🔍 Read the full analysis: OpenAI’s AI Mathematics And Its 722 Proofs: What’s The Bigger Picture? on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
OpenAI has published 722 mathematical manuscripts produced by an unnamed, unreleased model, covering 372 families of results drawn from about 4,000 problems. The manuscripts include claims about longstanding open problems, but the company says outside mathematicians have not yet confirmed them, and its repository warns that some results without formal verification may have issues.
OpenAI published 722 mathematical manuscripts on Monday, presenting work from an unnamed, unreleased model across 372 families of related results. The papers include claims about prominent open problems, but OpenAI chief executive Sam Altman said the results have not been confirmed by outside mathematicians, leaving their validity and broader importance unsettled.
According to OpenAI’s post and the project’s GitHub repository, the manuscripts cover areas including number theory, geometry, topology, operator algebras, theoretical computer science and mathematical physics. The work was selected from roughly 4,000 problems posed to the model; OpenAI said it filtered those problems for an appropriate level of significance. The source material reports an average of about three hours of ChatGPT Pro thinking compute per result.
The catalogue includes manuscripts claiming a proof of the Unique Games Conjecture, a solution to Hilbert’s tenth problem over the rationals, results on nonabelian free group factors, and a zero-free region for the Riemann zeta function to the right of Re(s) = 11/12. Other claims concern the Hodge conjecture for CM abelian varieties and conjectures in convex geometry. These are claims in the published manuscripts, not independently established solutions.
OpenAI released the materials under an Apache-2.0 license. The repository includes Lean formalizations for many, but not all, results, and cautions that some results lacking formalization could have issues. It also contains ten abridged reasoning summaries for the 372 families. The source reports two departures from the usual process: the write-up of the Riemann result was edited by humans for readability, and the Hodge result received different treatment. The selection and filtering were conducted by OpenAI, not external reviewers.
722 proofs, one question: will any of OpenAI’s AI mathematics actually lead anywhere?
An unreleased, unnamed model produced claimed proofs of results that would each define a career. Sam Altman calls them “claims not yet confirmed by outside mathematicians.” The real question isn’t whether it’s impressive. It’s whether answers nobody understands become discoveries anyone can build on.
Same day: Alon, Bloom, Gowers, Litt, Sawin post a digested, human-verified version. The model for success.
Connes rigidity counterexample challenged within a day — constructed groups fail the required condition. Three rival machine “counterexamples” from different labs now circulate.
~10,000 agents, 88 hours, est. ~$22M at retail. Priority dispute; 25 Fields Medalists sign “A Severe Misalignment” — not saying it’s wrong, saying it’s not understood.
Altman now hedges at announcement — a shift from September. Verification has barely started.
Humans extract the technique, write it up, build on it. This is where downstream discovery comes from.
The question is answered; nobody learns anything reusable. Closes a door without opening a field.
The proof breaks, or proves a statement that doesn’t match the conjecture as mathematicians mean it.
The Unique Games Conjecture is the clearest case. Results like the optimality of Goemans–Williamson for Max-Cut are proved assuming UGC. A correct proof converts them all — no understanding required. A zero-free strip for zeta works the same way for prime-distribution results. Free group factors, Kadison, Mahler would redirect whole programmes — but how depends on the method, which means digestion.
Technology. A Navier–Stokes blow-up proof doesn’t change how anyone designs aircraft; engineering turbulence models never depended on the answer. Near-term consequences are mathematical, not industrial. “AI will cure cancer next” skips several steps.
“Verification abundance, adjudication scarcity” — making proof-checking cheap doesn’t reduce the burden of deciding what’s true and what matters. 722 manuscripts land on a review system built for a trickle, filtered by a selection nobody outside OpenAI made.
Humans re-deriving results, like Alon–Gowers et al. in May
Other people’s work building on these manuscripts
How many unformalized results survive expert checking
Do the Lean statements match the real conjectures?
Do any survive peer review?
Some of it, yes — where a literature is waiting (UGC), a correct proof pays off immediately; where a proof carries a new technique humans digest, it can open a field. Most of it, probably not on its own: at 722 manuscripts with 10 reasoning summaries, the Four Colour pattern is the likely default unless mathematicians are funded and given time. And some will be wrong — OpenAI says so itself. It’s an industry pattern, not one company’s: the forced-Euler result came from an Anthropic researcher, and rival machine-generated Connes “counterexamples” circulate from different labs. The proofs arrived this week. The discoveries, if they come, will arrive at the speed of human understanding.
