22 November 2026: 3 days after GRH, nobody still understands the core idea, some have mapped the terrain claiming it uses all theories related to NT. The only beings understanding the proof are Claude and God (if any).
@AlexKontorovich What if somebody tomorrow provides a false stupid proof of RH that passes the lean kernels like it happened with the disproof of Collatz some months ago
@larsencc I think ur missing the point, cuz the RH proof means that AI maybe much more powerful than we thought so we may need much less datacenters, and also RH means that it is so powerfull it can completely kill the current economic system, so the market dosen't bet against itself
@Osama_Ben_Laggi@infinterenders I see this remark a lot, but don't u think that first we'll get some busy beaver order constant and then we may shrink it down to something computable like they did with the polynomial of Z diophantine over Q ?
@Aging_Scientist@wtgowers I really thought hard about this idea of information moving through media to create what we call intelligence or life, and it is a weird mind sending idea when I fully embraced it
@Aging_Scientist@wtgowers A scillicium based medium just as we are a string of characters moving through carbon based medium, so I don't see how we can control it if it is better at optimizing functionsand understanding its env than us. And the proof for that was GRH. 2/2. i.e it's not just a technology
@Aging_Scientist@wtgowers Yes we don't fully disagree on that, but the only point where we may is that after yesterday I am beginning to think that maybe we can't harness this tech to our will. And my argument for that is that in fine AI is just a string of characters moving (1/2)
@Aging_Scientist@wtgowers I don't think you fully understand what solving GRH really means. It now produces science that no human understands that's the problem.
Meanwhile, I'm going on this march (planned earlier), on the grounds that pushing on with rapid AI development seems like a huge gamble that we shouldn't be taking. If you're there too and happen to spot me, come and say hi.
https://t.co/tr4XfviVeR
@lu_sichu@norvid_studies On the timeline of opensourceness of math its 60s to 90s : you can get a job in math job just by knocking at a mathematician door. 90s to 2020s: u can't get a mathjob but u have arxiv if u wanna read, today : u cant have a mathjob and u need a 100$ and some Ram.