> Under a partnership, the Simons Foundation, Anthropic and OpenAI will commit $15 million in combined funding and resources. The foundation will manage and award the funding, including through an open call for proposals, to support workshops and initiatives that help researchers evaluate recent AI-assisted results, learn and use new tools, share perspectives and shape collective standards for AI use.
So charitable that it launches this initiative with ClosedAI and Misanthropic. There are probably accelerationist and transhumanist nutcases at this foundation.
Can you ever look behind the surface of ... anything at all? Only naive and childlike statements in favor of AI proponents. It is literally like saying: "Prince Andrew and Sarah Ferguson are esteemed members of the magnanimous and generous Royal Family."
This IS me going beyond the surface. The obvious question here is "who is this foundation and are they working with the AI labs". I answered both.
The headline in this piece made it look like maybe a foundation was trying to help mathematician's push back against the unwanted assault on their profession by the AI labs. The section I quoted shows that's not the case.
Also not sure how you interpreted "foundation started by the Renaissance Technology hedge fund billionaire" as a mindless boost.
Respectfully, you didn't answer either. Jim Simons was a brilliant mathematician, has done a lot for math and people involved in math. If you assume they are "pro math and pro math people", the why is simple too.
I assumed Hacker News commenters would find the extra contextual information useful in the same way that I did. I didn't assume they would come to the same conclusions, and I've not come to conclusions myself.
Simons Foundation also significantly funds arXiv, and also a large number of specific grants across maths and sciences. I don't think there's much agenda here.
The foundation was an early backer of machine learning for science, including funding the development of the Lean proof assistant, which has revolutionized how mathematicians work and made AI-assisted proofs possible. Additionally, the foundation has convened workshops, lectures and collaborations that have brought mathematicians together to explore and discuss AI’s role in the field.
Read "the man who solved the market". It's a remarkable story. He was a remarkable dude, extremely smart, has given a lot of money to promote math and science.
So, basically they created the recent situation in math where a great deals of top mathematicians criticized their methods, even called to boycott their dumped results. And the answer of the AI labs is buying these poor academic souls with 15$ millions? Most of them even as AI tokens? Why do I see a macabre irony here?!
The people here who depict him or whoever runs his funds now as a pure philanthropists are wrong. These tiny investments into academia pay off many times.
Genuinely the least they could do for what a pox they are on so many industries, on consumers of electronics, on local municipalities, on the environment...
Wonder if they realize these types of things help grow massive resentment against such companies? I'm astounded by how terrible the politics of SV actually is.
I think a lot about how Terrence Tao and the usual 30 (already well-funded) math guys will receive this cash while us early-career math students are left holding the bag.
It's akin somewhat to "donating 100 dollars to the node JS foundation, not the packages that appear when you run npm fund".
As much as this looks like a math problem, it's actually an everyone problem. Some hands-on (literally) have more way to go, but otherwise this is a "hop" that the entire society needs to cross. Honestly, I don't think there's going to be be any Math 2.0 as there won't be any other 2.0, at least not in any utilitarian way and the effort to find utility where there's none is misguided at best. The analogy I like to make is that with horses trying to find utility y in the Ford era.
It's also not an OpenAI, Anthropic or any specific name caused problem. This type of progress has always been happening, it's more or less coincidental (like Stigler's law). The problem though is we need to rethink how we operate as a society and too few (IMO) are trying to make this conversation happen.
too few are talking about reshaping society because those in power are not letting that happen. the amount of wealth concentrated at the top is staggering and allows them to fortify their position by bribing, buying governments and media.
I love how "society" is investing billions to develop a tech that makes its own citizens obsolete, but doesn't manage to provide basic healthcare, housing or infrastructure...
Obsolete citizens don’t need healthcare, housing or infrastructure. They will just die of hunger and exposure and then the rich and powerful can frolic unencumbered in paradise.
No, it's not an "everyone" problem yet. Each field is likely on its own timeline. Knowing how well AI is at proving theorems is of only limited use for predicting what will happen and when in other fields.
