OpenAI Releases 722 AI-Generated Math Solutions, Triggering Existential Crisis in Mathematics
24 Sources
[1]
OpenAI just changed math forever. Exactly how, no one yet knows
Dump of 722 AI-generated papers has mathematicians wondering: Has their field been decimated or invigorated? The shock was palpable. On the evening of 6 October, some of the world's leading mathematicians gathered in Princeton, New Jersey, to mark the opening of another academic year at the
[2]
OpenAI's math solutions aren't meeting the field's standards yet
When OpenAI released hundreds of claimed solutions to some of the world's hardest math problems this week, the frontier lab said that it had consulted an advisory group of elite mathematicians to avoid the controversy that came with the last time one of its models solved a long-standing problem in
[3]
New 'First Proof' exam reveals how AI-powered math is a pay-for-play sport
I agree my information will be processed in accordance with the Scientific American Inc. Privacy Policy. We leverage third party services to both verify and deliver email. By providing your email address, you also consent to having the email address shared with third parties for those
[4]
AI Came for Math. Now It's Chasing the Rest of Science
Last week, I attended a talk at MIT by Neil Thompson, who studies progress in computing and artificial intelligence. Over the course of just half an hour, Thompson made a convincing case that several scientific fields are about to be upended by AI. As mathematicians reel from the way AI is
[5]
Hannah Fry: How OpenAI has caused an existential crisis in mathematics | New Scientist
A few years ago, when I first started talking about mathematics in public, a professor of algebraic number theory pulled me to one side and gave me some sharp words of warning: "The world fears mathematicians, Hannah. Let's keep it that way." He was being, to put it mildly, a bit of a snob. Maths
[6]
AI edges closer to solving the Riemann hypothesis, one of math's hardest problems
On 6 October, James Maynard was nervously checking his phone under the table throughout dinner. He'd heard rumors of a tsunami heading for the shores of math. When he checked in one last time from bed, the swell struck: OpenAI had released more than 700 math papers, generated with an advanced
[7]
AI's Problem Solving Is a 'Profound Disruption' for the Math Community - CNET
Omar Gallaga has covered technology, digital culture and other topics for outlets including CNET, NPR, Wired, Texas Monthly, MSNBC, Consumer Reports, The Washington Post, the Los Angeles Times, The Atlantic and the Austin American-Statesman, where he was a longtime tech reporter, editor and
[8]
OpenAI's blockbuster Navier-Stokes proof isn't what we thought it was
I agree my information will be processed in accordance with the Scientific American Inc. Privacy Policy. We leverage third party services to both verify and deliver email. By providing your email address, you also consent to having the email address shared with third parties for those
[9]
OpenAI mistranslated mathematics into code for its Navier-Stokes proof | New Scientist
OpenAI appears to have made a subtle error when publishing its proofs of the Navier-Stokes problem, a team of mathematicians has claimed. The error doesn't mean that the proofs are incorrect or that OpenAI hasn't correctly solved the problem, but it does call into question whether mathematical
[10]
Can math stay human in the age of AI?
