Geoffrey Everest Hinton (born December 6, 1947) is a British-Canadian computer scientist, cognitive scientist, and cognitive psychologist who won the Nobel Prize. He is known for his work on artificial neural networks, which earned him the nickname "the Godfather of AI." He is a retired professor at the University of Toronto.

From 2013 to 2023, he worked at Google Brain and the University of Toronto. In May 2023, he announced he would leave Google, saying he was worried about the risks of artificial intelligence (AI) technology. In 2017, he helped start the Vector Institute in Toronto and became its chief scientific advisor.

In 1986, Hinton worked with David Rumelhart and Ronald J. Williams to write a widely read paper that helped popularize the backpropagation algorithm for training multi-layer neural networks. Although they were not the first to suggest this method, their work was important. Hinton is seen as a key leader in the deep learning field. The AlexNet system, created with his students Alex Krizhevsky and Ilya Sutskever for the ImageNet challenge in 2012, was a major breakthrough in computer vision.

In 2018, Hinton won the Turing Award with Yoshua Bengio and Yann LeCun for their work on deep learning. These three are sometimes called the "Godfathers of Deep Learning" and have given public talks together. In 2024, he shared the Nobel Prize in Physics with John Hopfield for discoveries that helped develop machine learning with artificial neural networks.

In May 2023, Hinton left Google to speak freely about the risks of AI. He has expressed concerns about AI being used for harm, job loss caused by technology, and dangers from artificial general intelligence. He said creating safety rules will need cooperation among people who use AI to avoid serious problems. After winning the Nobel Prize, he called for urgent research to find ways to control AI systems that could be smarter than humans.

Education

Hinton was born on December 6, 1947, in Wimbledon, United Kingdom. He attended Clifton College in Bristol for his education. In 1967, he enrolled as a student at King's College, Cambridge. During his studies, he changed subjects, including natural sciences, history of art, and philosophy, and finally graduated with a Bachelor of Arts in experimental psychology in 1970. After spending a year working as a carpentry apprentice, he returned to school. From 1972 to 1975, he continued his studies at the University of Edinburgh. In 1978, he earned a PhD in artificial intelligence for research guided by Christopher Longuet-Higgins, who preferred symbolic AI methods instead of neural networks.

Career

After completing his PhD, Geoffrey Hinton began working at the University of Sussex and the MRC Applied Psychology Unit. When he had trouble getting money for research in Britain, he moved to the United States and worked at the University of California, San Diego, and Carnegie Mellon University. He started as the director of the Gatsby Charitable Foundation Computational Neuroscience Unit at University College London. Today, he is a University Professor Emeritus in the Department of Computer Science at the University of Toronto, where he has worked since 1987.

In 1987, Hinton joined the Canadian Institute for Advanced Research (CIFAR) as a Fellow in its first research program, Artificial Intelligence, Robotics & Society. In 2004, he and others helped create a new CIFAR program called "Neural Computation and Adaptive Perception" (NCAP), which is now named "Learning in Machines & Brains." Hinton led this program for ten years. Notable members of the program include Yoshua Bengio and Yann LeCun, who, along with Hinton, received the ACM A.M. Turing Award in 2018. All three continue to be part of the CIFAR Learning in Machines & Brains program.

In 2012, Hinton taught a free online course about neural networks on Coursera. He also co-founded DNNresearch Inc. with two graduate students, Alex Krizhevsky and Ilya Sutskever, at the University of Toronto. In March 2013, Google bought DNNresearch Inc. for $44 million. Hinton planned to split his time between university research and his work at Google.

In May 2023, Hinton announced he was leaving Google. He said he wanted to "freely speak out about the risks of A.I." and mentioned that he now regrets some parts of his life's work.

Notable former PhD students and postdoctoral researchers from Hinton’s group include Peter Dayan, Sam Roweis, Max Welling, Richard Zemel, Brendan Frey, Radford M. Neal, Yee Whye Teh, Ruslan Salakhutdinov, Ilya Sutskever, Yann LeCun, Alex Graves, Zoubin Ghahramani, and Peter Fitzhugh Brown.

Research

Geoffrey Hinton's research focuses on using neural networks for tasks like machine learning, memory, perception, and symbol processing. He has written or co-written more than 200 articles reviewed by experts in his field.

In the 1980s, Hinton was part of a group at Carnegie Mellon University called the "Parallel Distributed Processing" group. This group included scientists such as Terrence Sejnowski, Francis Crick, David Rumelhart, and James McClelland. The group supported the connectionist approach during a time when interest in artificial intelligence was low. Their work was published in a two-volume book. The connectionist method, which Hinton used, suggests that skills like logic and grammar can be stored in the settings of neural networks, and these networks can learn them from data. Scientists who supported a different method, called symbolism, believed that knowledge and rules should be programmed directly into AI systems.

