A humanoid artificial intelligence robot has defeated professional human players in table tennis, marking a significant milestone in sports robotics and machine learning capabilities.
The robot, named “S,” developed by Sony AI, competed in a series of exhibition matches in Tokyo, where it managed to win three out of five games against experienced players.
Experts say the development is notable because table tennis requires extremely fast reaction times, precise timing, and the ability to read spin skills traditionally considered difficult for machines to replicate.
According to organizers, the matches were conducted under the official rules of the International Table Tennis Federation, ensuring competitive fairness and standardized play conditions.
The AI robot demonstrated advanced capabilities, including returning complex spins, controlling near-net shots, and generating backspin, while continuously improving its performance as the matches progressed.
Human players noted that the system adapts quickly during gameplay, making it increasingly challenging over time. The robot uses multiple high-speed cameras and sensor-based tracking to analyze ball speed, trajectory, and spin in real time.
While the robot won the majority of matches, it also lost two games to professional athletes, highlighting that human skill still retains an edge in certain competitive scenarios.
Researchers say the experiment reflects rapid progress in AI-driven robotics, suggesting that future developments could further narrow the gap between human and machine performance in sports.