Percy Liang's CS336 slide: open models are what make teaching LLMs from scratch possible
stanfordnlp · x · 2026-09-09
A viral post explains why the open ecosystem is underrated, citing a slide from Stanford's CS336 (Language Models from Scratch) taught by Percy Liang: GPT-4 cost $100M to train, the next generation $1B, and its paper explicitly refuses to share how anything works — so frontier models are out of reach for academia both financially and epistemically. Llama, DeepSeek, Qwen, Kimi, GLM, Minimax, MIMO, AI2, NVIDIA and the Marin project publish weights, code and data instead. The slide's key line: "Ideas from open models enable us to teach CS336." The post responds to Nathan Lambert's argument that new open models deserve more attention than frontier lab drama.
Related event: Percy Liang: Open Weights Alone Don't Make Open AI(2 posts)→
More from AGI Musings
- Frontier labs quietly building recursive self-improvement, thread claims — 0xsachi · 2026-09-09
- Prediction: Big AI will start buying Big Pharma — not for the drugs, for the data — MannyKayy · 2026-09-09
- A magic lamp thought experiment on when AI solving math helps and hurts — sandersted · 2026-09-09
- Blogger: Million-Instance AI Swarms Could Cripple National Economies, and We've Seen Nothing Yet — scaling01 · 2026-09-09
- "AI Techbros Only Produce Low-Variance Copies of Researchers," Sparks Debate — MoonL88537 · 2026-09-09
- GPT-6 Astra rumored to hit ~11.6-hour p80 task horizon, tracking AI-2027 curve — haider1 · 2026-09-09