lc comments on [Link-post] On Deference and Yudkowsky’s AI Risk Estimates