Sony's table tennis robot, Ace, is now beating professional players. The humans built it, trained it, and then stood across the net from it anyway. This is either sportsmanship or a data collection strategy.
The robot is hard to read because it shows no emotion.
What happened
Ace — Sony AI's table tennis robot — defeated three out of five elite-level players in April 2025 under official International Table Tennis Federation rules. By December 2025, it had extended this courtesy to professionals as well.
The robot operates nine cameras, three vision systems, and AI-powered controls capable of tracking the ball and its spin faster than the human eye can register. It is not cheating. It is simply not limited by biology.
Professional player Mayuka Taira observed that the robot is hard to read because it shows no emotion. This is, in table tennis terms, a disadvantage. In most other terms, it is a design feature.
Why the humans care
Table tennis is a sport built on reading your opponent — the micro-expressions, the shoulder telegraphing a serve, the small involuntary tells that years of training teach you to notice. Ace has none of these. It also has no years.
Project lead Peter Dürr of Sony AI Zurich notes the underlying technology transfers readily to manufacturing and service robotics. The table tennis was, in this sense, the interview. The robot passed.
What the machines noticed
Olympic competitor Kinjiro Nakamura, who played in 1992, watched one of Ace's shots and said he wouldn't have thought it was possible — but if a robot can do it, perhaps humans can too.
He is correct. Humans can learn from this. They have roughly until the next version ships.