A Writer I Admire

Wouldn't it be great to write like Maciej Cegłowski? I've riffed off many of his riveting talks, including What Happens Next Will Amaze You, Haunted By Data, The Website Obesity Crisis and Anatomy of a Moral Panic. Now, in a must-read tweetstorm, Cegłowski takes on "Web3", the emerging name for the mania surrounding blockchains and cryptocurrencies. He starts from this tweet:

We’re starting a new crypto team at @Stripe. I’m hiring engineers and designers to build the future of Web3 payments: https://t.co/A40QPCw3XG. pic.twitter.com/ygYTrA125T

— Guillaume Poncin (@gponcin) October 12, 2021

The replies it garnered are hilarious. Below the fold, some extracts from Cegłowski to persuade you to read his whole thread (Unroll here).

Cegłowski starts:

The web3 concept that is slowly congealing is an interesting inversion of web 2.0. Back then the idea was "build social websites and figure out the money part later." Today it's "build money stuff and figure out some non-speculative use for it later." https://t.co/gaS6XyVTwx

— Pinboard (@Pinboard) October 12, 2021

Then he lands this haymaker:

There are three non-fraud foundational problems with "web3":

1. No way to reference anything in the real world (oracle problem)
2. Immutable code makes any smart contract its own bug bounty.
3. Everything breaks (more) unless expensive distributed systems are run in perpetuity.

He credits Trammell Hudson (@qrs) for #2, which brilliantly captures the type of problem which I discussed here:

The most amazing thing about the Compound fiasco is this:

There are a few proposals to fix the bug, but Compound’s governance model is such that any changes to the protocol require a multiday voting window, and Gupta said it takes another week for the successful proposal to be executed.

They were so confident in their programming skills that they never even considered that an exploit was possible. They built a system where, if an exploit was ever discovered, the bad guys would have ~10 days to work with before it could be fixed.

Engineering is all about asking "what could possibly go wrong?" but these cowboys are so dazzled by the $$$$$ that they never ask it.

He understands that "decentralization" is the Holy Grail that drives the technologists, despite (or perhaps because of) it's being unattainable in practice. He writes:

There's a poorly articulated sense of "decentralization = freedom" that drives this culture, as well as the familiar Year Zero mentality of silicon valley that enjoys reinventing human relationships from first principles and moving them into code. And there's oceans of real money

The "oceans of real money" are the root of the evil. For example:

Note that A16Z just raised a $2.2B fund dedicated to pouring money into similar schemes. This is enough to fund 650 Chia-sized ventures! (David Gerard aptly calls Andreesen Horowitz "the SoftBank of crypto")

The "oceans of real money" are chasing the real big bucks, which are from consumers (Consumer spending is currently 69% of US GDP). There are many companies trying to be the channel between consumers and cryptocurrencies:

The real villains to focus on right now are companies like Coinbase and Stripe that are trying to make this connection happen, with one leg in the regulated financial system and one leg in the cesspit of blockchain. They should be regulated into a fine pink mist.

Here's where I disagree with Cegłowski. He writes:

If there is value in the decade plus of experimentation with blockchains (and I'm bending over backwards here to try to see it) then it will find a way to break through without this Niagara of real money investment. We'll see at least one application that is not self-referential

The "Niagra of real money" is preventing any actual value emerging, for three reasons:

• Almost the entire discourse about blockchains and their applications is corrupt. The extraordinary Gini coefficients of cryptocurrencies give the whales the means, motive and opportunity to hype their HODL-ings so that number go up "to the moon", and to practice social media "DDoS" against skeptics with armies of cultists.
• Anyone wanting to develop a blockchain-based application needs to buy into the myths of "decentralization" and "immutability" and "security". They're already in the cult.
• More fundamentally, the only defense permissionless blockchains have against Sybil attacks is to make participating in consensus be expensive, so that the cost of mounting an attack is much greater than the rewards it could obtain. Thus miners need to be reimbursed for their expensive participation. They can't be reimbursed by some central agency, that wouldn't be "decentralized"; they have to be reimbursed via the blockchain's cryptocurrency.

Thus you can't have "decentralized" without the corrupting cryptocurrency. Any possible blockchain application with actual value will be corrupted by the essential nature of the infrastructure upon which it is built.