Shmi comments on (Double-)Inverse Embedded Agency Problem