Nobel Prize in Physics 2024: John Hopfield and Geoffrey Hinton honoured for pioneering work in AI

The Royal Swedish Academy of Sciences has awarded the 2024 Nobel Prize in Physics to John J. Hopfield and Geoffrey E. Hinton for their foundational contributions to artificial intelligence through the development of machine learning with artificial neural networks. Their groundbreaking research, dating back to the 1980s, has significantly influenced the field of AI, helping to shape the technologies that power modern data processing and pattern recognition.

John Hopfield, a professor at Princeton University, developed an associative memory network that could store and reconstruct images and other patterns in data. His innovation, known as the Hopfield network, uses principles of physics, particularly the concept of atomic spin, to represent data and store it with low energy requirements. This network functions by adjusting its values to match input data, making it possible to reconstruct incomplete images and find patterns within noisy information.

Geoffrey Hinton, a professor at the University of Toronto, expanded upon Hopfield’s ideas with the Boltzmann machine. This advanced model uses principles from statistical physics, enabling the network to autonomously learn and identify distinct elements within data. The Boltzmann machine is not only capable of classifying images but also generating new data that matches learned patterns, proving foundational in image recognition and other AI applications.

These neural network models, inspired by the brain’s structure, represent a major leap in artificial intelligence. As Ellen Moons, Chair of the Nobel Committee for Physics, noted, "The laureates' work has already been of immense benefit," underscoring how artificial neural networks are now used in physics and material science to design new materials with specific properties.

The 2024 Nobel Prize in Physics, comes with an award of 11 million Swedish kronor to be shared equally between Hopfield and Hinton.