Related Articles
John J. Hopfield and Geoffrey E. Hinton received the Nobel Prize in Physics on Tuesday for discoveries that helped computers learn more in the way the human brain does, providing the building blocks for developments in artificial intelligence.
The award is an acknowledgment of A.I.’s growing significance in the way people live and work. With their ability to make sense of vast amounts of data, artificial neural networks already have a major role in scientific research, the Nobel committee said, including in physics, where it is used to design new materials, crunch large amounts of data from particle accelerators and help survey the universe.
The machine learning breakthroughs of Dr. Hopfield and Dr. Hinton “have showed a completely new way for us to use computers to aid and to guide us to tackle many of the challenges our society face,” the Nobel committee said.
Neural networks — systems that learn skills by analyzing data and are named after the web of neurons in the human brain — are a part of everyday internet services, including search engines like Google, talking digital assistants like Apple’s Siri and chatbots like OpenAI ChatGPT. These services are rooted in mathematics and computer science, not physics.
But research by Dr. Hopfield and Dr. Hinton in the late 1970s and early 1980s helped influence the development of the digital neural networks that have become part of the fabric of the modern internet.
“If there was a Nobel Prize for computer science, our work would clearly be more appropriate for that,” Dr. Hinton, a recipient of the 2018 Turing Award who has been called the “godfather of A.I.,” said in a phone interview with The New York Times. “But there isn’t one.”
Dr. Hinton left his job as a researcher at Google last year, in part so that he could freely discuss his concerns that the A.I. technologies he helped create could end up harming humanity. In a call during the Nobel announcement in Stockholm on Tuesday, Dr. Hinton expressed worries over machine learning and said it would have an extraordinary influence on society.
“It will be comparable with the Industrial Revolution,” he said. “Instead of exceeding people in physical strength, it’s going to exceed people in intellectual ability. We have no experience of what it’s like to have things smarter than us.”
While Dr. Hinton expressed his concerns, he also said that the advanced technology would bring much better health care. “It’ll mean huge improvements in productivity,” he said. “But we also have to worry about a number of possible bad consequences, particularly the threat of these things getting out of control.”
Speaking with The Times, he said that winning the Nobel Prize could bring more attention to his concern about the future of the technology. “Having the Nobel Prize could mean that people will take me more seriously,” he said.
In a news conference on Tuesday, Dr. Hopfield compared advances in A.I. with the splitting of the atom, which led to both deadly bombs and bountiful energy.
“One is accustomed to having technologies which are not only good or only bad, but have capabilities in both directions,” he said. But he added, “You want to have some idea of how you can control the system, and how you can prevent disasters from occurring.”
‘Ground zero’ for modern A.I.
Since it was first awarded in 1901, the Nobel Prize in Physics has honored research from the discovery of subatomic particles to gravitational waves and supermassive black holes. But in some years, the committee has acknowledged the application of physics to other disciplines, like in 2021 for work contributing to understanding of climate change.
For this year’s award, the committee emphasized the way that Dr. Hopfield’s and Dr. Hinton’s work in biology and computer science had roots in the physical sciences.
While it may seem an unusual fit under the umbrella of physics, Dmitry Krotov, a physicist with the Massachusetts Institute of Technology and IBM who has published several papers with Dr. Hopfield in recent years, said the boundaries between fields were “somewhat artificial,” adding that “what is nice about physics is that historically, it is always expanding.”
John J. Hopfield, a Chicago native, is an emeritus professor at Princeton known for seminal discoveries in computer science, biology and physics. He is 91, and the third oldest Nobel physics laureate.
He began his career at Bell Laboratories in 1958 as a physicist studying the properties of solid matter, but felt limited by the boundaries of his field. He moved to the University of California, Berkeley, as an assistant professor in 1961 and joined the physics faculty at Princeton in 1964. Sixteen years later, he moved to the California Institute of Technology as a professor of chemistry and biology, and in 1997, returned to Princeton, this time in the department of molecular biology.
In the 1980s, his work focused on how the processes of the brain can inform how machines save and reproduce patterns. He explained in an interview that his work came from an initial intrigue with the connections between physics and biology. “Biology is just a physical system, but a very complicated one,” he said.
In 1982, Dr. Hopfield developed a model of neural networks, today known as the Hopfield network, to describe how the brain recalls memories when fed partial information, similar to the method your brain uses to remember a word on the tip of your tongue.
This ability is called associative memory. In describing the Hopfield network’s nodes and their linkages, Dr. Hopfield’s work showed that their behavior resembled the physics that explains how the spins of nearby atoms affect one another.
He did not anticipate that his work on neural networks would ever be useful in machine learning. But there’s a “natural handshake” between questions in A.I. and biology, he said.
