The Obedient Machine: Why AI Alignment Is a Problem of Human Governance
AI safety asks whether a machine will do what its owner intends. It does not ask whether the owner answers to anyone. This thesis argues alignment is a governance problem, solved by enforceable rules, independent institutions, and citizens who own a stake in the tools and the gains.