29. Critiques and Skepticism: Hype, Scams, and Recentralization in Practice

a server is simply a computer, always on and connected to the internet, that stores data and answers requests from other computers — the machine "behind" a website or app. The dream of decentralization is that millions of users each run a small piece of the network so no single company controls it. Marlinspike's point is that almost nobody wants the chore of maintaining such a machine, so in practice they pay a convenient company to do it — and that company becomes a new central point of control, exactly the thing decentralization was meant to abolish.

decentralization is supposed to mean that no small group can dominate the network. In practice the people doing the confirming cluster into a few large organizations, because pooling resources is more profitable and predictable than going it alone. The decentralization is real at the level of the protocol's rules but thin at the level of who actually runs it.

at four different layers — who confirms transactions, who issues the stable money, who holds the coins, and who connects users to the chain — the supposedly decentralized system leans on a handful of powerful companies. None of this means the underlying ledger is fake; it means that the experience of using Web3 routes through chokepoints that look uncomfortably like the platform monopolies of Web2 that Web3 was supposed to replace (Chapter 6, The Discontents of Web 2.0).

wash trading is faking a market — selling something to yourself over and over to make it look popular and rising, so real buyers pile in. The unsettling finding is that, by one credible estimate, the majority of NFT trading in 2022 was this kind of fakery, and on some marketplaces almost all of it was. A market in which most of the reported activity is manufactured is not a healthy market discovering real prices; it is, in large part, a stage set.

Terra promised a digital dollar that would hold its value automatically, with no bank vault behind it — just a clever feedback loop with a sister token. When confidence wobbled, the feedback loop ran in reverse and devoured itself: the more people tried to escape, the faster the system printed the token meant to back them, until both were worthless. It was a bank run on a bank that had, by design, no reserves.

FTX was not undone by a flaw in any blockchain. It failed the way old-fashioned financial frauds fail — by secretly spending customers' money, keeping fake books, and collapsing when people asked for their deposits back. The lesson the critics draw is pointed: the worst disasters in "decentralized finance" were caused by the centralized, unregulated companies that sat on top of it. The technology's promise to remove the need to trust institutions was, in the cases that mattered most, irrelevant — because users were trusting institutions all along.

"solutionism" is the error of treating a hard human problem — say, why people in a country distrust their institutions — as if it were a software bug that the right app could patch. The blockchain version says: put it on a ledger and the trust problem disappears. The critics' reply is that trust never disappears; it just moves to whoever wrote and runs the ledger. And those new objects of trust — code, validators, exchanges — are often harder to hold accountable than the bank or government you started with, because there is no manager to call and no court that obviously has jurisdiction.

the most demanding critique is also the calmest. It does not say Web3 is a scam; it says the jury is still out on its biggest promises, and the burden of proof has not been met. After more than ten years, the things Web3 has clearly delivered are narrower than the things it has promised — and even its strongest advocates admit that much of the visible activity is gambling rather than the building of a new internet.