Regina Rini in TLS: ‘On October 6 (…) OpenAI published results for 372 outstanding problems in mathematical research, all generated by an unreleased AI model. One mathematician, Ken Ono, told the New York Times it felt like getting “punched in the stomach”. Another, Max Lieblich, told Science that the news raised questions of “what our discipline even is, what we’re trying to accomplish”.’

Also: ‘Mathematicians have options. They might decide to treat AI use as a shameful weakness, in the same genre as stealing unpublished results from a colleague’s office.’

And: ‘Kai Shaikh, a PhD student, told the New York Times that AI offers “proofs without stories”, taking from mathematics its “first glimpses of strange beasts”. We are on the edge of a territory that threatens to explore itself. Here be Dragons.’

AI is often described as a language machine, perhaps its true strength is the language of mathematics.

Poor mad geniuses.

(a sf 2282)