Stanford researchers have developed a time efficient and safer algorithm for autonomous cars that combines game theory and risk awareness. This algorithm computes approximate feedback Nash equilibria where all agents are risk aware, a novel approach.