deep reinforcement learning

In a landmark study, lab-grown neural networks learned to play Pong more efficiently than sophisticated AI, demonstrating the superior learning speed of biological systems.