Proposals for AI agent alignment and the motivations of value-driven AI - View it on GitHub