Can an AI Agent Safely Run a Government? Existence of Probably Approximately Aligned Policies

Open in new window