The years leading up to the Hopfield network were like an “A.I. winter,” Dr. Krotov said. But Dr. Hopfield’s work in 1982 “was the major driving force that ended that period,” he said. He continued, “It’s the ground zero for the modern era of neural networks.”
This point was affirmed by the neuroscientist and computer scientist Terry Sejnowski, now at the Salk Institute for Biological Sciences, who studied under Dr. Hopfield and later became a key collaborator of Dr. Hinton’s.
Work on the Hopfield network “drew many physicists into the machine learning field,” he said. “In many ways, it helped create the field.”
The road to chatbots
Geoffrey E. Hinton, born just outside London, has lived and worked mostly in the United States and Canada since the late 1970s. He is a professor of computer science at the University of Toronto.
Dr. Hinton, 76, began researching neural networks as a graduate student at the University of Edinburgh in the early 1970s, a time when few researchers believed in the idea.
In 1985, Dr. Hinton and his colleagues developed a new neural network they named the Boltzmann machine.
As with Dr. Hopfield’s research, the nodes in Dr. Hinton’s Boltzmann machine could be described with physics. But instead of spin, they used the Boltzmann equation, named for the statistical physics pioneer Ludwig Boltzmann. The equation describes the energy of a system.
Yann LeCun, the chief A.I. scientist at Meta, pointed out that the Hopfield network and Boltzmann machine were not used by modern A.I. technologies. But he said that modern technologies were very much influenced by these early, physics-related creations from Dr. Hopfield and Dr. Hinton.
Their work, he explained, inspired many scientists to begin exploring neural networks, which most academics had previously dismissed as a scientific dead end.
“It made the whole neural net field kosher again,” he said. “Before this, it was taboo.”
Following on the Boltzmann work, Dr. Hinton and his collaborators developed a new form of neural network based on a mathematical idea called “backpropagation.” He and others, including Dr. LeCun, nurtured this idea for the next few decades, largely at universities in Canada and Europe.
Ultimately, Dr. Hinton and two of his graduate students at the University of Toronto made a breakthrough with the technology in 2012, and he joined Google. More recently, he shared the Turing Award with Dr. LeCun as well as Yoshua Bengio, a professor of computer science at the University of Montreal whose research focuses on ensuring that A.I. is developed safely.
Packed inboxes, canceled appointments
Neither man was expecting to be named a Nobel physics laureate.
Dr. Hopfield said he was “astonished,” when asked in an interview with The Times how he felt about winning the Nobel Prize.
Currently in England, he was out getting a flu shot and having coffee during the announcement. He returned home to an inbox overflowing with congratulatory messages.
“I’ve never had so much email before in my life,” Dr. Hopfield said, adding that it took some digging for him to discover what, exactly, he was being congratulated for.
Dr. Hinton said he learned of the prize while staying in a “cheap hotel” in California.
“I was going to get an M.R.I. scan today, but I think I’ll have to cancel that,” he said.
Who received the 2023 Nobel Prize in Physics?
The prize was shared by Pierre Agostini, Ferenc Krausz and Anne L’Huillier for work that let scientists capture the motions of subatomic particles moving at impossible speeds.
Who else has received a Nobel Prize in the sciences this year?
On Monday, the prize in Physiology or Medicine went to Victor Ambros and Gary Ruvkun for their discovery of microRNA, which helps determine how cells develop and function.
When will the other Nobel Prizes be announced?
The Nobel Prize in Chemistry will be awarded on Wednesday by the Royal Swedish Academy of Sciences in Stockholm. Last year, the prize went to Moungi G. Bawendi, Louis E. Brus and Alexei I. Ekimov for discovering and developing quantum dots that are expected to lead to advances in electronics, solar cells and encrypted quantum information.
The Nobel Prize in Literature will be awarded on Thursday by the Swedish Academy in Stockholm. Last year, Jon Fosse of Norway was honored for plays and prose that gave “voice to the unsayable.”
The Nobel Peace Prize will be awarded on Friday by the Norwegian Nobel Institute in Oslo. Last year, Narges Mohammadi, an activist in Iran, was recognized “for her fight against the oppression of women in Iran and her fight to promote human rights and freedom for all.” Ms. Mohammadi is serving a 10-year sentence in an Iranian prison where her attorneys have raised concerns about her well-being.
Next week, the Nobel Memorial Prize in Economic Sciences will be awarded on Monday by the Royal Swedish Academy of Sciences in Stockholm. Last year, Claudia Goldin was awarded for her research uncovering the reasons for gender gaps in labor force participation and earnings.
All of the prize announcements are streamed live by the Nobel Prize organization.
Adam Satariano contributed reporting.
<
