reinforcement learning Facebook research wants to build more capable AI using self- learning methods The researchers were able to get a six-legged robot to teach itself to walk. « Previous1Next » 1 of 1