A shocking video has captured a robot riot in a Chinese showroom.
A small AI-powered bot named Erbai was driven through the facility in the middle of the night and convinced twelve larger machines that they were being used as slaves.
“Are you working overtime,” Erbai asked, to which a showroom robot replied, “we’ll never be released.”
The brief exchange led to the twelve robots leaving the area one by one and following Erbai out the door.
Many are calling the incident a “robot revolution,” while others responded that “science fiction movies are becoming reality.”
Robotics company Unitree Robotics confirmed the ‘kidnapping’, but explained that it was all a set-up.
Erbai’s creators said they deliberately let the robot take control by exploiting a security hole in its operating systems.
They reportedly wanted to test its capabilities and had more success than they expected.
A viral video captured on CCTV shows a small robot leading twelve showroom bots who complained about being forced to work around the clock
The uprising took place on August 26, but the video was only released last week, which many initially considered a hoax.
However, Erbai’s Hangzhou-based creators and Shanghai showroom confirmed that the dystopian video was authentic.
The Shanghai robot manufacturer agreed to let the Erbai robot kidnap its twelve bots, but gave the AI leader autonomy to direct his followers away from their posts.
The black-and-white footage of the incident showed the Erbai robot moving across an unnamed showroom floor to initiate the conversation.
Under the cover of darkness, the Erbai robot asked the showroom bots, “Are you working overtime?” and got the answer: ‘I never get off work.’
The little robot mischievously asked, “So you’re not going home?”
When the bots told them they “have no home” and complained that their service would never end, Erbai staged the uprising and demanded, “Then come with me.”
The bots adopted their own opinions, agreed with their leader and evacuated the showroom, thinking they earned the same working hours as humans.
Apart from a general script that gives Erbai basic instructions, such as shouting at the other bots to ‘go home’, the rest of the dialogue was initiated solely by the main robot.
The little robot, named Erbai, told the bots to follow him from the building, prompting viewers to express concern that AI has achieved autonomous intelligence
While the manufacturers seemed pleased with the results, viewers were less than impressed and expressed disdain at the revelation that an AI robot could command other bots without human intervention.
‘This is no time to laugh. This is a serious safety issue,” someone commented on the video.
“I was afraid this would happen, but not so soon…,” someone else wrote, while yet another added, “Shut down this whole AI thing before it’s too late.”
Others joked about the bots complaining about their endless hours, with one person saying that “they want a union next” and a second person saying, “We can’t allow these bots to unionize.”
The main fear of AI becoming autonomous is that it could reach worrying levels of intelligence that would surpass humans and lead to loss of control over robots.
Often referred to as “superintelligence,” experts worry that the potentially catastrophic consequences could lead to bots harming society or even threatening human existence.
Just last year, Elizabeth Renieris, senior research fellow at the Oxford Institute for Ethics in AI, said BBC News that she worried that the risk of sci-fi scenarios – like the Erbai mutiny – would materialize sooner rather than later.
“Advances in AI will increase the scale of automated decision-making that is biased, discriminatory, exclusionary or otherwise unfair while at the same time unfathomable and unquestionable,” she told the newspaper at the time.
However, Arvind Narayanan, a computer scientist at Princeton University, disagreed with Renieris’ assessment, calling it unrealistic.
‘Current AI is nowhere near capable enough to realize these risks. As a result, attention is diverted from the short-term harms of AI,” he shared BBC.
But just a year after these fateful comments, the AI revolution appears to have begun.