AI godfather-turned-doomer shares Nobel with neural network pioneer

First-ever awarded for contributions to artificial intelligence

If you needed another sign that we've well and truly entered the AI age, here it is: The first Nobel Prize has been awarded for contributions to artificial intelligence. 

AI "godfather" Dr. Geoffrey Hinton, and his intellectual predecessor in the realm of learning machines, Dr. John Hopfield, were jointly awarded the 2024 Nobel Prize for physics today "for foundational discoveries and inventions that enable machine learning with artificial neural networks," the Royal Swedish Academy of Sciences said

El Reg readers and AI watchers are likely familiar with Hinton's pioneering work on neural networks, and his high-profile departure from an advisory role at Google, driven by concerns over the potential dangers of the AI systems he helped create. Hopfield's work, on the other hand, is even more foundational to modern AI, and influenced Hinton's advancements.

According to a write up [PDF] of the reasons for the award, Hopfield's greatest contribution to AI came in 1982 when he created a neural network (named after himself) capable of storing multiple patterns and retrieving them from memory by distinguishing between them.

The Committee likened the "Hopfield network" to the brain's associative memory, where we search for and recall information, such as words. It described the network as a system of artificial neurons with varying connection strengths.

"Hopfield described the overall state of the network with a property that is equivalent to the energy in the spin system found in physics; the energy is calculated using a formula that uses all the values of the nodes and all the strengths of the connections between them," it explained.

By the time the entire network processes the data, it often reproduces the original image it was trained on, the Academy noted, but what made it truly special was its ability to store multiple pictures at the same time, and differentiate between them.

From recall to interpretation

While a Hopfield network and associative memory techniques can recall images as patterns in data, it can't interpret what they are. That's where Hinton came in. 

"When Hopfield published his article on associative memory, Geoffrey Hinton was working at Carnegie Mellon University in Pittsburgh, USA," the Academy said. "Along with his colleague, Terrence Sejnowski, Hinton started from the Hopfield network and expanded it to build something new, using ideas from statistical physics." 

"The states in which the individual components can jointly exist can be analyzed using statistical physics, and the probability of them occurring calculated," the Nobel awarding body said. Measure those probabilities and assign them to objects, and you have Hinton's Boltzmann machine

Another neural network, but a far more advanced one, the Boltzmann machine can learn from examples of data, recognize familiar traits across samples, and recognize new examples of an object by filtering known basics into various categories. 

Boltzmann machines are still used today to power recommendation engines and other basic AIs, and are frequently a part of larger machine learning networks. 

"The laureates' work has already been of the greatest benefit. In physics we use artificial neural networks in a vast range of areas, such as developing new materials with specific properties," Ellen Moons, Chair of the Nobel Committee for the physics prize said.

From interpretation to worry

It's the second major award Hinton's won for his contributions to AI, after sharing the 2019 Turing Prize - often called the Nobel of computing - with fellow "AI godfathers" Yoshua Bengio and Yann LeCun.

Since then, however, Hinton has become downright skeptical of the learning machines he helped create. After leaving Google in May 2023, Hinton expressed regret for his role in laying the foundation for modern AI, saying that when he looked at AI's growth to date and how it was likely to impact society in the future, the possibilities were "scary." 

"I console myself with the normal excuse: if I hadn't done it, somebody else would have," Hinton told the New York Times' Cade Metz, a former Register journalist, last year. 

Hinton was joined by Bengio in signing an open letter last year calling for regulation of AI to prevent future harms. The letter likened AI to the threat of climate change, saying that while we ignored those warnings, we could head off trouble with AI before it's too late. 

"There is a responsible path, if we have the wisdom to take it," the letter begs. Hinton reiterated his concerns about AI in an interview shortly after he found out he won the Nobel.

"I wish I had a sort of simple recipe that if you do this, everything's going to be okay. But I don't," Hinton told Nobel Prize Outreach's chief science officer Adam Smith. "We're a kind of bifurcation point in history where in the next few years we need to figure out if there's a way to deal with that threat [of AI running amok]."

"One thing governments can do is force the big companies to spend a lot more of their resources on safety research," Hinton added. "Companies like OpenAI can't just put safety research on the back burner."

Hinton and Hopfield will share a prize of 11 million Swedish kronor (about $1 million), and will receive their awards on December 10, the anniversary of dynamite inventor and prize namesake Alfred Nobel's death in 1896. ®

More about

TIP US OFF

Send us news


Other stories you might like