Sony AI's Ace robot beats elite table tennis players for first time — machines now outperform humans in competitive physical sport

Sony AI unveiled Ace, a robot that defeated elite human table tennis players in matches during December 2025, and published results in Nature journal. This marks the first time a robot has beaten professional-level athletes in a competitive physical sport requiring real-time decision-making and precise motor control.

Ace uses computer vision (cameras that interpret what it sees) and adaptive strategy software to read incoming ball spin, speed, and angle, then adjusts paddle rotation and arm position mid-swing. Unlike factory robots that repeat pre-programmed moves, Ace makes live tactical choices during play — deciding whether to attack, defend, or place shots strategically based on opponent behavior.

This demonstrates that robots can now master complex, unpredictable physical tasks where millisecond timing and split-second judgment matter. Sports engineers, manufacturing firms, and surgical robotics teams will watch whether Sony's approach — learning from live human opponents — becomes the template for building robots that handle real-world variability, not just controlled lab conditions.