Two Nobel prizes in one week go to the science behind modern AI
The 2024 physics prize honoured the inventors of neural network learning, and half the chemistry prize went to the creators of AlphaFold.
In 30 seconds
- Physics: John Hopfield and Geoffrey Hinton, for discoveries that enable machine learning with artificial neural networks.
- Chemistry: half to David Baker for computational protein design, half to Demis Hassabis and John Jumper for protein structure prediction.
- It is the first time the Nobel committees have so directly rewarded AI research.
In the second week of October 2024, the Royal Swedish Academy of Sciences awarded two of its three science prizes for work at the heart of today’s AI boom.
Physics: how machines learn
John Hopfield created a network in the early 1980s that can store and recall patterns, a kind of associative memory. Geoffrey Hinton built on it with the Boltzmann machine, which can learn the features of data on its own. Both ideas use tools from physics that describe how many small parts, like atoms in a material, settle into stable states.
These ideas led, through decades of further work, to the deep learning systems that now power chatbots, image tools and speech recognition.
Chemistry: the shape of life
Proteins are chains of amino acids that fold into complex 3D shapes. The shape decides what a protein does, but working it out in a lab can take years.
Demis Hassabis and John Jumper of Google DeepMind built AlphaFold, which predicts a protein’s shape from its sequence with accuracy close to lab methods. Its predictions for more than 200 million proteins are free online.
David Baker of the University of Washington went the other way: designing completely new proteins that do not exist in nature, for use in vaccines, sensors and medicines.
Why it matters for India
Protein structure prediction is now a free tool that Indian drug and vaccine researchers use daily. The prizes also signal that AI skills are becoming core science skills, which matters for how Indian universities design physics, chemistry and biology courses.
Our take
The physics prize surprised many physicists, because the work is now mainly used in computing. But Hopfield and Hinton borrowed their ideas from statistical physics, and the committee chose to honour that link. The chemistry prize was far less controversial: AlphaFold solved a 50-year-old problem.
Sources
We read these before writing this story. Go to them for the full detail.