A Ray comments on Reinforcement Learning Study Group