Post #2279898

@xameer@mathstodon.xyz There was that famous recent proof where the mathematician used chatgpt to explore and furnish their proof in some hol. There is also that benchmark people some of whom are around here put together for "whether the strongest llms can effect something like a week of mathematician work" (not close). I would say LLMs are more or less bad at math, but relative to anyone else slopping about, someone like Andrej Bauer interacting with a big one constitutes primary research. @andrejbauer@mathstodon.xyz