#ai #artificial-intelligence #existential-risk #ai-alignment #ai-safety Created at 211023 # [Anonymous feedback](https://www.admonymous.co/louis030195) # [[Epistemic status]] #shower-thought Last modified date: 211023 Commit: 0 # Related # human compatible stuart russel - The risk to humanity from advanced AI is a serious concern despite the uncertainty surrounding future progress in AI. - We need to ensure AI operates in line with human preferences. - Russell proposes three principles to ensure AI operates in line with human preferences. - If we are to ensure AI remains beneficial to humans in the long run, we may need to radically rethink its design. - Russell suggests that we can rebuild AI on a new foundation, according to which machines are designed to be inherently uncertain about the human preferences they are required to satisfy. - Such machines would be humble, altruistic, and committed to pursuing human goals. - The book has received positive reviews for its accessible style and dry humor, but some critics have pointed out its technical parts are too difficult, and its philosophical ones too easy.