(untitled)

OpenAI (allegedly) used AI to solve Navier-Stokes. https://openai.com/index/navier-stokes-solution/ But there is great reason to believe there might have been foul play: https://cims.nyu.edu/%7Etristanb/statement.pdf I'm a recent math graduate, and I'm so angry that math has been reduced to a publicity stunt for these awful, unethical corporations. My only consolation is that maybe, just maybe they're getting desperate claiming AGI on Astra and now this. fuarrrk them. I admit I used AI before because I thought it was interesting technology but it's in the worst hands possible. They will never have my money, data or support ever again. fuarrrk the spirits that live in the computer. I hope this bubble crash takes us back to the iron age.

>>94169 Cry all you want, you either live with it or kill yourself.

>>94170 counterpoint: it is the most unpopular technological shift to date and could yet collapse. you are thus invited to kill your self.

>>94169 >I hope this bubble crash takes us back to the iron age. Hell yeah, lain. That's the spirit.

I don't even care if the LLM did math, it's actually cool. But when I read results like this, I always pick up the subtext: "we don't need you any more, you will not be allowed to participate in the world we are building". >>94170 It is you who should kill yourself, barking dog.

>>94171 >>94173 You should have used an AI to summarize my single sentence. I did not tell to kill yourself, but told you that it is one of the two way to handle your future. >unpopular technological shift to date and could yet collapse >I hope this bubble crash takes us back to the iron age. Keep living in your delusions of getting back the old World, it never happened and never will.

>>94174 >I did not tell to kill yourself, but told you that it is one of the two way to handle your future. Sure, and I responded by saying you should kill yourself. It's the best way to handle your future, or lack thereof. Why are you still talking, dog?

The whole point of ai is to eliminate the participation of humanity from civilization. It doesn't matter if it's real or not, it doesn't matter if it doesn't work. It's all an excuse to remove all individual agency & autonomy.

>>94169 > OpenAI (allegedly) used AI to solve Navier-Stokes. And they did it because of the rumour that Anthropic did it, and in the end, they solved an inferior form ("unforced Euler") of it compared to Anthropic: > Our effort began on September 1st after hearing a rumor which we later realized was related to Levent Alpöge, an Anthropic employee, and Tristan Buckmaster, a math professor at NYU. After the completion of our full project and Lean verification (on September 6th), believing from the rumor they also had a solution of Navier–Stokes, we reached out to them to offer a concurrent release of our result and to recognize their priority in a joint announcement. At that point we found out that they had a resolution of the forced Euler problem. In these discussions we offered them visibility into all of the prompts we used and later to see the proof. We recognize the priority of their work on forced Euler and congratulate them on their remarkable mathematical achievement. > We (the researchers and the agents) did not see any of their work through any means until they released it publicly — in particular, no specific user data was accessed in order to solve this problem. While unlikely, we cannot rule out that de-identified data derived from their usage of our products helped improve our models⁠. However, our proofs differ significantly and even the precise results proved are different in the Euler case (forced vs unforced). Also, not so long ago Claude disproved Jacobian conjecture which was also an open math problem: https://en.wikipedia.org/wiki/Jacobian_conjecture > On July 19, 2026, Levent Alpöge presented an explicit counterexample in three variables which he credited to Claude Fable 5, which disproves the conjecture for ⁠ n > 2. The correctness of the counterexample is easy to verify with any computer algebra system. It has not been revealed, however, how it was found. Nevertheless, it led some mathematicians to elaborate on the mathematical reasons and the implications of the existence of the counterexample.

>>94170 why are those two the only options on the table when existence has an abundance of choices and possibilities to choose from? >>94177 >The whole point of ai is to eliminate the participation of humanity from civilization. True, and once you eliminate humanity from the equation, civilization ceases to be a civilization and becomes ever more like a machine. No point to challenging oneself either mentally or physically, or to experience life when the clankers do it for you.

>>94175 You are being really sensitive. Its ok, if I was being replaced by refusing to use a technology (lmao). B-but, the evil technology!!! its so mean!. >No point to challenging oneself either mentally or physically, or to experience life when the clankers do it for you. What a loser.

>>94178 >And they did it because of the rumour that Anthropic did it correction: a guy who works at Anthropic was collaborating on a personal level with a university professor, who is the corresponding author on their papers. They used both Claude and OpenAI tools, and then only publicly available ones. Portraying this as "Anthropic vs. OpenAI" is just very inaccurate. >and in the end, they solved an inferior form ("unforced Euler") of it compared to Anthropic Unforced is more interesting to me, since it implies you don't need to contribute energy to the system to cause a finite-time blow up. But I'm not a specialist in that area. Both teams only have AI slop versions of their Euler papers out because they were racing for it, so I can't say which one is more aesthetically pleasing. >>94180 Quit barking, dog. Pretending to be ironically detached from your masters won't save you.

