Geoffrey Hinton

Legacy & significance

Hinton kept faith with neural networks through two AI winters, and won. The training method he championed in 1986 and the 2012 breakthrough his lab produced turned a marginal idea into the technology behind modern AI. Then, at the peak of vindication, he quit Google to warn the world about what he had helped build, giving AI risk its most credentialed voice.

Major achievements & key works

• Learning representations by back-propagating errors (1986, with Rumelhart and Williams). Made multi-layer network training practical, the algorithm modern AI still runs on.
• AlexNet (2012, with Krizhevsky and Sutskever). Halved the ImageNet error rate and started the deep-learning gold rush; the lab was acquired by Google in 2013.
• Turing Award 2018 and Nobel Prize in Physics 2024. Computing's and science's highest honors for the same body of neural-network work.
• The 2023 resignation and risk advocacy. His departure from Google moved existential AI risk from fringe worry to boardroom and parliament agenda item.

Pivotal turning points

Year	Event	Impact
1986	Backpropagation paper in Nature	Deep networks become trainable
1987	Moves to Canada, later founding the Toronto ML group	Builds the school that produces the field's leaders
2012	AlexNet wins ImageNet	Deep learning goes mainstream overnight
2018	Turing Award with Bengio and LeCun	The 'deep learning conspiracy' canonized
2023	Resigns from Google to discuss AI dangers	Risk debate gains its most decorated insider
2024	Nobel Prize in Physics	Neural networks enter the scientific pantheon

Sources & further reading

Primary sources

• Learning representations by back-propagating errors, Nature 323 (1986)
• ImageNet Classification with Deep Convolutional Neural Networks (AlexNet), NeurIPS 2012

Recommended secondary sources

• 'The Godfather of A.I.' Leaves Google and Warns of Danger Ahead, The New York Times (May 1, 2023)
• Nobel Prize in Physics 2024, official citation