Sony's terrifyingly fast robot arm is finally beating pro-level table tennis players at their own game

If you're anything like me, you've laughed yourself sick at the various attempts robots have made to enter the arena of human sports. The World Humanoid Robot Games was, well, something of a disaster, and the Terminator-like uprising seems to be a long way away.

It was only a matter of time until one became genuinely good enough to challenge the pros, though, and by the looks of it, Sony's new robot arm is exactly that (via AP News). It's called Ace, it plays table tennis for a living, and it's seriously fast—so much so that it's capable of challenging (and occasionally beating) expert players on a regular basis.

The bot uses The Power of AI™ to not only track incoming balls with a mirror-based camera system, but to use reinforcement learning to improve its game, which it's been doing at an increasingly rapid rate. Sony says it's now the "first time a robot has achieved human, expert-level play in a commonly played competitive sport in the physical word"—and the footage is downright impressive.

Robots are normally either pretty slow, lumbering creations, or jittery little fellas, bouncing around all over the place and falling over. This robot arm, though, seems confident in its movements—and the turn of speed it's capable of reacting with looks to be something of a turning point for modern robotics.

Michael Spranger, president of Sony AI, seems to agree. "Speed is really one of the fundamental issues in robotics today, especially in scenarios or environments that are not fixed," says the AI chief.

"We see a lot of robots that are in factories that are very, very fast," says Spranger. "But they’re doing the same trajectory over and over again. With this technology, we show that it’s actually possible to train robots to be very adaptive and competitive and fast in uncertain environments that constantly change."

More than that, it's the fact that the robot plays the game in essentially the same way as a human player. "It’s very easy to build a superhuman table tennis robot," Spranger continues. "You build a machine that sucks in the ball and shoots it out much faster than a human can return it.

"But that’s not the goal here. The goal is to have some level of comparability, some level of fairness to the human, and win really at the level of AI and the level of decision-making and tactics and, to some extent, skill."

Cool. Also, mildly terrifying. Imagine two of these things hooked up to some tank tracks, and you've essentially got a Johnny No. 5 competitor who's been to the gym. Not that this sort of thing keeps me awake at night. No siree.