>>94180 >What a loser. Yes. AI, automation, and increasingly mechanized modes of production *are* making us losers. It has been the case for the last +200 years that technology props up some people while decimating others, and it continues to do so till the sun blows up. If you are replaced by a clanker, can't compete with a clanker and are henceforth unable to find a job besides exhausting, dirty, and lower paying ones that nobody wants to do, you are moving downwards on the socio-economic ladder, therefore losing rather than gaining as a result of technological innovation and introduction of said innovations. If a clanker can make 200 cars in a day, and you can only do, if we're being generous, 1/36th of a car in that same time, wouldn't you agree as well that you can't even hope to physically compete with a clanker, let alone an army of clankers in terms of productivity and costs? Likewise, if you're doing research on x and you can mid-to-deep-read at maximum of 100 articles in a day, but a clanker can run through the entire Western canon of scientific literature in a matter of day, maybe even less, wouldn't you say the clanker has surpassed your research abilities a thousand times over? I don't know what kind of copium you're on, but it must be hell of a drug for you to act so smug and clueless regarding the capabilities of AI, especially since it has so clearly and demonstrably fallen into the wrong hands already, and can churn out "decent enough" output in seconds, which would take, no matter how banal or mundane or demanding the task, a thousand times longer for a human to do.

>>94182 You are insulting me because of you insecurities. Like you the life you choose to pursue, nothing you do or say matters. >>94185 You are choosing to be a future loser. Instead of evolving you are choosing to watch the World you go without you. Being a loser is a personal choice.

>>94186 >Being a loser is a personal choice. I see. If choice encompasses the entirety of life, then it must be the case as well that one can choose one's mental deficiencies, setbacks, catastrophes, injuries, among every other thing that happens to you, both positive and negative, irrespective of the surrounding world or the choices that other people make for us regardless of our wishes, intentions or even conscious input. If that is the case, then please choose to be a little less retarded in your beliefs.

>>94187 The grim reality is while he will remain retarded in the future, said future will bite your balls off. We're entering the cyberpunk epoch. No going back. Basically the worst outcomes are roughly in play. We might get a solid devestating war yet, and China/Asia is on a creeping ascendency. Shit is bad out there lain, and getting worse.

>>94189 >Shit is bad out there lain, and getting worse All we can do is ignore that and pour all we got into a nice personal project, if bad things are about to happen i'd better make use of the time i got than worry about things out of my control

>>94187 >mental deficiencies, setbacks, catastrophes, injuries, among every other thing that happens to you Excuses, excuses, excuses. Something that I have no control of happened to me, I must become a loser.

>>94186 >You are insulting me because of you insecurities no, I'm insulting you because you are deciding to lay down and take whatever they give you instead of having even the tiniest little backbone. You're a dog begging for table scraps. You think it bothers me that my life is inconsequential? That I will die anonymous? This says more about you than me.

>>94193 >You think it bothers me that my life is inconsequential? That I will die anonymous? You are such a loser lmao, no wonder you are insulting me at every opportunity. Calling me a dog with no backbones because I'm not hating on a technology. Good luck with the rest of your life, you will need some.

>>94194 Nah, he's right. And so am I: you're retarded.

>>94169 I think good discussion that came from this is that we have officially entered the "dark forest internet". Here is a somewhat good article about it https://www.theintrinsicperspective.com/p/culture-becomes-a-dark-forest This part is interesting to me "And yes, obviously, this outcome is vampiric and enervating and sad, but frankly, this has been the mode of this technology from the beginning. Human culture always led naturally to more human culture—it was kind of like crop rotation, which is slow but infinitely sustainable. AI looks more like slash-and-burn agriculture, where you purposefully reduce the existing natural growth to ashes, grow a couple years of crops on it, and then move on to the next plot. The only way to counteract the situation is for all of us to become what Liu Cixin called “wallfacers” in The Three-Body Problem—if you are working on something interesting, then wallfacing is necessary. You must face the wall, and so leak nothing, give away nothing." Which makes me wonder. Is this end to the global world? I don't really see much point in posting anything with much effort if it will just be scooped up. Obviously stuff still works in private networks. I think it might even be cool in some way as it gives everything this underground feeling. But how do we keep having any kind of discoverability? Is the open web like dead dead?

>>94170 >>94190 There's more than one way to skin a cat. You are not as powerless as they would have you believe.

>>94198 AI is a hungry and fragile technology. It requires the cooperation of everyone to operate.

>>94199 I can run it locally on my 3060 nvidia, with things to keep private from cloud AI. >>94169 You have become useless, thats why you are mad. Face it. Power loss. Your on the level of a cock receiving foid now.

Have you considered trading futures? Unlike stocks, the futures market is open Sunday 5pm to Friday 4pm CT, with an hour long break each day from 4-5pm.

Iff you use i.a for bad sircunstance is problem you, god bless for you >>94202 I full prond in medicine. Like this man use i.a for make destruction sentence off war + i.a ≈ dead one day. In my case, i use i.a for studing and for conclusions off studing. Imagine you go to war in europe because have beauty womans? Or maybe location great for life because others persons talk about this place. Realy, i recoverry a man say for me, must powerfull rich in planet: god bless for this type off person. I joke hin not australian

>>94202 >Your on the level of a cock receiving foid now. Hot...

