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

On the Navier–Stokes Millennium Prize Problem

openai.com

Post reply on HN
Print view View original

881–890 of 1001 posts

Re: On the Navier–Stokes Millennium Prize Problem

#881

Post by doctorpangloss »

tedsanders wrote: post #621
Imnimo wrote: post #252

Earlier quoted context omitted.

>we did not read any private chats The question I am interested in is not "did we read private chats", but "was this new model trained using any of Tristan and Levent's chats, regardless of whether they were marked private". Can you comment on that?

If they opted out of training, then we definitely did not train on them. If they did not opt out, then I don't personally know if training signals came from their chats, and I don't think we'd be able to tell without their cooperation in identifying them. And even if signals were trained on in some manner, I highly doubt it made a difference to a problem as challenging as the NS proof. Reasons for my doubt: - I know…

> If they opted out of training, then we definitely did not train on them.

are opted-out-of-training chats ever paraphrased, and thus "de-identified" (in openai's own words)? what this means is, it would be hard to prove that a "synthetic" (but actually paraphrased) training instance came from a particular chat. except if the chat was about some esoteric math proof, of course.

Re: On the Navier–Stokes Millennium Prize Problem

#882

Post by hansvm »

eulgro wrote: post #848
hansvm wrote: post #733

Earlier quoted context omitted.

Not a counterpoint per se, but I burned $50k recently on a much more modest math problem (result already known, just thought I had a sketch of a more interesting proof), and the LLM thought it had proved it within those bounds but had instead subtly fucked up the Lean definition. Take from that what you will. Not to mention, it's still very much up in the air whether the model derived the answer of its own accord or…

How can you afford to burn $50k on an already solved problem?

Those are two separate questions:

> How can I afford?

Mostly not my money, tech makes one fabulously wealthy, I care more about math than fast cars or whatever (even as a pizza driver you can afford a fancy car if that's your primary motive), etc.

> Already solved

That's a very interesting insight into mathematics. It's absolutely just as interesting to prove that certain proof techniques are or aren't possible as it is to actually prove the main result, sometimes moreso, especially if they have any chance of improving attacks at other problems.

Re: On the Navier–Stokes Millennium Prize Problem

#883

Post by nullbio »

mewse-hn wrote: post #78

"we cannot rule out that de-identified data derived from their usage of our products helped improve our models ." What a landmine sentence to bury in this report, you can't rule out your models were spying on other researchers?

I've been saying it for a while now, but no one gives a fuck. Let me repeat it again.

THE BIG LABS CLEAN ROOM YOUR DATA (CREATE SYNTHETIC DATASETS ON IT), EVEN IF YOU OPT OUT, SO THEY CAN BYPASS COPYRIGHT LAWS AND THEIR OWN LOOSELY WORDED TERMS OF SERVICE.

"TOS: We don't train on your data" -> Correct. They train on the synthetic version of your data.

I guess we're just going to ignore this forever though. Who cares about the gaping hole that exists in copyright and contract law now that never existed before LLMs were a thing.

Re: On the Navier–Stokes Millennium Prize Problem

#884

Post by mmanfrin »

jdm2212 wrote: post #724
nozzlegear wrote: post #670

Earlier quoted context omitted.

> If you need privacy, then you are going to have to pay full price for those tokens (API). At this point, how can we even trust that they aren't accidentally training on those tokens too?

It'd be corporate suicide for them to be caught violating zero-data-retention commitments. But also if you're really paranoid you can just use ChatGPT on Azure or AWS, where nothing is flowing back to OpenAI at all.

> It'd be corporate suicide for them to be caught violating zero-data-retention commitments

Would it, though? Considering their entire business model is built on the agglomeration of data that isnt theirs.

Re: On the Navier–Stokes Millennium Prize Problem

#885

Post by combobyte »

mrbungie wrote: post #427
Fordec wrote: post #375

Earlier quoted context omitted.

