Wback in 2011 Marc Andreessen, a venture capitalist with aspirations to be a public intellectual, published an essay titled “Why Software Is Eating the World”, which predicts that computer code will take over large parts of the economy. Thirteen years later, it seems that software is now making its way through academia as well. In any case, this is one possible conclusion that can be drawn from the fact that computer scientist Geoffrey Hinton shares the 2024. Nobel Prize in Physics with John Hopfield, and that the computer scientist Demis Hassabis shares half the share Nobel Prize in Chemistry along with one of his DeepMind colleagues, John Jumper.
The award to Hassabis and Jumper was in a way predictable because they built a machine – AlphaFold2 – enabling researchers to solve one of the most difficult problems in biochemistry: predicting the structure of proteins, the building blocks of biological life. Their machine was able to predict the structure of almost all of the 200m proteins that researchers identified. So this is a big problem – for chemistry.
But Hinton is not a physicist. Indeed, he once was presented at an academic conference as someone who “failed physics, left psychology and then joined a field without any standards: artificial intelligence”. And he worked as a carpenter for a year after graduation. Yet he’s the guy who found the method (“back propagation”) that enables neural networks to learn, which was one of the two keys that unlocked machine learning and caused the current manic rise of AI. (The other was the invention of the transformer model by Google researchers in 2017).
But where is the physics in all this? It comes from Hopfield, with whom Hinton shares the prize. “Hopfield networks and a further development of them called Boltzmann machines were based on physics,” Hinton told the man of the New York Times. “Hopfield nets used an energy function, and the Boltzmann machine used ideas from statistical physics. So that stage in the development of neural networks did depend – a lot – on ideas from physics.”
So it’s okay then. But the media often describes Hinton as “the godfather of AI,” which has a vaguely sinister ring to it. In person, he is the opposite: tall, friendly, courteous, cerebral and endowed with a wry and sometimes caustic wit. When asked by Cade Metz for his reaction when he got the news of the award, he replied that he was “shocked and surprised and bewildered”, which I think is what most people are saying. But in 2018 he shared the Turing Award – computer science’s equivalent of the Nobel – with Yoshua Bengio and Yann LeCun for their work on deep learning. So he was always in the top league. It’s just that there isn’t a Nobel Prize in computer science. Given the way software is eating up the world, maybe that needs to change.
There’s an old joke that the key to being a Nobel Prize winner is to outlive your competitors. Hinton, now 77, clearly took notice. But really, the most admirable thing about him is the dogged tenacity with which he continued to believe in the potential of neural networks as a key to artificial intelligence long after the idea had been discredited by the discipline. Given how academia works, especially in a rapidly evolving discipline like computer science, this required extraordinary determination and confidence. Perhaps what kept him going in his darker moments was the thought that his great-grandfather was George Boole, the 19th-century mathematician who invented the underlying logic. everyone of this digital stuff.
One also thinks about the impact that the price has on people. When the news of Hinton’s prize came through, I thought of Seamus Heaney, who was awarded the literature prize in 1995. He described the experience as “like being hit by a mostly benign avalanche”. Note that “mostly”: one of the consequences of a Nobel is that winners immediately become public property, of which everyone and their dog wants a slice. “All I do these days is ‘show up,'” Heaney wrote resignedly to a friend in June 1996. “I’m a function of grids, not an agent of my own being. And it’s going to be that way for weeks and months to come… Whatever the Stockholm effect ends up being, its immediate consequence is a desire to quit the job and start over in propria persona (in my own person).”
So… Memo to Geoff: congratulations. And keep control of your calendar.
What I have read
Speak this way
Is a chat with a bot a conversation? is a fabulous New Yorker set by historian Jill Lepore about her interactions with GPT-4o, Advanced Voice mode.
Interesting times…
October 2, 2024. This particular issue from Heather Cox Richardson’s indispensable Substack blog is a gem.
A real page turner
The Elite college students who cannot read books are a intriguing report in the Atlantic Ocean by Rose Horowitch.