The yet compressing into now is the definition of singularity. I'd argue that it's "only of limited use" if you don't understand the problem well enough - so yes, it's limited for predicting anything for most people but that doesn't make it an irrelevant aspect at predicting in general.
Why bring up the singularity? It hasn't happened and might not happen the way anyone thinks. For example, the timelines in different fields might not all go vertical at the same time. So, you could get something like an Internet singularity without a corresponding robotics singularity. Also, war is different from civilian applications. The use of drones to bomb data centers doesn't imply that drones will be useful for civilian applications like delivering pizza.
There's no particular reason to believe that solving math theorems generalizes to engineering or robotics without a whole lot of additional work.
I see it more like a natural phenomenon close to the reality of some conditions than some surprising phenomenon. There are a few reasons and what we're experiencing are related. One of them is automating reasoning which enables at scale work (reasoning) across domains, regardless. The other is that domains are actually much more interconnected than it would seem looking from outside. While this has been the subject of surprising discoveries for decades (at least) I also don't see how it can't be otherwise. Nature doesn't have different laws based on how we chose to artificially split it in "domains".
> There's no particular reason to believe that solving math theorems generalizes to engineering or robotics without a whole lot of additional work
that "whole lot" is only a lot in human time (the typical "man-month") which is what is being sped up here.
You cannot use mathematics without understanding it.
It doesn't help you that AI understands it, when you don't. You are the one getting screwed over by others who understand statistics better than you do.
When you're rich enough, you can pay others (or AI) to protect you. But then you're dependent upon them.
So, under a transactional model of social relationships, you are in an arms race over raw intelligence and competency/data. With only one winner, who takes all in the end.
Logically, most people have already lost in that game, so why keep playing that nonsense?
I think you can understand* math in general. I think not understanding something it's a matter of presentation. Also I think the presentation can be improved. A good-quality AI can do both, hence it can get a presentation that you can understand.
Yet this has zero utility (other than your own satisfaction). Once you understand something, you will likely not even do a better job at explaining it to someone else, so the utility is strictly limited to yourself.
So even ignoring ability to understand the problem is the cost/efficiency of understanding (and applying it).
I agree with your other statements, but I feel you're suggesting the non-sense is scoped to just math. It's not it's almost everything that we do (exceptions are social interactions including care taking, etc.)
So in this frame your "nonsense" is (almost) everything that we do. Personally I'd like to keep doing things.
You can't use mathematics without understanding it?
Let's say GPT-23 Astra grinds away for a full month one day and out pops a warp drive. Our best physicists ask it to explain. The resulting stream of text is essentially incomprehensible. But, they build a scale model at its direction to test and sure enough it does the Kessel run in 12 parsecs.
Are you saying we shouldn't hop aboard the USS Enterprise and check out Alpha Centauri because we don't understand the math?
I think what's really happening here is that there's an ivory-tower academic class that feels entitled to make their living at public expense thinking about and writing about their topic of interest, and along comes a new tool that's changing that dynamic up. Meanwhile, in the real world, the value remains with people who can take whatever the tool of the day is and do useful things with it. Which, I think, is why despite all the talk the only professional job losses blamed on AI appear to largely be companies that had a failed strategy of some sort or another and used AI as a boogeyman to do the layoff that they needed to do anyway. To people living in the private sector this isn't seen as a dangerous invader, LLM's are productivity multipliers.
>Let's say GPT-23 Astra grinds away for a full month one day and out pops a warp drive.
In the real world, the frontier labs are still hiring lots of humans for business and software roles (https://openai.com/careers/search/) despite being better positioned than anyone to automate them away with their world-beating models and save a bunch of salaries.
That seems wrong unless you have an unusual definition of "understanding":
* Lots of people use math for calculations without having proven any theorems.
* You can understand a mathematical statement without understanding or even having a proof. There are proofs built on the axiom of choice and conjectured results.
