Ethics

Scott Alexander has a recent piece about "deontological bars" in the context of AI safety. He describes the state of the discourse like this: I’ve been thinking about this lately