It's a great idea, I hope you get traction with it :)
I am personally coming to the conclusion that having these vast repositories of knowledge that can actually talk to us is actually great. We have some issues to solve, but the end-state of having a global repository of all knowledge that can talk to us and answer questions is actually an amazing outcome.
We just need to solve those problems first; mostly getting past the AI bubble and the massive over-investment, and then solving the hallucination problems. I don't believe either of them are insoluble.
I do worry about how future generations move on from this, though. In the same way that 90's music is still effectively the zeitgeist, and we will never move on from that, because of the way that streaming services work. It's a rare new band that can compete with (e.g.) Nirvana when appealing to that segment of audience, a competition that Nirvana themselves didn't have. So we are effectively locking in Nirvana as The Disaffected Youth Grunge Band for the rest of eternity. So similarly, we are in danger of locking in the current state of the world to the training data, and never being able to move on from that, because any new zeitgeist has to compete with this one on unequal footing.
Imagine that AI lives up to its promise, gets captured by a handful of corporations, and ushers in a world of untouchable über-oligarchs that de facto rule the world. Would it have been worth it then? Can this problem even be solved?
I personally believe that it's more likely that the current batch of massively overleveraged AI companies go broke in the next year or so as the AI bubble bursts. Hopefully the USA manages to get its democratic shit together enough to not bail them out (as happened in 2008), and they just... stop.
The OSS models are almost at the same level, and are progressing fine. The technology continues to be useful, even if all the oligarchs go away.
My spidey sense goes off when we refer to open weight models from Chinese companies trained against frontier models as OSS... as if it was some Torvalds types maintaining a public git repo.
I think those open weight models exist because openai, anthropic, and google rule the roost. And I don't think things would continue as-is if those guys implode.
The technology has certainly moved faster because of all the money thrown at it recently, but it existed before that, and was progressing before that. So no, I think OpenAI, Anthropic only exist because of the open-weight models (that came before them) rather than the other way around. Google is a special case; they invented the transformer and then published it instead of commercialising it, so they actually weren't evil (for once), or they didn't realise what they'd done until too late.
I think things will continue perfectly fine if those guys implode. They didn't invent the technology, they're not the people who are advancing it most (because everything they do is proprietary and undisclosed), and their efforts at commercialising it are just getting in the way of the real research.
Imagine that AI lives up to its promise, gets captured by unaccountable bureaucrats and socialist dictators, ushering in a world of untouchable über-oligarchs that de facto rule the world. Would it have been worth it then? Can this problem even be solved?
I am personally coming to the conclusion that having these vast repositories of knowledge that can actually talk to us is actually great. We have some issues to solve, but the end-state of having a global repository of all knowledge that can talk to us and answer questions is actually an amazing outcome.
We just need to solve those problems first; mostly getting past the AI bubble and the massive over-investment, and then solving the hallucination problems. I don't believe either of them are insoluble.
I do worry about how future generations move on from this, though. In the same way that 90's music is still effectively the zeitgeist, and we will never move on from that, because of the way that streaming services work. It's a rare new band that can compete with (e.g.) Nirvana when appealing to that segment of audience, a competition that Nirvana themselves didn't have. So we are effectively locking in Nirvana as The Disaffected Youth Grunge Band for the rest of eternity. So similarly, we are in danger of locking in the current state of the world to the training data, and never being able to move on from that, because any new zeitgeist has to compete with this one on unequal footing.