Critic says those who dismissed deep learning as "curve fitting" should apologize

inductionheads · x · 2026-09-07

An AI community user argues that dismissing deep learning as mere "curve fitting" was obviously wrong at the time, claiming it biased people toward wrong 2D priors — and calls on François Chollet to apologize to his followers.

Original post →

More from AGI Musings

AGI Musings channel →