"A watershed statement": when math for AI research becomes cheaper than experiments

IgorCarron · x · 2026-08-18

Igor Carron reshared a claim by Dimitris Papailiopoulos, calling it "another watershed statement": when math for AI research becomes cheaper than running experiments (very close to that point), using it to probe reality will let us break through many walls.

The core argument: the falling cost of theoretical analysis relative to experiments may reshape AI research methodology.

Original post →

More from AGI Musings

AGI Musings channel →