Post #2920384
2026-05-10 20:16 UTC
@tao@mathstodon.xyz In other words, when thinking of new projects or paradigms, a lot comes from thinking how a method of proof, paradigm or construction can be applied to/considered in other situations that have little to do at first or even second sight with the one at hand. I don't think an AI will ever be able to do this.
Replies (1)
-
@alison@burningboard.net 2026-05-10 20:30
@jesusmargar@mastodon.social @tao@mathstodon.xyz The underlying questions in the discussion are, why bother to prove theorems? What is the purpose? And, do AI proofs contribute to that purpose?