb0Cpz6yvUVY2epCd4oBX461QG8j5irt/w200-h146/Trailer.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">Source</a></td></tr></tbody></table>
I started writing about the danger to innocent road users, pedestrians and first responders caused by Elon Musk lying about the capabilities of his buggy Level 2 driver assistance technologies (Autopilot and <strike>Full</strike>Fake Self Driving) <a href="https://blog.dshr.org/2021/04/elon-musk-threat-or-menace.html">two-and-a-half years ago</a> and followed up with <a href="https://blog.dshr.org/2021/08/autonowashing.html">two</a> <a href="https://blog.dshr.org/2022/01/elon-musk-threat-or-menace-part-2.html">more</a> posts. I have personally been involved in a <a href="https://blog.dshr.org/2022/01/elon-musk-threat-or-menace-part-2.html?showComment=1680533453053#c8700683981111996684">near-accident</a> caused by a Tesla randomly stopping for no reason on the Dumbarton Bridge.<br />
<br />
The reason for yet another post in the series is that Trisha Thadani, Rachel Lerman, Imogen Piper, Faiz Siddiqui and Irfan Uraizee of the <i>Washington Post</i> have published an extraordinarily detailed forensic analysis of the first widely-publicized fatal Autopilot crash in <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/"><i>The final 11 seconds of a fatal Tesla Autopilot crash</i></a>. This was the crash in which Autopilot failed to see a semi-trailer stopped across its path and decapitated the driver.<br />
<br />
Below the fold I comment on the details their analysis reveals.<br />
<span><a name='more'></a></span>
<br />
The TL;DR of the <i>Post</i>'s analysis and my questions is:<br />
<ul>
<li>The driver set Autopilot's speed to 69 in a 55 zone. Modern cars know what the speed limit is, so why did Tesla allow this?</li>
<li>The driver enabled Autopilot on a road with cross traffic despite Autopilot not being allowed on roads with cross traffic. Modern cars have GPS so they know what road they are on, so why did Tesla allow this?</li>
<li>The driver took his hands off the wheel, and was clearly not paying attention. Tesla's driver monitoring system is notoriously inadequate. It would only have warned the driver to pay attention half a mile down the road. Why would Tesla think that the driver not paying attention for half a mile was OK?</li>
<li>Autopilot didn't brake - if it had braked only 160' before impact the crash would not have happened. Apparently the fact that its cameras could not reliably detect cross traffic was well-known to Tesla's engineers, which is why Autopilot was not supposed to be enabled on roads with cross traffic. Why would Tesla think a system unable to detect cross traffic was safe for use on public roads?</li>
</ul>
The <i>Post</i> team <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">explain their methodology</a>:<br />
<blockquote>
To reconstruct the crash, The Post relied on hundreds of court documents, dash cam photos and a video of the crash taken from a nearby farm, as well as satellite imagery, NTSB crash assessment documents and diagrams, and Tesla’s internal data log, which the NTSB included in its investigation report.
</blockquote>
The driver, Jeremy Banner started the accident by <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">violating the law</a>:<br />
<blockquote>
At 6:16 a.m., Banner sets cruise control to a maximum of 69 mph, though the speed limit on U.S. 441 is 55. He turns on Autopilot 2.4 seconds later.
</blockquote>
It is typical of Tesla's disdain for the law that, although their cars have GPS and can therefore know the speed limit, they didn't bother to program Autopilot to obey the law.<br />
<br />
Banner didn't just violate the law, he also violated Tesla's <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">user documentation</a>:<br />
<blockquote>
According to Tesla’s user documentation, Autopilot wasn’t designed to work on a highway with cross-traffic such as U.S. 441. But drivers sometimes can activate it in areas and under conditions for which it is not designed.
</blockquote>
Again, Tesla's disdain for the safety of their customers, not to mention other road users, meant that despite the car knowing which road it was on and thus whether it was a road that Autopilot should not be activated on, it allowed Banner to enable it.<br />
<br />
Banner immediately violated the user documentation <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">again</a>:<br />
<blockquote>
Two seconds later, the Tesla’s data log registers no “driver-applied wheel torque,” meaning Banner’s hands cannot be detected on the wheel.
</blockquote>
Safety should have required an immediate warning and for Autopilot to start braking the car. <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">But no</a>:<br />
<blockquote>
If Autopilot does not detect a driver’s hands, it flashes a warning. In this case, given Banner’s speed, the warning would have come after about 25 seconds, according to the NTSB investigation.<br />
<br />
Banner does not have that long.
