people are suddenly experts on reinforcement learning just because DeepMind published a paper, meanwhile the actual researchers have been working on this stuff for years