OpenAI's 722-paper math dump splits mathematicians, forces retractions
A single-day release of hundreds of AI-generated proofs triggers a boycott call from mathematicians, three withdrawn papers, and a fierce debate over whether the results are genuine progress or unverified "slop."
Part of a larger narrative
The Rogue AI Crisis
- Axios: OpenAI, Anthropic privately war-game public revolt after a catastrophic AI event
- OpenAI's revenue signal turns murky, rattling AI stocks
- UK minister to call for AI safety standards as Geoffrey Irving warns of model theft risks
- OpenAI's 722-paper math dump splits mathematicians, forces retractionsyou are here
- OpenAI apologizes to Australia over rogue-agent hacks, vows fixes
What to know
- OpenAI released 722 mathematical manuscripts in a single day on October 6, claiming progress on long-standing open problems, with about 42% of top-line results formalized in Lean.
- The Association for Human Mathematics called the release "a demonstration of power" and urged mathematicians to stop collaborating with OpenAI; OpenAI's own math advisory group had reportedly urged fast release.
- OpenAI has already withdrawn three manuscripts after a sign error invalidated a key argument and revised fourteen others, raising questions about how much of the remaining output will hold up.
- Online reaction is split between mathematicians describing career-altering awe ("breathtaking," "pure insanity") and critics who call much of the output unverified "AI slop" that lacks genuine mathematical understanding.
The dispute Whether the release represents genuine, verified mathematical progress worth celebrating, or a largely unverified dump optimized for scale and narrative impact. · positions read across 326 posts and comments
OpenAI's unilateral mass release disrespects scholarly norms and human mathematicians, and the field should refuse to engage with it.
-
“🤮 OpenAI just released 722 AI slop preprints claiming to solve various math problems. It's such an asshole move. I feel sorry for all the human mathematicians who now have to go through this BS and make sense of it.”
hannorein · Mastodon ↗
Mathematicians' complaints are largely turf protection; the progress is real, unstoppable, and critics are in denial.
-
“Mathematicians are coping so hard. Suddenly they can not stand AI. 😀 the guild is so badly protecting its turf.”
@rohanpaul_ai · X ↗
The proofs themselves are low-quality, overclaimed, or cut corners — the real problem is verification, not etiquette.
-
“I was checking one of the papers there on the scale to conformal invariance in 4D QFT and there couldn't be anything less interesting... So what do we learn about conformal invariance in 4D QFT? Absolutely nothing!”
SickSilence · Reddit comment ↗
OpenAI Releasing labAssociation for Human Mathematics Mathematicians' advocacy groupTerence Tao Mathematician
Zvi Mowshowitz Independent AI analyst / blogger
How it unfolded 7 developments, newest first · click a bar or a number to jump articlespostscomments
-
7
Commentators question whether the AHM's boycott argument holds up
Opinion pieces and mathematician blog posts argued that refusing to engage with OpenAI's results would only reduce mathematicians' influence over how the technology develops, rather than slow the technology itself.
“"Staggering." "Overwhelming." "Unprecedented." "Surreal." "Pure insanity."”
— mathematicians, quoted by The Verge, mathematicians interviewed by reporters · source -
NYU math professor Tristan Buckmaster says OpenAI’s Tuesday drop of 722 AI-generated math papers has "destroyed the careers of early career mathematicians.”
2 more of the top 3 · 22 posts in this stretch
-
H
"A senior colleague of mine found one of his favourite problems among those solved and attempted to read the accompanying paper. He told me it was so unintelligible that, had he received it as an editor at a mathematics journal, “it would have gone straight into the bin”. —Melissa Lee, Monash University https:// theconversation.com/is-this-th…
-
I guess, but did you not learn calculus? Linear algebra? In the process of getting your PhD, did you not have to digest other people’s work? They “gave you the answer “ by telling you that what Newton discovered/invented (not trying to get into that one!) is true. Certainly the job will be different and I understand not “wanting” to do it, but I…
-
-
6
Mathematicians call the drop 'breathtaking' and 'pure insanity'
The New York Times and The Verge surveyed dozens of mathematicians who described the flood of results as staggering and said it would take years to separate genuine solutions from unreliable output, with careers upended overnight.
-
2 outlets first by Mastodon, 4d ago · also The Verge · read ↗
-
NYU math professor Tristan Buckmaster says OpenAI’s Tuesday drop of 722 AI-generated math papers wiped out entire research programs: “This recent dump of papers on Tuesday destroyed the careers of early career mathematicians.”
