Building Thoughtful AI Systems

i think the best courses involve syllabi that compel people to think and write about what they're reading, coding, and learning. that's why i'm so incredibly excited that karina and kevin's course on post-training, sponsored by a few of my favorite profs (narges norouzi, john denero, joey gonzalez!) in berkeley's eecs department, compels us to jot down our thoughts in a somewhat comprehensible, publishable manner.

here goes :)

[1/26/26] fundamentals: what do we owe machines?

[2/2/26] fundamentals: the lifecycle of a language model

[2/9/26] post-training: post-training foundations