That also means that you will need buy-in from a plurality/majority of that society to make that "hop". This is not going to break politically this election cycle, but it will show up. Barring a 2008 level event, it will likely be the issue with the greatest growth heading into 2028, and that will place the impetus on the "specific name"s to communicate how this isn't just trickle down benefits. Especially when it will be an easy sell to people that the "hop" means real people falling into the gap or getting left behind. They need to figure out a way to make the distribution of benefits more immediate and tangible to everyday folks, because the perception of the distributed costs is a very easy sell in this political environment, regardless of the reality. I can't imagine that getting better over the next two years.
Yeah, the tangibility of benefits really need to be felt, especially for the so called "deplorables" who would like to swing for a politician that can demonstrate that they will get them housing, healthcare, prosperity in their communities. High paying jobs across the educational ladder and in rural areas also wouldn't hurt.
Like my other comment, this might be a public-vs-private sector misconception. I can't help but notice my own employers IT and design departments, which both lean pretty right, openly use Codex to help with about everything. One group is even training their own model at the moment. And our marketing team, which is solid blue? They're the only team in the company allowed to use both Claude and ChatGPT because they claimed, repeatedly and loudly, that both were critical to their jobs because Claude was better at design (they say) and Chat for everything else (again, their claim, not mine).
No one wants a data center in their back yard, but in the private sector I think adoption isn't as political as some make it on Hacker News.
Maybe you get conditioned on how the rich usually treats other people, but this view isn't well supported by healthy discussions among the math community wrt recent AI 'disruption'
We're not so far from 'you never had some god given right to be alive, well fed and sheltered. You have a good body, make yourself useful in some industry.'
Math proofs are right in the sweet spot for training LLMs with RLVR. It isn't some general "LLMs are getting insanely smart!". Even most software development itself is miles away from being as well suited to training as math proofs are.
The majority of math research is akin to professional chess. Completely useless and the folks doing it are using their giant brains to play a game that no one outside themselves benefits from.
For what it's worth, most mathematicians are paid to teach, and the amount of research funding going into mathematics is relatively small. Plus, there are rare but important broader advancements due to pure and applied math that are hard to predict in advance (e.g., cryptography, compressed sensing).
Mathematicians also decide what to prove in their research - taste is not a skill we have RLVR'd away (yet, not claiming this is fundamental).
Of course, the field is being fundamentally transformed, it will probably end up looking more like experimental science in the future.
Do you think it’s possible to operate math research in a fully autonomous way (i.e. with no human expert involvement)?
As a software engineer, my mental model so far is that OpenAI made a gigantic (but potentially sound) pull request to the math community, and the mathematicians just did not have enough bandwidth to review and merge it.
If my model is correct, I guess it makes a lot of sense to fund mathematicians, because their lack of resource is the bottleneck here.
That works, as long as the math is not beyond the mathematicians. Besides that, since you make a pull request analogy: not all pull requests survive review, but generating them like this is cheap. Mathematicians are unlikely to submit pull requests that will be refused because they have a stake in the game: if they do so too often then their papers will no longer be reviewed as eagerly as when they were not submitting bs. The AI peddlers have no such limitations and will happily flood the receivers. As a result actual mathematicians papers run the risk of no longer being given enough consideration because there is this bottleneck you identified. So what will happen instead is exactly the same as what is happening in the world of code: they will engage AI to help with the reviewing. Before long the foundation of math will no longer be understandable by humans. This would be a loss imo.
Reading your comment, I’m most curious if AI models will eventually “fork” mathematics. (Sorry for beating the same cranky analogy)
Like, not expecting human experts to review and merge their results, but developing math branches on their own. I guess it’s at least doable for applied math. For example, if AI agents find a fast method of prime factorization, they can demonstrate it by cracking RSA encryption.
I’m less sure how it plays out for pure math, though. My intuition is fairly limited here.
The bulk of the people on this planet do not make their money in advanced math but in mundane jobs that are well within the capabilities of AI as it is today. It will take some time to figure out how to put it all together but all of those jobs are on borrowed time unless there is some kind of regulatory action to stop that from happening.
Perhaps. But as someone who uses AI to automate something that seems in the bucket of things you are talking about - it's harder than it seems, the edge cases are endless, claude code is worse at building real systems than people think and humans are doing a lot more than people realize once you start trying to automate.
