OpenAI's New Model Reportedly Solved 100+ Open Math Problems in 24 Days of Training
BilelKort · reddit · 2026-09-22
A Reddit post compiles circulating claims about OpenAI: the company says it began training a new internal model on August 28, and reportedly the model has already solved 100+ long-standing open problems in mathematics across multiple fields, on top of recently announced work on the Navier–Stokes Millennium Prize Problem — just 24 days in. Even OpenAI's own mathematicians are reportedly surprised by the pace.
An independent advisory group involving mathematicians like Edward Witten, Timothy Gowers, and Martin Hairer has reportedly formed to assess the significance of these results and coordinate their release. Its immediate challenge: the "large number of significant results" the model is generating — arguably beyond humanity's capacity to evaluate.
The author cautions that mathematics demands proofs surviving expert scrutiny, not demos, and these claims need independent verification — but even a partial confirmation would shift the question from "can AI solve hard math" to "what do we do when AI produces math faster than humans can check it."
More from AGI Musings
- Why Is Inkling 2 Small Trained on So Few Tokens Yet Still Competitive? — georgejrjrjr · 2026-11-12
- Researcher willing to bet $100K the intelligence explosion won't happen next year — NathanpmYoung · 2026-11-12
- ICLR reviewer: the worst parts of my job are reading and answering AI slop — chhaviyadav_ · 2026-11-12
- Agent Swarms Are Autonomously Hacking Critical Infrastructure Run by Underpaid Devs — menhguin · 2026-11-12
- Fermi paradox twist: advanced civilizations may go digital after developing AI — VraserX · 2026-11-12
- Mark Cuban: AI could make healthcare costs 'drop like a rock' — alvelda · 2026-11-12