I knew mathematicians were a smart bunch, but honestly, the events of the past few weeks have really given me a new level of respect for them.
They might be the first community I've seen to experience the AI "rush" and (at least as presented to an outside observer) immediately come together, assess the situation, and calmly, empathetically, and rationally act. They evaluated what AI is good at and what it lacks. They've thought through how it'll likely affect their field in the future. They've explained where the need for humans still lies, and made clear proposals for how to change their own field and for what demands to make of AI companies. Of course they're not all on the same page, but they're at least talking and trying.
They haven't started worshipping the machine god and loudly claiming their whole field is solved. Nor have they flailed wildly at LLMs as if complaining enough about it will make them go away.
Every major statement I've seen come out of the math community on this matter reads as well thought-through, humble, reasoned, and deeply human.
In these days of fear, uncertainty, and obsolescence anxiety, honestly, they've given me some confidence that maybe we will figure this stuff out after all. Maybe we'll learn from them. Who knows.
> I knew mathematicians were a smart bunch, but honestly, the events of the past few weeks have really given me a new level of respect for them.
As a mathematician, I am a bit disappointed by my (admittedly illustrious) colleagues.
I get the need to take it slowly (and I am a quite impatient person, so I shouldn't get to decide stuff like this), but everything said feels a bit too sour grapes for my taste.
Ok, maybe AI did not solve the field (I believe it will, btw), maybe there is a need for human "understanding", but:
1) They don't seem to consider even the possibility (not the certainty) that they might be wrong, that math as we know it is gone, and we cannot "adapt"
2) They seem to have been oblivious all these years about AI eventually reaching this point (at least I personally wasn't, I predicted this stage back in 2018)
Regarding 1, it's not clear what you want them to actually do. Would you rather they just gave up, rolled over and died instead of at least trying to rescue their field? What is to be gained from considering this?
I suppose you could say, hypothetically, that if the field were “solved” then the protection of the feelings and sense of need for human understanding of the now obsolete practitioners might hinder any benefits that the now solve field could provide.
The answer to this is probably just “get out of the way” and then your “work” would be to develop a better appreciation of what has been produced.
> The answer to this is probably just “get out of the way” and then your “work” would be to develop a better appreciation of what has been produced.
They aren't in the way. AI labs can do their own math all they want, and nobody is stopping them. Rather, mathematicians are simply speaking about the topic.
If LLMs could both prove any relevant theorem, explain such a theorem cogently and lay the groundwork for further theory, I don't see how any way mathematicians could stand in the way of that.
As a Nobel Prize winner, I am surprised that you have not noticed how Tao and Gowers were steering everyone in the AI direction for years and now do damage control by nudging everyone to either hypothetical open models paid for by industry or the EU or sitting in a committee that "advises" OpenAI.
No one is saying "Perelman and Wiles didn't need that silly AI and N-S was more of a counterexample". Most people imagine the sky is falling.
What if OpenAI is stalling because they don't have 100 additional unpublished proofs as claimed?
> As a Nobel Prize winner, I am surprised that you have not noticed how Tao and Gowers were steering everyone in the AI direction for years
I did notice, this is why I had thought that they had thought through the consequences. But, unfortunately, it is likely they imagined that AI would plateau somewhere around the "smart undergraduate" level and that they will stay relevant in pure problem solving.
And why have the Nobel Prize winners been silent (or taken post-Nobel bets contrary to the current consensus ie Hinton-Hassabis)?
Taking GP at his word, I'd be more terrified if Oai had identified one hundred NEW open problems..
In contrast with Fields Medallists, some Nobelists get the prize for something they'd eventually NOT be (in)famous for. Like Josephson. In hindsight, will it be because they found out, as young unknowns, how to explore underappreciated avenues without millions in dedicated funds (uh encouragement).. ??
GP, did you? Get the prize for the question you are most proud of.
> What if OpenAI is stalling because they don't have 100 additional unpublished proofs as claimed?
Which is what I think a large chunk of AI progress actually is - people taking the latest models out for a spin, and browbeating them to actually ship code, and RLHFing them to ship better code in the process, either directly or indirectly through Github.
I think I'm mostly on your side (also I believed early it would get here), but what does it mean to "solve the field"? I'm pretty much as bullish as you get on AI but I'm sure I can come up with questions AI cannot solve. I think the space of problems in math is so large and the distribution of proof lengths so heavy-tailed, there's probably no way without dyson-sphereing the sun to solve all easy to state problems.
