robotpepi

It's in OpenAI's first announcement that they had solved the problem.

> Solving a major maths problem that has resisted the best mathematicians for so long seems to prove otherwise (even if it were just a matter of remixing old ideas, which is not the case here).

I don't think we know enough about how they work to claim that. OpenAI said they had 10 THOUSANDS agents working on the problem, testing all ideas they found in the literature (including, it seems, the breakthrough of the guys who had it for the hypo viscose case).

I was referring mostly to

> This proof is just checking the boxes for mathematicians

There are already quite a lot of summaries of the story that lead to the solution, and the impact that the intermediate results have had.

I was referring to the "it's just mathematicians checking boxes" claim

fair enough

you're as sure of what you say as wrong about it.

> Because very few people actually have access to these logs, all access is monitored and recorded, and improper access will get you fired. It's not worth risking your job over something like this.

That's beyond naive. The money this would mean for OpenAI (and the money they've already spent)...

> just more user friendly language

you try to justify it in a logical way, but you fail to realize that this is inherently subjective

> 2) 'taking responsibility' is not 'vacuous corporate speak' - that's a terribly cynical view. It's literally the case - the leader takes respnosbility because they failed to set the conditions for the good outcome.

Tbh it seems like you're autistic and simply don't understand how such a "o take responsibility" message really looks like and what effect it'd have.

that's something a human needs to do, and it's non trivial, but it's a simple task compared to checking the correctness of the proof. in any case, most of the language is probably already defined in Lean and checked independently by many people.

you're missing the point...

> reduce the burden of refereeing new work.

As a professional mathematician, I rarely need to worry about the correctness of a paper. The main difficulty of writing a review is instead understanding what the results of the paper mean in its context, how the results are presented, etc.

there's less corruption than in the US nowadays.

that's precisely what the post says under the title.

> while remaining at the top.

isn't the linked post precisely an example of the opposite?

Tao is a good example of a (extremely good) problem solver that doesn't build theory.

do we have any concrete idea of how well the models are scaling now? I agree with these 10 results being impressive, and it is easy to think "wow, and last year the models were barely able to solve IMOs problems". But for me it is perfectly possible that a non sofic group could be found by 100 good IMO students working on all the different strategies that have been proposed (OpenAIs solution was based on "expander graphs", which were introduced to solve the problem some years ago), so it could be that current AI is simply many (say 1000) old models working in parallel. This is linear scaling, not exponential. It could be I'm completely wrong also, the problem is that we have little information.

> but proving a group was non-sofic was out of reach

a colleague was telling me that the base idea for proving that something is not sofic already appeared in the literature around 2019 or so (this is the "expanders graphs" that are mentioned in OpenAI s paper. no one had managed to find a concrete example though. this doesn't make the result less impressive in any case.

it could also be that they try every possible approach that has been proposed by humans. it seems that was the case for the non sofic group example. humans are not able to do the same at that scale. it's unfortunate that we don't know what's happening behind the hood with these models, and that's a huge danger also for the rest of us without access to them.

what are even "results" in mathematics if not understanding? for instance, what's the utility of the counterexamples built by openAI?

It seems you're fighting an hallucinated enemy.

that's an incredibly narrow view of what's happening.

it's still important. not everyone has access to 1 million USD. saying it "only" coat 2000 USD is highly misleading for the discussion and future. the concentration of power is a huge problem with AI.

Exactly! I have some many colleagues who typically say that they do "math for the beauty" while holding permanent research positions funded by the state.

He sounds as someone really pedantic who never understood what math is about or why it is important.

Chatgpt routinely cites and uses papers I don't have access to because they're behind a paywall. I don't think OpenAI is paying for all that copyright. That's in my opinion way more serious.

people here focus too much on the numbers, the exact values, the details, and end up missing completely the point.

so many people missing the point...

try to get out of the box

> managers will simply direct them

that Made me laugh. what you say won't happen. it's not that AI won't be sufficiently intelligent, it's that managers are not.

give some examples or real insights, otherwise it's difficult to take you seriously

and of everyone else, right? what service or product is only available to the US? Even with Chinese models lagging behind, the difference in capabilities is not much.