Quote

to allow those who could not safely transact for fear of their government to transact

It doesn't solve that problem either. For example, I sold 99.9% of my BTCs, because of new KYC/AML regulations, since 2025. I still have some coins, just to play with them, if needed, and Bitcoin is still a very interesting project, from the programming perspective, but it will be harder and harder to use it in practice, without having your ID checked by someone.

The problem, which is really solved, is just a double spending problem. If you have "Alice->Bob" and "Alice->Charlie" transaction, then blockchain can tell you, which one is valid, and which one was supported by the hashrate majority. And that's it. There's not that much beyond Byzantine Generals Problem, really (by the way: if you want to execute an example, described by Satoshi, where each General has a computer, then you can use OP_SIZE on DER signature, and lock it into a shared public key, for a given group of Generals, and then, the fastest General can create a transaction, which will claim the whole pot, and will be first broadcasted to the Bitcoin network).

Quote

And then programming would be dragged backwards in progress by about 40 years.

We are getting there, but not because of Bitcoin. Recently, the "magic box, that can do everything" is Artificial Intelligence. Blockchain-based ideas will soon hit their limits, if people will find out, that transactions have fees, and if you do a lot of computations, without batching anything, then you have to pay a lot of money for doing that. Also, AI-based models will hit a different limits, when people will notice, that "AI programmed by AI" is not the best way to do things, and can lead to more and more dumb models, which will just turn everything into some kind of echo chamber, where the same data will be just processed over and over again, without adding anything creative.

Quote

Really, how should that blockchain-based magic "removal of AI fake news" even work?

Fake news will never be removed, because if some alternative version of the reality is logical, and could happen in theory, then it is hard to falsify it, if you don't have enough data. If you have "Alice->Bob" genuine transaction, and "Alice->Mallory" fake transaction, then both can have valid signatures, and the news saying that "Mallory received coins from Alice" is called "fake news", only because the "hashrate majority" decided to do so. And because all AI models have their limits, and their models are simplified and optimized, to give answers fast enough, there will be always some room for "fake news". Because AI can predict your age, based on your age, with 99% accuracy (and there is always 1% left for fake answers, just because that's how AI models are trained, and that's what prevents them from "overfitting", when you introduce a penalty for producing "too complex solutions").