Mathematician (algebraic geometry, motives & friends, singularities in statistics and ML).
'Geometry is successful magic' (R. Thom)
University of Essen.
Lots of noise on here. A few things that should perhaps be more obvious than they are.
(1) The Lean-formalized proof of the Riemann hypothesis is exceedingly likely to be correct.
(2) No human understands it yet. There doesn't seem to be a meaningful obstruction to humans understanding it, though, and there's already a lot of activity in this direction.
(3) There are several competing claims by humans to have an alternate proof, or to have debunked the proof elicited from Claude by Alpöge. None of these claims seem to me to be credible. If something useful comes out of a claimed alternate proof that would be great of course.
(4) There's a huge amount of work to be done. Even for RH, there's a version of it for any system of polynomial equations, where in general we don't even know the relevant L-function converges on the critical line. RH is the case where the system of polynomials is empty.
Shout out to insane British planning law for being the reason the 250,000 page Anthropic Riemann hypothesis proof *isn't* the longest document I've ever heard of
I think it’s bad to mock people struggling with the changes AI brings to their industry or profession.
“Haha such cope!” Yea buddy, we’re all coping. What’s next, you’re going to kick someone’s dog and laugh when they get sad? There’s no need to be shitty.
@ObhishekSaha@littmath At the very least, it seems like a combination of topics where experts are used to work together on big collaborative projects ?
https://t.co/x7EIhYLhAB
This is exactly how one should view this kind of AI proof: as part of nature that remains to be understood by humans in such a way that it advances our understanding of broad area of mathematics.
@PatrickHeizer Surely the first factor is much important than the second - after all, spaces of mathematical statements/proofs are also unfathomable large!
A mathematician begins reading the 250,000-page proof of the Riemann hypothesis.
“What’s the main proof technique?” a colleague asks.
“Proof by exhaustion.”
“Of all possible cases?”
“No. Of all possible readers.”
@ElliotGlazer Do you also predict that labs will spend less resources on math internally, or will they continue to use open problems as internal benchmarks?
@thecampeterson@danrobinson But many other problems in number theory and arithmetic geometry did, and a lot of the tools that were developed in the process seem to be used in this proof of RH.
@readwithai@AlanMCole What do you mean "it didn't matter before"? That would certainly be news to, say, generations of graph theorists who have been looking for a human-understable proof of the 4 colour theorem...
@AlanMCole A fundamental function of formal proofs is to lead to *new* fun/elegant/useful ideas which are not contained in the pair (statement,truth value) but only appear in the course of (searching for) the proof. Different proofs of the same statement are thus not equally valuable.
@ColmezPierre I think this was an understandable reaction in 1974, and perhaps until Weil 2/BBD/Laumon's Fourier transform, which put Weil 1 in a larger context?
i took some time to go over RH proof and understand the main themes and techniques and how it is constructed. it is really cool. [link to some claude slop website]