Top mathematician littmath: interrogating AI's proofs feels little different from solving myself
littmath · x · 2026-09-30
Mathematician Daniel Litt (littmath), in a discussion with Dominik Peters, shares his take on using AI models for math research. He says literature-style work was always a small part of his intellectual labor, though sometimes time-consuming — invoking Grothendieck's dictum that a problem isn't solved until it's "ripe."
If a model does answer such a question, he notes, one still has to interrogate and turn over the solution oneself — and his experience is that this isn't so different from solving it himself. A rare front-line data point on how AI actually fits into a working mathematician's workflow.
More from AGI Musings
- 2011 HN Comment Predicted Memory Bandwidth, Not FLOPS, Was the Real Bottleneck for Strong AI — Darpinian · 2026-10-01
- Kerala's 'Onlooker Fees': Knowledge Workers May Charge to Watch AI Work — deepakns · 2026-10-01
- Blog essay: "I became a cognitive empty nester thanks to AI agents" — vivekhaldar · 2026-10-01
- Ethan Mollick: frontier labs can ship half-built AI products because models improvise — emollick · 2026-10-01
- Researcher Lucius Caviola Warns of Social Conflict Over Diverging Beliefs in AI Sentience — cccalum · 2026-10-01
- threepointone: I stopped writing about what AI erases and want to build what's worth keeping — threepointone · 2026-09-30