[nvm0n1]

Yes, fully agree. There are a lot of unstated but unintuitive assumptions and intuitions going on in the AI risk/ethics community. It's useful to surface those.