A.I. Godfather Geoffrey Hinton ‘Made Up a Meeting’ to End a Call with Elon Musk Early

Elon Musk and Geoffrey Hinton, a pioneering artificial intelligence (A.I.) academic often referred to as the “Godfather of A.I.,” are seemingly aligned on the threats posed by the emerging technology to humanity. But when Musk, who last year launched the A.I. startup xAI, recently asked Hinton to join an advisory board, the billionaire’s “stream-of-consciousness” ramble caused Hinton to make up an excuse to leave the call 20 minutes in, the British-Canadian computer scientist told The Globe and Mail in an interview this week.

“I made up a meeting to get him off the phone,” said Hinton. “He’s not going to like that. Because him being the center of attention, that’s what he wants. I don’t think he’s my friend anymore.”

Through an academic career focused on neural networks—a machine learning process modeled on neurons of the human brain that enables computer programs to learn skills by analyzing data—Hinton played a critical role in shaping the foundations of A.I. systems powering the likes of OpenAI’s ChatGPT. But since leaving his role at Google (GOOGL) in 2023 after a decade-long tenure, he’s turned into one of the most vocal critics of how Big Tech companies are developing the technology.

Musk, too, has long warned of A.I.’s dangers and reiterated Hinton’s cautionary claims, tweeting last year that the academic “knows what he’s talking about.” But Musk himself, alongside powerful figures like Peter Thiel (a key investor in numerous A.I. labs) and Mark Zuckerberg, are part of the problem, according to Hinton.

While discussing how A.I. training data must be curated to avoid bots being used without a moral compass, Hinton pointed towards the need to regulate A.I. leaders to achieve such goals. “The only thing that could possibly keep Elon and Peter Thiel and Zuckerberg under control is government regulation,” he told The Globe and Mail. “That’s the only thing strong enough to do the job.”

Earlier this month, Hinton also endorsed an open letter signed by more than a dozen former OpenAI employees that claimed current whistleblower protections are insufficient and argued that A.I. companies like OpenAI are clamping down on criticism from workers championing safety concerns. The letter followed the departure of several top OpenAI employees like Ilya Sutskever, who previously led a failed ouster of OpenAI’s CEO Sam Altman.

Geoffrey Hinton has mentored many A.I. power figures

Sutskever and computer scientist Alex Krizhevsky were both former students of Hinton’s at the University of Toronto and in 2012 worked alongside the professor to make breakthroughs in neural networks. The following year, Google shelled out $44 million to acquire a company started by the trio and brought all three on as employees. Hinton subsequently received the Turing Award, known as the “Nobel Prize of Computing,” alongside fellow academics Yoshua Bengio and Yann LeCun for their contributions towards neural networks and A.I.’s development.

Hinton doesn’t necessarily regret his involvement in the field. “If I hadn’t done it somebody else would have,” he said, adding that “it might have all happened a few weeks later.” But the computer scientist is continuing to ring the alarm on the technology’s dangers and his belief that unless preventative measures are taken, A.I. machines will eventually become smarter than humans and go rogue. A.I. agents designed merely to organize holidays and buy packages for humans, for example, “will quickly realize that the more control they can get, the more efficient they can be at doing it,” said Hinton. “They control the world so that they can achieve things. So in that sense, these machines are like little kids. And we’re like the parents who have no clue how to parent them.”

For Hinton, the dire realities of A.I. have only become more pressing. While he previously believed that the technology had the potential to wipe out humanity in 30 to 50 years, that figure has since shifted to between five and 20 years. “There’s a 50-50 chance A.I. will get smarter than us,” he said. “When it gets smarter than us, I don’t know what the probability is that it will take over, but it seems to me quite likely.”