leogao comments on A Rocket–Interpretability Analogy

[leogao]

For ML, yes. I’m deriving this from the bitter lesson.