THE HOOK
OpenAI just handed mathematicians a reading list that'll keep them busy until spring.
THE GIST
OpenAI published 722 manuscripts covering 372 result families from an unreleased frontier model, posted to GitHub on 6 October. Claims include the four-dimensional Kakeya conjecture and progress toward the Riemann hypothesis, with many proofs already checked in Lean.
THE TAKE
This is the most impressive flex in AI this year, and it's being handled like a product drop. Lean-checked proofs are hard to argue with, so the sceptics will shift the fight to originality and credit. They should. A model that eats decades of human papers owes those authors more than a footnote, but three hours of Pro-level thinking per result is cheap genius.
OPINION
Scientific American reports OpenAI says almost every result came from a single agent with a single prompt, a long way from the costly 10,000-agent swarm behind its Navier-Stokes work. If that holds up, frontier-grade maths stops being a lab luxury, which is exactly why the field wants the model and not just the PDFs.
- Story type: Important story
- Impact: High
- Confidence: High (official announcement)
- Audience: Researchers
- What changed: OpenAI released 722 AI-generated maths manuscripts across 372 result families from a model it hasn't shipped.
- What to watch: Independent mathematicians' verdicts on which proofs are genuinely new, and whether OpenAI ever releases the model.
- Reading time: 1 minute
