Mathematician gives CMU talk on OpenAI's proof of a non-sofic group

littmath · x · 2026-09-04

A mathematician (@BooleanAnalysis) gave a talk at Carnegie Mellon covering OpenAI's recent proof of the existence of a non-sofic group — a notable moment of AI-generated math entering serious academic discourse.

Thomas Bloom, in amplifying it, argues we should normalize great mathematicians and expositors giving talks about AI-generated work that isn't their own, as the best way to get these ideas "inside the minds of humans."

Original post →

More from Research

Research channel →