Post #2644272

2026-05-02 15:32 UTC

So, if you manage to solve an Erdős problem using Claude or whatever, it's not going to mean jack shit unless there's a new idea in there, and *you can explain* the new idea. If you can't explain the new idea, then whoever explains and generalises it is going to be the one who is rewarded. In this case, the loser who operates Claude is going to treated like the apple that fell on Newton's head. And rightfully so.

Replies (5)

@filipb@mathstodon.xyz 2026-05-02 15:59

@jonmsterling@mathstodon.xyz (An amateur's opinion:) More generally, beyond just being validity certificates, I've always felt the value of proofs is more in bridging fields, building new frameworks (like you say), and on the more constructivist side, providing a methodology to solve a problem (I'm primarily thinking of proof = algorithm in compsci). Proofs by an LLM I would conceptually consider as a stochastic proof tree generation process, based on existing patterns in the literature. So a random collage that links axioms / lemmas / theorems to the goal statement. While random meanderings through bits of existing work might hit upon new ideas, I very much agree there is the challenge and burden to ontologically fix (and explain) that tree.

Open ##2644273
@oantolin@mathstodon.xyz 2026-05-02 17:59

@jonmsterling@mathstodon.xyz You probably missed it because you blocked him 😛, but Tao wrote about this recently: https://mathstodon.xyz/@tao/116477351524980995

Open ##2644274
@m3tti@functional.cafe 2026-05-02 18:31

@jonmsterling@mathstodon.xyz i was just talking to a friend that the llm just proposed something in another direction without or falsly proving it. What a scam headline. I hate this bubble. Shitheads!

Open ##2644275
@EJWagenmakers@mastodon.social 2026-05-02 19:05

But is the goal of mathematics to be rewarded, as an individual, and get credit for one's math skills? Maybe this is how it works in maths, but in other fields I think the purpose is to gain knowledge, and if that's the goal then the means are irrelevant. If an LLM finds the cure to cancer, we don't care about how much credit the LLM operator gets...

Open ##2644276
@christianjaeh@mathstodon.xyz 2026-05-03 09:31

@jonmsterling@mathstodon.xyz That's a great image with the apple. Gonna steal that. :newton:

Open ##2644277