Mathematicians launch MathHelm to crowd-review OpenAI's 722 AI-generated math papers
fortnow · x · 2026-10-23
A nonprofit forum called MathHelm, backed by a scientific committee led by Xinyi Yuan (Peking University) and including Yi Ni (Caltech), has launched to collectively review the 722 mathematical papers OpenAI generated with its internal GPT model on October 6, 2026.
Key points:
- The papers span most areas of mathematics and claim significant results, but their correctness is unverified and most have not been formalized in Lean.
- Valuable expert commentary on these papers (errors, astonishing methods, reliance on prior human ideas, alleged plagiarism) is scattered across social media; the forum aims to centralize it.
- Mathematicians worldwide can register under real names to post objective, citation-backed reviews of correctness, readability, originality, key ideas, and more, and can also discuss other AI-generated papers.
More from AGI Musings
- Lightspeed podcast: agent-era cybersecurity and why models may be more interpretable than thought — theamberyang · 2026-10-24
- Treat freedom of alignment like free speech: billions of identical agents risk ethical monoculture — robleclerc · 2026-10-24
- AI-driven vulnerability discovery will push cybercrime losses onto society — mimi10v3 · 2026-10-24
- End of 'good enough software': formal verification becomes the missing AI demand curve — bookwormengr · 2026-10-24
- Revisiting Dario's 'Big Blob of Compute': RL distributions teach no learnable alignment lesson — 1a3orn · 2026-10-24
- How would you verify a US-China AI deal? Podcast digs into inference verification tech — HaydnBelfield · 2026-10-24