AI/LLMs Navier-Stokes "Solved" by AI

[Umbran]

Perhaps you missed my previous post on the subject.

If I missed it, I obviously wouldn't know if I had or not.

But... I was agreeing with you, merely expanding with a reason. I'm sorry if that offended.

Please, either place me on two way ignore or stop responding to my posts. (Or help me place you on my ignore list, which I understand is not possible?)

I cannot ignore anyone. I'll do my best to abide by your request.

 

[Stalker0]

Perhaps you missed my previous post on the subject. Please, either place me on two way ignore or stop responding to my posts. (Or help me place you on my ignore list, which I understand is not possible?)

You can choose not to respond to someone. but with respect this notion of "stop reading and responding to my posts" is not how productive debate is conducted.

 

[Umbran]

You can choose not to respond to someone. but with respect this notion of "stop reading and responding to my posts" is not how productive debate is conducted.

Let's please not argue over this.

 

[Justice and Rule]

And in such cases, at least with US criminal courts, we lean on the concept of “innocent until proven guilty”

I'm not a court, nor is the "court of public opinion". The reason you are "innocent until proven guilty" in a court of law is that it can actually adversely affect you beyond your reputation: they can sanction you, fine you, arrest you, etc. The burden is much higher than because the effect is much deeper.

I don't need "innocent until proven guilty" to understand when someone is a liar: I can put together facts and simply observe their previous actions. Since I'm not a court and can't out a binding judgement on them, I can use a lesser standard... though that doesn't necessarily make it any less valid. While someone might find a loophole around a law, it doesn't mean I need not consider them a cheat, nor would that loophole invalidate that opinion.

Yeah, I think these are well-articulated reasons to be distrustful of OpenAI. That said, for me it isn't enough to get from those reasons to the belief that they are lying about reading the prompts. I could be wrong, but that in particular feels to me like a step further, and the pace of their response makes sense for me.

And gosh, I hesitate to bring this up, because I think it is unfair to Buckmaster. But it is worth noting that Kasparov's response on losing to Deep Blue was to accuse IBM of cheating. There is a sort of precedent, resonance here.

Sure, sure. But yeah, I dunno if that's the case. OpenAI has started to do this sort of "academic claim jumping" to try and build hype more recently, and this isn't the only instance of this. This blog post points out one that I hadn't even heard of, but probably for obvious reasons: instead of trying to fight it, this professor simply published early and incomplete:

This isn’t even the first time they’ve done this this month: they attempted to do the same thing to Dr. Julia Stadlmann, when it became clear that her work on bounded gaps between primes was nearing completion. Every analytic number theorist on the planet knows she was working on this problem, and many of them have a rough sense of the way she planned to proceed.

• 08/31: her preprint is released, and it contains the following ominous line: “As such, we believe that our method could achieve bigger improvements, but due to limited resources and limited time, we have decided to present this modest improvement as a proof of concept.”
• 09/03: OpenAI releases a preprint, following roughly her methods, lowering the bound to 186.

That sort of stuff, to me, indicates a concerted effort to overshadow academics in the field to get PR. We are only really hearing about it because Buckmaster actually stood his ground, rather than either acquiescing as some have or rushing as others. That blog post is also really good about the destructive nature of these sorts of practices on academics and the understanding of math as a whole, if anyone is interested in reading the full blog post (which I would recommend).

I also think that it'd be weird to think that the AI companies, which are desperate for PR wins to hide how bad the economics of this sort of technology currently is, would hold back any sort of resource or trick in their repertoire. I just don't see how suddenly they'd have an ethical line-in-the-sand about using their own information, and the opacity that they've created around how these things are done means that they can give denials with no actual proof that we can check ourselves; we are forced to take them at their word or not.

So it's certainly possible that they do have some sense of scruples or that they have some lines they won't cross, but I'm not really convinced of it given the "grey at best, incredibly unethical at worst" nature of academic compute achievements.

 

[Stalker0]

I'm not a court, nor is the "court of public opinion". The reason you are "innocent until proven guilty" in a court of law is that it can actually adversely affect you beyond your reputation: they can sanction you, fine you, arrest you, etc. The burden is much higher than because the effect is much deeper.

I don't need "innocent until proven guilty" to understand when someone is a liar: I can put together facts and simply observe their previous actions. Since I'm not a court and can't out a binding judgement on them, I can use a lesser standard... though that doesn't necessarily make it any less valid. While someone might find a loophole around a law, it doesn't mean I need not consider them a cheat, nor would that loophole invalidate that opinion.

Mostly off topic but you raise the interesting point of "the burden is much higher because the effect is much deeper".


Could it be argued that in modern times because a hit to ones reputation can have such a castastrophic impact to your success online (with the rise of cancel culture and the like)....that the burden of even criminal matters is not actually that much higher?

Anyone just kind of a bemusing thought, not really relevant to our discussion here.

 

[The Firebird]

That sort of stuff, to me, indicates a concerted effort to overshadow academics in the field to get PR. We are only really hearing about it because Buckmaster actually stood his ground, rather than either acquiescing as some have or rushing as others. That blog post is also really good about the destructive nature of these sorts of practices on academics and the understanding of math as a whole, if anyone is interested in reading the full blog post (which I would recommend)..

