Tag: AI alignment

I take a second look at the Carlsmith report on risks from power-seeking artificial intelligence, focusing on Carlsmith’s argument for instrumental convergence.