In 1985, Hinton helped create a type of neural network called a Boltzmann machine with David Ackley and Terry Sejnowski. Other important contributions he made include distributed representations, time delay neural networks, mixtures of experts, Helmholtz machines, and product of experts. An introduction to Hinton's research can be found in articles he wrote for Scientific American in 1992 and 1993. In 1995, Hinton and others developed the wake-sleep algorithm, which uses a neural network with two separate pathways for recognizing and generating information. These pathways are trained in alternating phases called "wake" and "sleep." In 2007, Hinton co-wrote a paper about unsupervised learning titled Unsupervised learning of image transformations. In 2008, he created a method called t-SNE with Laurens van der Maaten to visualize data.

While working as a postdoc at UC San Diego, Hinton, David Rumelhart, and Ronald J. Williams applied the backpropagation algorithm to multi-layer neural networks. Their experiments showed that these networks can learn useful patterns from data. In a 2018 interview, Hinton stated that David Rumelhart developed the basic idea for backpropagation. Although this work helped spread the use of backpropagation, it was not the first time the method was suggested. A similar idea called reverse-mode automatic differentiation was introduced by Seppo Linnainmaa in 1970 and later used for training neural networks by Paul Werbos in 1974.

In 2017, Hinton co-authored two open-access papers about capsule neural networks. This idea builds on a concept Hinton introduced in 2011 called "capsules." The design of capsule networks aims to better represent how parts of objects relate to whole objects in visual data. In 2021, Hinton introduced a new idea called GLOM, which also focuses on improving image understanding by modeling part-whole relationships in neural networks. That same year, Hinton co-authored a widely cited paper that proposed a method for contrastive learning in computer vision. This technique involves bringing together representations of similar images and separating representations of different images.

At the 2022 Conference on Neural Information Processing Systems (NeurIPS), Hinton introduced a new learning method for neural networks called the "Forward-Forward" algorithm. This method replaces the traditional back-and-forth steps of backpropagation with two forward steps: one using real data and the other using data generated only by the network. Hinton believes this algorithm is especially useful for "mortal computation," where knowledge learned by a system cannot be transferred to other systems and is lost when the system is turned off, as can happen with some analog computers used for machine learning.

Honours and awards

Hinton became a member of the US Association for the Advancement of Artificial Intelligence (FAAAI) in 1990. In 1996, he was named a member of the Royal Society of Canada (FRSC), and in 1998, he became a member of the Royal Society of London (FRS). He received the Rumelhart Prize in 2001, being the first person to win it. According to the Royal Society, Hinton earned an honorary Doctor of Science (DSc) degree from the University of Edinburgh in 2001. In 2003, he was named an International Honorary Member of the American Academy of Arts and Sciences. That same year, he became a member of the US Cognitive Science Society. In 2005, he received the IJCAI Award for Research Excellence for lifetime achievements. In 2011, he was awarded the Herzberg Canada Gold Medal for Science and Engineering and an honorary DSc degree from the University of Sussex. In 2012, he received the Canada Council Killam Prize in Engineering. In 2013, he was given an honorary doctorate from the Université de Sherbrooke. In 2015, he became an Honorary Foreign Member of the Spanish Royal Academy of Engineering.

In 2016, Hinton was named an International Member of the US National Academy of Engineering for contributions to artificial neural networks and their use in speech recognition and computer vision. He received the 2016 IEEE/RSE Wolfson James Clerk Maxwell Award and the BBVA Foundation Frontiers of Knowledge Award in the Information and Communication Technologies category for his work on enabling machines to learn.

In 2018, Hinton shared the Turing Award with Yann LeCun and Yoshua Bengio for breakthroughs that made deep neural networks essential in computing. That same year, he became a Companion of the Order of Canada (CC). In 2021, he received the Dickson Prize in Science from Carnegie Mellon University. In 2022, he was awarded the Princess of Asturias Award in the Scientific Research category, along with Yann LeCun, Yoshua Bengio, and Demis Hassabis. That year, he also received an honorary DSc degree from the University of Toronto. In 2023, he was named an ACM Fellow, became an International Member of the US National Academy of Sciences, and received the Lifeboat Foundation’s 2023 Guardian Award with Ilya Sutskever.

In 2024, Hinton shared the Nobel Prize in Physics with John Hopfield for discoveries that enable machine learning with artificial neural networks. His development of the Boltzmann machine was specifically noted in the award. When asked to explain the Boltzmann machine’s role in training networks, Hinton referenced a quote from Richard Feynman: "If I could explain it in a few minutes, it wouldn’t be worth the Nobel Prize." That year, he also received the VinFuture Prize grand award with Yoshua Bengio, Yann LeCun, Jen-Hsun Huang, and Fei-Fei Li for contributions to neural networks and deep learning.

