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

501–510 of 872 posts

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#501

Post by TZubiri »

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…

> If the answer is no, then this seems fair game.

Yes, fair game, but innacurate to sell it in the media as an advancement of AI as some sort of artificial intelligence, and telling people to use the smart AI, when in actuality the mechanism by which the discovery was found was hybrid human/machine, and telling people to use this tool will result in the discoveries being sniped by the vendor.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#503

Post by ComplexSystems »

red_green_yell wrote: post #402

If you're smart enough to solve this Navier-Stokes problem, you're smart enough to read a TOS and recognize that OAI is a highly untrustworthy company. Putting cutting edge research that could lead to a $1M prize into a cloud LLM with a TOS that allows training on your chats is really just asking for it. Given Tristan doesn't explicitly say he was using the API, and given he doesn't mention anything about the API TOS…

If it's unethical, then there is reason to call it out as Tristan is doing. I see no reason to blame the victim.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#504

Post by dogomatic »

m00x wrote: post #494
20k wrote: post #451

Earlier quoted context omitted.

That doesn't make it fine. We should not excuse this behaviour just because its rampant already, especially when it comes to such a serious prize

You're getting a massive discount because you're helping to train the model. If you want to have ZDR, you have to pay API rates. This is well-known to anyone in the industry.

If you think this through, it becomes a little classist

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#505

Post by anonymousDan »

Another problem I foresee for academia given the behaviour of AI companies is that even if they don't share their research with ChatGPT, as soon as they submit it for publication many reviewers likely will. Especially if the initial submission is rejected they then risk getting scooped. Possibly uploading preprints to arxiv could help.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#506

Post by CrankyBear »

shiandow wrote: post #423
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…

If he didn't opt out I'm not sure I'd agree that it was fair game. I'm pretty sure it would be considered plagiary amongst colleagues and it is a terrible precedent if we just let OpenAI steal any good idea they can get their hands on if they think it is profitable. You'd effectively sign away any and all rights to anything built with AI if OpenAI chooses to reengineer it before you.

"If we just let OpenAI steal any good idea they can get their hands on." Ah, that's all AI does.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#507

Post by unfoundedacc »

qnleigh wrote: post #17

> It was also said that if OpenAI posted after us, they would say that we deserved the Clay Prize, and that we were the “closest humans to the problem”. I declined both offers. I said that if OpenAI released its result in the way proposed I would go public with what happened. The reply was, “Why would you ruin your career?” I replied that I am an academic, and asked why he thought going public would ruin my career. T…

How careful you are. Instead of just saying what a piece of s..t this Shmubeck is, and what kind of even worse people likely pushed Shmubeck to act as he did.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#508

Post by namuol »

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…

> this is ambiguous even to OpenAI

I took their words as “can neither confirm nor deny”, in the that they are _presenting_ it as ambiguous, but I suspect it’s… less ambiguous to OpenAI.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#509

Post by hellohello2 »

modeless wrote: post #73
AJRF wrote: post #62

Earlier quoted context omitted.

> leaving aside the idea that OA might've used data from the researchers Codex sessions Why leave that aside? That is _the_ story. If a Chinese research lab did this we'd call it espionage.

Because they didn't do that. Tristan doesn't specifically claim that they did, and Anthropic employees don't think they did either. https://x.com/_sholtodouglas/status/2097218240397410733

Its very easy for OpenAI to answer, yes or no, if the model they used trained on their chats.

Re: Navier-Stokes – Tristan Buckmaster [pdf]

#510

Post by mmanfrin »

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…

> If the answer is no, then this seems fair game

Wildly disagree. "Training data" should not imply 'we can look at exactly what you are doing and then do it quicker and get the flowers for it', even if the terms allow for it.

Post reply on HN