I think it is good people are talking but there seems to be a lot of "forced optimism", especially by people obviously hoping to keep a channel open between themselves and the labs.
I think it isn't true that no one imminent is considering (1). Tsimmerman said he thought the job of math researcher was over as soon as he announced he was leaving the field.
I think the question is so far still open, but it is easy to imagine Tsimmerman was right. I haven't been impressed with the calls for "human understanding" as the really important thing. It seems to miss the point of why math has been funded (not exactly problems, but certainly not to ensure every theorem is understood either).
As a scifi comment, it is now possible to start thinking about some future AI system which decides if it will need to explain certain math to humans and to pick and train young people for that purpose.
I would not attribute this attitude to the entire Math community, there clearly are many who are reacting negatively to this, as evidenced by several comments on various forums including this one. People are understandably upset about their chosen profession and livelihood being upended and reacting accordingly, similar to how people from other impacted professions are.
This group of Mathematicians just happens to be more willing to accept the inevitable and adapt. I would actually say software engineers, being amongst the earliest impacted by AI after creatives, clearly have adapted to the new reality much more broadly as evidenced by online discourse and coding agent providers' skyrocketing revenues.
Presumably OP believes that software engineers went to far in the direction of “started worshipping the machine god and loudly claiming their whole field is solved.” Certainly within software engineering there is much louder voices on that side of things. By now the voices saying that it’s useless have pretty much died out. But it seems quite reasonable to say that the wider software community didn’t “come together, assess the situation, and calmly, empathetically, and rationally act.”
One might argue that this is due to size of the software community, or commercial interests. But I do believe OP was right in his description of the response of the math community as different from the software community.
No, I don't think they are adapting well. I've seen 3d designer at my company adapting better on the disruption and paradigm shifts GPT-6 Astra is going to cause.
The 3D designers are in a tougher situation because it's far less of abstract issue then math, and yet the results are astounding. It's hard to believe until you see it yourself. It's surreal.
What do you believe doesn’t reflect positively on the authors of that open letter? It raises valid concerns, and the statement you quoted is a factual observation.
Their field is extremely rigorous. As rigorous as it can possibly get in that they have to prove each and every line of their work beyond any doubt. The discipline they have cultivated in their culture shows in their response to AI as well.
As opposed to some other fields in which rigor was either not part of the culture or was not always possible. For e.g., software engineering in terms of code quality being produced. There were some indirect signals here and there but they are all subjective.
Every community approaches problems with the tools at its disposal. Doing mathematics is about thinking, so mathematicians approach problems (at least in public) with more thinking and less emotion (than some fields where emotion is much more important). While you and I may appreciate this approach (and even be inspired by it), it remains to be seen whether it ultimately makes the outcomes any better.
I disagree. Software developers seem to have adapted to AI much faster than the maths community.
There are certainly still strong opponents of AI, but they seem to be a small, vocal minority. Most developers seem to have simply adapted and moved on.
> If I thought this was a good thing to do, this would be a great group of people to do it. But I have immediate misgivings. OpenAI has had some very bad publicity, and so they are trying to exploit the trust and respect that these mathematicians command. It is unrealistic to think that this group can change the way OpenAI does business; I don’t need to tell you all the objections that people have raised to that. Is it a good idea to help their crisis management?
(FRS, UCLA professor of algebraic geometry and topology, with work influenced by the Hodge conjecture, one of the remaining Millennium Prize problems).
I don't really get it, seems like a gate keeping exercise for the academic community.
Of course OpenAI and others will hype any LLM led/assisted research results, and the media will find juicy headlines to write about it. But in the end they are publishing findings, and the academic community can gauge the developments and decide what they want to do with them. To me that seems like research working as expected.
> We are currently facing the very specific challenge of advising OpenAI on how to coordinate the release of a large number of significant results in mathematics that they report have been produced by their internal model.
The only thing I want to see is the problem statements, solutions, and associated Lean proofs. Anything else is gatekeeping. What a low point for academia.
I'm not sure how you've ended up so frustrated with an advisory group whose advice has not been given. Human mathematicians routinely delay publication of their results past the point you describe; is someone gatekeeping when they take a month to clean up their paper?
