A new reinforcement-learning algorithm has learned to optimize the placement of components on a computer chip to make it more efficient and less power-hungry.3D Tetris: Chip placement, also known as c...