When AI companies value competition over collaboration, we may lose out on understanding "Exponentially exploding," mathematician Thomas Bloom told me by phone, describing the flood of new results in his field. Little did he know that hours later, OpenAI was going to post 722 mathematical
[11]
'Pure insanity': Mathematicians will need years to make sense of OpenAI's latest drop
"Staggering." "Overwhelming." "Unprecedented." "Surreal." "Pure insanity." Those were among the descriptions more than three dozen mathematicians reached for in conversations with The Verge as they tried to make sense of the flood of mathematical results OpenAI abruptly dropped on the field this
[12]
AI causes uproar in mathematics as experts warn against sharing research with LLMs after OpenAI shares 722 manuscripts and proofs -- professor argues that the field must focus on building community over raw solutions
Last week, OpenAI shared over 400 results for longstanding mathematical formulas that have troubled humans for decades; now researchers say it could take years for humans to validate and prove all the AI-driven work. But while analysts try to check the work of the unnamed OpenAI internal frontier
[13]
Is this the 'mathocalypse'? Why OpenAI's latest results dump has left mathematicians in shock
Earlier this week, OpenAI published a trove of hundreds of mathematical results it claimed could "push the frontier of human knowledge". Produced largely by an unreleased artificial intelligence (AI) model, the 722 papers relate to 372 open problems covering everything from algebra and geometry to
[14]
The most exciting claims from OpenAI's heap of new proofs
I agree my information will be processed in accordance with the Scientific American Inc. Privacy Policy. We leverage third party services to both verify and deliver email. By providing your email address, you also consent to having the email address shared with third parties for those purposes. On
[15]
OpenAI Just Carpet-Bombed Mathematics
If mathematicians have no reason to pursue the mysteries of their field, the rest of us will stumble into a less certain world. In the growing chorus of mathematicians revolting against AI companies, Hugo Duminil-Copin, who in 2022 received the Fields Medal, math's highest honor, has perhaps put
[16]
'Breathtaking,' 'Devastating': Mathematics Reels After New OpenAI Release
Sign up for Science Times Get stories that capture the wonders of nature, the cosmos and the human body. Get it sent to your inbox. "If a human did this, it would be an instant Fields Medal, no questions asked." That's how Alex Kontorovich, chair of the department of mathematics at Rutgers,
[17]
Mathematicians marvel, and grumble, at OpenAI's trove of new results
I agree my information will be processed in accordance with the Scientific American Inc. Privacy Policy. We leverage third party services to both verify and deliver email. By providing your email address, you also consent to having the email address shared with third parties for those
[18]
OpenAI publishes 377 math results, tackles long-standing problems
OpenAI has released 377 AI-generated mathematical results, including work on unsolved problems at the frontier of research. The findings span algebra, number theory, theoretical computer science, mathematical logic and topology. Released Tuesday, the results came from an advanced model that
[19]
Has OpenAI started the 'mathpocalypse'? Experts say AI's solving of iconic math problems may cause huge issues in the long run
* OpenAI has solved or partially solved 372 outstanding mathematical challenges * Some of the solutions, while clear, cannot be understood by some humans, and were produced with an internal, unreleased AI model * Renowned mathematician Terence Tao has called for colleagues working with OpenAI to
[20]
OpenAI math release wipes out early-career researcher programs
Mathematicians are raising concerns about academic credit, collaborative norms, and whether unpublished work fed OpenAI's models OpenAI's release of more than 700 AI-generated mathematical manuscripts has wiped out research programs belonging to early-career mathematicians, according to Tristan
[21]
Mathematicians React With Fury as OpenAI Releases Hundreds of New AI-Generated Proofs
More information Adding us as a Preferred Source in Google by using this link indicates that you would like to see more of our content in Google News results. It's safe to say that when it comes to mathematicians, OpenAI isn't exactly winning hearts and minds. On Tuesday, the ChatGPT maker
[22]
'Breathtaking,' 'devastating': Mathematics reels after new OpenAI release
"If a human did this, it would be an instant Fields Medal, no questions asked." That's how Alex Kontorovich, chair of the department of mathematics at Rutgers, responded to an AI-generated proof that OpenAI, the artificial intelligence giant, released on Tuesday. And that single proof -- worthy,
[23]
OpenAI Just Changed What It Means to Do Math
OpenAI's recent advancements in mathematics have sparked a shift in how complex problems are approached and solved. As highlighted by Two Minute Papers, one notable breakthrough involves solving the "irrational triangular billiards" problem, a long-standing puzzle about how light rays behave when
[24]
OpenAI Unveils 377 New Mathematical Results, Giving Researchers Fresh Problems To Explore
OpenAI's New Mathematical Push: OpenAI has published 377 new mathematical results on GitHub, marking another step in its growing focus on AI-assisted mathematical discovery. The release gives researchers and enthusiasts access to fresh results, problems, and ideas that could inspire further
Share
Copy Link
OpenAI dropped 722 AI-generated mathematical proofs covering 372 distinct results, including solutions to decades-old problems. But the release has sparked controversy over accuracy, readability, and what it means for human mathematicians when machines can solve problems no one yet understands.
