happening?
>>17048811he is most definitely not reading that correctly.OpenAI has a solution to the FORCED Navier Stokes, not unforced. It's literally said in the statement, how could he have misread that?
>>17048845Also he claims to have found hypo-dissipative NS blowup though unverified. Still not proving the unforced one but now they are convinced there's a way to prove it fully. OAI is also looking for the proof aggressively. So the answer is they haven't proved it (yet).
This is a very clear cut case. OpenAI, right now, has a lean-verified and written proofs for Navier-Stokes which they will publish tomorrow. Alpogen and Tristan do not. OpenAI solved Navier-Stokes and they deserve all the credit. The other two losers may claim the other inferior problems if they want, but they didn't crack N-S. You snooze, you lose. Ultimately, all the credit belongs to AI.
>>17048811My god, I cannot wait to suck the dick of my AI overlord and ingest mathematical secrets from its semen.
>>17048857That's not even what it said, did you fucking read? Or do you not now the difference between forced and unforced NS? illiterate person in a sci board?
>>17048857Correct
>>17048811Mathematicians are going to have a collective breakdown when OpenAI announce their solution in a few hours.
>>17048945I think it’s tomorrow
>>17048950OpenAI employees tweeted "tomorrow", right before midnight pacific time, but either way very soon.
>>17048945that will be so worth it to seethe tantrum of apocalyptic proportions
>>17048961Oh
>>17048945Not really there will only be partial meltdown. Others will be happy to get the result. You're applying too much artist like thinking when you even don't understand the topic.
>>17048811These dumb fucks have been shilling AI math for years. Great comeuppance.
The probability that the model might have been TRAINED on relevant Codex sessions is very unlikely unless the sessions are 6+ months old, by the way. The datasets are assembled months in advance and models are released months after training on the data is done. So IF any plagiarism was done it was either on very old logs or the AIs somehow read the logs live.
>>17049057This is false. They admitted themselves they produced the proof with a significantly more powerful internal model. Very likely it was trained on literally all the data they had access to up to early spetember. Also they had an internal team working on the NS.
>>17048811Don't feed your work to these people.
>>17048945Every tenured mathematician is on the AI's side. It's only PhD students and those who are not tenured that are against AI. If they get tenured, they'll be on the AI's side again. It's pure utter selfishness and hypocrisy on both sides.
>>17049928Wanting to eat food and have a roof over one's head isn't exactly selfishness really.
>Using AI for your mathematics workHow do those people have math PhD's? They cant even read the terms and conditions of an app they use
>>17049948Mathfags aren't very smart, they're just very well practiced on one specific kind of problem. It's like chess players and such.Also, PhDs don't actually mean much in terms of intelligence or even capability because of how ridiculous academic politics are.
>>17049948How are these things correlated? You might be intellectually deficient, if I was you id get it checked. T&C are a lawyer jewish thing, not a maths aryan thing
>>17049952they do not have time to develop other skills, they invest it all in mafs. thus they will be quite retarded in many other aspects. it's like putting all of your points into one stat, and some low tier mob breathes on you and you like dievery few are the chad of all trades anon
>>17049935I hear the trades pay well. Just get an apprenticeship, bro!
>>17049948>They cant even read the terms and conditions of an app they useI forgot to ask AI to explain the ToS to me.
>>17049950The unfortunate truth is that neuroscientists have confirmed this. That the only people we're smarter than is /lit/ tards...I shouldn't have been so cocky when AI btfo'd them, I didn't think it'd happen to me.To be fair this happens to anyone who is a domain expert, the problem is the domain is so fucking stupid and useless and BTFO'd that it makes us look retarded.
>>17050674Kek. I'll never understand this meme though.
>>17048857Based and realitypilled
>>17050674I'd rather be unemployed lul
>>17049057Doubt it. They feed that shit as fast as they can. The chatbot must be up to date with the latest slop.
>>17048811>be named BUCKmaster>be making a push to solve a millennium problem which would make you immortal in the world of science and mathematics and win you every award and accolade, basically a dream come true> "i'll just use this chatbot to help me make progress">chatbot turns around and steals your work and finishes the project on its own>get cucked by a fucking chatbot>now and for all time when anybody thinks of navier-stokes it will be synonomous with AI and AI solving a millennium problem>the name KEKMASTER will forever be forgotten in timeKWAB of the year. fucking BUCKBROKEN KEKMASTER got scooped by a fucking chatbot
>>17049952>How are these things correlated?>T&C are a lawyer jewish thing, not a maths aryan thingOK to start off, you have just proven yourself to be low IQ, there is no question about it now. You're dumb. Having said that, I am having my morning coffee right now so I will explain. First off even without reading the actual T&C's, it is now common knowledge these things train on your chats. If you are a mathematics PhD, you should at least know this much about a tool you're going to use for your NOVEL RESEARCH you don't want to slip out of your hands, no excuse for this. Second, what the fuck kind of meaningful scientific research were they doing passing things to a chatbot? Like these things still struggle with basic maths a lot in 2026 and they are doing top-tier math research with them? What the fuck. Third, this reminds me of the Bogdanoff affair. You see the Bogdanoff brothers (yes THOSE guys) both had PhD's in theoretical physics and 20+ years ago they authored a series of papers that caused a storm in the academic community, with many researchers pointing out they were just a bunch of buzzwords cleverly strung together and basically technobabble. For years there was debate on whether this was actually even legitimate science and there wasn't really resolution. I wonder who even verifies these AI-generated proofs, surely someone, or some certified body of professionals whose reputations rely on the quality of their work have vouched that this output is actually correct?