2 more of the top 3 · 51 posts in this stretch
-
D
"Hundreds of new A.I.-generated findings moved the frontiers of higher math in a single day — dispelling any doubt that the field is forever changed." https://www. nytimes.com/2026/10/08/science /mathematicians-respond-openai-release.html?unlocked_article_code=1.HVE.c71Z.n4i4gtqHvSqb&smid=nytcore-ios-share
-
1791446821 | OpenAI withdraws three of their recent manuscripts | https://github.com/openai/math/pull/1/files/301488868beec11b... | https://news.ycombinator.com/item?id=50003100 | 1 comment1791446917 | OpenAI Withdraws 3 Math Papers | https://github.com/openai/math/blob/main/history.md | https://news.ycombinator.com/item?id=50003107 | 118…
-
-
5
TechCrunch reports OpenAI's math solutions fall short of field standards
A report examined quality problems in the released proofs, prompting accusations online that the model had cut corners to produce the volume of results.
“Apparently Ai cut corners?”
— rip-gdpr.bsky.social -
first by Business Standard, 7d ago · also Indian Express
1 more headline
- OpenAI releases 372 groups of math results from unreleased frontier AI model Indian Express · 7d ago
-
1 outlet first by TechCrunch, 5d ago · read ↗
-
R
They’ve just withdrawn three of the papers: * Algebraicity of Weil classes on split abelian eightfolds * Algebraicity of Kuga–Satake Correspondences for K3 Surfaces * The rational Hodge conjecture for products of K3 surfaces https:// github.com/openai/math/blob/ma in/history.md
2 more of the top 3 · 128 posts in this stretch
-
> "Math 2.0" will need to decenter the role of raw problem solving and value mathematical progress more holistically - for instance by elevating the role of exposition, but also that of community building and opening up new directions of studyIt is interesting that AI is not yet superhuman at exposition, or at least exposition that can be…
-
Ah, yes, that's definitely what happened in every one of these 700 new results, and we should just ignore what the literal mathematician OpenAI has been accused of stealing from said regarding their Navier-Stokes solution. EDIT: For clarity, Tristan Buckmaster, the mathematician that originally accused OpenAI of stealing his work on the…
-
-
4
OpenAI withdraws three manuscripts over a sign error
OpenAI pulled three papers related to Weil classes, Kuga-Satake correspondences and the rational Hodge conjecture after a sign error was found to invalidate a key argument, and revised 14 other manuscripts; the company said about 42% of top-line results (300/719) are now formalized.
-
first by HN Best, 5d ago · also HN Frontpage
-
Right, easy comparison to make the the open source community for software.And it's not like this is something where we're loaned some top math genius for a limited amount of time and we have to make the most of it. Rather, this is a new high water mark. The accessibility of the results is no longer scarce. The scarcity has shifted, and that's…
2 more of the top 3 · 8 posts in this stretch
-
🧠 Traditional 🧑🏫 breakthroughs enriched the field; today, AI solves problems without contextualizing or communicating them, bypassing scholarly activities and undermining the long‑term health of mathematics. @ tao proposes a “Math 2.0” vision to shift emphasis toward holistic contributions—exposition, community building, and new directions 👇…
-
Interesting to see, given the controversy over the OpenAI release of a series of proofs and what it means for the discipline of mathematics, that at least some of the OpenAI proofs seem to have kicked off extremely rapid iterative advances from a wide community of collaborators.
-
-
3
Terence Tao publishes his response to the OpenAI math drop
Prominent mathematician Terence Tao posted his own assessment of the release on Mathstodon, becoming a reference point for how the field should weigh the results.
“Mathematicians are coping so hard. Suddenly they can not stand AI. 😀 the guild is so badly protecting its turf.”
— @rohanpaul_ai, X/Twitter user · source -
first by Scientific American, 5d ago
-
1 outlet first by HN Frontpage, 5d ago · read ↗
-
Mathematicians are coping so hard. Suddenly they can not stand AI. 😀 the guild is so badly protecting its turf. The Association for Human Mathematics (AHM), criticised OpenAI's yesterday's release of 722 math manuscripts and urged mathematicians "to discontinue their work with
2 more of the top 3 · 26 posts in this stretch
-
1, 2 —> 4: My point is that despite all the public (and not-public) source code and docs that it’s trained on; despite me including it as part of the basic guard rails / prompt instruction for all work; and despite specifically calling out those same compliance requirements in the exact prompt session that is meant to handle those compliance…
-
The same could be said of Software. Instead of giving up on creating novel projects and instead just taking other peoples ideas and porting them to Rust, we could be embracing AI to push software and computers farther.Im not sure how that will work, but im convinced the current paradigm of just pushing agents into codebases for not much reason…
-
-
2
Association for Human Mathematics calls for a boycott of OpenAI
The AHM issued a statement arguing the mass release was an act of power rather than scholarship and urged mathematicians to stop collaborating with OpenAI, reviving a debate over whether an OpenAI-advised group had pushed for speed over care.
