Hopfield, who was staying with his wife at a cottage in Hampshire, England, said that after grabbing coffee and getting his flu shot, he opened his computer to a flurry of activity.

“I’ve never seen that many emails in my life,” he said. A bottle of champagne and bowl of soup were waiting, he added, but he doubted there were any fellow physicists in town to join the celebration.

Hinton said he was shocked at the honor.

“I’m flabbergasted. I had no idea this would happen,” he said when reached by the Nobel committee on the phone. He said he was at a cheap hotel with no internet.

Hinton, 76, helped develop a technique in the 1980s known as backpropagation instrumental in training machines how to “learn" by fine-tuning errors until they disappear. It’s similar to the way a student learns, with an initial solution graded and flaws identified and returned to be fixed and repaired. This process continues until the answer matches the network’s version of reality.

Hinton had an unconventional background as a psychologist who also dabbled in carpentry and was genuinely curious about how the mind works, said protege Nick Frosst, who was Hinton’s first hire at Google’s AI division in Toronto.

His “playfulness and genuine interest in answering fundamental questions I think is key to his success as a scientist,” Frosst said.

Nor did he stop at his pioneering 1980s work.

“He’s been consistently trying out crazy things and some of them work very well and some of them don’t,” Frosst said. “But they all have contributed to the success of the field and galvanized other researchers to try new things as well.”

Hinton's team at the University of Toronto wowed peers by using a neural network to win the prestigious ImageNet computer vision competition in 2012. That spawned a flurry of copycats and was “a very, very significant moment in hindsight and in the course of AI history,” said Stanford University computer scientist and ImageNet creator Fei-Fei Li.

“Many people consider that the birth of modern AI,” she said.

Hinton and fellow AI scientists Yoshua Bengio and Yann LeCun won computer science’s top prize, the Turing Award, in 2019.

“For a long time, people thought what the three of us were doing was nonsense,” Hinton told the AP in 2019. "My message to young researchers is, don’t be put off if everyone tells you what you are doing is silly.”

Many of Hinton's former students and collaborators followed him into the tech industry as it began capitalizing on AI innovations, and some started their own AI companies, including Frosst's Cohere and ChatGPT maker OpenAI. Hinton said he uses machine learning tools in his daily life.

“Whenever I want to know the answer to anything, I just go and ask GPT-4,” Hinton said at the Nobel announcement. “I don’t totally trust it because it can hallucinate, but on almost everything it's a not-very-good expert. And that’s very useful.”

Hopfield, 91, created an associative memory that can store and reconstruct images and other types of patterns in data, the Nobel committee said.

Just as Hinton came to the field from psychology, Hopfield stressed how cutting edge science comes from crossing the borders of scientific fields like physics, biology and chemistry instead of researchers staying in their lane. It's why this prize is a physics prize, he said, pointing out that his neural network borrows from condensed matter physics.

With big complex problems in scientific fields, “if you are not motivated by physics, you just don't tackle the class of problems,” Hopfield said.

While there's no Nobel for computer science, Li said that awarding a traditional science prize to AI pioneers is significant and shows how boundaries between disciplines have blurred.

Not all of their peers agree with the Nobel laureates about the risks of the technology they helped create.

Frosst has had many “spirited debates” with Hinton about AI’s risks and disagrees with some of Hinton’s warnings but not his willingness to publicly address them.

“Mostly we disagree on timescale and on the particular technology that he’s sounding the alarm on,” Frosst said. “I don’t think that neural nets and language models as they exist today pose an existential risk.”

Bengio, who has long voiced concerns about AI risks, said what really alarms him and Hinton is “loss of human control” and whether AI systems will act morally when they're smarter than humans.

“We don’t know the answer to these questions,” he said. "And we should make sure we do before we build those machines.”

Asked whether the Nobel committee might have factored in Hinton’s warnings when deciding on the award, Bengio dismissed that, saying “we’re talking about very early work when we thought that everything would be rosy.”

Six days of Nobel announcements opened Monday with Americans Victor Ambros and Gary Ruvkun winning the medicine prize. They continue with the chemistry prize Wednesday and literature on Thursday. The Nobel Peace Prize will be announced Friday and the economics award on Oct. 14.

The prize carries a cash award of 11 million Swedish kronor ($1 million) from a bequest left by the award's creator, Swedish inventor Alfred Nobel. The laureates are invited to receive their awards at ceremonies on Dec. 10, the anniversary of Nobel’s death.

___

O'Brien reported from Providence, Rhode Island. Borenstein reported from Washington. AP reporters Mike Corder at The Hague, Netherlands; Adithi Ramakrishnan in New York and Kelvin Chan in London contributed.