Stuart_Armstrong comments on One-step hypothetical preferences