x
[Intro to AI Alignment] 1. Goal-Directed Reasoning and Why It Matters — LessWrong