All I'm saying is the math advances of late really hasn't changed much outside of math as far as AI is concerned.
Agreed, in the meantime, try finding work as (a) a translator (b) a junior programmer (c) a copywriter (d) a proofreader. And so on. There are so many job categories that are already feeling the impact of AI that I think it is wrong to focus on the plight of the mathematicians, I think their jobs are - for the most part anyway - still safe. It's the low end of the information processing where the real damage is.
Sure. I guess the reason some people don't feel so bad for the math folks is the people in those careers you highlight tend to not have the asset a math researcher has - a giant brain that is valuable outside their little world of math proofs.
Like bankruptcy, automation will happen gradually and then suddenly. This time last year LLMs hadn't solved a single open mathematics conjecture of note. Within months they solved a millennium prize problem.
Robotics is likely to follow a similar trajectory. You can say that robotics is not clinical like mathematics is, yet it kind of is, you can create simulated training environments of any complexity.
It's a bad analogy. Math proofs lend themselves to just throwing more compute at the problem both in training and at inference, you can verify when you are done. Most tasks don't lend themselves well to training or inference.
Robotics isn't really a coherent grouping. Some things that would be termed "robotics" are already "solved", other things aren't even close and need actuating and sensing capabilities that are miles off or so cost prohibitive that it doesn't make any sense.
I'm curious what you as an early-career math student would want the Simon Foundation to do differently from what it says it will do. I hope there is a way for students like to you to get a slice of the pie.
That's like saying "if you want to work as a chef, DoorDash delivery offers the ability to order lots of meals". The new jobs are both fewer and shittier.
I hope the future you built for others is imposed on you too.
>I think a lot about how Terrence Tao and the usual 30 (already well-funded) math guys will receive this cash while us early-career math students are left holding the bag.
Never understood this sentiment. Theses top researchers could do almost anything else with their skills outside Academia and make far far more money. You don't get rich off grants... They could all presumably walk into Anthropic or OpenAI for near $1M a year in TC.
If you yourself are as cracked as you think you could also do the same...
Isn't that basically all of AI for white collar workers?
I haven't escape note that I can do tons of things with AI because I spent the last 10 years building a bunch of side projects and now I can complete those side projects because I know all the magic words.
Gonna need to unionize because the gerontology is just getting stronger.
> > Under a partnership, the Simons Foundation, Anthropic and OpenAI will commit $15 million in combined funding and resources.
OAI and Ant got blowback for essentially strip mining mathematical discoveries into PR wins for their models, accidentally laying waste to the research field in process. Throwing ~$5M each into PR whitewashing (partially in token credits), seems self-serving.
Kudos to the Simons Foundation for seizing the moment to offer the labs a fig leaf of respectability. Might have been the fastest "Yes" on a grant proposal ever.
OpenAI's market cap or valuation estimate went up by about 230 billion dollars within a few weeks after announcing that they solved NS. It can't be known how much of that value change was because of that announcement, but must be at least a good chunk of it.
I’m in the funding markets around this stuff. Nobody took this into account. It’s more for PR with government and potential hires (and maybe some corporate marketing).
Softbank's rally began in Japanese markets on 3 September and American markets on 4 September. OpenAI announced on 8 September. When OpenAI delayed its IPO plans on 13 September, Softbank sank back to its 7 September level [1].
>... essentially strip mining mathematical discoveries into PR wins for their models, accidentally laying waste to the research field in process.
Jeeze, that's an almost impressively charged way of saying "private company performing SOTA research on open problems that anyone has been welcomed to approach for over a century" lol
I think it does a disservice to the dialogue to strip away any nuance here and just paint these companies as evil for doing what everyone has been anticipating companies like them doing well before LLMs even existed. It's clearly just trying to fit these events into a specific narrative which does nothing to actually address the inevitability of everything that comes next. Feel free to pressure OpenAI and Anthropic to stop performing math research so that it can just be picked up by whoever catches up and doesn't care about bad PR. I'm sure Musk with Grok will be extra considerate of pure math researchers when he sees an opening to pick up those scraps lol
how is that different than what's been happening forever? Mathematics has ALWAYS been a "charity". The only thing that changes over time is who gives the money, first it's the church, then it's princes, then government, it's been corporations for a while, this is nothing new.
