François Fleuret: math and CS modularity explains AI's IMO success

francoisfleuret · x · 2026-09-22

AI researcher François Fleuret argues that AI's success in mathematics and computer science (e.g., IMO) stems largely from the modularity of constructions in these fields: something that "works" can serve as infinitely load-bearing building blocks.

This property greatly benefits random search, since locally valid pieces compose into globally valid solutions. He notes this doesn't hold in fields like engineering, which explains why AI can't replicate the same success there.

Original post →

More from AGI Musings

AGI Musings channel →