x
A Logically Self-Consistent Framework for AI Alignment: The Guardian AI and Existential Incompatibility — LessWrong