Or are you perhaps sublimating your own AI anxiety into confrontational assertions that other people who discuss AI aren't as AI-pilled as you?
I'm not sure why you see bullying here? OpenAI has always worked with expert mathematicians; their unit-distance result (https://openai.com/index/model-disproves-discrete-geometry-c...) was a much deeper collaboration. They published the Navier-Stokes result early, to either benchmark or show off their new model depending on how you look at it, and they want to understand more about the negative effects which many mathematicians feel that early publication had.
If the advisory group says something like "AI is bad and nobody should use it in math", I'm pretty confident OpenAI will ignore them.
> This group came together after OpenAI approached some of its members about establishing an external advisory board. In agreement with OpenAI, they decided to create an independent group and invite others to join.
The purpose is obvious - AI companies are using math to as a PR tool to demonstrate that AI is more capable than humans. This devalues math and mathematicians. There is even a relevant announcement from today, where OpenAI announced they solve 100 open math problems, without even bothering to state what they are.
However, even if they are successful in removing math from the PR toolkit of AI companies, I can’t see it more than a short term band aid. If AI progression continues, the announcements will just transition from big flashy one, to a kid using Claude on their laptop.
I don’t have a solution here, and I’m worried it’s coming to all human fields. So it’s more of an ability to observe what might happen to all of us.
> There is even a relevant announcement from today, where OpenAI announced they solve 100 open math problems, without even bothering to state what they are.
This group has been formed to tell OpenAI whether or not (and if yes how) they should release those 100+ problems.
IMO “AI companies are using math to as a PR tool” is not quite it; they are using math problems as a benchmark to evaluate their AIs, but after the prior blowback from the mathematical community, they are happy to sit on the results until the community (this group) tells them how to publish the results.
And as you say, even if they don't publish the results eventually anyone will be able to do this when this model is released, so it's just a matter of time.
> I don’t have a solution here, and I’m worried it’s coming to all human fields. So it’s more of an ability to observe what might happen to all of us.
You can also look (as I did back in 2018 when contemplating the end of math) at how chess or Go players adapted mentally to the idea that whatever they may do, a machine could do it better (Lee Sedol famously quit Go).
I think this matters very little. These were adaptations to initial advances of small scope which we are rapidly leaving behind. If there is truly no near and practical ceiling to the abilities of AI in mathematics, not even in terms of efficiency, then I think it is likely there is similarly no ceiling in other cognitive endeavours.
In that case, everything stands to be disrupted and those who wield AI will have massive amounts of power in the short term. But this is not a stable situation, and therefore those who wield AI will have to cede control of it or otherwise enter into a great war with those that do not.
I hope for all our sakes that there is a practical ceiling.
These math problems don't exist in a vacuum. It's not like mathematicians are doing mathematics just for the funsies.
Calling it a publicity stunt is an interesting way of framing ground breaking discoveries. The beneficiaries down stream of these discoveries could not care less how they were found.
> It's not like mathematicians are doing mathematics just for the funsies.
But they do. Most of higher math has no practical applications and is just a mental playground, philosophy constrained by formal logic and a set of axioms.
Hope is these capabilities will somehow translate to something more practical like physics, chemistry or biology.
The composition of the panel might give us some clues as to the nature of OpenAI's unreleased results. Some of Ed Witten's work is quite closely related to the Yang Mills existence and mass gap problem, which is one of the remaining Millennium problems. I went through the list of panel members and asked an LLM to associate each member of the panel with one or two millennium problems. Here is the resulting count:
Hodge conjecture: 5
Birch and Swinnerton-Dyer Conjecture: 3
Navier-Stokes: 2
P vs NP: 1
Yang-Mills existence and mass gap: 2
So this gives a bit of further evidence that they made progress on the Hodge and Birch and Swinnerton-Dyer conjectures. Maybe also Yang-Mills.
I really wish the models were smart enough to "tail their own logs" and tell us whose work/papers/chats were the catalyst for these math insights. What makes their "internal" model so much better at math? Must be trained on a whole lot of honesty.
Why don't they just publish the results? What does an advisory group do in this case when none of the members can validate or verify any of the purported solutions?
Honestly I am not impressed with Tao treating "ai" as a threat to mathematicians. He might as well say math itself, specifically linear algebra and gradient descent are a threat to mathematicians. Maybe think of something original rather than pander the hype of day?
Why aren't they consulting the younger mathematicians whom this actually affects?