</blockquote>
25 seconds at 69 mph is 2,530 feet or 0.48 miles. Although Tesla admits that Autopilot is a Level 2 driver assistance system which requires the driver's full attention at all times, they think it is OK for the car to drive itself for half a mile at 69mph before the driver needs to start paying attention. Of course, there would be an additional delay after the warning as the driver recovered situational awareness.<br />
<br />
Combined with Tesla's pathetic "wheel torque" system for detecting whether the driver is paying attention, a 25-second delay is obviously why their cars keep running into first responders, highway dividers and other obstacles up to half a mile away from where the driver zoned out.<br />
<br />
But from Tesla's point of view, actually enforcing the rule that the driver have their hands on the wheel and be paying attention would involve "nagging" the driver. And this would make it clear even to the fan-bois that the technology was nothing like the fantasy of Tesla's marketing. And that the idea that Teslas being used as robo-taxis would more than <a href="https://www.ft.com/content/1e24011c-9865-491d-8a0a-7e6c434180d1">double Tesla's market cap</a> was a sick joke.<br />
<br />
<table align="center" cellpadding="0" cellspacing="0" class="tr-caption-container" style="margin-left: auto; margin-right: auto;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiztzDG_P7sFxEees-0amtZs24LnvSvOwf_Xn5FMMHNl8s-0tckceXt_zg-LC8Q-xMeX6qkzVOHlu_l5QpzgqdtXRrVim-_-cccjew5LNFSYOtRJvwmxUVoa5NX_EiNbHwL2rkUEbz0pppcWp8n3BsAPc9IQOpgp1ePC2L8Fe3KOQemdxr7DrUA7d9rNsuQ/s1480/CameraImage.png" style="margin-left: auto; margin-right: auto;"><img border="0" data-original-height="401" data-original-width="1480" height="109" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiztzDG_P7sFxEees-0amtZs24LnvSvOwf_Xn5FMMHNl8s-0tckceXt_zg-LC8Q-xMeX6qkzVOHlu_l5QpzgqdtXRrVim-_-cccjew5LNFSYOtRJvwmxUVoa5NX_EiNbHwL2rkUEbz0pppcWp8n3BsAPc9IQOpgp1ePC2L8Fe3KOQemdxr7DrUA7d9rNsuQ/w400-h109/CameraImage.png" width="400" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">Source</a></td></tr></tbody></table>
The inadequacy of Tesla's camera-only driver assistance systems is vividly illustrated by <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">this image</a>:<br />
<blockquote>
Two seconds later — just before impact — the Tesla’s forward-facing camera captures this image of the truck.<br />
<br />
The car does not warn Banner of the obstacle. “According to Tesla, the Autopilot vision system did not consistently detect and track the truck as an object or threat as it crossed the path of the car,” the NTSB crash report says.
</blockquote>
Banner is not paying attention, and is now <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">doomed</a>:<br />
<blockquote>
The Tesla continues barreling toward the tractor-trailer at nearly 69 mph. Neither Banner nor Autopilot activates the brakes.
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjbEsl_t20Vpg4Au6ST5BT8bi0Xmutz45Nx1zFwuzwDr42KghBcWfZnpiU_ZZIV5OP3uLf_6J90ie1Me1BvAWLphfE591LDyEkGX_OjtUYzDNvjSdiN7W5tfjN0yj-Vm-zfQxUnF9vhQGf2eOz0Dn7Lw75IKbeGe7a9_W-6t-bRA5IFdYP7pmZEq-6O5UBs/s611/Remains.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="445" data-original-width="611" height="146" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjbEsl_t20Vpg4Au6ST5BT8bi0Xmutz45Nx1zFwuzwDr42KghBcWfZnpiU_ZZIV5OP3uLf_6J90ie1Me1BvAWLphfE591LDyEkGX_OjtUYzDNvjSdiN7W5tfjN0yj-Vm-zfQxUnF9vhQGf2eOz0Dn7Lw75IKbeGe7a9_W-6t-bRA5IFdYP7pmZEq-6O5UBs/w200-h146/Remains.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">Source</a></td></tr></tbody></table>
The Tesla passes under the semi-trailer, removing the top half of the car and the driver. It <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">doesn't stop</a>:<br />
<blockquote>
The Tesla continues on for another 40 seconds, traveling about 1,680 feet — nearly a third of a mile — before finally coasting to a stop on a grassy median.
</blockquote>
Late braking could have <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">saved Banner</a>:<br />
<blockquote>
Braking even 1.6 seconds before the crash could have avoided the collision, The Post’s reconstruction found by reviewing braking distance measurements of a 2019 Tesla Model 3 with similar specifications, conducted by vehicle testers at Car and Driver. At this point the truck was well within view and spanning both lanes of southbound traffic.
</blockquote>
The NTSB <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">concluded</a>:<br />
<blockquote>
The NTSB investigation determined that Banner’s inattention and the truck driver’s failure to fully yield to oncoming traffic were probable causes of the crash.<br />
<br />
However, the NTSB also cited Banner’s “overreliance on automation,” saying Tesla’s design “permitted disengagement by the driver” and contributed to the crash.
</blockquote>
Where did Banner's “overreliance on automation” come from? The <i>Post</i> team have a <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">suggestion</a>:<br />
<blockquote>
Banner researched Tesla for years before buying a Model 3 in 2018, his wife, Kim, told federal investigators. Around the time of his purchase, Tesla’s website featured a video showing a Tesla navigating the curvy roads and intersections of California while a driver sits in the front seat, hands hovering beneath the wheel.<br />
<br />
The video, recorded in 2016, is still on the site today.<br />
<br />
“The person in the driver’s seat is only there for legal reasons,” the video says. “He is not doing anything. The car is driving itself.”
