David Mathers🔸 comments on On Deference and Yudkowsky’s AI Risk Estimates