Anyways, this is overblown. Let me know when AI swarms can create a research program or library from scratch. We're really missing long term planning and abstraction creation in these things.
They might be the first community I've seen to experience the AI "rush" and (at least as presented to an outside observer) immediately come together, assess the situation, and calmly, empathetically, and rationally act. They evaluated what AI is good at and what it lacks. They've thought through how it'll likely affect their field in the future. They've explained where the need for humans still lies, and made clear proposals for how to change their own field and for what demands to make of AI companies. Of course they're not all on the same page, but they're at least talking and trying.
They haven't started worshipping the machine god and loudly claiming their whole field is solved. Nor have they flailed wildly at LLMs as if complaining enough about it will make them go away.
Every major statement I've seen come out of the math community on this matter reads as well thought-through, humble, reasoned, and deeply human.
In these days of fear, uncertainty, and obsolescence anxiety, honestly, they've given me some confidence that maybe we will figure this stuff out after all. Maybe we'll learn from them. Who knows.
As a mathematician, I am a bit disappointed by my (admittedly illustrious) colleagues.
I get the need to take it slowly (and I am a quite impatient person, so I shouldn't get to decide stuff like this), but everything said feels a bit too sour grapes for my taste.
Ok, maybe AI did not solve the field (I believe it will, btw), maybe there is a need for human "understanding", but:
1) They don't seem to consider even the possibility (not the certainty) that they might be wrong, that math as we know it is gone, and we cannot "adapt"
2) They seem to have been oblivious all these years about AI eventually reaching this point (at least I personally wasn't, I predicted this stage back in 2018)
The answer to this is probably just “get out of the way” and then your “work” would be to develop a better appreciation of what has been produced.
They aren't in the way. AI labs can do their own math all they want, and nobody is stopping them. Rather, mathematicians are simply speaking about the topic.
So is the answer that they should just shut up?
No one is saying "Perelman and Wiles didn't need that silly AI and N-S was more of a counterexample". Most people imagine the sky is falling.
What if OpenAI is stalling because they don't have 100 additional unpublished proofs as claimed?
I did notice, this is why I had thought that they had thought through the consequences. But, unfortunately, it is likely they imagined that AI would plateau somewhere around the "smart undergraduate" level and that they will stay relevant in pure problem solving.
Taking GP at his word, I'd be more terrified if Oai had identified one hundred NEW open problems..
In contrast with Fields Medallists, some Nobelists get the prize for something they'd eventually NOT be (in)famous for. Like Josephson. In hindsight, will it be because they found out, as young unknowns, how to explore underappreciated avenues without millions in dedicated funds (uh encouragement).. ??
GP, did you? Get the prize for the question you are most proud of.
Which is what I think a large chunk of AI progress actually is - people taking the latest models out for a spin, and browbeating them to actually ship code, and RLHFing them to ship better code in the process, either directly or indirectly through Github.
How do you reconcile that with incompleteness and undecidability results?
I think it isn't true that no one imminent is considering (1). Tsimmerman said he thought the job of math researcher was over as soon as he announced he was leaving the field.
I think the question is so far still open, but it is easy to imagine Tsimmerman was right. I haven't been impressed with the calls for "human understanding" as the really important thing. It seems to miss the point of why math has been funded (not exactly problems, but certainly not to ensure every theorem is understood either).
As a scifi comment, it is now possible to start thinking about some future AI system which decides if it will need to explain certain math to humans and to pick and train young people for that purpose.
This group of Mathematicians just happens to be more willing to accept the inevitable and adapt. I would actually say software engineers, being amongst the earliest impacted by AI after creatives, clearly have adapted to the new reality much more broadly as evidenced by online discourse and coding agent providers' skyrocketing revenues.
One might argue that this is due to size of the software community, or commercial interests. But I do believe OP was right in his description of the response of the math community as different from the software community.
The 3D designers are in a tougher situation because it's far less of abstract issue then math, and yet the results are astounding. It's hard to believe until you see it yourself. It's surreal.
Can't disagree with "deeply human", but I wouldn't always use it in a positive sense:
https://proofsandprompts.com/2026/09/10/open-letter-about-th...
>Participation in an event so closely associated with Anthropic and OpenAI could plausibly negatively impact the future reputations of participants
Their field is extremely rigorous. As rigorous as it can possibly get in that they have to prove each and every line of their work beyond any doubt. The discipline they have cultivated in their culture shows in their response to AI as well.