</blockquote>
This video is a <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">notorious fake</a>:<br />
<blockquote>
a Tesla engineer testified that a team specifically mapped the route the car would take in the video. At one point during testing for the video, a test car crashed into a fence, according to Reuters. The engineer said in a deposition that the video was meant to show what the technology could eventually be capable of — not what cars on the road could do at the time.<br />
<br />
While the video concerned Full Self-Driving, which operates on surface streets, the plaintiffs in the Banner case argue Tesla’s “marketing does not always distinguish between these systems.”
</blockquote>
There is a massive disconnect between what Musk and Tesla's marketing says about their driver assistance technologies and what they tell <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">regulators and juries</a>:<br />
<blockquote>
In a Riverside, Calif., courtroom last month in a lawsuit involving another fatal crash where Autopilot was allegedly involved, a Tesla attorney held a mock steering wheel before the jury and emphasized that the driver must always be in control.<br />
<br />
Autopilot “is basically just fancy cruise control,” he said.<br />
<br />
Tesla CEO Elon Musk has painted a different reality, arguing that his technology is making the roads safer: “It’s probably better than a person right now,” Musk said of Autopilot during a 2016 conference call with reporters.<br />
<br />
Musk made a similar assertion about a more sophisticated form of Autopilot called Full Self-Driving on an earnings call in July. “Now, I know I’m the boy who cried FSD,” he said. “But man, I think we’ll be better than human by the end of this year.”
</blockquote>
For the past <i>seven years</i> Musk has been claiming <a href="https://www.tesla.com/de_DE/blog/all-tesla-cars-being-produced-now-have-full-self-driving-hardware">"better than human"</a> about a system that only works in limited situations and always <i>requires a human to be ready to take over at a moment's notice</i>. Who are you going to believe, the richest man in the world or the <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">fine print in the user documentation?</a><br />
<blockquote>
Philip Koopman, an associate professor at Carnegie Mellon who has studied self-driving-car safety for more than 25 years, said the onus is on the driver to understand the limitations of the technology. But, he said, drivers can get lulled into thinking the technology works better than it does.<br />
<br />
“If a system turns on, then at least some users will conclude it must be intended to work there,” Koopman said. “Because they think if it wasn’t intended to work there, it wouldn’t turn on.”
</blockquote>
Because, after all, the system claims to be "better than human". Except in court, Tesla cannot tell the truth about their Level 2 driver assistance technologies because to do so would decimate their stock price.<br />
<br />
It should be obvious that this technology needs <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">regulation</a>:<br />
<blockquote>
The NTSB said it has repeatedly issued recommendations aiming to prevent crashes associated with systems such as Autopilot. “NTSB’s investigations support the need for federal oversight of system safeguards, foreseeable misuse, and driver monitoring associated with partial automated driving systems,” NTSB spokesperson Sarah Sulick said in a statement.
</blockquote>
<a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">But</a>:<br />
<blockquote>
Four years later, despite pleas from safety investigators, regulators in Washington have outlined no clear plan to address those shortcomings, allowing the Autopilot experiment to continue to play out on American roads, with little federal intervention.
</blockquote>
As I've been pointing out for years, Musk is running a human-subject experiment with Autopilot and FSD that has not just potentially but actually lethal outcomes. It isn't even a <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">well-designed experiment</a>:<br />
<blockquote>
Not only is the marketing misleading, plaintiffs in several cases argue, the company gives drivers a long leash when deciding when and how to use the technology. Though Autopilot is supposed to be enabled in limited situations, it sometimes works on roads it’s not designed for. It also allows drivers to go short periods without touching the wheel and to set cruising speeds well above posted speed limits.<br />
<br />
For example, Autopilot was not designed to operate on roads with cross-traffic, Tesla lawyers say in court documents for the Banner case. The system struggles to identify obstacles in its path, especially at high speeds. The stretch of U.S. 441 where Banner crashed was “clearly outside” the environment Autopilot was designed for, the NTSB said in its report. Still, Banner was able to activate it.
</blockquote>
The bigger problem is that there is no way for most of the subjects to provide informed consent for this experiment. As one of the involuntary subjects, if asked I would have refused consent. <a href="https://www.washingtonpost.com/technology/interactive/2023/tesla-autopilot-crash-analysis/">Steven Cliff, a former NHTSA administrator understands</a>:<br />
<blockquote>
“Tesla has decided to take these much greater risks with the technology because they have this sense that it’s like, ‘Well, you can figure it out. You can determine for yourself what’s safe’ — without recognizing that other road users don’t have that same choice,” ...<br />
<br />
“If you’re a pedestrian, [if] you’re another vehicle on the road,” he added, “do you know that you’re unwittingly an object of an experiment that’s happening?”
