Problems associated with current day Reinforcement learning approaches