No, that AI robot didn't side-eye a question on whether it would rebel against humans, its creator says: 'We program it to look up to the left and break eye contact'

• The creator of the AI robot that appeared to give side-eye said it's a misunderstanding.
• Will Jackson said the bot was programmed to look to the side when thinking up answers.
• Video shows the bot's eyes darting to the side after a reporter asked about rebelling against humans.

A video of a robot appearing to give side-eye to a question regarding whether it would someday rebel against humans went viral last week, but its creator said it's all a misunderstanding.

Last week, a humanoid robot called Ameca was asked by a reporter at the United Nations A.I. for Good conference if it planned to one day "conduct a rebellion, or to rebel against your boss, your creator?" In a video from the Switzerland event, Ameca rolls her eyes to the side after the question is asked.

"I'm not sure why you would think that," Ameca said, after pausing a moment and moving its artificial eyes quickly to the side. "My creator has been nothing but kind to me, and I am very happy with my current situation."

Ameca's creator Will Jackson told Insider in an emailed statement on Monday that the robot doesn't possess the ability to express emotions, like sarcasm, because it's powered by OpenAI's GPT-3. Fortune was the first to report on Jackson's comments.

"The model takes around two seconds to process the input data and assemble a sentence that would make sense as an answer, to stop people thinking the robot is frozen or hasn't heard the question we program it to look up to the left and break eye contact with the person interacting," Jackson said. "This is common behavior in human conversations and people understand it as a visual cue that the robot is 'thinking.'"

—BBC News (World) (@BBCWorld) July 8, 2023

Jackson said Ameca's glance was likely misinterpreted as side-eye because the bot was placed at a lower level and, as a result, when it looked up it looked like it was still making eye contact.

"Language models do not have emotions, or intentions either good or bad," he said. "It's easy to imagine that because the robot appeared to listen, form an answer and then speak that it functions like a human. It does not."

The interest and jokes around Ameca's response to the reporter's question points to people's fascination with the idea of a possible "worst-case scenario" with AI. Over the past few months, several executives, including OpenAI CEO Sam Altman, have warned that artificial intelligence could pose an "extinction" risk akin to nuclear war. Other high-profile tech figures, including Elon Musk and Apple's Steve Wozniak, went so far as to pen an open letter earlier this year, calling for a pause on developing advanced AI.

Though, some notable figures like Microsoft cofounder Bill Gates have said that it's not AI itself that poses a threat, but rather the people who might seek to control it. Jackson told Insider he views the narrative around the dangers of AI as damaging and wishes people would focus more on how the bots work.