Human Misalignment An immediate danger from AI: getting what we want? I have very mixed feelings about AI. I find it fun to play around with, and sometimes useful. Uncertainty about future developments (and their social effects) feels extremely unsettling. But much current discourse on the topic is pretty frustrating, simultaneously dismissively neglecting the significance of worst-case scenarios while exaggerating silly objections in order to push for blanket ideological taboos. I’d like to...