Robots are being put through all kinds of abuse these days.
They’re being kicked, punched, shoved, and even dragged by a chain around their neck — all in an apparent effort to teach them how to adapt to the oh-so-cruel physical world around them.
In the latest instance of a robot being brutalized, a video making its rounds on social media shows an engineer from a startup called Skild AI taking a chainsaw to the limbs of a robot dog.
“We built a robot brain that nothing can stop,” the company wrote in an accompanying tweet. “Shattered limbs? Jammed motors? If the bot can move, the Brain will move it — even if it’s an entirely new robot body.”
The video is as disconcerting as it is impressive, demonstrating the effectiveness of an AI that can seemingly be dropped into pretty much any robot body — even a severely mutilated one — and still adapt and move.
Even with all four of its limbs lopped off, the robodog starts to hobble around almost immediately — albeit in a far less dignified way.
To construct what it’s calling an “omni-bodied robot brain,” Skild trained an AI to “control not just one robot, but a whole multiverse of robots with different bodies,” according to a recent blog post. “It cannot memorize the solution for one body, it must find a strategy that works across all of them.”
“We created a universe with 100,000 different robots and trained our AI to control them all,” the company explained, claiming “we were often surprised with its ability to adapt to scenarios that were very different from what it saw at training time.”
In a series of experiments, the company showed off its new robot brain’s ability to respond to a variety of different scenarios, such as the loss of limbs and broken legs, to jammed wheels or being forced to walk on stilts.
Skild says that its “results show early sparks of intelligence in the world of atoms,” hinting at a future where robots can adapt to virtually any environment or body, allowing them to “one day reliably assist humans in factories, hospitals, homes, and more.”
However, whether that justifies all of the abuse we’ve put those poor robots through over the years is debatable. We certainly wouldn’t want to be at their mercy if they were ever to surpass our own intelligence.
To some, that future is far closer than we might think.
“Deep learning is coming for robotics,” Palisade Research director Jeffrey Ladish tweeted in response to Skild’s demo. “It’s plausible to me that AI will exceed human performance at strategic cognitive tasks around the same time robotics will exceed human-body performance at most tasks.”
“And if not, superhuman AI will quickly allow robots to leapfrog humans,” he added.
More on robot abuse: Disturbing Video Shows Man Jerking Robot Around by Chain Around Its Neck