Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Navier-Stokes – Tristan Buckmaster [pdf]

cims.nyu.edu

Post reply on HN
Print view View original

611–620 of 872 posts

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#611

Post by eaurouge »

CoolestBeans wrote: post #460
highfrequency wrote: post #412

From OpenAI: > While unlikely, we cannot rule out that de-identified data derived from their usage of our products helped improve our models This is the crux of it. If Tristan's work and insights were not used to train OpenAI models, then this just looks like a case of hyper-competitive academic sniping that has been going on for decades (check out Watson and Crick!) accelerated by AI as a tool. The fact that this is…

I think this might be a red herring. All it takes is someone to get an inkling that someone is working on a new approach and seeing some success for OpenAI to fire the AI cannon at the problem. The community seems fairly small (from this outsider's point of view). The idea that the data made it into the training set and that's how the bot figured it out is definitely possible, but I would want to rule out the simpler…

That's pretty much it. This is a bit like (but worse imo) running a vc firm, listening (formally or informally) to idea pitches, then spinning out and funding competitors with millions of dollars, to outcompete the originators of those ideas. Yes, the idea is not secret in this case, but the unfair advantage is massive and there's only one (a couple at most) positive outcome.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#612

Post by WD-42 »

davesque wrote: post #532
hkmaxpro wrote: post #336

Earlier quoted context omitted.

Both Sam Altman and Sebastien Bubeck admitted they only want Buckmaster to be the lead author on a rewrite of the OpenAI proof. https://x.com/sama/status/2097385167002415140 https://x.com/SebastienBubeck/status/2097379411691516310 A wake up call for using OpenAI models. If you discover something with their model and you work for a competitor, they “felt it would be inappropriate” for you “to author OpenAI’s work”.

Honestly this whole thing is so fucking weird. I feel like there's an argument that absolutely no one involved in the final crossing of the finish line to the proof actually did any work (other than just intelligently directing an LLM) and deserves any credit. As the author of this doc mentions, the mathematicians who did the actual work that led to the formulation of this approach (without the use of LLMs; just good…

Yup! I wanted to side with the mathematician on this one but I read the statement only to discover that they were also pushing an llm on someone else’s idea producing mountains of slop.

Have LLMs actually improved anything? Is mathematics better off than if these slop proofs didn’t exist? Who or what is actually benefiting here.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#613

Post by needfish »

ryan_n wrote: post #359
smcg wrote: post #354

Earlier quoted context omitted.

you massively collapsed what AI companies have been doing by comparing it to old internet-scraping. Facebook flat-out admitted that they scanned copyrighted books for their AI. The image generators most definitely trained on copyrighted images.

LAION and Common Crawl both scraped copyrighted images. From what I can tell (I'm not an expert in this domain at all), the main difference between those two and frontier labs is in how they stored and used the data. CC and LAION seem to be actually open (unlike "Open"AI) and are more centered around publicly sharing the data they scrape to support research and innovation. OpenAI et al also stole everything from ever…

Personal two cents, I have friends whose music work posted on YouTube were scraped to be in LAION-DISCO-12M, so yeah not very open.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#614

Post by kccqzy »

antonmks wrote: post #16

There will be a lot of hurt and pain in mathematician's community. It is hard to accept that major discoveries are now just a function of spent token $$.

What good is a math result if there’s no human understanding behind it? Unlike many other fields where there’s value to an artifact even if there’s no human understanding, the whole point of mathematics research is just gaining insight and understanding.

On the Navier–Stokes issue specifically, it has long been suspected that such a blow up would exist, and an AI telling you it indeed exists doesn’t contribute any new understanding to the field. And this problem seems like one that would be solved by humans anyways even if AI didn’t exist; accelerating the result by a few months/years using AI doesn’t mean much.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#615

Post by stevemk14ebr »

dekhn wrote: post #400
Semkas wrote: post #94

Earlier quoted context omitted.

"using the same approach that Buckmaster and Alpoge had been exploring" is imo mealy wording: it seems fairly likely that OA heard Buckmaster and Alpoge were close to a breakthrough, and decided to use their unlimited compute to quickly prompt based on their assumptions about B&As work.