Except that even the Cosa Nostra is less anti-human than the current tech bros and their foundation fronts.
You really cannot see any difference between the government and corporations supporting something? This "things have always been that way" talking point is getting out of hand.
Highly doubt it. Remember how cursor would never save or train or customer’s sessions and then got acquired by SpaceX because of their large coding training dataset ?
Aside from the fact that $15M is couch change for these companies, and aside from the fact that this project is hypocritical and self-serving (as explained by many other posts), the whole idea of naming a supposedly large sum of money in a headline in order to impress is just silly. Whether a project is successful or worthwhile has little to do with how much funding may be lined up, and throwing numbers around is often just a smokescreen for poor planning and execution.
I really can't empathize with treating mathematicians as some sort of AI refugees who need coddling and support. AI is a tool, and competent mathematicians should learn to use it like everyone else. I find it hard to believe that AI doesn't represent a net potential benefit to human understanding of mathematics. Yes it can be misused, but any tool can be misused. Just don't misuse it...
Uhm...I fear you may have made the mistake of just reading the comments here rather than the article. Normally that works out OK but this thread has attracted the highest density of idiots I've ever seen on an HN thread.
The point of the initiative in the article is to help do the things you say mathematicians should be doing. Mathematicians recognize AI has reached the point that it can make significant advances, and are now trying to figure out how to most effectively use it.
Because it's not for the benefits of those receiving the funding. It's money to pay people willing to improve how AI does math so AI companies can be even more effective and rendering human endeavors unprofitable. Math is a hot target right now and they're going after it.
It's completely selfish and won't help anyone except those at the top.
So they want mathematicians to keep prompting and giving feedback to the models for clarifications on their proofs. Then write it all up and publish it in a journal and mention their AI usage. Then suddenly the models end up producing readable and understandable proofs through the "emergence" of the skills because of the scaling laws.
Is mathematics deprived of the associated knowledge that comes from a human first deriving a proof, how exploring the problem domain including the dead-ends has changed the mathematician's thinking, when a proof is given by AI? It seems that the solving rather than just the solution gave much to mathematics.
Can you explain to me what "navigation" means if AI actually succeeds at automating mathematics? Or is the bet here that the AI companies will fail to deliver?
In chats with models like deepseek the LLM makes many basic math wrong claims. Maybe more advanced models can in special circumstances, guided by the work of humans and Lean, prove theorems. But just like in RL, the brute force approach of failing a million times may not be enough when the reward is far.
In math the reward is finding the decomposition that reduces the problem to simpler cases, and in many cases that reward is very far. Furthermore, the LLM may process the many cases fast, but the decomposition often comes from a human, and the model takes the credit for the proof.
This foundation could provide the LLMs with the decompositions, generated by human ingenuity, that they need. But once supplied, the model can run the cases and get the credit. Hence more credits for LLMs and fewer jobs and worse prospects for humans, even if it isn't an intended consequence.
Nearly everyone rejected the idea of a CERN-like project for mathematics when it was recently proposed. It should have been more seriously considered.
Now that things are settling a bit after the Navier-Stokes and theorem-dump shocks, something like what mathematics will look like post-AI is starting to emerge. Of course, things should be conditioned on the assumption that capabilities will continue to improve, but imagine something like this--
An institute with a tremendous amount of compute, either rented or owned, built on the Humboldtian model of the unity of teaching and research. The main draw for big names would be the access to compute, free of the fear of being scooped. New forms of attribution and collaboration would develop, and we'd gain some clarity for students and PhDs on what a future career looks like.
