tweetindex

François Chollet @fchollet · 03 Sep 2026

Benchmarking AI systems is a continual process that co-evolves with the models. New benchmarks challenge AI capabilities with emerging questions to shape the directions and feedback signal of the research process. Then they adapt as models progress, targeting the residual between
29 244Views
263Likes
10Reposts
13Replies
1Quotes
19Bookmarks

Is that a lot?

1.00×vs this author's median29 244 views is typical
40Percentile for this authorof 5 recent posts
4.76×vs 100K–1M median6 139 views is typical
4.04%Reachviews ÷ followers
0.98%Engagement rateof viewers reacted

Compare with the benchmark table →

Open on X →