double gain leads to debate about scientific areas

Nobel committees recognized the transformative power of artificial intelligence (AI) in two of this year’s prizes: honoring pioneers in neural networks with the physics prize, and developers of computational tools to study and design proteins with the chemistry prize. But not all researchers are happy.

Moments after the Royal Swedish Academy of Sciences in Stockholm announced the winners of this year’s Physics Nobel, social media lit up, with several physicists claiming that the science underlying machine learning would be celebrated in the awards given to Geoffrey Hinton and John Hopfield were awarded, actually it wasn’t. physics.

Chemistry Nobel goes to developers of AlphaFold AI that predicts protein structures

“I am speechless. I love machine learning and artificial neural networks as much as anyone else, but find it difficult to understand that this is a physics discovery,” wrote Jonathan Pritchard, an astrophysicist at Imperial College London, on the social media platform X. “Guess the Nobel Prize hit by the AI ​​hype.”

The research by Hinton, at the University of Toronto in Canada, and Hopfield, at Princeton University in New Jersey, “falls into the field of computer science,” says Sabine Hossenfelder, a physicist at the Munich Center for Mathematical Philosophy in Germany. “The annual Nobel Prize is a rare opportunity for physics – and physicists – to be in the spotlight. It’s the day when friends and family remember they know a physicist and might ask him or her what this recent Nobel Prize is about. But not this year.”

Bringing fields together

However, not everyone was concerned: many physicists were happy with the news. “Hopfield and Hinton’s research was interdisciplinary, bringing together physics, mathematics, computer science and neuroscience,” says Matt Strassler, a theoretical physicist at Harvard University in Cambridge, Massachusetts. “In that sense it belongs to all these fields.”

Anil Ananthaswamy, a science writer based in Berkeley, California, and author of the book Why machines learn (2024) points out that while the research cited by the Nobel Committee may not be theoretical physics in the purest sense, it is rooted in techniques and concepts from physics, such as energy. The Boltzmann machine and the Hopfield network invented by Hinton “are both energy-based models,” he says.

Physics Nobel Prize won by machine learning pioneers

The connection to physics became weaker in subsequent developments in machine learning, Ananthaswamy adds, especially in the “feed forward” techniques that made it easier to train neural networks. But physics ideas are making a comeback and helping researchers understand why increasingly complex deep learning systems do what they do. “We need the way of thinking we have in physics to study machine learning,” says Lenka Zdeborová, who studies the statistical physics of computation at the Swiss Federal Institute of Technology in Lausanne.

“I think the Nobel Prize in Physics should continue to spread to more areas of physics knowledge,” said Giorgio Parisi, a physicist at Sapienza University of Rome, who shared the 2021 Nobel Prize in Physics. “Physics is becoming broader and encompasses many areas of knowledge that did not exist or were not part of physics in the past.”

Not just AI

Computer science seemed to have taken its Nobel prize the day after the physics prize was announced, when Demis Hassabis and John Jumper, co-creators of the AlphaFold, an AI tool for predicting protein folding, presented the prize at Google DeepMind in London. won half the prizes. chemistry Nobel Prize. (The other half was awarded to David Baker of the University of Washington in Seattle for his work on protein design that did not involve machine learning.)

The award was a recognition of the disruptive power of AI, but also of the steady accumulation of knowledge in structural and computational biology, said David Jones, a bioinformatician at University College London, who worked with DeepMind on the first version of AlphaFold. “I don’t think AlphaFold represents a radical change in the underlying science that didn’t exist before,” he says. “It’s just the way it was put together and designed so seamlessly that allowed AlphaFold to reach these heights.”

These five scientific areas win the most Nobel Prizes

For example, one of AlphaFold’s key inputs are the sequences of related proteins from different organisms, which can be used to identify amino acid pairs that tend to co-evolve and therefore can be physically close to each other in the 3D structure of a protein . Researchers were already using this insight to predict protein structures at the time AlphaFold was developed, and some even began embedding the idea in deep neural networks.

“It wasn’t just that we went to work and pressed the AI ​​button, and then we all went home,” Jumper said during a press conference at DeepMind on October 9. “It was really an iterative process where we developed, we did research, we tried to find the right kind of combinations between what the community understood about proteins and how we build those intuitions into our architecture.”

AlphaFold would also not have been possible without the Protein Data Bank, a freely available repository of more than 200,000 protein structures – including some that have contributed to previous Nobel Prizes – determined using X-ray crystallography, cryo-electron microscopy and other experimental techniques. methods. “Every data point is someone’s effort over many years,” says Jumper.

Since the Nobel Prize was first awarded in 1901, Nobel Prizes have often emphasized the impact of research on society and rewarded practical inventions, not just pure science. In that respect, this year’s prices are not outliers, says Ananthaswamy. “Sometimes they are given for very good engineering projects. This also includes the prices for lasers and PCR.”