What parts of math will resist AI the longest?

MickeySteamboat · x · 2026-07-21

The author argues that the disproof of the Jacobian conjecture suggests a new way to think about which parts of mathematics are most resistant to AI.

Rather than treating it as a simple “AI can now do everything” signal, the post asks what kinds of math problems may remain hardest for agents, and why some areas might be more or less formalizable than others.

Related event: Claude 'Solving' Math Conjecture Sparks AI Community Memes(73 posts)→

Original post →

More from AGI Musings

AGI Musings channel →