Zamaamiro··on Deep reinforcement learning is a waste of time (2019)It’s not useless; it just doesn’t get us closer to AGI. That doesn’t mean it’s useless.