Physix Frontier · News Briefing Card (QbitAI · Oct 7, 2026)

OpenAI Releases 722 Math Manuscripts, Covering Millennium Problems

KEY FACTS

OpenAI has released 722 math manuscripts, grouped into 372 results, produced by an unreleased internal model.
The model attempted roughly 4,000 research problems, with each result averaging compute equivalent to about 3 hours of ChatGPT Pro thinking.
The manuscripts touch on the quasi-Riemann hypothesis, the Hodge conjecture, the BSD conjecture, and others, spanning 17 areas of mathematics.
Papers, source code, and some Lean proofs have been uploaded to GitHub; some results still lack formal proofs.
An independent advisory group of nine scholars stated that it does not endorse these results or the process that produced them.

KEY DATA

722Released manuscripts

372Result groups

~4,000Problems attempted

~3 hours of ChatGPT Pro thinkingAverage compute

PHYSIX OBSERVATION

AI is mass-producing mathematical results at a speed that far outstrips the human community's ability to digest them. The advisory group's rush to disclaim endorsement shows that academia has yet to establish a trust mechanism for AI proofs. What truly matters is not the quantity, but whether these manuscripts can be verified and cited; otherwise they are just a mountain of papers piled up by compute.

Source: QbitAI report