That's true. OpenAI has some of the best talents. In my experience, domain experts get the most benefits from the models. They can work much faster, catch false positives, and stir the model in right direction.
I wish they take a bit of more time to communicate the findings effectively.
There are good reasons not to delay publishing at all:
> They should release all their results immediately. (Imagine working on one of the problems they already solved.)
This is the most popular answer to a question regarding AI advisory group and immediate access on a popular website for professional mathematicians: https://mathoverflow.net/a/515442/473286
The whole debate regarding the behaviour of OpenAI is a red herring. Mathematics need to redefine their profession and how they work (like us software developers too). There are very good reasons to believe mathematics has an important role to play. If they could just stop talking about OpenAI and get back to work - they are very much needed, in particular now!
> If they could just stop talking about OpenAI and get back to work - they are very much needed, in particular now
This kind of phrasing sounds particularly empty. We are not in WWII researching the nuclear bomb. What are they so urgently needed for to drop everything and work on understanding openai's proof on partition principle and axiom of choice?
Where did I say they should "drop everything"? I hope to read more about how they envision their future, instead of all this regretting and whaling how one company (that I very much dislike too) published a large amount of proofs. They will get more of them, very soon - if they like it or not -, and I wish that would be the primary subject of the discussion. And yes, I'd also hope they engage with the published proofs. The more raw these proofs are, the better. If AI companies start selecting mathematicians to write nice expositions of their proofs, this is doomed to become a very elitist science.
Says who? The professional mathematician writing the article disagreed. Why should people start dancing the tune that openai wants to play for their own reasons and interests? And I do not see how taking the time and effort to write a proper exposition makes it "a very elitist science" when this exact effort and time is needed to actually get other experts understand and build on a result. Unless you equate spending time and effort learning math as "elitism", which is the ai-shilling moto some time now with everything time and effort related. I cannot see how spending time and effort to understand a field and then spend time and effort to make a proper exposition so that other people can also understand it as "elitist" vs throw everything out there "in raw form".
This was my opinion. But anyone arguing to publish "results immediately" is likely to imply something like it. I guess in chemistry we have the situation you envision - for different reasons: Laboratories holding back their data, until their scientists have published their papers or developed their products. There is a real danger AI companies will do something similar too.
Who do you expect they will select for the exposition?
On which basis do you want a (likely US based) AI company to decide who is to untangle a proof that their latest internal model has just spit out?
Do you expect this to fall to an aspiring, but still unknown mathematician at - say - the mathematics department of Nairobi university?! This is what I meant with my rather unclear "elite" reference: The first publication will always show the name of a mathematician already known to the field, more likely than not to come from the same country as the company ("Our message ... is, you’re a great American company, but you’ve got to hire great American workers"). Are you not worried at all? Don't you think it would be good, if anyone in mathematics had a chance to write that first paper on a new proof?
There are parts of mathematics where the result is the important part. That's not what we're seeing here. Knowing whether the partition principle implies the axiom of choice doesn't meaningfully shape downstream knowledge and decisions. For these more foundational problems, clever proof techniques and the exposition around them are literally the point. Without that, neither humans nor AI can take this slop and derive anything useful.
And if AI can do that then great. I don't care about being elitist or not, and I'm fine with AI taking over math. That's not what it's done though, at least not yet.
> Mathematics need to redefine their profession and how they work (like us software developers too). There are very good reasons to believe mathematics has an important role to play. If they could just stop talking about OpenAI and get back to work - they are very much needed, in particular now!
The root issue is OpenAI et al.'s thoughtlessness in their engagement with a field.
OpenAI has resources.
That they fail to allocate enough of those to cleaning up pre-print papers (that seem to be a corporate PR priority for them to release) so they can be consumed and engaged with by the field they're targeting is... acting like a jackass?
It's the same "Meta / Alphabet can't vs won't hire more human reviewers" problem.
OpenAI could, at an immaterial salary level to them, pay a ton of PhD students and mathematicians just to clean up their proofs and papers.
Not doing so is a leadership and financial choice.
I'd prefer AI companies don't decide who's "cleaning up pre-print papers". This should remain the job of mathematicians at universities, which I am happy to pay with my taxes. Ideally there was something like a Bermuda Principles declaration for mathematics (https://en.wikipedia.org/wiki/Bermuda_Principles). This gave mathematicians even at poor universities and beyond the chance to participate in mathematical progress.
What do you think is gained, if AI companies manage "cleaning up"? Tax money?
No one is "responsible". If the paper is read, depends on the interest of the individual scientist. Many mathematicians interested in a theorem proven/disproven in a new AI paper will be curious, even if it is utterly cumbersome to extract the relevant line of thought. But don't you agree that the job can only be done by a mathematician anyway - be she/he paid by the company or by a university?!
What you are arguing for is somewhat like the "proprietary period" in astronomy (e.g. see https://www.scientificamerican.com/article/nasas-plan-to-mak... ). But there is no mathematician who asked for the run of AI, i.e. there is no mathematician, who can be regarded as the owner of the result, even if ownership was temporary. Complex proofs might take years to explain in detail (think of the ternary Goldbach problem, https://en.wikipedia.org/wiki/Goldbach%27s_weak_conjecture ). It's just unfair, if AI companies held back with their data until a proof has been put into a nicely readable article, and at the same time mathematicians elsewhere are spending all their time trying to solve it.
I'm not. To the extent my taxes are used for maths research (which is minute as a proportion of them) I want them to be used for the development of human mathematical understanding, culture and education, not trawling through a mountain of slop mechanically generated by a Silicon Valley startup that's about to IPO for trillions in the hope there might be some nuggets of insight hiding in there. If the latter is an activity with some value the startup in question can pay for it.
I wish they take a bit of more time to communicate the findings effectively.