He helps make Canada a high-tech hotbed
GEOFFREY HINTON A PIONEER IN AI RESEARCH
As an undergraduate at Cambridge University, Geoffrey Everest Hinton thought a lot about the brain. He wanted to better understand how it worked but was frustrated that no field of study — from physiology and psychology to physics and chemistry — offered real answers.
So he set about building his own computer models to mimic the brain’s process.
“People just thought I was crazy,” said Hinton, now 69, a Google fellow who is also a professor emeritus of computer science at the University of Toronto.
He wasn’t. He became one of the world’s foremost authorities on artificial intelligence, designing software that imitates how the brain is believed to work. At the same time, Hinton, who left academia in the United States in part as a personal protest against military funding of research, has helped make Canada a high-tech hotbed.
Dictate a text on your smartphone, search for a photo on Google or, in the not too distant future, ride in a self-driving car, and you will be using technology based partly on Hinton’s ideas.
His impact on artificial intelligence research has been so deep that some people in the field talk about the “six degrees of Geoffrey Hinton” the way college students once referred to Kevin Bacon’s uncanny connections to so many Hollywood movies.
Hinton’s students and as- sociates are now leading lights of artificial intelligence research at Apple, Facebook, Google and Uber, and run artificial intelligence programs at the University of Montreal and OpenAI, a nonprofit research company.
“Geoff, at a time when AI was in the wilderness, toiled away at building the field and because of his personality, attracted people who then dispersed,” said Ilse Treurnicht, chief executive of Toronto’s MaRS Discovery District, an innovation centre that will soon house the Vector Institute, Toronto’s new public- private artificial intelligence research institute where Hinton will be chief scientific adviser.
Hinton also recently set up a Toronto branch of Google Brain, the company’s artificial intelligence research project. His tiny office there is not the grand space filled with gadgets and awards that one might expect for a man at the leading edge of the most transformative field of science today. There isn’t even a chair. Because of damaged vertebrae, he stands up to work and lies down to ride in a car, stretched out on the back seat. “I sat down in 2005,” said Hinton, a tall man, with uncombed silvering hair and hooded eyes the colour of the North Sea.
Hinton started out under a constellation of brilliant scientific stars. He was born in the United Kingdom, growing up in Bristol, where his father was a professor of entomology and an authority on beetles. He is the greatgreat- grandson of George Boole, the father of Boolean logic. His middle name comes from another illustri- ous relative, George Everest, who surveyed India and made it possible to calculate the height of the world’s tallest mountain that now bears his name.
Hinton f ol l owed t he family tradition by going to Cambridge in the late 1960s. But by the time he finished his undergraduate degree, he realized that no one had a clue how people think.
“I got fed up with academia and decided I would rather be a carpenter,” he recalled with evident delight, standing at a high table in Google’s white-on-white café. He was 22 and lasted a year in the trade, although carpentry remains his hobby today.
Hinton then heard about an artificial intelligence program at the University of Edinburgh and moved there in 1972 to pursue a Ph.D. His adviser favoured the logicbased approach, but Hinton focused on artificial neural networks, which he thought were a better model to simulate human thought.
His study didn’t make him very employable in Britain, though. So, Ph.D. in hand, he turned to the United States to work as a postdoctoral researcher in San Diego with a group of cognitive psychologists who were also interested in neural networks. They began working with a formula called the back propagation algorithm that allowed neural networks to learn over time and has since become the workhorse of deep learning, the term now used to describe artificial intelligence based on those networks.
Hinton moved in 1982 to Carnegie Mellon University in Pittsburgh as a professor, where his work with the algorithm and neural networks allowed computers to produce some “interesting internal representations.”
Here’s an example of how the brain produces an internal representation. When you look at a cat light waves bouncing off it hit your retina, which converts the light into electrical impulses that travel along the optic nerve to the brain. The brain reconstitutes those impulses into an internal representation of the cat, and if you close your eyes, you can see it in your mind.
At that point, Hinton was becoming disillusioned with the politics of the United States in the Reagan era.
Canada beckoned with a research position at the Canadian Institute For Advanced Research. He moved to Toronto and eventually set up a program at the institute that is now called Learning in Machines & Brains. By 2012, computers had become fast enough to allow him and his researchers in Toronto to create those internal representations as well as reproduce speech patterns that are part of the translation applications we all use today.
GEOFF, AT A TIME WHEN AI WAS IN THE WILDERNESS, TOILED AWAY AT BUILDING THE FIELD AND BECAUSE OF HIS PERSONALITY, ATTRACTED PEOPLE WHO THEN DISPERSED. — ILSE TREURNICHT, CHIEF EXECUTIVE OF TORONTO’S MARS DISCOVERY DISTRICT