Surprising AI breakthroughs raise soul-searching questions for mathematicians | Letter
TL;DR
Whether mathematics thrives or perishes in the age of AI depends on what it is that society values in the human intellect, writes Dr Henry Bradford I share Kasra Rafi and Bruce Schneier’s impression that recent mathematical breakthroughs by AI consist in clever recombination of existing ideas, not development of truly novel theory (No, AI doesn’t mean the end of mathematics – at least not yet, 25 August). The question is: what happens to mathematics if this changes?
Nauti's Take
The progress is notable: if OpenAI's Astra really proved the existence of non-sofic groups, AI moves from recombining known ideas toward original theory work. The limit sits in the sourcing, because the account comes from a single reader letter and has not been independently verified.
For AI teams this belongs on the watchlist rather than in product claims about autonomous research.
Sources