I like this title. Mathematicians can already explore an idea with computer programs, and it seems like ML (tho maybe not gpt) could be used in this capacity. Sort of like fuzz testing in software development. Theorem proving might or might not be a fruitful path to pursue with ML.
The strategy for the very highest levels of maths is; if you get stuck, see if you can interest Terrance in your problem as he probably has an idea of what to do next.
4 comments
[ 5.7 ms ] story [ 19.7 ms ] threadThe strategy for the very highest levels of maths is; if you get stuck, see if you can interest Terrance in your problem as he probably has an idea of what to do next.