First reactions. Telephone interview, October 2024

“I’m still somewhat in shock”

“You have to build up from the bottom.” In this interview shortly after the announcement, John J. Hopfield talks about how he found out about the prize when he was going through his e-mails. ”It didn’t sink in until I got to the fourth e-mail!” From his cottage in the village of Selborne in England, John J. Hopfield reflects on how to tackle big questions, such as how the mind works. He discusses how to choose good problems, his fears for the future of machine learning and the need for interdisciplinary approaches. 

Interview transcript

John J. Hopfield: Hello?

Adam Smith: Oh, hello. Is this John Hopfield?

JH: This is John Hopfield speaking. Yes.

AS: Oh, hello, my name is Adam Smith. I’m calling from the website of the Nobel Prize, and Mary very kindly set up this time to talk to you. Is that okay?

JH: Yes.

AS: Are you on speaker phone?

JH: How is this at your end?

AS: That’s absolutely perfect. That’s great. Thank you very much indeed. First of all, congratulations on the award of the Nobel Prize.

JH: Oh, thank you. Thank you.

AS: Mary tells me, you find yourself in Hampshire today.

JH: That’s right.

AS: It’s quite a good place to hear the news of the Nobel Prize because you’re slightly hidden.

JH: We’re off on our own, as it were, in a tiny town of less than a thousand people.

AS: It gives you some solitude on such a busy day.

JH: I don’t think there’s another physicist in the town of Selborne, so that things slowly leak out over the news. But there’s no marching in the street here.

AS: How did you actually learn the news that you’d been awarded the Nobel Prize?

JH: I had been out doing things with my wife, flu shot, a cup of coffee somewhere, came back here and there was this enormous list of emails on my computer, which I did not expect at all. And reading into the first two or three of them, you realized there must be a Nobel Prize there. And it was just astounding. My first reaction was they’ve announced the Nobel Prize because he described it without actually managing to connect me and the Nobel Prize in the same sentence. And so I thought it was sort of an email to me about the Nobel Prize to somebody. And it wasn’t until I got down to about the third one that I realised, no, it was to me, that the leading ones on top were just ticklers. I didn’t sink in until I got down to about the fourth email.

AS: I like the idea of those teasers. The prize is given for enabling machine learning and artificial neural networks. But I think I’m right in saying that you didn’t embark on this work in order to create the tools, but rather to understand how mind arises from the wiring in the brain.

JH: That’s right. I, my motivation was really coming from seeing that something does work, the brain, and understanding more about how the brain works would be necessary to understand thought consciousness or what have you. And that it somehow was related to collective phenomena in networks. And I slowly wove my way from an interest in how the brain functioned to a question of how could hardware or software, or whatever you want to call it, wetware, produce such a thing. And the centre of gravity of my knowledge and understanding moved slowly from much more physics oriented to the neurobiological one. And somewhere along the line, this connection between AI, networks, neural networks and physics developed.

AS: You’ve looked at a number of different questions in biology over the years using the lens of physics. I wondered what, what tempts you, what makes a good problem for you as a physicist?

JH: Yes. In a good physics problem, you have a system which is well defined and where you can understand something about how collectively it may work in a way which is more robust than the individual little bits and pieces. You don’t leap into a problem overall saying, I want to understand how mind works. You have to build up from the bottom. If you were doing weather, you would say, well, I want to understand what storms are without going back to interacting air nitrogen molecules. 

You have to have the right level of question. And it isn’t obvious what the level of question should be. You get your hands rather dirty in trying to work on several things which don’t pan out.

AS: Yes. I suppose there’s a long history of physicists turning their attention to the brain, to consciousness. People like Francis Crick or Don Glaser, and it is all about getting the level of the question right, isn’t it?

JH: I’d read some of the things that Don Glaser wrote, for example, and they’re imaginative physics, they’re not quite such good biology. There was a consensus that said you had to be able to reach out from physics and get to some of these things you’d like to, but then you have to know enough about the biology that the whole thing makes sense. And you really have to present things in such a way that a community develops. I didn’t realize that at the time, but certainly one of the important things of what I did had to do with enabling people who came from physics, or who came from biology, become a community, working on not just one little problem or piece, but somehow collectively working together toward trying to get an understanding.

AS: Yes. It catalysed the community and the Hopfield network was a huge advance for people that they could latch onto and develop. Let me ask you one other thing that your co-laureate, Geoffrey Hinton, is very vocal in speaking about his fears about machine learning and its potential. Do you share his worries?

JH: Yes, I share his worries. You always worry when things look very, very powerful and you don’t understand why they are, which is to say you don’t understand how to control them, or if control is an issue, or what their potential is. If you don’t really understand and can’t explain how they work without saying, if you go deeply enough in the mathematics they’ll work. That’s not a satisfactory answer. I would like to have more understanding of how the microscopic physics gives rise to the interesting properties of the larger system.

AS: Do you hope that this Nobel Prize will send some message? It’s the first prize in artificial intelligence, if you like.

JH: I think that the prize is recognizing, in part, the fact that understanding the deep problems of things like mind is not going to come forth in some simple way like Newtonian physics. It really requires much more understanding of the relationship between structure and properties, and structure dynamics and properties. And that’s a mixture of some corners of physics, some corners of chemistry, some corners of biology, coming together to understand and create an area of study.

AS: Thank you. Very nicely put. Let me just finish by commenting that I realize you are hearing this news in Selborne, which was the subject of Gilbert White’s The Natural History of Selborne.

JH: Oh, you’ve discovered Gilbert White! Good for you.

AS: But it’s nice for Selborne that it gets to have a Nobel Prize announced in its midst, given that it has such a deep, ancient association with natural science.

JH: Well, Gilbert White was an astute observer.

AS: Yes. It’s been an enormous pleasure speaking to you. Thank you very, very much. And let me again add our congratulations on today’s news.

JH: Thank you. I know it’s not simple to try to interview me when I’m still somewhat in shock.

AS: Very understandable. It’s been fascinating, and I look forward to a longer conversation when all the dust settles in the future. Thank you.

JH: Right, bye, bye.

AS: Bye, bye.

Did you find any typos in this text? We would appreciate your assistance in identifying any errors and to let us know. Thank you for taking the time to report the errors by sending us an e-mail.