• open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.

A frontier-level agentic coding score reached purely through post-training on an existing base, at a third of a competitor's parameter count, resets what teams should assume about the cost of catching up.

• articleI wrote an AI textbook — how long until AI can do it better?
• article5 useful things you'll learn in my new post-training textbook (shipping now!)
• articleLessons from the hacks
• articleOpen models recap: more on Kimi K3, Qwen 3.8, Xi's WAIC speech, distillation, the open-closed gap, and what's next

• articleGLM-5.2Nathan Lambert
• postCongrats to @Zai_org on GLM-5.3. It massively beats every American open model: NSemiAnalysis_
• articleOpen models recap: more on Kimi K3, Qwen 3.8, Xi's WAIC speech, distillation, the open-closed gap, and what's nextNathan Lambert

Sign in to comment.

Loading comments…