Sony AI’s Ace robot defeated top-level human table tennis professionals, with results published in Nature.
Key Takeaways
The robot is named Ace, built by Sony AI, and its results appear in a peer-reviewed Nature paper (s41586-026-10338-5).
Performance represents a step-change: roughly one year ago, Google DeepMind’s table tennis robot was SOTA but only matched non-players.
The setup requires very bright lighting to track the ball, indicating a specialized controlled environment rather than general-purpose conditions.
Speed of progress mirrors software AI gains: from amateur-level to beating professionals in a single year.
Hacker News Comment Review
Dominant reaction is a Deep Blue analogy: commenters expect table tennis robots, like chess engines post-1997, to become permanently uncontestable by humans within years.
A contested fairness point: human players rely on reading opponent body language and kinematics to anticipate shots; a robot’s alien movement profile may neutralize trained human prediction skills, making the matchup asymmetric rather than a clean benchmark.
Skeptics argue the milestone is overstated until a humanoid robot operating under human kinematic constraints achieves the same result, distinguishing specialized arm systems from general physical intelligence.
Notable Comments
@dmurray: traces the exact one-year gap from DeepMind’s amateur-level SOTA to Ace beating professionals, flagging it as an unexpectedly rapid jump for physical robotics.
@aslushnikov: “This feels to me very much like a Deep Blue moment” – argues the field quietly crossed a threshold most dismissed as still years away.
@janalsncm: links the Nature paper and flags that the arena must be lit very brightly, surfacing an under-discussed hardware constraint.
