@sairahul1 - Stanford CS229 Teaches Building
@sairahul1 on X A post promoting a free Stanford CS229 lecture, 1 hour 44 minutes long, that walks through how to build large language models from scratch. The author argues it explains how ChatGPT and Claude actually work.
- Anthropic reportedly pays $750,000+ a year for engineers who can build LLMs from scratch.
- Stanford released the full lecture from CS229 for free, running 1 hour 44 minutes.
- The author claims the lecture teaches more about how ChatGPT and Claude work than most people at top AI companies learn over their careers.
The post is a promotional tweet, and most of its punch comes from two claims: that Anthropic pays engineers $750,000 or more a year for LLM-building skills, and that a Stanford lecture teaches the same material. The salary figure is presented without a source or a link to a job posting, so treat it as the author's framing rather than a verified fact. The lecture part is checkable. The author says it runs 1 hour 44 minutes, is free, and comes from CS229, Stanford's machine learning course, which is a real and well-known course with publicly posted lectures. Nothing in the post itself links to the video, though, so you would need to find it on Stanford's YouTube channel or the CS229 site yourself.
The substance the author is pitching is a from-scratch build of a language model: tokenization, the transformer architecture, attention, training loops, and the basics of how a model learns next-token prediction. That is a useful thing to understand, and it is a reasonable recommendation for anyone who wants to see the mechanics behind ChatGPT or Claude. The claim that it teaches more than most people at top AI companies learn in their careers is hyperbole. Plenty of people at those companies work on infrastructure, evaluation, safety, or product, and a single lecture won't cover what they actually do day to day.
For a read-it-later queue, the sensible takeaway is this: the post points to a legitimate, free lecture on how modern language models work, and that is worth an evening if you have the background in linear algebra and basic calculus to follow it. Skip the salary number and the "more than their careers" line when you judge the value, and judge the lecture on what you find when you watch it.