Robot Defeats Elite Players at Table Tennis

During a rally, Ace's advantage becomes clear. Unlike a human who relies on two eyes and instinct, the eight-jointed robotic arm draws on multiple cameras positioned around the court to track the ball's position, speed, and spin simultaneously. By zooming in on the ball's logo, the system estimates the axis of rotation within milliseconds — then selects the best return shot based on tactics refined through roughly 3,000 hours of simulated games. Yet Sony deliberately kept Ace fair. Michael Spranger, president of Sony AI, explained that building a machine capable of overpowering humans with sheer speed would be easy, but the true goal was to win through superior decision-making and genuine skill — keeping the robot's capabilities comparable to an athlete who trains twenty hours per week.

The achievement extends a long tradition of using games to test AI, from chess to video games, but represents a significant leap by moving intelligent machines from simulated environments into the unpredictable physical world. Spranger described the past year as a "ChatGPT moment for robotics," suggesting the technology could benefit manufacturing and other industries. Not all experts are equally impressed, however. John Billingsley, a retired mechatronics professor who pioneered robot table tennis research in 1983, argued that Sony's multi-camera system gives the machine an unfair advantage over two-eyed humans. Players who faced Ace reported a different kind of challenge: the robot has no eyes to read, no body language to interpret, and feels no pressure in tight games — an eerie reminder that competing against a machine is unlike facing any human opponent.