robotpepi

> My recent take on AI is that people haven't adapted to the new huge potential of the tool, and don't dream big enough (yet).

I guess you can dream big enough only if you have access to the really interesting models. The 20$ subscription won't get you there, and the 100$ one is already out of reach for so many people.

You shouldn't be so cofindent then, specially since your opinion is based on OpenAI's PR.

> This stuff reads as cope by scientists who spent their entire careers chasing these problems.

This is their only defense so far: deflecting valid criticism by suggesting, without evidence, that the scientific community is gripped by fear.

To be honest this doesn't feel as such impressive as the others. It was recently proven in 3 dimensions, so it could be "just" an extension of that. I haven't read the paper anyways so this is pure spectulation.

> but incorrect in macro/long run

That you don't know. Nobody knows what AI is going to be capable of in 2 years. It could continue improving, or not. There's no strong evidence in any direction. You're just ignoring what doesn't fit your narrative.

people here really struggle with metaphors.

> that is most people's experience with LLMs.

citation needed. my (university) students don't seem to learn as much as before.

your reasoning is sloppy. so confidently incorrect throughout this thread!

doesn't it use special relativity or some modern physics?

We don't know what models are going to be able to do in 2 years. If they don't improve much but people in charge simply decide to reduce the number of mathematicians, then we'll end up with a dead math community and no progress. That's obvious to everyone in position in power. Arguing that he's not emotional ready is very naive.

> All the arguments about the future mathematicians are imagined, and emotional.

Even if you don't like, at this moment, AI cannot replace mathematicians.

It doesn't make sense to expect such a positive reaction from the community when OpenAI publishes these papers without proper citations, scoop mathematicians, and copy-paste publicly available preprints. These are points I raised above and you didn't address. It is obvious that the reaction is distrust: OpenAI spends millions in prompting and subsequent self-publicity, but proper citation is too expensive?

> FWIW, in many online platforms, many actual mathematicians who referee papers, have said the quality of the papers are not outstanding, but they are still better than average.

I don't know what "online mathematicians" you're talking about, but I'm a professional one, do reviews and discuss regularly with other mathematicians (some of them of high profile); no one thinks that these AI papers are above average in terms of writing. This is what even in the most well-known online platforms has been repeated over and over. It seems like you're biased on this one.

It seems like you didn't answer the question.

Of course, everyone is curious about these results, we're mathematicians after all. But by incentives I refer to those things that really sustain the system: obtaining permanent positions, grants for future projects, ideas for PhD theses, etc.

> Do you disagree with this?

Absolutely. If they cared about properly citing the bibliography, sharing how their models work, etc, it would be so much easier to make an assessment of the situation, of what will be math in the future, and all those deeper questions. What we got instead? from those 400 papers, there are already 4 that have been found to be a copy of recent publicly-available papers. They spend millions for "the good of science", but they can't afford basic scientific ethics?

> but also presenting them with "enjoyable proofs". > Or maybe _these_ mathematicians are afraid that _other_ mathematicians will do it. Just elitism and gate keeping.

Ok, I don't see the point of discussing with you. It's clear that you decided what to believe in and no evidence will convince you that reality is more complex. The proof is that you ignored all the nuances expressed here by simply sticking to your simplistic interpretation, without any explanation of why such nuances are invalid.

> The alternative you are proposing implicitly is even crazier. OpenAI should not release a proof that is most likely correct so that it doesn’t burden others.

The alternative is that they do the work to properly present the results. They spend billions of dollars in AI training and inference but can't afford to even cite the literature properly? They're doing the bare minimum because they're inly interested in doing a PR stunt.

> But it also sucks when mathematicians, who are considered experts on a particular problem, refuse to engage with breakthrough results about that problem.

For me the problem is that rigth now the structure of incentives that has been built (e.g. you publish more = you get a grant; good exposition < solving a conjecture) is now broken. So, for instance, you would be very irresponsible if you throw your student into one of those AI papers, it's too much the risk. This part is mathematician's responsability, they need to change this incentives structure.

In any case, OpenAI is being a dickhead here. They throw millions of dollars at these problems, but they can't afford basic literature reviews (the drafts barely cite previous work)? Or checking that Lean's formalizations really correspond to what they claim to prove (even for Navier-Stokes they made this mistake)? It's obvious that for them this is just a PR stunt.

> There's a new generation of younger, hungry mathematicians that are highly interested in figuring out why the result is true

The problem is that right now mathematicians don't have the economical incentive to read these AI generated results. Even if you love mathematics and all that, it's always more important to get a job, and for that it doesn't seem like a good idea to invest time around problems that AI touches because you can't compete with it and you don't know if tomorrow they'll improve by x10 the sota.

These takes forget that OpenAI is doing all this as a PR stunt. They're utilizing the field without worrying about any consequence to it.

You're off by an order of magnitude at least. A mathematician does much more than just prove theorems. Let's start by teaching and administration duties.

And then you're off an order of magnitude of the efficacy of OpenAI. We don't know how much work of mathematicians is behind all these results, what's the real cost (which includes those problems they couldn't solve), etc.

Most mathematicians don't use gigawatts in a 48 hour window either, so...

How much lean code do you need to read to check if NAvier-Stokes was correctly formalized?

> OpenAI investigated and confirmed that their research was not in the model’s training data.

You mean when they say it was impossible to confirm anything but one day after it was 100% confirmed that there was no theft? And here I'm not even talking about all the ethical problems related to trying to scoop another group when you hear they're close to success, or how current solutions follow extremely closely human-generated ideas, or about the lack of relevant citations in OpenAI's paper.

Believing that OpenAI's claims have any substance cannot be explained by naivety alone.

matmul <= 2.25 is no big surprise tbh. unique games and mul < n log n are much bigger.

> they intend to be scientific infrastructure.

beyond naive.

which you an run IF you have the hardware. who knows how heavy these models are.

After reading your comment one could even think that LLM's invented math from the ground up.

> HN struggles with truth-seeking on these topics.

The facts are much more nuanced than how you're presenting them here.

> This is significant progress and released without all the drama.

I feel gaslighted.