OpenAI Drops 722 Mathematical Proofs, Shocking the Field
On October 6, OpenAI released an unprecedented collection of 722 preprints covering 372 distinct mathematical results, sending shockwaves through the global mathematics community
1
. The announcement came during a gathering at Princeton's Institute for Advanced Study, leaving leading mathematicians stunned by the scale and timing of the release1
. Among the claimed solutions were approaches to famous unsolved problems including the 166-year-old Riemann hypothesis and results that human researchers had been racing to complete1
5
.The release followed OpenAI's controversial Navier-Stokes breakthrough from the previous month, which required 10,000 agents working continuously for 88 hours at an estimated cost of millions of dollars
1
. This time, the company used an unreleased ChatGPT model to attempt approximately 4000 problems, with most solutions requiring just 3 hours of ChatGPT Pro thinking time on average1
. The dramatic reduction in computational costs signals that AI-powered math is accelerating rapidly, though questions about the quality and comprehensibility of these solutions persist.Human Mathematicians Struggle With Unreadable AI Proofs
The presentation of OpenAI's work has proven deeply problematic for the mathematical community. Steven Sivek from the Association for Human Mathematics found papers closest to his research area running 140 pages and described them as "completely unreadable"
1
. Edward Witten, a mathematical physicist at the Institute for Advanced Study, expressed confidence that humans who understand these results could write them more clearly1
.The lack of human understanding of mathematical results at the point of release represents a fundamental shift in how mathematics progresses
2
. Prominent mathematician Terence Tao criticized the approach, noting that "problems are being solved autonomously by AI prompters who have no interest in the broader field itself once their initial target is 'solved', and do not understand the AI output well enough to answer questions on the result, give talks, or otherwise interact with the rest of the field"2
. This disconnect threatens the traditional process where mathematicians take responsibility for their discoveries through papers, talks, and seminars that increase understanding and allow knowledge to be applied practically2
.Verification Problems and Withdrawn Papers Raise Accuracy Concerns
OpenAI has already withdrawn three manuscripts and revised 14 others from the original release
1
. Of the 719 remaining papers, only 300 have been formalized in Lean, a proof-checking language that verifies logical steps1
. However, formalization provides no guarantee of truth, as Lean can only confirm a proof establishes the statement given in code, not whether that statement matches what the paper claims in English1
.A preprint from mathematicians at the University of Cambridge and King's College London documented discrepancies between the natural language proof and Lean code in OpenAI's Navier-Stokes solution
2
. These translation errors highlight risks in relying on AI models to formalize their own solutions without human oversight2
. The researchers concluded that "because of the phenomenon of mistranslations, the NL proof by OpenAI and other autoformalised Lean proofs should not prima facie be trusted without the same peer review process and scrutiny that other proofs are subjected to"2
.OpenAI Falls Short of Advisory Group Standards
The Advisory Group on Mathematics and Artificial Intelligence (AGMAI), hosted by Princeton University's Institute for Advanced Studies, released guidelines for frontier labs solving math problems in late September
2
. OpenAI's release failed to meet several key recommendations. The group's first request was "to stop testing advanced mathematical problems on proprietary models," yet OpenAI explicitly evaluated its proprietary AI models using open research problems in mathematics2
.Only 10 of the 719 manuscripts included releases of the model's chain of thought, and 42% of proofs had not undergone formalization despite AGMAI recommendations
2
. Most critically, it remains unclear whether OpenAI is taking responsibility for ensuring human understanding will follow, as the principles suggest funding the work of human mathematicians required to make these solutions meaningful2