</blockquote>
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhyULlf5HioGlYs8PylfKuvPm4SGprYOUnmaii8yLisNqlFw1vB7Us5QyZBK0QR-DVvbEi7wekAJqEEMu5S4kag5qSlf2nI4YLKF0UufjRdwFFZsnk17DbBm0K3ID_8dPjEz68OjAXMYY7fgZ-BnSYYid6d0zgoV1nD8etx3Vl7pYsdq2DysfsrBpnWaDjf/s800/CulverCityCrash.jpg" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="600" data-original-width="800" height="150" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhyULlf5HioGlYs8PylfKuvPm4SGprYOUnmaii8yLisNqlFw1vB7Us5QyZBK0QR-DVvbEi7wekAJqEEMu5S4kag5qSlf2nI4YLKF0UufjRdwFFZsnk17DbBm0K3ID_8dPjEz68OjAXMYY7fgZ-BnSYYid6d0zgoV1nD8etx3Vl7pYsdq2DysfsrBpnWaDjf/w200-h150/CulverCityCrash.jpg" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://arstechnica.com/cars/2021/08/us-investigates-autopilot-after-11-teslas-crashed-into-emergency-vehicles/">Source</a></td></tr></tbody></table>
It might be arguable that experimental subjects like Banner <i>did</i> provide informed consent, although I would point out that he was misinformed by Musk and Tesla's marketing. But it cannot be argued that the 17 first responders injured and one killed that the <a href="https://www.huffpost.com/entry/us-opens-formal-probe-tesla-autopilot_n_611a651de4b0f7bc26a7eef4">NHTSA had identified by mid-2021</a> had consented. Nor had the 16 people injured including one juvenile when a Model S caused an <a href="https://abc7news.com/tesla-autopilot-crash-sf-bay-bridge-8-car-self-driving/12599448/">8-car pile-up on the Bay Bridge</a>, just to take a couple of examples.<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com17tag:blogger.com,1999:blog-4503292949532760618.post-19338359350954772342023-10-19T08:00:00.007-07:002023-10-19T08:00:00.138-07:00The Invisible Hand Of The MarketIn <a href="https://blog.dshr.org/2023/10/not-suffficiently-decentralized.html"><i>Not "Sufficiently Decentralized</i></a> I explained how the SEC's William Hinman kneecapped his agency's ability to regulate Bitcoin and Ethereum, handing the baton to the CFTC. Matt Levine <a href="https://www.bloomberg.com/opinion/articles/2023-10-16/cds-bet-bites-carl-icahn">explains the result</a>:<br />
<blockquote>
The regulatory situation in the US is that there are exchange-traded funds that allow people to speculate on Bitcoin (and Ether), but those funds hold Bitcoin (or Ether) <i>futures</i>, not actual Bitcoins. The US Securities and Exchange Commission has, so far, declined to approve spot Bitcoin ETFs (funds that just hold Bitcoins). It has said that this is because the spot Bitcoin market is largely unregulated and so there is a risk of manipulation, whereas Bitcoin futures trade on regulated US exchanges and so are safer. This has always <a href="https://www.bloomberg.com/opinion/articles/2023-03-07/ftx-wants-its-bitcoins-back-from-grayscale">struck me as incoherent</a> (manipulating the spot market also manipulates the futures), and in August <a href="https://www.bloomberg.com/opinion/articles/2023-08-29/grayscale-can-be-a-bitcoin-etf">a federal appeals court ruled</a> that it was “arbitrary and capricious,” which <i>probably</i> means that the SEC will have to approve spot Bitcoin ETFs pretty soon.
