Commenter: AI math proofs stack huge technique heaps — math isn't 'easy' yet

erikphoel · x · 2026-11-20

erikphoel offers a measured take on the AI math breakthrough wave: his personal bar was whether AI could produce "new and creative intellectual ideas," and progress in math now seems well past that line.

Still, he notes the complexity: many of these proofs look hard because finding them required stacking an immensity of techniques with current tools, and math problems have well-defined solution spaces. Leaning on those caveats to declare "math is easy!" would be wrong — it isn't. A useful qualifier to the triumphant narrative.

Related event: Anthropic Mathematician Claims Claude-Assisted Proof of Riemann Hypothesis(124 posts)→

Original post →

More from AGI Musings

AGI Musings channel →