JP Addison🔸 comments on Principles for AI Welfare Research