“Clay Pavor predicts the rate-limiting factor in software development will shift from writing code to reviewing code, and eventually to deciding what is worth building.”