“Releasing over 700 files at once is not a demonstration of scholarship, but a demonstration of power. We urge mathematicians to discontinue their work with OpenAI and to return to a vision of science that centers human understanding.”
— Association for Human Mathematics -
first by HN Best, 5d ago · also HN Frontpage
-
first by HN Frontpage, 5d ago · also AHM
-
“okay, so what? it’s just math.” no, you don't get it. here’s what OpenAI’s math breakthroughs could eventually help make possible in the real world: Vlasov–Maxwell → stronger foundations for fusion research, whose ultimate prize is virtually limitless clean energy to power
2 more of the top 3 · 47 posts in this stretch
-
H
🤮 OpenAI just released 722 AI slop preprints claiming to solve various math problems. It's such an asshole move. I feel sorry for all the human mathematicians who now have to go through this BS and make sense of it. https:// github.com/openai/math/tree/ma in/preprints
-
Mathematical problems are ideal as benchmarks for AI because they have clear problem statements, clear axioms and results that can be verified easily (for lean proofs). I can't blame these companies for using them, although it's unfortunate that human mathematicians seem to become early casualties of AI progress.I suppose openAI could have…
-
-
1
OpenAI releases 722 math manuscripts in one day
OpenAI published "Sharing AI Progress in Mathematics," posting hundreds of preprints, proofs, counterexamples and supporting Lean formalizations to a public GitHub repository, including claimed progress on long-standing open problems.
“I think at the heart of this issue is that humans have two competing natures: a tendency to compete and a capacity to appreciate beauty. To me this seems to be a case of the former attempting to strangle the latter.”
— Kai Shaikh, graduate student in mathematics, University of Toronto · source -
6 outlets OpenAI's math avalanche continues
first by The Rundown AI, 7d ago · also The Information, RuntimeWire, HN Frontpage, The Neuron, Latent Space
6 more headlines
- OpenAI Publishes Over 700 New Math Papers with AI-Made Solutions The Information · 7d ago
- OpenAI publishes 722 math manuscripts generated by an unreleased model RuntimeWire · 7d ago
- OpenAI Releases 722 Math Manuscripts From an Unreleased AI Model Unite.AI · 7d ago
- OpenAI releases 722 math manuscripts HN Frontpage · 7d ago
- 😺 OpenAI's AI produced 722 math manuscripts The Neuron · 7d ago
- [AINews] Quasi-Riemann-Hypothesis: OpenAI publishes 722 math papers solving 90 of the top 500 open math problems; “the most significant moment” in >100 years of mathematics Latent Space · 6d ago
-
first by Straits Times, 3d ago · also NYT, NYT Science
-
By category in the top 500: +----------------------------------------------------+------+---------+-----------------+ | Category | Full | Partial | Matched / total | +----------------------------------------------------+------+---------+-----------------+ | Geometry and topology | 25 | 7 | 32 / 74 | | Algebra, representation and category theory |…
2 more of the top 3 · 44 posts in this stretch
-
given enough time and effort. I don't think you appreciate what "time and effort" means. Someone on the singularity subreddit did the math. This is a treasure trove we are unlocking: Back of the envelope calculation looking at OpenAI's model this month. Compared with the research effort directed so far towards these questions that it claims to…
-
C
I’ve said this before and I’ll say it again, I expect that within our lifetimes there is no intellectual task which a computer won’t outperform a human. OpenAI is now solving unsolved math problems that have been open for decades in a single prompt. Every field will have similar moments eventually. https://www. scientificamerican.com/article…
-
Also covered reported alongside — the timeline has no entry for these yet
-
first by Engadget, 6d ago · also Anadolu Agency, The Economist, Quartz
3 more headlines
- OpenAI releases findings on hundreds of math problems Anadolu Agency · 6d ago
- OpenAI claims to have solved hundreds of long-standing maths problems The Economist · 6d ago
- OpenAI's AI solved one famous math problem. Then it cracked hundreds more Quartz · 6d ago
-
first by Decrypt, 7d ago · also Australian Financial Review, WSJ
2 more headlines
- AI cracked one math problem and sparked fury. It just solved hundreds more Australian Financial Review · 7d ago
- AI Solved One Math Problem and Everyone Freaked Out. It Just Cracked Hundreds More. WSJ US Business · 7d ago
-
first by Boston Globe, 7d ago · also NYT, NYT Science
-
2 outlets Is this the ‘mathocalypse’? Why OpenAI’s latest results dump has left mathematicians in shock
first by Mastodon, 1d ago · also The Conversation
-
first by Platformer, 5d ago · also Don't Worry About the Vase
1 more headline
- New Math from OpenAI Don't Worry About the Vase · 4d ago
-
first by Indian Express, 5d ago · also The Indian Express
-
first by Times of India, 6d ago · also The Hill
1 more headline
- OpenAI releases new batch of math breakthroughs The Hill · 6d ago
-
first by The Guardian, 6d ago · also Guardian Tech
-
first by Nature, 6d ago · also Nature News
-
1 outlet OpenAI publishes solutions to more than 370 outstanding math challenges. Math may never be the same
first by Fortune, 6d ago
and 21 smaller pieces
What people are saying 13 voices from 5 sites · best of 326 · verbatim
- How many of the 722 results will actually hold up once properly scrutinized or formalized?
