AI ping pong robot beats top human players, but don’t freak out yet

Sony researchers said the Ace robot arm, trained with reinforcement learning in simulation and guided by nine cameras, could challenge elite table tennis players and win some matches.

The result marks a major step for autonomous physical AI, especially in fast perception and real-time reaction.

But human players quickly found exploitable weaknesses, including simple tactics like a knuckle serve, showing the system is not yet fully robust.

The work highlights both the promise and the limits of advanced robotics, along with concerns about possible security or battlefield uses.