I actually really miss all the research being done on having (reinforcement learning) AIs beat Atari games and the like. Or the one that stopped at a TV playing random images instead of continuing through the level. Has there been any progress in that field? It seems like LLMs came around and all the projects stopped completely.