12 ms·
> I am dreading the inevitable onslaught in a year or two of language model “proofs” of the Riemann hypothesis which will just contain claims which are vague or
by jebarker 2y ago
> I am dreading the inevitable onslaught in a year or two of language model “proofs” of the Riemann hypothesis which will just contain claims which are vague or inaccurate in the middle of 10 pages of correct mathematics which the human will have to wade through to find the line which doesn’t hold up.
I wonder what the response of working mathematicians will be to this. If the proofs look credible it might be too tempting to try and validate them, but if there’s a deluge that could be a hug time sync. Imagine if Wiles or Perelman had produced a thousand different proofs for their respective problems.
- bqmjjx0kac 2y agoMaybe the coming onslaught of AI slop "proofs" will give a little bump to proof assistants like Coq. Of course, it would still take a human mathematician some time to verify theorem definitions.
- Hizonner 2y agoDon't waste time on looking at it unless a formal proof checker can verify it.
- kevinventullo 2y agoHonestly I think it won’t be that different from today, where there is no shortage of cranks producing “proofs” of the Riemann Hypothesis and submitting them to prestigious journals.