Profile

Geoffrey Hinton, born in London in 1947, is a British-Canadian computer scientist and cognitive psychologist. He earned a bachelor’s degree in experimental psychology at King’s College, Cambridge, and a PhD in artificial intelligence at the University of Edinburgh. After research posts in the UK and the US, including Carnegie Mellon University, he settled at the University of Toronto for most of his career.

In 2013 Google acquired DNNresearch, the company he had started with his students, and he spent close to a decade doing research at Google part-time. He left in 2023, saying he wanted to speak more freely about the risks AI could pose.

Key contributions

Through decades when neural networks were out of fashion, Hinton remained one of their most persistent champions.

In 1986 he published a paper with David Rumelhart and Ronald Williams showing how backpropagation could train multi-layer networks to learn useful internal representations. The method is still the foundation of how deep learning models are trained.

The Boltzmann machine, developed with Terry Sejnowski and others, brought ideas from statistical physics into neural network learning, and it is a central part of why he later received the Nobel Prize in Physics.

Around 2006 he and collaborators introduced deep belief networks and layer-wise pre-training, showing that deep networks could be trained effectively and helping popularise the term “deep learning.” He also contributed to work on dropout regularisation, knowledge distillation and capsule networks.

Notable work

In 2012 Hinton and his students Alex Krizhevsky and Ilya Sutskever presented AlexNet, a convolutional neural network that beat every other entry in the ImageNet image recognition challenge by a wide margin. The result quickly turned academia and industry toward deep learning and is often treated as the start of the current AI wave.

In the same period his group worked with several organisations to apply deep neural networks to speech recognition, with clear gains that pushed the technology into industrial use.

Honours

  • 2018 Turing Award, shared with Yoshua Bengio and Yann LeCun, for conceptual and engineering breakthroughs that made deep neural networks a critical part of computing.
  • 2024 Nobel Prize in Physics, shared with John Hopfield, for foundational discoveries and inventions that enable machine learning with artificial neural networks.
  • 2022 Princess of Asturias Award for Technical and Scientific Research.
  • 2025 Queen Elizabeth Prize for Engineering.
  • Fellow of the Royal Society.

Influence

Hinton is widely called one of the “godfathers of deep learning.” His influence runs through his papers and through the students and postdocs he trained, many of whom became leading figures in research and industry.

In recent years he has devoted much of his public voice to AI risk, warning that advanced systems could slip out of human control or be misused, and urging more investment in safety research. That a researcher who spent his life building neural networks now asks the world to be careful with them is one of the more thought-provoking footnotes of the AI era.