Builders
The people building AI — what they're saying, shipping, and arguing about, with editorial framing anchored to primary sources.
About Builders
BeatThe signal in AI mostly moves through the people building it. Karpathy, Simon Willison, founders, indie devs, researchers at the labs — their X threads, blog posts, and offhand interviews land harder than press releases.
Builders is our coverage of the people whose work is shaping the field. Every post here links to the primary source. The job is not aggregation — it's reading the tea leaves of what these people are saying, doing, and disagreeing about, and writing back the version that matters.
-
Karpathy: the eval game has lapped the model game
By Nizam Alavi ·
Karpathy's latest thread argues frontier labs are now optimizing for benchmarks more than capability — and that the way out is harder, weirder evals nobody can game.
-
Simon Willison's quiet tool shipping is doing more for indie AI dev than any frontier release
By Nizam Alavi ·
While the labs argue about benchmarks, Simon is shipping small CLIs that make the actual day-to-day of working with LLMs unrecognizable from a year ago.