Tell Us

ChatGPT can now control Boston Dynamics robots. Is it time to panic?

How worried are you about the rapid rise of artificial intelligence?

A Boston Dynamics robotic dog waits backstage at UMass Lowell Chancellor Julie Chen’s Inauguration in April 2023. Jessica Rinaldi/Globe Staff

By Peter Chianca

May 5, 2023

1 minute to read

OK people, this isn’t funny anymore. It seems that the Boston Dynamics robots — which are off-putting enough when they’re being controlled the usual way (er, via algorithms, probably, or maybe a very sophisticated clicker) — can now be controlled by ChatGPT. Yes, that ChatGPT, the one known for doing middle schoolers’ homework for them and making stuff up.

According to the Boston Globe, a software outfit in Florida called Levatas has figured out how to communicate with the robots using the recently released artificial intelligence chatbot. They say it will allow everyday workers to control the robots by using “natural language commands.”

“It can help factories maintain productivity … We’re literally keeping humans safe in an industrial setting,” Levatas founder Chris Nielsen told the Globe, never once using the phrase “new robot overlords.”

But nobody is fooling Gary Marcus, professor emeritus in psychology and neural science at New York University, who asked hypothetically, “What could possibly go wrong? A lot.”

MORE ON BOSTON DYNAMICS:

“ChatGPT still lacks reliable common sense,” Marcus told the Globe. “Put that in a robot and you could easily wind up with a robot doing dangerous things, or just not doing what you expect.”

It’s a concern that apparently played into “godfather of AI” Dr. Geoffrey Hinton’s recent decision to step down from his post at Google, and even tell the New York Times that he has regrets about his life’s work.

“It is hard to see how you can prevent the bad actors from using it for bad things,” Hinton told the Times, noting also that, according to the newspaper, “he fears a day when truly autonomous weapons — those killer robots — become reality.” But other than that Mrs. Lincoln, how was the play?

So how about you? Do you share Marcus and Hinton’s concerns that AI could spiral out of control? Or do its benefits outweigh its risks? Answer the Boston.com poll below, or e-mail [email protected], and we may use your response in an upcoming article.