Is that necessarily wrong, so long as the original innovators get a citation credit?

Yes. Quite simply yes. It's unethical. And it's a dick move

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#616

Post by za_creature »

fc417fc802 wrote: post #557
anonymousDan wrote: post #517

Earlier quoted context omitted.

This is just a nonsense line of reasoning. Training based on the solution to the problem (or the key insight behind the problem) is clearly a form of plagiarism.

What about my line of reasoning is nonsense? I made no claim either in support of or contrary to yours. Rather I pointed out that by this logic literally everything that an LLM spits out is plagiarism of the vast majority of the entire body of human literature in existence. Can you offer meaningful refutation of that observation of mine?

Many do indeed hold the position that all LLM output is uncopyrightable plagiarism. They're probably right, but there's an even stronger argument here:

Science papers of a phd level must contain:

1. one or more novel insights

2. a long list of citations to contextualize them and

3. some work to prove that the insights are in fact meaningful

---

In this context, consider a prompt based diffusion model which, when asked, will happily produce a few pictures of a horse in orbit. You then tell it "silly robot, horses can't breathe in space" to which it adds the necessary space suit in a follow up image.

That image is twice plagiarized:

1. the model did not come up with the original idea of putting a horse in space, nor with insight that horses need a space suit

2. the model failed to cite where it pulled the "horse" and "space" concepts from.

It merely did the work (3) to combine the concepts using the user provided insight.

---

The implied accusation here is that OpenAI used the insights from an existing prompt to train a new model that was able to one shot "a horse race in space" picture, and they were all wearing space suits.

This is still academic plagiarism, even if you disagree that all LLM outputs are.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#618

Post by rakejake »

spongebobstoes wrote: post #551
rakejake wrote: post #327

Earlier quoted context omitted.

OAI doesn't need to mention Buckmaster's name directly in a prompt. They just need to select a basket of sessions that is guaranteed to contain Buckmaster's and then direct the LLM to attack only a specific method/angle. This is trivial to do while maintaining plausible deniability about not using his work.

what reason do we have to believe that they did this? both things were proved by AI, isn't it logical that they could have very similar approaches? it is common that multiple people essentially simultaneously prove/invent the same thing I see zero evidence of wrongdoing

OAI started working on this only after they found out it was close to being solved. They threw a team of researchers who spent sleepless nights + a ton of compute. This is not exactly healthy academic competition - it's like if you spend a year hunting for oil fields and finally find a very promising area to be explored, only to find that Exxon tapped their entire exploration unit to go all in and and find it overnight just to stake claim to the discovery. Tao said it right - math should not be treated as a non-renewable resource to be mined.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#619

Post by hellohello2 »

dekhn wrote: post #604
hellohello2 wrote: post #599

Earlier quoted context omitted.

Every university I know has access to clusters with fresh GPUs. Not sure when you graduated but you'd be surprised how much money is getting poured in I think!

Sure, they have clusters. We often call them "closet clusters". Nothing has changed. Some institutions have larger systems (some extremely large) but none of them have demonstrated running warehouse-scale systems. I'm talking one to two orders of magnitude (and the storage and networking to make sure all those systems don't stall waiting for data).

Perhaps I am underestimating the size of these frontier models then, how big of a cluster do you think would be required to serve a university?

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#620

Post by mapontosevenths »

Catloafdev wrote: post #546
m00x wrote: post #531

Earlier quoted context omitted.

It would be very difficult to say. It confirms that Tristan's data is likely part of the data the models use, but a lot of filtering, pruning, and transform goes into training. Data has to be determined to be signal and not just noice, then it could go through processes of generating questions/answers from that data, then it RLHF's over this. OpenAI have petabytes of data, all anonymized. It could take months to say…

Frankly, I don't buy this difficulty argument. They know which model was used to come up with that particular idea. A text search over the corpus of user data used in the training set can only take so long.

> A text search over the corpus of user data used in the training set can only take so long.

Did you notice the line in the article that says the models had access to an offline copy of THE INTERNET. Like all of it.

Post reply on HN