top of page

The Nobel Prize in Physics: 2024

Writer's picture: Mishkat BhattacharyaMishkat Bhattacharya

The 2024 Nobel Prize in physics was awarded to John Hopfield and Geoffrey Hinton, “for foundational discoveries and inventions that enable machine learning with artificial neural networks”. Below we address some of the questions raised by this prize, keeping in mind that this is not my area of specialization.


What was it about?


The awardees used models from physics to simulate the human brain, leading to the creation of artificial neural networks (ANNs). Their work turned out to be seminal for machine learning, which has had numerous benefits to science and society, enabling image recognition, language generation, data processing, etc. The work of this year's laureates has led to new pharmaceuticals, vaccines, nanomaterials and sensors, among other applications.


Was it unexpected?


From the reaction on social media, there is some surprise associated with the selection this year, for reasons to be discussed below.


Was it physics?


The fact that many people are asking this question means the award was not given for one of the well-established (canonical) branches of physics. Few probably would ask the question if the prize were to be awarded, for example, for particle physics or atomic physics.


Some colleagues felt the work was top notch computer science, but not appropriate for physics. In my opinion this may be considered as a prize awarded for biological physics. I remember arriving in graduate school in 1995 and being briefly fascinated by ANNs through my roommate, who was a physicist and working on those systems. I was especially attracted by the abilities of these systems to display phenomena like memories and dreams.


Hopfield is definitely a physicist. In fact he is a past president of the American Physical Society (which says the Nobel was awarded for application of physics to network theory; here is an APS post that addresses the controversy) and a recipient of the Oliver E. Buckley Condensed Matter Physics Prize in 1969 (a heavyweight laurel). His work is based on models of interacting magnets and leans on statistical mechanics, which is a core area of physics.


Hinton, on the other hand, has been exclusively employed by computer science departments in his academic career. He is one of the acknowledged heavyweights of AI (got the Turing award in 2018), but the connection to physics is less clear in his case.

What does this mean for the Nobel prize going ahead?


This probably means that we have to acknowledge contributions from relatively new areas in physics. It could also mean, as some have suggested, that a separate Nobel prize should be set up for computer science. In any case, noting that the Chemistry prize was awarded for computational protein design and prediction, this year AI has received a strong nod from Stockholm (the joke is that the Nobel Committee has been taken over by an AI system). Whether this is going to develop into a trend will be interesting to see in the coming years.


Afterword


No post this Friday - have a great weekend!

Recent Posts

See All

Comentarios


Responsible comments are welcome at mb6154@gmail.com. All material is under copyright ©.

© 2023 by Stories from Science. Powered and secured by Wix

bottom of page