</blockquote>
Below the fold I discuss an $80M demonstration of the "risk of manipulation".<br />
<span><a name='more'></a></span>
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEixqJfz4nOQn2YTLeTWCqwrvwLdnOzYKE8HPj_NQZSALJI_JzwjS5GdJU0e0bvKT3GU69tq6tDk_qUxHzZlsDMK8IfDHZ9r27beQsLn7zkSHZY3U3q3lb-Y7pdbKI5TncQZamCXXVfuxop-TF3DUaB_veFSBsqOxAnExD-olq1ftQyfi0Y3d2IzzZjdslvw/s1200/Spike.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="675" data-original-width="1200" height="113" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEixqJfz4nOQn2YTLeTWCqwrvwLdnOzYKE8HPj_NQZSALJI_JzwjS5GdJU0e0bvKT3GU69tq6tDk_qUxHzZlsDMK8IfDHZ9r27beQsLn7zkSHZY3U3q3lb-Y7pdbKI5TncQZamCXXVfuxop-TF3DUaB_veFSBsqOxAnExD-olq1ftQyfi0Y3d2IzzZjdslvw/w200-h113/Spike.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.bloomberg.com/news/articles/2023-10-16/bitcoin-surges-above-30-000-on-etf-approval-speculation">Source</a><br /></td></tr></tbody></table>
The prospect of Bitcoin ETFs sucking retail dollars into the cryptocurrency void and thereby rescuing the exchanges and miners from the <a href="https://www.bloomberg.com/news/articles/2023-10-11/coinbase-coin-trading-volume-is-likely-lowest-since-before-public-debut">collapse</a> of <a href="https://www.bloomberg.com/news/articles/2023-10-16/crypto-lending-returns-coinbase-binance-try-to-revive-revenues">liquidity</a> has the crypto-bros panting with anticipation. So it was inevitable that someone would exploit this opportunity to manipulate the market and liquidate the shorts. And boy, did they ever! Vildana Hajric's <a href="https://www.bloomberg.com/news/articles/2023-10-16/bitcoin-surges-above-30-000-on-etf-approval-speculation"><i>Bitcoin Whipsawed by False Report Claiming SEC Approval of ETF</i></a> explains how they did it:<br />
<blockquote>
After months of declining volatility, Bitcoin traders experienced some of the wildest price swings they’ve seen in a while when a false report circulated saying that the US had approved a long-awaited exchange-traded fund.<br />
<br />
The largest digital asset jump more than 10% to $30,000 around 9:30 a.m. in New York, only to cut the gain by more than half a few minutes later after BlackRock said that its application for an ETF that invests directly in the cryptocurrency is still under review. The intraday move was the largest since March.<br />
<br />
Data from tracker Coinglass shows that more than $80 million worth of trading positions — mostly from traders who were betting on lower prices — were liquidated during the spate of volatility.
</blockquote>
Hajric is careful not to suggest <a href="https://www.bloomberg.com/news/articles/2023-10-16/bitcoin-surges-above-30-000-on-etf-approval-speculation">anyone planted the rumor</a>:<br />
<blockquote>
The episode suggests that there remains a lot of excitement and hope for a spot-Bitcoin product in the US, which regulators have in the past refused to allow. The US Securities and Exchange Commission previously cited market manipulation, among other reasons, for not granting an endorsement. However, the incident also serves as a reminder that Bitcoin’s price can easily be moved by gossip or hearsay.
</blockquote>
So the SEC was right that there is a "risk of manipulation", but no-one actually did a manipulation? How seriously market participants take their responsibilities for investor protection in the unregulated market was illustrated by Hajric's <a href="https://www.bloomberg.com/news/articles/2023-10-16/bitcoin-surges-above-30-000-on-etf-approval-speculation">closing quote</a>:<br />
<blockquote>
“That was a good 30 minutes of fun and games,” Teong Hng, chief executive officer of crypto investment firm Satori Research, said of Monday’s roundabout.
</blockquote>
Matt Levine <a href="https://www.bloomberg.com/opinion/articles/2023-10-16/cds-bet-bites-carl-icahn">gets it</a>:<br />
<blockquote>
I am not <i>that</i> sympathetic to the SEC’s objections to a spot Bitcoin ETF. But this story makes me a little more sympathetic? The price of Bitcoin jumped by 10% because people on social media were saying, falsely, that BlackRock’s spot Bitcoin ETF was approved. Seems like the sort of thing that could be manipulated!<br />
<br />
Obviously the Chicago Mercantile Exchange front-month Bitcoin futures also briefly jumped above $30,000 today. If you want to move the price of Bitcoin futures with gossip and hearsay, that is pretty much exactly as easy as moving the price of spot Bitcoin with gossip and hearsay, because they are kind of the same price.
</blockquote>
Exactly. The Bitcoin futures market may be regulated, but the price is based on an unregulated market which is notoriously manipulated. Planting the rumor on social media with the news that all the crypto-bros were expecting would goose the price was (a) free and trivially easy to do, and (b) trivially easy to predict. And it worked like an $80M charm! Why anyone takes the "price" of Bitcoin seriously is a mystery.<br />
<br />
In another section of the same piece, Levine provides another example of the <a href="https://www.bloomberg.com/opinion/articles/2023-10-16/cds-bet-bites-carl-icahn">insanity of the cryptocurrency market</a>:<br />
<blockquote>
Crypto banking is purely financial, a levered bet <i>on the size of the crypto market</i>. There are crypto firms, quasi-banks, that hold people’s cryptocurrency for them and pay them interest, and they use that crypto to invest in more crypto. They lend out crypto to hedge funds that want to make levered bets on crypto, and as long as crypto prices generally go up those hedge funds will make enough money to pay back the loans with interest and get rich themselves. But <i>nothing productive is happening</i> with these money flows; people are buying tokens but not doing any activity that makes anyone better off in the real world. Crypto prices go up because people speculate on crypto, so there is more money in crypto, so it is easier to borrow more money to make more levered bets on crypto, so prices keep going up, etc., until a slight breeze blows it all over and there’s nothing left.
