Title: Redefining Intelligence as the Rate of Learning Efficienc...
URL Source: https://www.bestblogs.dev/status/2032899704913276985
Published Time: 2026-03-14 19:20:33
Markdown Content: Skip to main content Toggle navigation menu Toggle navigation menuArticlesPodcastsVideosTweetsSourcesNewsletters
⌘K
Change language Switch ThemeSign In
Narrow Mode
Redefining Intelligence as the Rate of Learning Efficiency ==========================================================
Redefining Intelligence as the Rate of Learning Efficiency ==========================================================  ### François Chollet
@fchollet
Knowing how to do a task is not proof of intelligence, but in reverse not knowing how to do it isn't proof of lack of intelligence. Intelligence is the rate of increase of your competence as a function of your resources/training, and that rate is independent of absolute levels.
Mar 14, 2026, 7:20 PM View on X
12 Replies
6 Retweets
102 Likes
3,356 Views  François Chollet @fchollet
One Sentence Summary
François Chollet defines intelligence as the rate of competence acquisition relative to resources, rather than static knowledge or task performance.
Summary
In this tweet, François Chollet, the creator of Keras and ARC-AGI, provides a theoretical definition of intelligence. He argues that intelligence is not demonstrated by simply knowing how to perform a task, nor is it disproven by a lack of specific knowledge. Instead, he posits that intelligence is the 'rate of increase' in competence as a function of training resources. This perspective emphasizes learning efficiency and generalization over the absolute level of skill, which is a cornerstone of his work on measuring Artificial General Intelligence (AGI).
AI Score
84
Influence Score 23
Published At Today
Language
English
Tags
Intelligence Definition
AGI
Machine Learning Theory
ARC-AGI
François Chollet HomeArticlesPodcastsVideosTweets
Redefining Intelligence as the Rate of Learning Efficienc... ===============