The current melange of math university departments all following their own paths through the crisis without consolidating resources makes the transition more difficult. Think of the path physics took beginning with scattered cyclotrons (the basement of the Niels Bohr institute was even dug out to put one in), then the larger labs that followed. That transition was gradual because physics led it from within, whereas with mathematics every university department is suddenly at a massive compute disadvantage with respect to labs that aren't following the same model, and are driven by different incentives.
It could be this simple--maybe we don't need to completely throw away the old model, attribution practices, and so on, and maybe we can preserve the transfer of knowledge to the next generation of mathematicians, if we simply give them lots of compute.
An analogy to physics to highlight how the divergent interests of the labs could be managed in a state of "peaceful coexistence" might be: physicists got well-funded accelerators and labs that followed standard academic practices. These were separate from "war work" (Los Alamos) that followed government priorities, even with the crossover of personnel. You could have the same thing with a "math accelerator" working on the academic model, while the AI labs follow their own priorities, and some subset of people cross between.
It might be objected that this is merely a conservative attempt to preserve the existing academic conventions and practices currently under threat, but tenure itself already does that, and is arguably the very thing that exacerbates the problem for young mathematicians.
These days I use AI for development and it feels like my coding skills are getting worse, so I sometimes need rehab, and I'm a bit worried that mathematicians will go through the same thing. Later, since the proof structure is built by AI, I think there will be cases where mathematicians themselves can't touch it (because if they do, they can't tell where and how it will break).
Truth about the Simons Foundation (it was an early AI booster) is flagged here. Jim Simons gets a hagiography by a couple of shills. A group of utterly vile people frequents this site.
111 comments
[ 1.2 ms ] story [ 47.1 ms ] thread[1] https://www.simonsfoundation.org/news/from-trust-to-verifica...
[2] https://lean-lang.org/fro/
Quanta magazine is the propaganda arm.
The headline in this piece made it look like maybe a foundation was trying to help mathematician's push back against the unwanted assault on their profession by the AI labs. The section I quoted shows that's not the case.
Also not sure how you interpreted "foundation started by the Renaissance Technology hedge fund billionaire" as a mindless boost.
The foundation was an early backer of machine learning for science, including funding the development of the Lean proof assistant, which has revolutionized how mathematicians work and made AI-assisted proofs possible. Additionally, the foundation has convened workshops, lectures and collaborations that have brought mathematicians together to explore and discuss AI’s role in the field.
https://www.simonsfoundation.org/flatiron/
He has invested in AI for a long time:
https://www.wealthmanagement.com/high-net-worth/family-offic...
The people here who depict him or whoever runs his funds now as a pure philanthropists are wrong. These tiny investments into academia pay off many times.
How shameless can you be?
It's akin somewhat to "donating 100 dollars to the node JS foundation, not the packages that appear when you run npm fund".
It's also not an OpenAI, Anthropic or any specific name caused problem. This type of progress has always been happening, it's more or less coincidental (like Stigler's law). The problem though is we need to rethink how we operate as a society and too few (IMO) are trying to make this conversation happen.
There's no particular reason to believe that solving math theorems generalizes to engineering or robotics without a whole lot of additional work.
> There's no particular reason to believe that solving math theorems generalizes to engineering or robotics without a whole lot of additional work
that "whole lot" is only a lot in human time (the typical "man-month") which is what is being sped up here.
It doesn't help you that AI understands it, when you don't. You are the one getting screwed over by others who understand statistics better than you do.
When you're rich enough, you can pay others (or AI) to protect you. But then you're dependent upon them.
So, under a transactional model of social relationships, you are in an arms race over raw intelligence and competency/data. With only one winner, who takes all in the end.
Logically, most people have already lost in that game, so why keep playing that nonsense?
Yet this has zero utility (other than your own satisfaction). Once you understand something, you will likely not even do a better job at explaining it to someone else, so the utility is strictly limited to yourself.
So even ignoring ability to understand the problem is the cost/efficiency of understanding (and applying it).
I agree with your other statements, but I feel you're suggesting the non-sense is scoped to just math. It's not it's almost everything that we do (exceptions are social interactions including care taking, etc.)