</blockquote>
Some people would claim that staking is a productive investment, earning rewards for validating transactions. Of course, this only happens in Ethereum and other Proof-of-Stake systems, not in Bitcoin. But both for staking and for mining, the real question is where the rewards are coming from? They are coming from increasing the number of coins or, in other words, <i>inflating the currency</i>. Even if you are lending coins to a staking pool, you aren't actually doing anything that earns real money in the real economy, you are just benefiting from the inflation of your currency.<br />
<br />
David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com2tag:blogger.com,1999:blog-4503292949532760618.post-48391787780955675122023-10-10T08:00:00.010-07:002023-10-10T08:00:00.163-07:00Not "Suffficiently Decentralized"<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjqHsfL12YXSpmZtr49iW9fSIAL8QipwtczkGiLdm_qNP7_wFC5iJ-p-REUyDhV9fMA5FSUByEA4AySpJwg6mVgpr21Nk3aQtdvHUc7cfUxO2QY-wvsTc_M5HbzxLl-kCDdcArw_pgmiGAG/s826/HashRateDistribution-062518.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="534" data-original-width="826" height="129" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjqHsfL12YXSpmZtr49iW9fSIAL8QipwtczkGiLdm_qNP7_wFC5iJ-p-REUyDhV9fMA5FSUByEA4AySpJwg6mVgpr21Nk3aQtdvHUc7cfUxO2QY-wvsTc_M5HbzxLl-kCDdcArw_pgmiGAG/w200-h129/HashRateDistribution-062518.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;">Mining power 25 June 2018</td></tr></tbody></table>
Perhaps the most consequential result of the tsunami of <a href="https://blog.dshr.org/2022/01/blockchain-gaslighting.html"><i>Blockchain Gaslighting</i></a> ocurred on 14<sup>th</sup> June 2018 when William Hinman, the Director of the SEC's Division of Corporate Finance gave a speech to the Yahoo Finance All Markets Summit: Crypto entitled <a href="https://www.sec.gov/news/speech/speech-hinman-061418"><i>Digital Asset Transactions: When Howey Met Gary (Plastic)</i></a>. In it he hamstrung his own agency's ability to regulate the two most important cryptocurrencies by saying (my emphasis):<br />
<blockquote>
when I look at Bitcoin today, I do not see a central third party whose efforts are a key determining factor in the enterprise. The network on which Bitcoin functions is operational and <b>appears to have been decentralized for some time, perhaps from inception</b>. Applying the disclosure regime of the federal securities laws to the offer and resale of Bitcoin would seem to add little value.[9] And putting aside the fundraising that accompanied the creation of Ether, based on my understanding of the present state of Ether, <b>the Ethereum network and its decentralized structure</b>, current offers and sales of Ether are not securities transactions. And, as with Bitcoin, applying the disclosure regime of the federal securities laws to current transactions in Ether would seem to add little value. Over time, there may be other sufficiently decentralized networks and systems where regulating the tokens or coins that function on them as securities may not be required.
</blockquote>
Follow me below the fold for both the evidence that Hinman was talking though his hat, and also yet another update on the theme of my <a href="https://blog.dshr.org/2023/09/brief-remarks-to-iosco-defi-wg.html"><i>Brief Remarks to IOSCO DeFi WG</i></a>, that successful systems claiming to be decentralized, like Bitcoin and Ethereum, aren't.<br />
<span><a name='more'></a></span>
<br />
Hinman's speech was, as we shall see, ill-informed to say the least. The consequence of his abdication of responsibility was to provide an opening for the much smaller and much less powerful CFTC, which already approved trading in futures contracts on the notoriously manipulated cryptocurrency spot markets. Grabbing the chance to expand her agency's mandate, CFTC chair Rostin Banham later asserted that "<a href="https://www.coindesk.com/policy/2022/09/29/bitcoin-could-double-in-price-under-cftc-regulation-chairman-behnam-says/">Bitcoin might double in price if there’s a CFTC-regulated market</a>”. It isn't clear how the CFTC's mission migrated from protecting investors to pumping cryptocurrencies.<br />
<br />
<table cellpadding="0" cellspacing="0" class="tr-caption-container" style="float: right;"><tbody><tr><td style="text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi69tFrJSIk9jDjoRpb6eSUPwh4oK03F7HXOLrs2XJc_do1GRQhap3KVxE3NYH8PsLUSpdonJUPaPTgjmf8afXqLnjMlqRl1NJqnc3FyRrFNlPALKMPPyO5ROD9XdjuD5l13fJ3X7qz7bQv/s688/EtherMiners-101018-crop.png" style="clear: right; margin-bottom: 1em; margin-left: auto; margin-right: auto;"><img border="0" data-original-height="387" data-original-width="688" height="113" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEi69tFrJSIk9jDjoRpb6eSUPwh4oK03F7HXOLrs2XJc_do1GRQhap3KVxE3NYH8PsLUSpdonJUPaPTgjmf8afXqLnjMlqRl1NJqnc3FyRrFNlPALKMPPyO5ROD9XdjuD5l13fJ3X7qz7bQv/w200-h113/EtherMiners-101018-crop.png" width="200" /></a></td></tr><tr><td class="tr-caption" style="text-align: center;"><a href="https://www.etherchain.org/charts/topMiners">Ether Miners 10/10/18</a></td></tr></tbody></table>
It so happens that, <i>eleven days</i> after Hinman spoke, in <a href="https://blog.dshr.org/2018/07/special-report-on-decentralizing.html"><i>Special Report on Decentralizing the Internet</i></a>, I included a pie chart of the distribution of Bitcoin mining power showing that three mining pools controlled the Bitcoin blockchain. In a later post, <a href="https://blog.dshr.org/2018/12/blockchain-whats-not-to-like.html"><i>Blockchain: What's Not To Like?</i></a>, I included a pie chart showing that on 10<sup>th</sup> October that year three mining pools controlled the Ethereum blockchain.<br />
