Sony's AI-powered robot Ace has achieved a landmark victory by defeating elite table tennis players, marking a significant milestone in physical AI applications. The robot uses a network of cameras and advanced algorithms to achieve rapid planning and reaction times, beating three out of five expert players with over ten years of training experience.

According to research published in Nature, Ace scored 48 points versus 70 in matches against professionals, demonstrating remarkable adaptability in handling spin and unexpected trajectory changes. Since the paper's submission, the robot has played four additional matches, winning two against elite players and one out of two against professionals. The system employs nine cameras and three gaze-control systems positioned around the table to track ball movement with precision.

The breakthrough represents a shift from AI's dominance in cognitive games like chess and Go to physical sports, following recent robot achievements in running competitions. While experts acknowledge the technical impressiveness of the achievement, they note that table tennis's structured environment differs from real-world scenarios where robots must operate with incomplete information and potential safety risks.