Live data from Hacker News

On the Navier–Stokes Millennium Prize Problem

openai.com

651–660 of 1001 posts

Re: On the Navier–Stokes Millennium Prize Problem

#651

What is going to become of life for those of us who do not work at AI labs and are unlikely to be hired by AI labs, despite all the years we put into learning math, coding, etc? Those of us who made the mistake of studying anything other than machine learning. How will we make a living? (We don't live in a world that seems likely to distribute gains widely instead of largely to the handful of already mega-rich.)

Most jobs don't involve formally verifiable outputs. Lots of things involve judgment, nuance, parsing ambiguity and indeed just being a human who can be in a meeting and explain themselves. Maybe those jobs will go too, eventually, but it's not purely a function of applying 10,000 agents to the problem.

Much though jobs may involve those things, it has been rare for me to be in a position where management has valued those things to an extent where they would discern between me and a frontier reasoning LLM's capabilities on those same decisions.

Re: On the Navier–Stokes Millennium Prize Problem

#652

From my reading of the announcement: - There are at least two versions of a model more powerful than Astra at OpenAI at the moment. - The less capable version was used to solve the unforced Euler problem (while the one solved by Levent Alpöge and Tristan Buckmaster was forced Euler) with 100 agents. - The more improved version was used to solve Navier-Stokes, given the results of the unforced Euler problem from their…

The entire drama is that OpenAI sniped a Millennium Prize Problem from an Anthropic-affiliated research team who had been working on the problem for nearly a year. In just 5 days. I don't think that can be understated.

Re: On the Navier–Stokes Millennium Prize Problem

#653

the context here is super important, for those who haven't seen it yet. OAI maybe just trained on a real researchers solution and then celebrated having scored the goal unassisted save for the brief commentary at the bottom of this blog post. Here's the other side. https://x.com/rynorhn/status/2097223532438487463

>maybe

Maybe that happened. What we know for sure is that this is definitely how ChatGPT works to the point where the possibility of this happening exists at all.

Don't get distracted by what may have happened, focus on the facts that we know, ChatGPT trains on user conversations, if you use ChatGPT to create something of value, you are not using the one true ring.

Re: On the Navier–Stokes Millennium Prize Problem

#654

Earlier quoted context omitted.

There's no evidence that Anthropic did it first.

Yes there is - that OpenAI PR says that it was Levent Alpöge, an Anthropic employee, and Tristan Buckmaster, a math professor at NYU.

They appear to have solved sub-problems (Euler problem) not this one...

Re: On the Navier–Stokes Millennium Prize Problem

#655

What is going to become of life for those of us who do not work at AI labs and are unlikely to be hired by AI labs, despite all the years we put into learning math, coding, etc? Those of us who made the mistake of studying anything other than machine learning. How will we make a living? (We don't live in a world that seems likely to distribute gains widely instead of largely to the handful of already mega-rich.)

The only way forward that is not large-scale misery is a fundamental reorganization of our socioeconomic system.

I talked about this at some length here, including a diagnosis of the structural issue we're facing as well as a path forward: https://news.ycombinator.com/item?id=49461333

Re: On the Navier–Stokes Millennium Prize Problem

#656
post #458
post #46

Earlier quoted context omitted.

If they could then it wouldn't be de-identified data...

Well you could search for elements similar to the proof / problem in the training data, even if it's de-identified, right? OpenAI can probably do better than Ctrl-f "Navier Stokes".

There are probably thousands of serious academics taking a crack at millennium problems using AI every day. All those attempts are in the training data. And in fact the two researchers benefited from those attempts as well.

Re: On the Navier–Stokes Millennium Prize Problem

#657
post #312
post #53

Earlier quoted context omitted.

This is going to be dramatic in so many different ways. - First off, to reiterate, WOW. - Second of all, when does this end? Are we at the dawn of the singularity now? - People are saying OpenAI "stole" this from the work of an OpenAI user. If so, that's pretty fucked - how can we trust them? - Time to think about retiring from any knowledge work or business? This could be winner-take-all where a leading lab can butt…

>- Do "normies" even know what's happening? Absolutely not. Even to a lot of techy/nerdy people it's still just a chatbot that they sometimes use to help them at work. Even on here people will do whatever they can to downplay. The lack of fucks given is staggering.

How many fucks should be given, in your estimation?

Re: On the Navier–Stokes Millennium Prize Problem

#658
Fuck OpenAI. Fuck everyone who works there. Like seriously, to all the people who gift their life's work to this monstrosity, do you actually think something good will come of any of this?

Not in a happy-go-lucky "if we just ignore the problem of politics and resource allocation for a bit" world, but in ours. Do y'all really think this will make the world a better place?

Maybe stop building the Torment Nexus, you numbskulls.

Re: On the Navier–Stokes Millennium Prize Problem

#659
post #621

Earlier quoted context omitted.

All the ai labs are open about training on prompts. The question is if buckmaster had disabled that with the toggle openAI provides.

That does not make it ethical

Isnt the entire history of academic progress iterating on work that other academics shared with you? Obviously this situation is spicy but openAI cited their work no?

Re: On the Navier–Stokes Millennium Prize Problem

#660
post #311
post #52

People had joked a couple years ago "Well if they solve a Millenium problem it's AGI"... Well here we are.

Yeah well, its easy to do if you steal someone elses work and then try to threaten them into staying quiet about it Edit: OpenAI have now admitted they were training on prompts at the time they made their breakthrough: https://mastodon.social/@tristanbuckmaster/11723647135247030...

What is that supposed to prove? OpenAI is almost always going to be training new models.
Post reply on HN