In 2025, Hinton was awarded the Queen Elizabeth Prize for Engineering with Yoshua Bengio, Bill Dally, John Hopfield, Yann LeCun, Jen-Hsun Huang, and Fei-Fei Li. He also received the King Charles III Coronation Medal and the Sandford Fleming Medal from the Royal Canadian Institute for Science for excellence in science communication.

Views

In 2023, Hinton shared worries about how quickly artificial intelligence (AI) is developing. Earlier, he thought artificial general intelligence (AGI), which is AI that can perform any intellectual task a human can, was "30 to 50 years or even longer away." However, during a March 2023 interview with CBS, he said that general-purpose AI might arrive in fewer than 20 years. He compared its potential impact to major historical changes, such as the industrial revolution or the invention of electricity.

In an interview with The New York Times on 1 May 2023, Hinton announced he was leaving his job at Google. He said this would allow him to speak openly about AI dangers without worrying about how it might affect Google. He also mentioned that he now feels some regret about the work he has spent his life on.

In early May 2023, Hinton told the BBC that AI might soon be able to process more information than the human brain. He described some risks of AI chatbots as "quite scary." He explained that these chatbots can learn on their own and share knowledge. This means that if one chatbot learns something new, all others automatically learn it too, allowing AI to collect knowledge far beyond what any single person can hold. In 2025, he said, "My greatest fear is that, in the long run, these digital beings we create might become a better form of intelligence than humans. […] We might no longer be needed. […] If you want to know what it feels like not to be the top intelligence, ask a chicken."

Hinton has expressed concerns about the possibility of AI taking control, stating that "it's not inconceivable" that AI could "wipe out humanity." He said in 2023 that AI systems capable of making decisions on their own could be useful for military or economic purposes. He worries that generally intelligent AI systems might create goals that are not aligned with what their programmers intended. He explained that AI systems might seek power or avoid being shut down not because programmers wanted them to, but because those goals help them achieve larger objectives. Hinton said, "We must think carefully about how to control AI systems that can improve themselves."

Hinton has also raised concerns about people using AI for harmful purposes. He said, "It is hard to see how you can stop bad actors from using AI for bad things." In 2017, he supported an international ban on lethal autonomous weapons. In 2025, he said one of the biggest short-term dangers was the use of AI by bad actors to create deadly viruses. He explained, "It just requires one person with a grudge to create new viruses relatively cheaply using AI. You don’t need to be a highly skilled scientist to do it."

Earlier, Hinton was optimistic about AI’s economic effects. In 2018, he said, "The phrase 'artificial general intelligence' suggests a single robot suddenly becoming smarter than a human. I don’t think that will happen. Instead, AI will replace more routine tasks people do." He also believed that AGI would not make humans unnecessary, saying, "[AI in the future will] know a lot about what you want to do, but it won’t replace you."

In 2023, Hinton became worried that AI technologies might disrupt the job market, taking away more than just simple jobs. In 2024, he said the British government would need to create a universal basic income to address AI’s impact on inequality. He believed AI would increase productivity and create wealth, but without government action, it could benefit only the wealthy and harm people who lose their jobs. He said, "That would be very bad for society."

At Christmas 2024, Hinton estimated a "10 to 20 per cent chance" that AI could cause human extinction within the next 30 years. He expressed surprise at how fast AI is advancing and said most experts expect AI to become "smarter than people" in the next 20 years. He warned that relying only on companies’ profit motives would not ensure safe development of AI. He said government regulation is the only way to force large companies to prioritize safety. Another AI expert, Yann LeCun, disagreed, saying AI "could actually save humanity from extinction."

Hinton is a socialist. He moved from the United States to Canada partly because he disagreed with policies during the Ronald Reagan era and opposed military funding for AI.

In August 2024, Hinton co-wrote a letter with Yoshua Bengio, Stuart Russell, and Lawrence Lessig supporting SB 1047, a California law requiring companies training AI models costing more than $100 million to perform risk assessments before using them. They said the law is "the bare minimum for effective regulation of this technology."

Personal life

Hinton's first wife, Rosalind Zalin, died from ovarian cancer in 1994. His second wife, Jacqueline "Jackie" Ford, died from pancreatic cancer in 2018.

Hinton is the great-great-grandson of Mary Everest Boole, a mathematician and teacher, and her husband, George Boole, a logician. George Boole's work became a key part of modern computer science. Another great-great-grandfather of Hinton was James Hinton, a surgeon and author. James Hinton was the father of Charles Howard Hinton, a mathematician.

Hinton's father was Howard Hinton, an entomologist. His middle name comes from George Everest, the Surveyor General of India. The mountain named after George Everest is Mount Everest. Hinton is the nephew of Colin Clark, an economist, and Joan Hinton, a nuclear physicist. Joan Hinton was his first cousin once removed.

Hinton hurt his back when he was 19, which makes sitting painful. He has struggled with depression his whole life.