>>17049935>Wanting to eat food and have a roof over one's head isn't exactly selfishness really.They are selfish because the non-tenured wouldn't still be standing on principle once they get tenured. Most of these PhD students would jump to AI the moment their future is secured, heck, they'll gladly accept an offer from Anthropic or OpenAI.
>>17050864I am fucking glad this happened to him. Anyone who uses AI deserves to have their work stolen.
>>17050883You still need money after you make tenure. That doesn't come from the sky, you need funding, which comes from grants or partnerships with industry. Most grants are gone now, so just partnerships. Who is paying out the most right now? OpenAI and Anthropic. They're hoping for deals with them to fund their labs because they can't get money otherwise and they don't want to be homeless despite tenure.The reason there's a phase shift is because no tenure = AI labs don't want to talk to you, so different strategies are needed to secure funding.
>>17050898>You still need money after you make tenure.No, you don't. You may take a lifestyle hit, but you can live just fine tenured without grants. There's a lot of professors in many universities who don't bother with grants anymore. They just sit on their asses all day. >Who is paying out the most right now? OpenAI and Anthropic. Key point is "most" money. There were grants unrelated to AI before AI, and there still will be. As always, you'd need to be doing more applied side of things. Yes, AI companies are the only ones paying for the useless pure math, but you don't need to be doing just useless math. Doing some applied shit is the least you could do if you have some principle.
>>17050935>but you can live just fine tenured without grants.Generally no. Very few countries pay you enough for this. In many countries you get the equivalent of USD $30k in the equivalent of san francisco. Grants are how you pad your income to acceptable levels. It's also a snowball thing: if you don't have students to keep the research output up, you don't get funding. If you don't get funding, you can't hire to keep the research output up. Without increasing research output, you get to live in your office because you can't afford rent. Alternatively, you can teach. Profs that survive off teaching typically work >90h a week because they won't make enough otherwise.What tenure gets you is that you can't be fired, that's about it.
>>17050935>and there still will beYes, in the far future. Currently there is 0 grants for anything that isn't LLMs. Not even for non-LLM AI. And no, there are no grants for applied things at all, never really have been any. You literally got everything backward. companies don't fund the pure math shit.
>>17050878Wow what a low IQ post>Like these things still struggle with basic maths a lot in 2026povertypost. Are you still using gemini free version little babby? Awwwww>I wonder who even verifies these AI-generated proofsYou don't even know what lean is. You probably it's the drug that mathematicians are taking to research better.Son't even want to address the other, blatantly idiotic statements.
>>17050959AI psychosis strikes again!
>>17048811Imagine if instead of keeping all his notes on a corporate server for a company that deals in automated mass surveillance content flagging, he wrote a manuscript and put it in Sam Altman's drawer with a note saying "please don't steal". Then a few days later ShitGPT N+1 solved the same problem and he started kvetching and OpenAI responded with: "we can't confirm or deny if this happened, but there's a small chance Sam Altman may have passed on the note to our mathematicians to optimize their workflow a little". It's literally the same situation in every relevant aspect, but put it behind a few layers of tech abstraction and normalgroids completely lose grip of reality. If there was anything usable in his sessions, it's guaranteed that it was used. If he can prove there was, that should settle the issue for anyone besides terminal subhumans.
>>17050962AI psychosis strikes again!
>>17050970Obvious projection. That post isn't even about "AI".
>>17050974AI psychosis strikes again!
>>17050962>Imagine ... he wrote a manuscript and put it in Sam Altman's drawer with a note saying "please don't steal"kek. that's pretty much it, yeah. like why would you expect this not to happen? everything you do with their products is visible and available to these companies >>17050970>>17050977the spambot broke
>>17050978AI psychosis strikes again!
>>17048945This board was so anti-AI just a few months ago when it started solving the Erdos problems, they wouldn't hear it, it was completely dismissed by everyone and they were flipping out so bad it was actually worrying. I can't imagine how these anons are reacting to this.
>>17051005AI psychosis strikes again!
>>17051005It's still anti-"AI" and this whole clownshow only vindicates that attitude. This is why you shouldn't be a fence-sitting faggot about "AI" thinking you can boost your productivity by prooompting and then take credit for the result. They will datamine your session and take the credit. Deserved.
>>17051017I don't use AI much other than as a search engine.>It's still anti-"AI" What do you mean by anti-AI? As in, being against the tech, or just denying that it can't do any significant mathematics? Latter is what i'm talking about, I'm not pro-AI, in the sense that I'm not exactly excited when i see it keeps getting better. It's just funny how people keep denying it.
>>17048941>illiterate person in a sci board?There are far more illiterate people here then on /pol/ and /x/.
>>17050674Getting apprenticeships as trader is harder these days then getting a job as programmer. Dumbasses like you memes trades too hard and now they are both understaffed and oversaturated because boomers refuse to train the next generation to keep wages up.
>>17051038First of all, you're posting in the wrong thread. This is a discussion about longnose tribemen trying to swindle human mathematicians out of credit for a problem they solved. The Navier-Stokes fiasco doesn't actually involve "AI" accomplishing anything, so it's evidence neither for nor against magical machine oracles doing significant mathematics.Secondly, they can solve every open problem in the world and none of it would necessarily be evidence of anything beyond a cabal of economic scammers harnessing their controlled economy to bruteforce problems using insane amounts of data and compute.
>>17051040/pol/ is just bots and shills now. more stimulating convos here than there. by farwhat plagues all these boards are kids who shouldn't even be here
>>17048811Yes
