I don’t think that is correct. For consistently good results, you need to set the right goals and be able to judge the quality of your outcomes. Even if you don’t do the hard work in between some thinking is required for non trivial work.
Why on earth would open AI pay for that though? There is no value in it anymore. The pointlessness of this destruction of math really is depressing to me.
But they aren't more successful. They can solve this because of standing on a huge body of work made by working mathematicians. Without it AI would get nowhere. Now AI are burning the minds that fed it for short term gain.
Except they don't too the full work. If the prompt also verified that what was proven is what what asked for, and that the proof is correct, and then write it up in an understandable fashion, then the work is done. As it is now it is more like a spoiler for a movie than the actual movie.
Obviously use the appropriate distribution for your problem. But the Gaussian is used a lot for very good reasons, the central limit theorem being a particularly strong one.
My point is that there is a real difference between massive ai company budgets and research mathematicians driven it, it is not an argument against use of ai which seems inevitable.
But on one hand there are labs that pushes tens of millions $ to mine for publicity, and on one hand mostly underfunded researchers trying to improve general understanding. Big ai may seriously harm math and when Pr value diminishes down nobody is there to keep pushing.
I am not arguing that mathematicians would never use AI systems, they do. I am arguing that if we stop having mathematicians we would stop solving deep mathematical problems as there would be no one to ask the right questions.
> Here, I disagree. What makes you think the AI won't be able to ask the right questions without prompting?
But why would they ask the right questions without prompting? The LLMs have no intrinsic goal to research math problems, it would just be very costly to run queries without a reason to run them... arguably a human in that loop would be a mathematician, but then we do need them after all.
Our AI system don't just do things. They do it because someone asks. A lot of the boom in AI maths results are solving human formulated problems mostly for PR benefit. Once the novelty passes, would Anthropic or OpenAI keep spending? And without mathematicians to ask the right questions and able to appreciate the results, why would AI driven research continue?
Just my local regular non hallucinogenic toads makes dogs drewl like crazy if they try to catch on (or more commonly find a toad chip road kill on the gravel roads)