lukstafi comments on Reinforcement Learning Study Group