Donald Hobson comments on On decision-prediction fixed points