Post by @maxpool@mathstodon.xyz

2026-08-10 11:15 UTC

@atif@mathstodon.xyz @tao@mathstodon.xyz I don't see evidence that machines outpace humans in all aspects of the development of mathematics. Using the skills vs. intelligence distinction from https://arxiv.org/abs/1911.01547 used as the basis for the ARC Prize https://arcprize.org/, it's clear that machines master many skills and techniques that human mathematicians already know, and then can tirelessly search for solutions using these existing skills. But could they develop a completely new and novel technique or a new field in mathematics in order to solve a very hard problem? Not yet, I think. What would be an ARC Prize-style problem for mathematics?

Replies (1)

@atif@mathstodon.xyz 2026-08-10 18:39

@maxpool@mathstodon.xyz yes i share this view: https://mathstodon.xyz/@atif/117072533937598929 i think it’s plausible that we will see machines come up with novel ideas in the near future, but i speculate those will be new techniques or tricks for problem solving. which is still very far away from coming up with a new field, or creating a new theory. no doubt that these (possibly interesting and useful) techniques and tricks will be used for another marketing wave: “we just saw a glimpse of ais capabilities to come up with novel ideas!” while in reality, those may be the ceiling of capabilities of this (transformer etc) technology. i think it’s important that we remind ourselves of the fact that evidence doesn’t suggest such leapy reasoning, a kind of reasoning which has become rather ubiquitous. (thanks for the links!) @tao@mathstodon.xyz

Open ##5161035