Scott Garrabrant comments on Formal Open Problem in Decision Theory