So in this frame your "nonsense" is (almost) everything that we do. Personally I'd like to keep doing things.
John von Neumann
Let's say GPT-23 Astra grinds away for a full month one day and out pops a warp drive. Our best physicists ask it to explain. The resulting stream of text is essentially incomprehensible. But, they build a scale model at its direction to test and sure enough it does the Kessel run in 12 parsecs.
Are you saying we shouldn't hop aboard the USS Enterprise and check out Alpha Centauri because we don't understand the math?
I think what's really happening here is that there's an ivory-tower academic class that feels entitled to make their living at public expense thinking about and writing about their topic of interest, and along comes a new tool that's changing that dynamic up. Meanwhile, in the real world, the value remains with people who can take whatever the tool of the day is and do useful things with it. Which, I think, is why despite all the talk the only professional job losses blamed on AI appear to largely be companies that had a failed strategy of some sort or another and used AI as a boogeyman to do the layoff that they needed to do anyway. To people living in the private sector this isn't seen as a dangerous invader, LLM's are productivity multipliers.
In the real world, the frontier labs are still hiring lots of humans for business and software roles (https://openai.com/careers/search/) despite being better positioned than anyone to automate them away with their world-beating models and save a bunch of salaries.
* Lots of people use math for calculations without having proven any theorems.
* You can understand a mathematical statement without understanding or even having a proof. There are proofs built on the axiom of choice and conjectured results.
No one wants a data center in their back yard, but in the private sector I think adoption isn't as political as some make it on Hacker News.
Math proofs are right in the sweet spot for training LLMs with RLVR. It isn't some general "LLMs are getting insanely smart!". Even most software development itself is miles away from being as well suited to training as math proofs are.
The majority of math research is akin to professional chess. Completely useless and the folks doing it are using their giant brains to play a game that no one outside themselves benefits from.
Mathematicians also decide what to prove in their research - taste is not a skill we have RLVR'd away (yet, not claiming this is fundamental).
Of course, the field is being fundamentally transformed, it will probably end up looking more like experimental science in the future.
As a software engineer, my mental model so far is that OpenAI made a gigantic (but potentially sound) pull request to the math community, and the mathematicians just did not have enough bandwidth to review and merge it.
If my model is correct, I guess it makes a lot of sense to fund mathematicians, because their lack of resource is the bottleneck here.
Like, not expecting human experts to review and merge their results, but developing math branches on their own. I guess it’s at least doable for applied math. For example, if AI agents find a fast method of prime factorization, they can demonstrate it by cracking RSA encryption.
I’m less sure how it plays out for pure math, though. My intuition is fairly limited here.
All I'm saying is the math advances of late really hasn't changed much outside of math as far as AI is concerned.
Robotics is likely to follow a similar trajectory. You can say that robotics is not clinical like mathematics is, yet it kind of is, you can create simulated training environments of any complexity.
Robotics isn't really a coherent grouping. Some things that would be termed "robotics" are already "solved", other things aren't even close and need actuating and sensing capabilities that are miles off or so cost prohibitive that it doesn't make any sense.
Even if you want to work in research, AI offers a lot of opportunities to do even more research, and less grunt work
I hope the future you built for others is imposed on you too.
Never understood this sentiment. Theses top researchers could do almost anything else with their skills outside Academia and make far far more money. You don't get rich off grants... They could all presumably walk into Anthropic or OpenAI for near $1M a year in TC.
If you yourself are as cracked as you think you could also do the same...
I haven't escape note that I can do tons of things with AI because I spent the last 10 years building a bunch of side projects and now I can complete those side projects because I know all the magic words.
Gonna need to unionize because the gerontology is just getting stronger.
Various historians agree with the idea credentialism is largely just a tool of aristocracy to manipulate the public.
OAI and Ant got blowback for essentially strip mining mathematical discoveries into PR wins for their models, accidentally laying waste to the research field in process. Throwing ~$5M each into PR whitewashing (partially in token credits), seems self-serving.
Kudos to the Simons Foundation for seizing the moment to offer the labs a fig leaf of respectability. Might have been the fastest "Yes" on a grant proposal ever.
