richard_ngo comments on Some thoughts on deference and inside-view models