⌘K

Change language Switch ThemeSign In

Narrow Mode

Extreme Generalization and Symbolic Compression in Science

Extreme Generalization and Symbolic Compression in Science

@fchollet

Science went from the initial observation of radioactivity to a working atom bomb over 47 years via only about 9 distinct key experiments -- extremely few data points -- and symbolic models concise enough they would fit on a single page.

This is what extreme generalization looks like, and it powered entirely by symbolic compression. Turn a handful of data points (deliberately collected) into a tractable plan to completely reshape reality, by reverse-engineering the causal symbolic rules behind the data.

Apr 5, 2026, 4:20 PM View on X

22 Replies

32 Retweets

316 Likes

19.8K Views  François Chollet @fchollet

One Sentence Summary

François Chollet argues that scientific breakthroughs rely on extreme generalization and symbolic compression, contrasting this with simple curve fitting.

Summary

Chollet uses the development of the atom bomb as a case study for 'extreme generalization.' He posits that true scientific progress is driven by symbolic compression—turning minimal data points into causal symbolic rules—rather than the curve-fitting approach often associated with current deep learning paradigms.

AI Score

88

Influence Score 71

Published At Today

Language

English

Tags

AI Philosophy

Symbolic AI

Generalization

Deep Learning

Scientific Discovery HomeArticlesPodcastsVideosTweets

Extreme Generalization and Symbolic Compression in Scienc...