STOCKHOLM — Two artificial intelligence pioneers – John Hopfield and Geoffrey Hinton – won the Nobel Prize in physics on Tuesday for helping to create the building blocks of machine learning that revolutionized the way we work and live but also created new threats to humanity, one of the winners said. .
Hinton, who is known as the Godfather of Artificial Intelligence, is a Canadian and British citizen who works at the University of Toronto, and Hopfield is an American who works at Princeton.
“These two men are truly pioneers,” said Nobel physics committee member Mark Pearce. “They … did fundamental work, based on physical understanding that led to the revolution we see today in machine learning and artificial intelligence.”
The artificial neural networks he pioneered are used throughout science and medicine and “have also become part of everyday life, such as in facial recognition and language translation,” said Ellen Moons, a member of the Nobel committee at the Royal Swedish Academy of Sciences. .
Hinton predicts that AI will have a “huge impact” on civilization, improving productivity and healthcare.
“It will be compared to the Industrial Revolution,” he said in an open call with reporters and officials of the Royal Swedish Academy of Sciences.
“Instead of surpassing people in physical strength, it will surpass people in intellectual ability. We have no experience of what it would be like to have something smarter than us. And it will be very good in many aspects,” said Hinton.
“But we also have to worry about some bad consequences, especially the threat of these things getting out of control.”
The Nobel committee that honors the science behind machine learning and AI also expressed fears about the possible flipside. Moons said that while there are “enormous benefits, rapid development also raises concerns about our future. Collectively, humanity bears the responsibility to use this new technology in a safe and ethical way for the greatest benefit to humanity.”
Hinton shared that concern. He quit his role at Google so he could speak freely about the dangers of the technology he was helping.
“I’m concerned that the overall consequence of this could be a system that’s smarter than we are ultimately in control of,” Hinton said.
On Tuesday, he said he was surprised by the award.
“I was surprised. I didn’t know this would happen,” he said when contacted by the Nobel committee on the phone. He said he was in a cheap hotel without internet.
There was no immediate reaction from Hopfield.
Hinton, now 76, in the 1980s helped develop a technique known as backpropagation that has been instrumental in training machines to “learn” by looking for errors until they disappear. It is similar to how students learn from teachers in the beginning. The solution is evaluated and defects are identified and returned to be corrected and corrected. This process continues until the answer corresponds to the version of the network.
A team at the University of Toronto then stunned their peers by using neural networks to win the prestigious ImageNet computer vision competition in 2012. The win led to many imitators, giving birth to modern AI.
Hinton and fellow AI scientists Yoshua Bengio and Yann LeCun won the top prize in computer science, the Turing Award, in 2019.
“For a long time, people thought that what these three guys did was pointless,” Hinton told The Associated Press in 2019. at the same time. My message to young researchers is don’t get discouraged if everyone tells you what you’re doing is ridiculous.
And Hinton himself uses machine learning in his daily life.
“Whenever I want to know the answer to something, I just ask GPT4,” Hinton said in the Nobel announcement. “I don’t really believe it because I can hallucinate, but almost all of them are not very good experts. And that’s very useful.”
“Twenty years ago, I think people would have been happy to agree that a system with the capabilities of GPT-4 or (Google) Gemini had achieved general intelligence comparable to humans,” Hinton told the AP this spring. “You can answer any question in a sensible way and you’ve passed the test. But now that AI can do it, people want to change the test.
Hopfield, 91, created associative memory that can store and reconstruct images and other types of patterns in data, the Nobel committee said.
“The thing that fascinates me the most is still the question of how the mind came from the machine,” Hopfield said in a video posted online by The Franklin Institute after awarding the physics prize in 2019.
Hinton used the Hopfield network as the basis for a new network that uses a different method, known as the Boltzmann machine, which the committee said can learn to identify characteristic elements in certain types of data.
Six days of Nobel announcements opened Monday with Americans Victor Ambros and Gary Ruvkun winning the medicine prize for the discovery of a tiny piece of genetic material that acts as an on and off switch inside cells that helps control what the cell does and when it does it. If scientists can better understand how it works and how it lies, there may be powerful treatments for diseases like cancer.
The physics prize carries a cash award of 11 million Swedish kronor ($1 million) from the legacy left by the award’s creator, Swedish inventor Alfred Nobel. The winners were invited to receive the award at a ceremony on December 10, the anniversary of Nobel’s death.
The Nobel announcement continues with the chemistry prize on Wednesday and the literature prize on Thursday. The Nobel Peace Prize will be announced on Friday and the economic award on October 14.
___
Corder reported from The Hague, Netherlands and Borenstein reported from Washington. Associated Press reporter Matt O’Brien contributed from Providence, Rhode Island.
