This project explores the principles of human motor control by modeling them through the lens of deep reinforcement learning. We aim to demonstrate that complex, efficient, and robust human-like control strategies can emerge from training an agent with simple, well-defined objectives in a physically realistic simulation.