Verification Will Shape the Impact
The release matters because some of the claims, if proved correct, could affect fields well beyond the individual problems. The Unique Games Conjecture, for example, underpins many conditional results in theoretical computer science about the limits of approximation algorithms. A valid proof could prompt researchers to revisit conclusions that depend on the conjecture. At present, however, the manuscript’s appearance in OpenAI’s catalogue does not establish that the conjecture has been settled.
Mathematical value also depends on what a proof teaches. A result can be correct yet offer few reusable ideas, or it can provide techniques that open new lines of work. The source material contrasts the Erdős unit-distance result, which mathematicians turned into a human-readable and verified account, with the computer-assisted Four Colour Theorem, whose proof settled a question but was difficult to survey. For the new collection, the key test is whether researchers can check the arguments, identify useful methods and build on them.
The scale of the release makes that work substantial: 372 families require scrutiny, while the publication offers only ten abridged reasoning summaries. The result could be an expanded research tool if mathematicians can validate and explain the work. If the proofs are faulty, prove a different statement than intended, or resist meaningful review, the headline count alone will say little about mathematical progress.
As an affiliate, we earn on qualifying purchases.
Earlier Releases Offer Caution
This is described in the source material as OpenAI’s fourth major mathematics release of the year. The earlier releases provide relevant context but do not settle the status of the new manuscripts. In May, OpenAI’s model produced a counterexample to the Erdős unit-distance conjecture. Five mathematicians, including Noga Alon, Thomas Bloom, Tim Gowers, Daniel Litt and Will Sawin, posted a human-verified account of the result the same day. That process showed how machine output can become assessable mathematical work through human checking and exposition.
An August release called “Ten Advances” had a more contested result: a claimed counterexample to Connes’s rigidity conjecture was challenged within a day. A critique said the groups constructed did not meet the condition required by the conjecture. In September, OpenAI announced a Lean-formalized result concerning finite-time blow-up for the Navier–Stokes equations, generated using about 10,000 concurrent agents over 88 hours, according to the source. That announcement prompted debate about research priorities and the role of AI in mathematics. The history underscores why formalization, independent scrutiny and clear exposition matter alongside the number of announced results.
As an affiliate, we earn on qualifying purchases.
Independent Checks Remain Pending
No outside verification of the 722 manuscripts is established in the supplied material. It does not say whether independent mathematicians have begun reviewing each result, how long that review may take, or which manuscripts will receive priority. Nor does it provide enough information to assess the full arguments behind the most prominent claims.
It is also unclear how many of the manuscripts have complete Lean formalizations, what level of checking those formalizations received, and whether the model’s reasoning can be reconstructed in a way researchers find useful. The catalogue’s selection process was internal to OpenAI, so the criteria and decisions behind the chosen problems and included results have not been independently assessed here. Until the arguments are checked, claims about major conjectures should be treated as unverified research claims, not settled mathematical results.
formal verification software for math
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Mathematicians Must Test the Claims
The next step is independent mathematical review: researchers will need to examine individual arguments, test formalized results where available and determine whether each proof establishes the stated theorem. For results that withstand scrutiny, readable explanations will help the wider field assess what is new and whether the methods can be reused.
OpenAI has published the manuscripts and repository, but the supplied source gives no timetable for external review, no named independent review panel and no schedule for further releases. The practical measure of the project will emerge result by result: which claims survive checking, which can be explained in human terms, and whether any lead to follow-on work. Until then, the size of the catalogue is confirmed; its mathematical record is still to be established.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did OpenAI publish?
OpenAI published 722 mathematical manuscripts, grouped into 372 families of related results and produced by an unnamed, unreleased model, according to the source material.
Have mathematicians verified the claimed proofs?
The supplied material says the claims have not been confirmed by outside mathematicians. OpenAI’s repository also warns that some results without formalization could have issues.
What is the Unique Games Conjecture claim?
One manuscript claims a proof of the Unique Games Conjecture, an important open problem in theoretical computer science. The claim remains unverified in the supplied material.
Why does a correct proof need to be understandable?
Researchers need to check that the argument works and to understand whether its methods can support further results. A proof may settle a problem without providing techniques others can use.
What happens next?
Mathematicians must review the individual manuscripts and formalizations. The source material provides no review timetable, so it remains unknown when the headline claims may be confirmed, challenged or clarified.
Source: ThorstenMeyerAI.com
Evergreen bestsellers Picks
bestsellers
As an affiliate, we earn on qualifying purchases.