All reasonable, and I agree with this part of your post. The main reason I don't think they would steal prompts or chats is because it is a much worse PR disaster than say, playing pretty rough with academia. Throwing compute at open problems and solving them shows AI Investment is necessary to stay competitive. But if they can't be trusted with data, they lose enterprise and lose academia and there is not much left. Of course, they could choose to run that risk and lie. I don't see their situation as so desperate at the moment, and the solution makes sense given the progress that has been made.

They announced a week ago significant progress on another millennium problem. It will be interesting to see how/what that looks like. Rumor has it Anthropic has major progress on a third.

You might be interested in this article from a math professor who is working through the effects on education.

 

[Cadence]

Sure, sure. But yeah, I dunno if that's the case. OpenAI has started to do this sort of "academic claim jumping" to try and build hype more recently, and this isn't the only instance of this. This blog post points out one that I hadn't even heard of, but probably for obvious reasons: instead of trying to fight it, this professor simply published early and incomplete:

A quote from the blog:

"To those who still seem to think this is about mathematics: wake up please? The AI-in-mathematics boom has never been about you, about us; nor indeed is it about mathematics. These companies do not give a flying rat’s ass about anything we value. They never have, they never will. We are pawns for OpenAI and Anthropic in their game of mutual extermination, nothing more. This technology will do its best to erase a 2500 year tradition of mathematical study, all for the sake of winning the race towards monopoly."

 

[Zardu]

Not only are the (privately owned) AI companies attacking the creative industry but they're now expanding the assault onto academia. This represents an existential threat to public human thought and expression - and its an entirely deliberate one. That the AI-boosters here are now weaponising the ignore function to silence engagement with and direct criticism of their arguments (look up what Sabine Hoffenfeleer, cited earlier, had to say about the diving-to-suicide of Jason Arday for example, as well as her attacks on academic rigour) and it makes the coming impacts painfully clear to me, at least.

If ENWorld is to retain any value at all, it has to ban AI discussion, otherwise it'll just be one of the later dominoes to fall. It's deeply depressing, honestly.

 

[Justice and Rule]

All reasonable, and I agree with this part of your post. The main reason I don't think they would steal prompts or chats is because it is a much worse PR disaster than say, playing pretty rough with academia. Throwing compute at open problems and solving them shows AI Investment is necessary to stay competitive. But if they can't be trusted with data, they lose enterprise and lose academia and there is not much left. Of course, they could choose to run that risk and lie. I don't see their situation as so desperate at the moment, and the solution makes sense given the progress that has been made.

They announced a week ago significant progress on another millennium problem. It will be interesting to see how/what that looks like. Rumor has it Anthropic has major progress on a third.

You might be interested in this article from a math professor who is working through the effects on education.

It's really hard to know because we just have so little information on how any of this actually operates. There is danger in lying, but since all this is basically a black box to anyone outside the company, how could you know they were lying? They can just say "Oh, no we didn't look at anything, this is just the power of our compute!", after all. It becomes a "He said, she said", and I suspect a lot of people before Buckmaster were not willing to get into that sort of conflict; most people are conflict-averse and just don't want to have that sort of big public row with a massive company with billions of dollars and a massive PR department.

Which is what makes this specific instance unique and telling to me. I think OpenAI would be more likely to do what they wanted and think that they are so big that anything tossed their way would just roll off. Instead, they found a guy who is dedicated and tenured enough to fight, and then Bubeck tried using the same sort of business-style polite aggression and veiled threats (as the conversation seems to have gone) that you get out of these sorts of situations. Given that they lied in private and then got silent when challenged on if they had used his work (Which I think is something that they had not encountered yet and spooked them), it just pushes me to think worse.

And again, I know this is my opinion, my vibe, my feelings. But there is only one side that could actually show me wrong and really enlighten all of us, and that's OpenAI. Instead, they've preferred to keep things vague and handwave-y, despite the backlash and damage we are seeing in academia. I'm just inclined to call a spade a spade and believe a desperate, unethical, hyper-opportunistic company will act like a desperate, unethical, hyper-opportunistic company. I just don't see a reason why OpenAI, of all the AI companies (save Grok, obviously), would have a limit when they can muddy the waters on any aspersions that are cast against them.

 

[The Firebird]

A quote from the blog:

"To those who still seem to think this is about mathematics: wake up please? The AI-in-mathematics boom has never been about you, about us; nor indeed is it about mathematics. These companies do not give a flying rat’s ass about anything we value. They never have, they never will. We are pawns for OpenAI and Anthropic in their game of mutual extermination, nothing more. This technology will do its best to erase a 2500 year tradition of mathematical study, all for the sake of winning the race towards monopoly."

I can appreciate the authors feelings. That said, Alpoge works for Anthropic, as does John Jumper (nobel, Alphafold). Jacob Tsimerman (Fields medal, 2026) is now at OpenAI. Dario Amodei was a scientist. Just to pick some prominent examples. The labs are full of very accomplished mathematicians and scientists down the line.

I can see an argument that like, the logic of competition means these companies act as impersonal entities that cannot care for the things we value and will necessarily oppose those things. That seems defensible. But I think many people at the companies care, deeply, about those things. With perhaps a difference of opinion on how to achieve those goals.