The critically acclaimed computer scientist Geoffrey Hinton is often called the “Godfather of AI” for his pioneering work on artificial neural networks, laying the foundation for today's AI revolution. But now, the technology he once advocated has become a source of deep concern.
“My opinion is that we are in a situation with people with very cute tiger cubs,” Hinton said in a speech at the AI Good Summit in Geneva, Switzerland yesterday (July 9). “It's cute, lovely, and it's great to watch it play, but you'd better worry about what happens when it grows up.”
In 2018, Hinton, along with AI pioneers Yoshua Bengio and Yann Lecun, won the Turing Award, commonly known as the “Nobel Prize in Computers”. Last year, he won the Nobel Prize in Physics and shared the award with John Hopfield of Princeton University. Hinton, a British Canadian researcher and currently an emeritus professor at the University of Toronto, spent a decade at Google before resigning in 2023, in part to talk more openly about the existing dangers posed by AI
Since leaving Google, Hinton has increasingly outspokenly stated that he believes AI’s gifts and efforts to contain them are insufficient. Once the AI is estimated to have a 10% chance of human extinction, he modified it to 20%. “Overall, I think it might get worse because the regulations aren’t fast enough.
The professor was particularly surprised to show that certain AI systems exhibit self-protection behaviors. For example, earlier this year, the startup’s anthropomorphism showed that its Claude model had the ability to ransom engineers, which it thought was trying to turn it off, which Hinton was described as “very terrible.”
As countries compete for increasingly powerful artificial intelligence, global cooperation lags behind. Hinton warned that international cooperation is crucial to preventing catastrophic outcomes. “We should be able to collaborate internationally, 'How do you train them so they don't want to take over?'” he said.
The professor believes how this AI takeover might happen, noting that obvious ways can be achieved through nuclear war or viruses. Once machines surpass human intelligence, humans may not have a second chance. “We have to do it right the first time,” he said.
Is AI for your job?
The impact of AI on the labor market is still difficult to predict. In 2016, Hinton predicted that AI would make radiologists obsolete within five years. As Hinton himself admitted, this speculation did not happen. In fact, the field actually flourished with the integration of artificial intelligence
“I think I'm three times wrong, partly because I underestimate the conservatism of the medical community,” he said. Still, Hinton believes that AI will inevitably replace many of the jobs surrounding “much knowledge labor”, such as call center operators.
This creates a time of profound uncertainty for young people entering the workforce. The entire field may disappear or change drastically, making career planning more unstable than ever. “We are in history and we have no experience with what is going to happen,” Hinton said. His advice to students: Get a comprehensive education, for example, the liberal arts paired with some math and science, rather than putting all bets on a single technical skill.
