Post #2522100
2026-05-10 10:08 UTC
I happened to read, quite recently, about a burst of proof proposals or proof claims made with LLMs in relation to conjectures by Erdös and more generally in the field of combinatorics.
It sounded like there were not enough mathematicians to examine those submissions.
I didn't realize my question might sound overly general.
@kentpitman@climatejustice.social @screwlisp@gamerplus.org @albinowax@infosec.exchange
Replies (1)
-
@screwlisp@gamerplus.org 2026-05-10 11:53
@vnikolov@ieji.de The recent thing was covered by @tao@mathstodon.xyz . The gist is that where there used to be a couple of Erdos problem proof attempts to review per year, there are currently about 1000 ai generated ones. Of the ones that have been formally checked, only the Erdos problem 728 proof has been anything like "useable" (which was produced "mostly by chatgpt by itself" (by a mathematician).). They are all lean proofs. @kentpitman@climatejustice.social @bagder@mastodon.social