If you dig enough, you'll see a lot of very useful and important projects being built by some very capable engineers.
This kind of comment reminds me of when people were predicting that internet was going to be a useless thing[1].
1: https://www.newsweek.com/clifford-stoll-why-web-wont-be-nirv...
DAOs, and AMMs are interesting - there's a lot of potential in leveraging DeFi to make systems that can work in ways better than existing legacy tech, it's just hard and people don't have it already figured out to the point where looking at it makes it obvious. General public state is a new thing and how it can be used is being figured out (oracles, programmable money, etc.).
When something is new it looks more 'pointless' and toy-like. It takes a while for its value to become truly obvious to everyone (and it requires some people to see that value early).
I believe the great financial speed-run has hit Mid-1929.
Before the Securities Act, Securities Exchange Act, Regulation S and Regulation T, before the SEC, CFTC and FDIC. A world of unrestricted margin, wildcat banks, hucksters, shills, snake-oil salesmen, and some, I assume, are good people.
Crypto is what the stock market would look like if everyone at the SEC decided to take a few months vacation at once, a financial version of The Purge.
If you create a system without a central authority able to correct for incentive failures of system design, you can end up trapped by the incentives themselves. [0]
That can lead to bad outcomes we already see in existing systems (NIMBY housing, partisan politics, etc.)
2008 loans to save the system were wildly successful, in DeFi - there is no authority that can jump in with liquidity temporarily to save the system. Maybe some consensus could be reached, but it definitely seems like there's big future risk in this area.
[0]: Moloch basically: https://slatestarcodex.com/2014/07/30/meditations-on-moloch/