If humans can figure out to optimize to circumvent bottlenecks, I have no doubt each new bottleneck will also get routed around, just now automated.

We are not in an everything-has-an-API world yet, and it'll for sure take some time to get there.

I'd argue we've been in an "everything-has-an-API" world for a long time now — it's just that discoverability of said APIs is still crap.

Re: On the Navier–Stokes Millennium Prize Problem

#886

Post by JumpCrisscross »

tristanj wrote: post #761
biophysboy wrote: post #109

Earlier quoted context omitted.

Why is it unlikely?

Because the models are trained on hundreds of billions of user conversations, across more than a billion different humans. The conversations are anonymized and not easily traceable back to a specific user. It's unknowable and not possible to prove if any one specific conversation contained the insights for solving Navier–Stokes. We also don't know if the authors unintentionally provided data to OpenAI through alterna…

> It's unknowable

Knowability and likelihood are almost orthogonal here. If I commit a crime and perfectly destroy the evidence, my deed may be unknowable. That doesn’t make it more or less likely.

> not possible to prove if any one specific conversation contained the insights for solving Navier–Stokes

It may be. We haven’t seen the researchers’ transcripts. We don’t know what Buckmaster or his co-author uploaded to OpenAI or with what permissions (or if OpenAI actually respects those toggles).

Re: On the Navier–Stokes Millennium Prize Problem

#887

Post by IshKebab »

swat535 wrote: post #684
biesnecker wrote: post #614

Earlier quoted context omitted.

I came here to say this, like... wow. I'm pretty sure at at least a few of the places I've worked that would be grounds for immediate termination.

They probably should have added the disclaimer: opinions are my own..

Yeah that isn't a magic get-out clause. I don't think I would have been immediately fired for this from anywhere I work at, but that's partly because I live in the UK.

Every company I've worked at has said very clearly not to comment about work things on social media. I would definitely have been in serious trouble for this. I imagine some strongly worded emails from marketing are flying around OpenAI right now.

Re: On the Navier–Stokes Millennium Prize Problem

#888

Post by p1esk »

armchairhacker wrote: post #371
reducesuffering wrote: post #319

Earlier quoted context omitted.

Practical?! The goalposts will keep moving until morale improves (narrator: it doesn't)

The goalposts for the singularity have always been that AI improves itself fully autonomously. AFAIK OpenAI is heavily using AI but still employs human researchers and developers.

They said they will automate AI researchers by March 2028. I personally think it will happen by March 2027.

Re: On the Navier–Stokes Millennium Prize Problem

#889

Post by ex-aws-dude »

stephbook wrote: post #411
pu_pe wrote: post #119

OpenAI thinks of this as a scoop, and it is, but the possibility that they trained the model on the prompts of the other mathematicians they were competing with will leave a terrible taste on every scientist's mouth. Seems like yet another advantage of using open models right here.

> they trained the model on the prompts of the other mathematicians they were competing with How would they have gotten that mathematician's progress though? Did that guy also use OpenAI? If that's the case, it only strenghtens their claims lol. If mathematician decide to use OpenAI's model to do the work, that only reiterates how strong their models are.

The guy did use OpenAI

Re: On the Navier–Stokes Millennium Prize Problem

#890

Post by JumpCrisscross »

dekhn wrote: post #849
octoberfranklin wrote: post #841

Earlier quoted context omitted.

> https://cims.nyu.edu/~tristanb/statement.pdf This really need to be a top-level story on HN.. I asked whether the model had been trained on, or had access to, our sessions in Codex, into which we had been putting all our drafts for the whole of this project. I was told the model did not look up user data. I asked again, about training, and I did not get an answer. This whole episode is more horrific than "AI is eat…

We don't have any proof of that at all. Please stop rushing to judge without data.

> We don't have any proof of that

It’s fair to give benefit of doubt to Buckmaster given OpenAI is currently being very credibly sued by Apple for openly stealing others’ original work in another context.

Post reply on HN