CNN
—
The 2024 Nobel Prize in physics has been awarded to John Hopfield and Geoffrey Hinton for their “foundational discoveries” that enable machine learning with artificial neural networks.
“Although computers cannot think, machines can now mimic functions such as memory and learning. This year’s laureates in physics have helped make this possible,” the Nobel Committee said in a statement.
The committee announced the prestigious honor, seen as the pinnacle of scientific achievement, in Sweden on Monday. The prize carries a cash award of 11 million Swedish kronor ($1 million).
Hopfield at Princeton University and Hinton at the University of Toronto were praised for laying the foundations for the machine learning that powers many of today’s products.
“Using fundamental concepts and methods from physics, they have developed technologies that use structures in networks to process information,” the committee said. This has caused the development of machine learning to “explode” over the past two decades, it added.
Hinton, who has been dubbed the “godfather” of artificial intelligence (AI), said he was “flabbergasted” to receive the prize.
Asked by reporters about the potential significance of the technology his research has helped to develop, he said AI will have a “huge influence” on our societies.
“It will be comparable with the industrial revolution. But 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,” he said in a phone interview after the announcement.
Hinton predicted the technology would revolutionize things such as healthcare, leading to a “huge improvement in productivity.”
“But we also have to worry about a number of possible bad consequences, particularly the threat of these things getting out of control,” he cautioned.
AI has become shorthand for machine learning using artificial neural networks. This technology – developed by Hopfield and Hinton – is based on the structure of the brain.
Whereas a brain has neuronsan artificial neural network has nodes with different values. Whereas the brain’s neurons communicate with each other through synapsesartificial nodes influence each other through connections. You can train an artificial neural network by developing stronger connections between the nodes, just like you can train the brain.
Just as we can wrack our brains for a particular word or fact we rarely use and only dimly remember, artificial neural networks can also search back through the patterns it has saved – thanks to the invention of the Hopfield network in 1982.
“The network that Hopfield built has nodes that are all joined together via connections of different strengths. Each node can store an individual value – in Hopfield’s first work this could either be 0 or 1, like the pixels in a black and white picture,” the committee said, detailing the duo’s work.
After Hopfield published his research, Hinton expanded it using ideas from statistical physics and developed the earliest form of machine learning, called the “Boltzmann machine.”
Since the 1980s, the networks have swelled in size. Whereas Hopfield used a network with just 30 nodes – with fewer than 500 parameters linking them – today’s networks can contain more than one trillion parameters.
Unlike traditional software, which is akin to following a recipe to bake a cake, an artificial neural network is able to learn by example – drawing on prior knowledge to create new recipes.