. OpenAI has stated it plans to fund workshops and other events to further understanding and is in talks with potential collaborating institutions1
.Proprietary AI Models Dominate as Open-Weight Alternatives Lag
The First Proof project's third evaluation revealed a massive gap between proprietary AI models and open-weight alternatives in mathematical reasoning
3
. The best performer, ProofCouncil—combining OpenAI's GPT 6 Astra with other checking agents—solved nine out of 10 research problems acceptably3
. This represents dramatic progress from February when top models solved just a few problems, and June when winners got six or seven correct3
.In stark contrast, Kimi K3, the leading open-weight model, solved exactly zero problems
3
. Mohammed Abouzaid from Stanford University noted that "the gap between the open-weight models and frontier models is very large in research mathematics—much larger than in other domains"3
. Lauren Williams from Harvard University urged the community to "invest a lot into improving" open models, as many mathematicians seek alternatives to proprietary systems due to privacy concerns following recent controversies3
.Related Stories
AI in Science Expands Beyond Mathematics
Research by Neil Thompson at MIT suggests the disruption spreading through mathematics will soon impact other scientific fields
4
. Between mid-2024 and mid-2026, the best AI models improved from solving approximately 2% of theoretical problems in probability, statistics theory, economic theory, high-energy theory, gravitation and cosmology, and quantum physics to solving about 35% of such problems4
. Thompson's study tested OpenAI models including o1, o3, GPT-5, GPT-5.1, GPT-5.4, and GPT-5.5 on problems from papers published after the models' training cutoff dates, ensuring they couldn't rely on memorized answers4
.Thompson expects "a lot of theory questions to be totally transformed in terms of AI's ability to do them" within just a few years
4
. While AI-driven breakthroughs may accelerate theoretical advances that drive practical progress, concerns remain about verification. Gabriel Manso, a PhD student on Thompson's team, emphasized that "we also need ways to verify what was done, because it's very easy for a model to tell you 'this is the right answer,' and when you look at the entire solution, you start to see maybe it is not"4
.An Existential Crisis for Mathematics
Mathematician Hannah Fry described the current moment as an existential crisis in mathematics, comparing it to chess grandmaster Gary Kasparov's experience with Deep Blue and Go champion Lee Sedol's encounter with AlphaGo
5
. The Navier-Stokes solution is "almost hilariously ugly"—a 166-page proof accompanied by hundreds of thousands of lines of formal code that AI labs "slapped down on the table and left for the mathematicians to pick apart, while the AI labs move on to the next thing"5
.The field now faces a proof almost no one wants to read, a $1 million prize no one wants to collect, and no one who can explain why this particular solution works
5
. Olga Holtz from UC Berkeley noted that mathematicians are forming groups to work through the claims, with goals extending beyond certification to "discover what, if anything, is new and useful in the mathematics"1
. The distinction between a claimed result, an independently verified result, and one absorbed into mathematical culture has never been more critical1
.References
Summarized by
Navi
[3]
Related Stories
OpenAI Forms Math Advisory Group After AI Resolves More Than 100 Open Problems
22 Sept 2026•Policy and Regulation
OpenAI's AI Solves Navier-Stokes Problem But Ignites Existential Crisis in Mathematics Community
15 Sept 2026•Science and Research
OpenAI's AI Maths Breakthrough Triggers Existential Crisis Among Mathematicians
20 Sept 2026•Science and Research
Recent Highlights
1
OpenAI's AI Models Disprove 27-Year Exponential Time Hypothesis, Rendering Decades of Research Obsolete
Science and Research
2
Google DeepMind unveils Genie 4 world model with audio and 2-minute interactive simulations
Technology
3
Microsoft AI Diagnostic Tool Beats Doctors With 83% Accuracy on Complex Medical Cases
Health
Recent Highlights
Today's Top Stories
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.
