JAX / Flax feels like it would feel more at home in a language other than Python. There's a lot to like in how it views the overlap of functions and data but Python's object model just kinda runs over everything and makes it confusing.

A short list of tips for keeping a clean, organized ML codebase for new researchers: eugenevinitsky.com/posts/quick-...

Not quite sure how parse projects like Rick Rubin’s AI backed interviews with dead people (like this one with Feynman pca.st/episode/4151...). It is an interesting way to organize archival material but I want to know the creative process used. How is it edited, what liberties are taken?

Has anyone made a Robotics feed yet? There are already a few for other disciplines, about time for ML and Robotics folks