Nathan Lambert

Indexed context

The hosts mention that Nathan has an RLHF book available for preorder with a full digital preprint, and one host is preparing the physical edition to be well organized and attractive. The book is framed alongside other recommended ML books as a valuable resource and part of the hosts' broader respect for authors who help people build models from scratch.

Indexed context

The book is mentioned as a highly recommended, practical resource for learning by doing—specifically building a large language model end-to-end. The speakers emphasize that implementing working code is the most reliable way to understand models because code can be verified, unlike potentially misleading figures or explanations.

Indexed context

The hosts mention Sebastian Raschka's new book as a strong recommendation for learning by implementing models directly, praising the hands-on approach. They argue that building models from scratch (with working code) is one of the most powerful and reliable ways to truly understand how models work.

Indexed context

The book Apple in China was mentioned to illustrate extreme work cultures and the human cost of rapid industrial progress, particularly how engineers endured grueling hours to build supply chains. The speakers use the book's examples (including 'saving marriage' programs and deaths) to compare historical practices in China to modern AI industry grind like 996 and startup intensity.