Regarding mass training / supercomputers. Can the true answer be peer networks? Like was the case with SETI@home, blockchains / Ethereal? Probably, even same hardware can be repurposed. Way more potential than contract / currency. Any "Satoshi Nakomoto" for distributed AI?

Schoolkids can learn Python even before university. Perhaps, university should give more fundamental knowledge. Thus - Racket is ok, plus C, plus MLs and/or Haskell. Python can be learned in other subjects like statistics / machine learning.

If it's possible to navigate on the Moon with AI and visual signs, why it's not possible to fuse even more signals to compensate for the loss of GPS? There surely are technologies to see through clouds and then technologies to recognize landscape or even stars above.

If as a result humanity will get more stupid, that is definitely the thread many others stem from. As AI rephrased it nicely to me: "there is a possibility that humanity might become more reliant on technology, leading to a decrease in individual intellectual capacity. This reduction in cognitive abilities could have far-reaching implications for the overall progress and development of society. ... While there are undoubtedly numerous benefits to be gained, such as improved automation, increased efficiency, and enhanced problem-solving capabilities, the potential trade-off of decreased human intelligence cannot be overlooked."

Designed product flows, which cause things to break right after their warranty expires, as well as artificial barriers to repair things, make damage in many ways: - more precious resources are spent - throw-away mentality multiplies the effect on consumerism - there can be more jobs if goods could be repaired - people can become smarter if they had a chance to repair things IMHO the problem is that nobody wants to start making repairable things, because first-movers [and here I also mean companies, which tried to make high quality long-lasting goods] can go bankrupt quickly. I see the solution is a new legislation, world wide, which will affect everyone. As long as everyone will need to play by these rules, it will not cause problems to corporations as they are more concerned of unfair advantages. Sooner or later, this change will happen under pressure of the consumers. Like when consumers will understand it and start to choose ecologically and socially friendly goods.

That is where most of the goods come from anyway. China embraces capitalism, then will suffocate it by capitalism's own rules. You will be free from gadgets, and ownership of anything, and happy to join on-site work at rare earth mines (to have food - can it be rented too?) so comrades can have their gadgets. This is how communism finally wins.

Earth civilization is much more advanced, and uses composite and carbon fiber materials in aerospace already. These guys are still stuck with iron and nickel? Or maybe it's some luxury space yacht?

The most beautiful thing about Python is it's syntax. While it may be true about CPython not so good for mobile applications or games or WebAssembly, but let remember that we are dealing with very high level language. It makes perfect DSL (domain-specific language) and if paired with Ai techniques... (wait? it's already there?) it can serve as language for a near-zero coding. So it may well be that strong AI affiliation will actually bring something making a huge paradigm shift to the programming as we know it. It's for sure programming itself will change in the near future. It will be boosted with AI. And Python is at the center of that domain. Armin Ronacher sees the need for change from the old paradigm, where programmers really write applications in some language, which at best is transpiled to something else. I rather see programming as a craft/art/science will change and Python to be one of the first to make next breakthrough possible. I do not really know what needs to be changed in the Python itself for what I described in my vision above, it may be just needs some genius to understand how incredible boost can be seen from natural-language programming (or maybe natural-though programming?) / near zero-code programming for routine things. Or maybe a Hindley–Milner type system needs to be added there to make type-checking to be a breeze? Let's leave it to that genius. Enough to say in Python one can write with the speed of thought - and if that thought is amplified by some problem domain ontologies and ensured with "invisible" type-checking - it will be really unstoppable. But lets that genius to surprise us!

Interesting, that Velemir Khlebnikov, Russion poet, predicted Internet almost 100 years ago. If you replace the work Radio with Internet in his "Radio of the Future" poem, you will get what we have today (and even more) in the way which Internet is used for where even "remote villages" participate. I can't think of another example for predicting information society with such a detail.