As opposed to some other fields in which rigor was either not part of the culture or was not always possible. For e.g., software engineering in terms of code quality being produced. There were some indirect signals here and there but they are all subjective.
There are certainly still strong opponents of AI, but they seem to be a small, vocal minority. Most developers seem to have simply adapted and moved on.
Don't give Dario any ideas please.
> next week: Mathematicians are solved [214 points, 891 comments]
> If I thought this was a good thing to do, this would be a great group of people to do it. But I have immediate misgivings. OpenAI has had some very bad publicity, and so they are trying to exploit the trust and respect that these mathematicians command. It is unrealistic to think that this group can change the way OpenAI does business; I don’t need to tell you all the objections that people have raised to that. Is it a good idea to help their crisis management?
https://terrytao.wordpress.com/2026/09/21/advisory-group-on-...
Of course OpenAI and others will hype any LLM led/assisted research results, and the media will find juicy headlines to write about it. But in the end they are publishing findings, and the academic community can gauge the developments and decide what they want to do with them. To me that seems like research working as expected.
There are typically processes to ensure rigor in the findings and to weed out crap.
It doesn't always work, but it tends to work much better than any other community
The only thing I want to see is the problem statements, solutions, and associated Lean proofs. Anything else is gatekeeping. What a low point for academia.
Or are you perhaps sublimating your own AI anxiety into confrontational assertions that other people who discuss AI aren't as AI-pilled as you?
If the advisory group says something like "AI is bad and nobody should use it in math", I'm pretty confident OpenAI will ignore them.
"This is changing the power we have, let's create a new power structure where we're still at the top, and in control."
No surprise here.
well terry tao no longer has this problem ;)
However, even if they are successful in removing math from the PR toolkit of AI companies, I can’t see it more than a short term band aid. If AI progression continues, the announcements will just transition from big flashy one, to a kid using Claude on their laptop.
I don’t have a solution here, and I’m worried it’s coming to all human fields. So it’s more of an ability to observe what might happen to all of us.
That announcement is about exactly this group: https://openai.com/index/advisory-group-on-mathematics-and-a...
This group has been formed to tell OpenAI whether or not (and if yes how) they should release those 100+ problems.
IMO “AI companies are using math to as a PR tool” is not quite it; they are using math problems as a benchmark to evaluate their AIs, but after the prior blowback from the mathematical community, they are happy to sit on the results until the community (this group) tells them how to publish the results.
And as you say, even if they don't publish the results eventually anyone will be able to do this when this model is released, so it's just a matter of time.
You can also look (as I did back in 2018 when contemplating the end of math) at how chess or Go players adapted mentally to the idea that whatever they may do, a machine could do it better (Lee Sedol famously quit Go).
https://archive.is/m9BYI
In that case, everything stands to be disrupted and those who wield AI will have massive amounts of power in the short term. But this is not a stable situation, and therefore those who wield AI will have to cede control of it or otherwise enter into a great war with those that do not.
I hope for all our sakes that there is a practical ceiling.
Calling it a publicity stunt is an interesting way of framing ground breaking discoveries. The beneficiaries down stream of these discoveries could not care less how they were found.
But they do. Most of higher math has no practical applications and is just a mental playground, philosophy constrained by formal logic and a set of axioms.
Hope is these capabilities will somehow translate to something more practical like physics, chemistry or biology.
Hodge conjecture: 5 Birch and Swinnerton-Dyer Conjecture: 3 Navier-Stokes: 2 P vs NP: 1 Yang-Mills existence and mass gap: 2
So this gives a bit of further evidence that they made progress on the Hodge and Birch and Swinnerton-Dyer conjectures. Maybe also Yang-Mills.
Notably absent is the Reimann hypothesis.
Worthless then isn't it? Next month there could be another 1,000 solutions to handle.
This is more pointless than Facebook's Oversight Board - at least that one Facebook pretends has power.
Incredible international representation...
Advisory Group on Mathematics and Artificial Intelligence (OpenAI post)
https://news.ycombinator.com/item?id=49790389
Why aren't they consulting the younger mathematicians whom this actually affects?
Anyways, this is overblown. Let me know when AI swarms can create a research program or library from scratch. We're really missing long term planning and abstraction creation in these things.