Nobel physics laureate Francis Halzen told reporters in Turin, Italy, on Wednesday about a little-known chapter of his career: back in 1991, he wrote a paper proposing to use neural networks to analyze physics experiment data. The 82-year-old laureate couldn't hide his pride: "In fact, the first neural nets appeared in the late 1980s, and I am very proud that I wrote a paper in 1991 proposing to use AI."
Halzen had been named the sole winner of the 2026 Nobel Prize in Physics just a day earlier (October 6), honored "for decisive contributions to the IceCube Neutrino Observatory and the discovery of high-energy neutrinos of astrophysical origin." Visiting Turin the day after the announcement, he volunteered the 35-year-old story when asked about artificial intelligence. His team used neural nets only occasionally afterward, he said, until far more powerful versions arrived a few years ago and changed everything. Neural networks — computer programs that mimic how the human brain recognizes patterns in data — are now the core of modern AI.
The timeline is striking: neural networks were barely a concept in the late 1980s, and AI as we know it was still a distant shadow in 1991, when Halzen proposed using the technique on "part of the classical physics experiment" — a remarkably forward-looking move. The irony runs deep: AI is now changing physics itself. The 2024 Nobel Prize in Physics went to AI pioneers Geoffrey Hinton and John Hopfield for foundational work on artificial neural networks, and the 2025 prize honored John Clarke, Michel Devoret and John Martinis for experiments demonstrating quantum physics in action.
Halzen's prize-winning achievement is a telescope "built" into Antarctic ice — the IceCube Neutrino Observatory. Born in Belgium and based at the University of Wisconsin–Madison since 1972, Halzen first proposed the idea in 1988: a vast network of sensors in the polar ice sheet to capture neutrinos from deep space. A precursor project, AMANDA, was completed in 2000 but proved too small; the expanded IceCube was finished in December 2010, funded by the National Science Foundation, with 5,484 optical modules embedded in the ice. In 2013, IceCube detected the first neutrinos from beyond the solar system.
This year's physics prize has a single recipient, with the full 12 million Swedish crowns (about $1.2 million) going to Halzen. "His tenacity and scientific vision has paved the way for a new kind of astronomy," said Nobel committee chair Mark Pearce. The medals will be presented by Sweden's King Carl XVI Gustaf in Stockholm on December 10. This was the second Nobel unveiled this year, following Monday's medicine prize, with chemistry next on Wednesday.
Neutrinos are often called "ghost particles": they have almost no mass and no electrical charge, and can travel through stars, planets and entire galaxies unchanged. When a neutrino collides with an atomic nucleus it produces a fleeting flash of light, and IceCube's sensors trace those flashes back to where the particle came from. "These tiny, high-energy particles from space transmit information which allow scientists to better understand distant galaxies and exploding stars," said Nobel committee member Eva Olsson. Calling in from Italy after the announcement, Halzen said it was a "great surprise": "When we started this project, everybody realized this was maybe a good idea but very few thought it would work, including myself."