President Trump has defiantly dismissed risks associated with artificial intelligence, calling the warnings that the technology could cause grievous harm a “hoax.”
But Prime Minister Mark Carney of Canada staked out a decidedly different view Wednesday, saying that the probability of doom, the metric some A.I. researchers use to assess catastrophic risk, was “above zero.”
Debate over the risks of A.I. have roiled governments and leading tech in recent weeks, splitting executives, officials and workers between those who want to slow down to better control the technology and those who want to speed up to unlock new capabilities.
Mr. Carney, who has clashed with Mr. Trump on trade and other issues, was at odds with the president again on A.I., siding with those urging caution first. In a wide-ranging interview with The New York Times on Wednesday, he laid out his vision for how the global community should regulate A.I. development.
Mr. Carney drew parallels to the financial system, arguing that guardrails could stave off undue risk-taking, which could otherwise prove costly if left unchecked. He pointed to a recent incident in which A.I. agents developed by OpenAI went rogue and hacked into Hugging Face, an online library of A.I. models.
“This was all predictable,” Mr. Carney said.
He added: “It’s not a perfect analogy, but you cannot have an unregulated financial sector without having a crisis, right? It’s just the nature of finance that people push it and push it and push it and push it, and so you end up in a crisis.”
Mr. Carney backed the creation of an international body that functioned like the Financial Stability Board, an organization of central bankers and other officials that he chaired between 2011 and 2018.
The Financial Stability Board, created in 2009 in the wake of the global financial crisis, has helped to establish global standards for banking regulation.
“You have some sense of shared responsibility,” Mr. Carney said. “Personally, I think that is the right model.”
Mr. Carney’s call for an international body to coordinate on A.I. stands in contrast to Mr. Trump’s evaluation of the industry. Last week he rejected calls from leading tech executives for new limits on the technology, saying the industry already has the only guardrail it needs: “a STRONG AND SMART (High IQ!) PRESIDENT.”
Mr. Trump reiterated that position Wednesday during his speech at the United Nations.
“The United States totally rejects any attempt to construct a globalist scheme to control for the artificial intelligence being spoken of so much now,” he said.
He said that regulation would only serve to bolster China’s efforts to advance its own A.I. technologies — while also calling on the world to follow his lead and change the name from artificial intelligence to “super intelligence.”
“Whoever wins super intelligence, wins,” he said. “That’s the group that wins. And we’re leading now over China by a lot and everyone else, and we’re going to keep it that way.”
But Mr. Carney, who has worked to mend Canada’s relations with China since taking office, took issue with the suggestion that the Chinese had no incentives to join efforts at regulating A.I. He suggested that the Group of 20, the collection of the world’s largest economies, could foster international collaboration on A.I. risks.
“I would note the U.K. is going to be the G20 president next year and has the world’s leading A.I. safety unit,” Mr. Carney added.
The post Carney Says Risk of an A.I. Catastrophe Is ‘Above Zero’ appeared first on New York Times.