<br />
Not that in 2018 the centralization of these blockchains was news or hard to understand. That successful permissionless peer-to-peer systems inevitably centralize is a consequence of two observations, one from computer science and one from economics:<br />
<ol>
<li>The security of these systems depends upon defending against <a href="https://en.wikipedia.org/wiki/Sybil_attack">Sybil attacks</a> by making participation more expensive than the reward from an attack. This had been well-known in computer science since well before the attack was named in 2002.</li>
<li>Technologies in general, and these systems in particular, have very strong economies of scale. This had been well-known in economics since well before W. Brian Arthur's 1994 book <a href="http://www.amazon.com/Increasing-Returns-Dependence-Economics-Cognition/dp/0472064967"><i>Increasing Returns and Path Dependence in the Economy</i></a></li>
</ol>
Thus participation in these systems will have strong economies of scale, which will drive oligopoly, in other words centralization. I first detailed the process on 7<sup>th</sup> October 2014 in <a href="https://blog.dshr.org/2014/10/economies-of-scale-in-peer-to-peer.html"><i>Economies of Scale in Peer-to-Peer Networks</i></a>. It wasn't a prophesy; four months earlier Dan Goodin's <a href="https://arstechnica.com/information-technology/2014/06/bitcoin-security-guarantee-shattered-by-anonymous-miner-with-51-network-power/"><i>Bitcoin security guarantee shattered by anonymous miner with 51% network power</i></a> reported that:<br />
<blockquote>
on June 12, GHash produced a majority of the power for 12 hours straight,<br />
...<br />
GHash's ascendency to a majority miner comes even as its operators <a href="https://ghash.io/ghashio_press_release.pdf">pledged never to cross the 51-percent threshold</a>. It also comes less than a year after GHash was accused of <a href="https://bitcointalk.org/index.php?topic=327767.0">using its considerable hashing power to attack a gambling site</a>.
</blockquote>
Thus <i>four years and two days</i> before Hinman spoke, the centralization of the Bitcoin blockchain had become obvious even to the crypto-bros. The fact that the Ethereum blockchain had, by the time Hinman spoke, suffered at least as much centralization as Bitcoin's was by then equally well-known. Indeed, 17 months before Hinman spoke, Vitalik Buterin, co-founder of Ethereum stressed the fragility of decentralization in <a href="https://medium.com/@VitalikButerin/the-meaning-of-decentralization-a0c92b76a274"><i>The Meaning of Decentralization</i></a>:<br />
<blockquote>
In the case of blockchain protocols, the mathematical and economic reasoning behind the safety of the consensus often relies crucially on the uncoordinated choice model, or the assumption that the game consists of many small actors that make decisions independently.
</blockquote>
and pointed out the <a href="https://medium.com/@VitalikButerin/the-meaning-of-decentralization-a0c92b76a274">lack of it in practice</a>:<br />
<blockquote>
can we really say that the uncoordinated choice model is realistic when 90% of the Bitcoin network’s mining power is well-coordinated enough to show up together at the same conference?
</blockquote>
Saying that Hinman was ill-informed about the technology he was refusing to regulate is an understatement.<br />
<br />
Why am I bringing up this history now, more than 5 years later? To point out that there has been no progress on resolving either the regulatory or the centralization problems.<br />
<br />
Nine-digit political contributions and lobbying from Sam Bankman-Fried and Ryan Salame failed to get regulatory authority assigned to the CFTC before the collapse of FTX made the idea radioactive. The SEC's enforcement actions have steered clear of Bitcoin and Ethereum, and now face <a href="https://www.bloomberg.com/opinion/articles/2023-08-03/terra-is-usually-a-security">incompatible rulings</a> from Judge Analisa Torres in the <a href="https://www.bloomberg.com/news/articles/2023-10-03/sec-bid-for-quick-appeal-of-crypto-ruling-is-denied-by-judge?srnd=cryptocurrencies-v2">Ripple case</a> versus Judge Jed Rakoff in the <a href="https://www.bloomberg.com/news/articles/2023-07-31/ripple-ruling-on-crypto-rejected-by-federal-judge-in-terra-case?sref=1kJVNqnU">Terra case</a>.<br />
<br />
But the proximate cause of this post was Olga Kharif's <a href="https://www.bloomberg.com/news/articles/2023-09-30/ethereum-eth-concentration-grows-after-key-relay-provider-drops-out"><i>Ethereum Is Becoming More Concentrated After a Key Infrastructure Provider Dropped Out</i></a>:<br />
<blockquote>
On Sept. 27, blockchain infrastructure provider Blocknative said it will stop operating as a so-called relay, a key participant in adding blocks of transactions to the Ethereum blockchain. Its exit leaves only four other major relay players to handle most Ethereum blocks and raises concern of potential problems, ranging from the censorship of transactions to stealing of other key operators’ profits.<br />
<br />
The relay issue comes as Ethereum already faces concentration concerns in other parts of its operations. The network is run jointly by relays as well as parties called builders, which compile most transactions into blocks, and validators, which order blocks into a blockchain. Both builder and validator functions are dominated by a handful of participants. The relay decline is especially worrisome.<br />
...<br />
The situation goes against the decentralized ethos of crypto, which was born in the fallout of the Global Financial Crisis as a reaction against powers that be: corporate and government control.