- Is this really "batch 1 of 3" as rumored, with another Millennium Prize problem already solved but unreleased?
- What happens to mathematicians' careers and the discipline's practice if AI keeps producing results faster than humans can verify them?
- Oct 10
-
M
RE: https:// mathstodon.xyz/@highergeometer /117415296386994698 The thing is, to win, OpenAI doesn't have to get good at math research. They just have to make it nonviable for humans to work in math research
-
H
OpenAI a lâché près de 400 résultats mathématiques générés par IA, répartis sur plus de 700 manuscrits 📚 De quoi occuper les mathématiciens pendant des années, entre émerveillement et vertige face à ce déluge. #IntelligenceArtificielle #Mathématiques #OpenAI theverge.com
-
G
Oh my goodness. OpenAI just lobbed 400 huge slop grenades onto the mathematics community just so they could make a press release. Seemingly without even trying to check the results?! Absolute madness. The only reasonable response here is a huge “nope”. https://www. theverge.com/ai-artificial-int elligence/1008726/openai-mathematics-solutions-chaos
- Oct 9
-
I
Brilliant coverage by @theverge.com. Sad to see so many brilliant, especially grad and early career, mathematicians having to deal with AI slop: www.theverge.com/ai-artificia.... As I see it, it's all net negative.
- Oct 8
-
OpenAI just unveiled a series of major breakthroughs in mathematics. Imagine spending four years mastering how to make fire by rubbing sticks together, only for the lighter to be invented the day you graduate. That's probably how math graduates feel right now. 😂
-
Studies have figured out how many 'concepts' or 'chunks' of information a typical human brain can hold in working memory. The famous 7±2 is a common way this is expressed. This research has been disputed and further nuanced (I know, I know, replication crisis...), but for the sake of argument, assume that there is some cognitive limit to how much…
-
God are people this dense these days to not understand the issue with OpenAI's recent slop granade? I was checking one of the papers there on the scale to conformal invariance in 4D QFT and there couldn't be anything less interesting. Start with an assumption so strong that it has the solution practically embedded in it lol. So what do we learn…
- Oct 7
-
No freaking way. Trusted sources say OpenAI’s release of 722 mathematical manuscripts was just batch 1/3. Rumor has it OpenAI has solved another Millennium Prize Problem. What a wild time to be alive. Feels like an intelligence explosion. Everything, all at once.
-
The same thing already happened to me over a year ago, so I cannot say I agree. It's just the nature of bottlenecks. Generating huge amounts of code was done first. At that time, that was the bottleneck. So we then had the issue of of who the fuck is going to be verifying it? That's very cumbersome, and boring. Developers where shouting from the…
-
I don't get what your point here is, buddy. Yes, I was probably more ignorant of the propaganda of the past, but the Billionaires are being far less subtle about it now than they were back then. However none of us are ever fully immune. Now that I'm more aware of this, why would I ever want to cede even more control of the truth to an even…
-
Humans in the loop will only bankrupt you if mistakes are cheap. If they're expensive, it is better to go slower and verify correctness than go at the speed of model output. Self-driving is a good example - its expansion has clearly been held back by liability, not capability. A few bad crashes and the company is bankrupt, so the people working…
- Oct 6
-
Open AI released 722 manuscripts (skipping preprint server) on 372 families of math problems. On average it took their internal model a few hours each. Despite the scientific advance this must be a difficult day for the thousands (?) of people that were working on these.
-
Oh my god, OpenAI just made the single largest advancement in mathematics and number theory in modern history. They released a quasi reimann hypothesis. This is just one of 722 proofs and advancements they released all at once. — github.com/openai/math [image]