I’m in the funding markets around this stuff. Nobody took this into account. It’s more for PR with government and potential hires (and maybe some corporate marketing).
Softbank's rally began in Japanese markets on 3 September and American markets on 4 September. OpenAI announced on 8 September. When OpenAI delayed its IPO plans on 13 September, Softbank sank back to its 7 September level [1].
[1] https://asia.nikkei.com/business/markets/softbank-group-plun...
Jeeze, that's an almost impressively charged way of saying "private company performing SOTA research on open problems that anyone has been welcomed to approach for over a century" lol
I think it does a disservice to the dialogue to strip away any nuance here and just paint these companies as evil for doing what everyone has been anticipating companies like them doing well before LLMs even existed. It's clearly just trying to fit these events into a specific narrative which does nothing to actually address the inevitability of everything that comes next. Feel free to pressure OpenAI and Anthropic to stop performing math research so that it can just be picked up by whoever catches up and doesn't care about bad PR. I'm sure Musk with Grok will be extra considerate of pure math researchers when he sees an opening to pick up those scraps lol
You really cannot see any difference between the government and corporations supporting something? This "things have always been that way" talking point is getting out of hand.
On ChatGPT: Settings -> Data controls -> Turn off “Improve model for everyone”
On Claude: Settings -> Privacy -> Turn off “Help improve our AI models”
They can scrape your work after you publish it.
The point of the initiative in the article is to help do the things you say mathematicians should be doing. Mathematicians recognize AI has reached the point that it can make significant advances, and are now trying to figure out how to most effectively use it.
(Yes I know it's not the group directly receiving the founding)
He was also a major backer of autism research, because his family member had it. People spend money to where their passion is.
It's completely selfish and won't help anyone except those at the top.
The whole discourse of past few weeks has been mathematicians claiming the opposite.
In math the reward is finding the decomposition that reduces the problem to simpler cases, and in many cases that reward is very far. Furthermore, the LLM may process the many cases fast, but the decomposition often comes from a human, and the model takes the credit for the proof.
This foundation could provide the LLMs with the decompositions, generated by human ingenuity, that they need. But once supplied, the model can run the cases and get the credit. Hence more credits for LLMs and fewer jobs and worse prospects for humans, even if it isn't an intended consequence.
Now that things are settling a bit after the Navier-Stokes and theorem-dump shocks, something like what mathematics will look like post-AI is starting to emerge. Of course, things should be conditioned on the assumption that capabilities will continue to improve, but imagine something like this--
An institute with a tremendous amount of compute, either rented or owned, built on the Humboldtian model of the unity of teaching and research. The main draw for big names would be the access to compute, free of the fear of being scooped. New forms of attribution and collaboration would develop, and we'd gain some clarity for students and PhDs on what a future career looks like.
The current melange of math university departments all following their own paths through the crisis without consolidating resources makes the transition more difficult. Think of the path physics took beginning with scattered cyclotrons (the basement of the Niels Bohr institute was even dug out to put one in), then the larger labs that followed. That transition was gradual because physics led it from within, whereas with mathematics every university department is suddenly at a massive compute disadvantage with respect to labs that aren't following the same model, and are driven by different incentives.
It could be this simple--maybe we don't need to completely throw away the old model, attribution practices, and so on, and maybe we can preserve the transfer of knowledge to the next generation of mathematicians, if we simply give them lots of compute.
An analogy to physics to highlight how the divergent interests of the labs could be managed in a state of "peaceful coexistence" might be: physicists got well-funded accelerators and labs that followed standard academic practices. These were separate from "war work" (Los Alamos) that followed government priorities, even with the crossover of personnel. You could have the same thing with a "math accelerator" working on the academic model, while the AI labs follow their own priorities, and some subset of people cross between.
It might be objected that this is merely a conservative attempt to preserve the existing academic conventions and practices currently under threat, but tenure itself already does that, and is arguably the very thing that exacerbates the problem for young mathematicians.
Which is indirectly forcing the mathematicians to use AI?