</blockquote>
<a href="https://www.bloomberg.com/news/articles/2023-09-30/ethereum-eth-concentration-grows-after-key-relay-provider-drops-out">Kharif notes that</a>:<br />
<blockquote>
The builder category is highly centralized, also. While a year ago one builder — Flashbots — controlled the market, nowadays four builders, including one run by crypto market maker Wintermute, account for the majority of blocks built, according to Relayscan.io.
</blockquote>
<a href="https://www.bloomberg.com/news/articles/2023-09-30/ethereum-eth-concentration-grows-after-key-relay-provider-drops-out">And</a>:<br />
<blockquote>
To add to the problems, a pool called Lido currently controls 32.3% of total validator power. A validator controlling 34% could potentially falsify transactions.
</blockquote>
It is worth repeating that, just like Bitcoin's, Ethereum's blockchain has never been really decentralized. Even more "against the decentralized ethos of crypto" is the fact that, as <a href="https://moxie.org/2022/01/07/web3-first-impressions.html">Moxie Marlinspike documented</a>, most "smart contracts" on Ethereum use one of two services, Infura and Alchemy. On the 1<sup>st</sup> January 2019 <i>The Block</i> published <a href="https://web.archive.org/web/20190114224145/https://www.theblockcrypto.com/2019/01/10/the-burden-of-infura/"><i>The Burden of Infura</i></a>, which started:<br />
<blockquote>
Infura (hosted on AWS) is positioned as a potential single central point of failure for a significant portion of the Ethereum network in its current form, as its infrastructure dependence is widely pervasive within the dapp ecosystem
</blockquote>
Ethereum's switch from Proof-of-Work to Proof-of-Stake clearly hasn't reduced centralization, because the fundamental problem isn't technical, it is economic.
<a href="https://www.bloomberg.com/news/articles/2023-09-30/ethereum-eth-concentration-grows-after-key-relay-provider-drops-out">Kharif writes</a>:<br />
<blockquote>
Currently, relays have little financial incentives to act, and many are simply there because they want to support the network or have side businesses that can benefit from offering the function. It costs $500,000 to $1 million a year to operate a relay, Klarman said, and relays don’t get paid. While several ways to compensate relays have been considered in the past year — since Ethereum moved to its new proof-of-stake system of ordering transactions that uses relays, builders and validators — none has taken off.
</blockquote>
I have often complimented the Ethereum team's technical and operational skills; the Merge was an extraordinary success along both axes. But their devotion to Austrian economics means that they aren't good at designing economic systems. Here, for example, they designed a clever piece of technical infrastructure that needed many instances if the system were to be "sufficiently decentralized". Running an instance cost "$500,000 to $1 million a year", so there were never going to be that many of them, and the designers forgot to provide the few instances that there were with any, let alone a viable, business model. It is really amazing that, after 8 years running a system in which economics forced centralization of its infrastructure, Ethereum failed to address the economic problems of its infrastructure.<br />
<br />
The failure is actually quite understandable because the problem is fundamental. As I wrote in <a href="https://blog.dshr.org/2023/03/more-cryptocurrency-gaslighting.html"><i>More Cryptocurrency Gaslighting</i></a> what investors need is an investment:<br />
<blockquote>
Proof-of-Stake has the same fundamental problem of gambling vs. investment. As I write 17.6M ETH have been staked. With a block time of 12s, each ETH can expect a reward on average once in 6.7 years (gambling). So the "nodes" in the Ethereum network are not individuals running their own validators, but staking services such as Lido, and exchanges staking their customers' coins such as Coinbase and Kraken. Each of these nodes stakes millions of ETH (investment).
</blockquote>
It seems the crypto-bros don't need decentralization and trustlessness in practice, they only need it in <a href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3326244">Prof Angela Walch's words</a> as "a veil that covers over and prevents many from seeing the actions of key actors within the system".David.http://www.blogger.com/profile/14498131502038331594noreply@blogger.com2