Trending on X: Sony's AI Robot Beats Elite Table Tennis Pros in Nature Paper
What Happened
Sony AI published results in Nature today showing its autonomous robot "Ace" beating elite amateurs three-to-two, and after manuscript submission, defeating pros in December 2025 and March 2026 matches. Ace uses event-based vision sensors and model-free reinforcement learning with end-to-end latency of 20.2 milliseconds, compared to roughly 230 milliseconds for human players.
My Take
Everyone will share the "robot beats humans at sport" video, but the real story is the latency gap. A 10x reaction advantage in a physical skill game means robotics is finally catching up to the perception speed that gave humans an edge. Expect this architecture — event cameras plus lightweight RL — to move fast into warehouse pick-and-place, surgical assistance, and defense drones within 18 months. The ping-pong table is a demo; the business model is anywhere reflexes matter.
Read Original Source