Humane Conversation #12: Getting the AI Regulation Right: a Research Agenda