Before AI can train itself, it needs more than compute: it needs context. Doris Xin argues that the real barrier to autonomous model development is semantic understanding: knowing what fragmented data means, where it came from, and when human judgment still matters.

The next important milestone for AI research is to automate model development. Every advance in reasoning, language, and perception is, in some sense, a step toward that goal. However, the path to model automation requires solving a set of foundational challenges that must be cracked first.