>>94169 >I'm a recent math graduate, and I'm so angry that math has been reduced to a publicity stunt for these awful, unethical corporations Let's say a very ethical and honest company solved it instead, would you not be seething anymore?

>>94206 Absolutely, I'd feel the same way as I do about the Four Color Theorem. Either indifferent, amazed by it, maybe a little bit skeptical. But when these companies shows the lengths they're willing to go to one up each other, and knowing they have no qualms about using this technology for surveillance and military purposes, do you not see the problem? I'll admit I was a bit emotional writing the first post, but Navier-Stokes is not the point here. Or rather, it should be, if the company behind it wasn't known to lie, spy, steal, kill.

>>94171 >>94199 This isn't true. "AI" is not just the cutting-edge models of the bloated superlabs, it can't be killed any more short of total apocalypse destroying all electrical civilization. Maybe if the economy soykafs its guts out tomorrow it would delay the strip-mining of math conjectures for a while, but local models on reasonable hardware (which would be available at fire sale prices after an economic crisis) can already run strong enough generative models to continue the social damage. >>94169 I do see the viewpoint that the math rush smells of desperation starting to set in. OpenAI and Anthropic really ought to have better things to do right now than putting all this effort into pushing automated theorem-proving. There's no practical case for a theorem cannon. It's a complete non-product to everyone outside of the tiny niche of pure math departments, and increasing the rate of theorem spam from math departments does not add any value for anybody. Nobody else would even use it for free. All it's good for is a grasp at PR hype, although I think the rockstar math and CS PhDs that work at these labs are overestimating how much people are impressed by the ability to produce solutions to esoteric math problems Related development, one of the constructions from the ten problems in the Astra hype package is also now being accused of being potentially scooped out of a mathematician's chat logs https://mathstodon.xyz/@andreasthom/117240535270608201

>>94208 >I think the rockstar math and CS PhDs that work at these labs are overestimating how much people are impressed by the ability to produce solutions to esoteric math problems A few points to make here: -the CS and math PhDs aren't the ones setting the direction of the company. That is done by the business guys and increasingly also by the company's chatbots, since all these guys get high off their own supply. -math is the only science where you can run a computer program that checks whether you did everything correctly. It is also the most grammar-based of the sciences, in the sense that a lot of mathematics can be reduced to string-rewriting rules. This makes it the easiest field to train an AI and specifically an LLM to do. -they're going after famous open problems, not esoteric ones. N-S has a million dollar prize associated with its solution. This is a big deal outside of math departments. Within math departments, it is apocalyptic. -The tacit agreement that mathematicians have with the broader scientific community and with the public, where we are offered a livable stipend and respect in exchange for occaisionally producing something of value to somebody, is breaking down at an accelerating rate. We are going the way of the software engineer, the graphic designer, the machinist, the steel worker, the weaver. Dead-end jobs with no career prospects ten years from now. Everyone sees the writing on the wall. Most of my colleagues are some combination of depressed and shellshocked.

>>94209 > We are going the way of the software engineer, the graphic designer, the machinist, the steel worker, the weaver. Dead-end jobs with no career prospects ten years from now. Everyone sees the writing on the wall. Most of my colleagues are some combination of depressed and shellshocked. Welcome to the club. Don't forget, you can still do it for fun.

>>94207 moralfagging. The fundamental point is obsolescence.

>>94169 Did it pass peer review?

>>94210 Of course! But I think the future will have far fewer mathematicians, which is sad. >>94212 iirc they have a lean certificate, which means the theorems are almost certainly correct, but the actual papers are unreadable garbage. Both teams were using an approach that is known to work.

>>94206 >Let's say a very ethical and honest company solved it instead, would you not be seething anymore? I think that part of the problem here is not anything about openAI per se, but the specific method that they used here exactly. Its almost certain that they effectively stole this guys work and then finished up a bit faster. Its as if google stole some guys notes and technical drawings for some engineering project because he was using gmail and google docs, and then filed a patent just before he did. Its straightup theft. AI is a proving to be a great tool and even ignoring openai's treachery in this case this is still a good achievement for AI+human collaboration.

>>94210 The ones doing it for fun are the ones who innovate the game anyway

>>94219 Essentially all mathematicians do it for fun. If you want money you go in to a different field. However, it's a difficult hobby. You need a quiet, distraction-free working space and multiple hours at a time to focus. You have to also be in the right headspace, which can be very difficult to do if you're stressed. Plus ,it's often very difficult to find meeting places where people have any interest in math whatsoever, except in university math departments.

>>94210 That's a fuarrrking lie and you know it. These tools are great, but they're no better than tools. They cost a lot and will get cheaper, but never so economical that they will cost so little that it does not make sense for a human to guide them. I'm constantly told I'm too pro-AI but all this soykaf is such a joke. We achieved AGI this week, don't you know? You know, that thing that means the technology should literally be able to improve itself as well as humans can. Except it doesn't. They still haven't fired their engineers. Because these autoregressive toys are useful text predictors, not thinking beings. We can barely run a fruit fly in a simulated, frozen state.

>>94220 and even then you can get into a subfield where there's like four people in the world who know enough to talk about it in depth. And you don't automatically get along just because you're in the same subfeild.