字幕列表 影片播放 列印英文字幕 (leaves rustling) As fall breaks out in Canada, I'm reminded of all the beauty, innocence and gun-free fun available from our neighbors to the North. (majestic music) There's the majesty of Toronto, vast hockey rinks, spectacular batches of poutine, and gallons of maple syrup that you can chug openly and guilt-free for this maple syrup is pure and nourishing. The changing of the seasons also happens to be the perfect time to encounter one of Canada's most prized creatures, the artificial intelligence nerd. (resolute music) Not too long ago, these beings were rare and hidden away in university dungeons. But today they flourish. They primp with instinctual grace. They wave their hands impressively to assert their intellectual dominance. They carb-load like overpaid professional athletes. And this makes some sense because they're among the best paid professionals in the world. Together these creatures did something truly remarkable. Without anyone paying much notice, they gave birth to an AI revolution. They turned Canada, yes, Canada, into one of the great AI superpowers. This is the story of how all this came to be. It's the story of one nation's quest to teach computers to think like humans. It's the story of what this science experiment will mean for all our lives and for the future of the human species. So if you're a human, or something trying to imitate one, you'll wanna pay attention. Ever since people first came up with the idea of computers, they've dreamed of imbuing them with artificial intelligence. I am a smart fellow as I have a very fine brain. That's the most remarkable thing I've ever seen. AI is just a computer that is able to mimic or simulate human thought or human behavior. Within that there's a subset called machine learning that it's now the underpinning of what is most exciting about AI. By allowing computers to learn how to solve problems on their own, machine learning has made a series of breakthroughs that once seemed nearly impossible. It's the reason computers can understand your voice, spot a friend's face in a photo, and steer a car. And it's the reason people are actively talking about the arrival of human-like AI. And whether that would be a good thing or a horrific end of days thing. Many people made this moment possible, but one figure towers above the rest. I've come to the University of Toronto to see the man they call the godfather of Modern Artificial Intelligence. Geoff Hinton. (calm music) Because of a back condition, Geoff Hinton hasn't been able to sit down for more than 12 years. I hate standing. I much rather sit down, but if I sit down I have a disc that comes out. Well at least now standing desks are fashionable. Yeah, but I was ahead. (laughter) I was standing when they weren't fashionable. Since he can't sit in a car or on a bus, Hinton walks everywhere. The walk says a lot about Hinton and his resolve. For nearly 40 years, Hinton has been trying to get computers to learn like people do, a quest almost everyone thought was crazy or at least hopeless, right up until the moment it revolutionized the field. Google thinks this is the future of the company, Amazon thinks this is the future of the company, Apple thinks this is the future of the company, my own department thinks it's just probably nonsense and we shouldn't be doing any more of it. (laughter) So I talked everybody into it except my own department. You obviously grew up in the UK and you had this very prestigious family full of famous mathematicians and economist, and I was curious what it was like for you. Yeah, there was a lot of pressure. I think by the time I was about seven I realized I was gonna have to get a PhD. Did you rebel against that or you-- I dropped out every so often. I became a carpenter for a while. Geoff Hinton, pretty early on, became obsessed with this idea of figuring out how the mind works. He started off getting into physiology, the anatomy of how the brain works, then he got into psychology, and then finally he settled on more of a computer science approach to modeling the brain and got in to artificial intelligence. My feeling is if you wanna understand a really complicated device, like a brain, you should build one. I mean you could look at cars and you could think you could understand cars. When you try and build a car you suddenly discover this is stuff that has to go under the hood, otherwise it doesn't work. Yeah. As Geoff was starting to think about these ideas, he got inspired by some AI researchers across the pond. Specifically this guy, Frank Rosenblatt. Rosenblatt, in the late 1950s, developed what he called a Perceptron, and it was a neural network, a computing system that would mimic the brain. The basic idea is a collection of small units called neurons, these are little computing units but they're actually modeled on the way that the human brain does its computation. They take incoming data like we do from our senses and they actually learn so the neural net can learn to make decisions over time. Rosenblatt's hope was that you could feed a neural network a bunch of data like pictures of men and women and it would eventually learn how to tell them apart, just like humans do. There was just one problem. It didn't work very well. Rosenblatt, his neural network was a single layer of neurons and it was limiting what it could do, extremely limited. And a colleague of his wrote a book in the late '60s that show these limitations. And it kinda put the whole area of research into a deep freeze for a good 10 years. No one wanted to work in this area. They were sure it would never work. Well, almost no one. It was just obvious to me that it was the right way to go. The brain's a big neural network and so it has to be that stuff like this can work 'cause it works in our brains. There's just never any doubt about that. What do you think it was inside of you that kept you wanting to pursue this when everyone else was giving up, just that you thought it was the right direction to go? I know that everyone else was wrong. Okay. Hinton decides he's got an idea of how these neural nets might work, and he's gonna pursue it no matter what. For a little while, he's bouncing around research institutions in the US. He kinda gets fed up that most of them are funded by the defense departments and he starts looking for somewhere else he can go. I didn't wanna take defense department money. I sort of didn't like the idea that this stuff was gonna be used for purposes that I didn't think were good. He suddenly hears that Canada might be interested in funding artificial intelligence. And that was very attractive, that I could go off to this civilized town and just get on with it. So I came to the University of Toronto. And then in the mid '80s, we discovered I had to make more complicated neural nets so they could solve those problems that the simple ones couldn't solve. He and his collaborators developed a multi-layered neural network, a deep neural network. And this started to work in a lot of ways. Using a neural network, a guy named Dean Pomerleau built a self-driving car in the late '80s, and it drove on public roads. Yann LeCun, in the '90s, built a system that could recognize handwritten digits and this ended up being used commercially. But again they hit a ceiling. They didn't work quite well enough because we didn't have enough data, we didn't have enough compute power. And people in AI, in computer science, decided neural networks was wishful thinking basically. So it was a big disappointment. Through the '90s into the 2000s, Geoff was one of only a handful of people on the planet who are still pursuing this technology. He would show up at academic conferences and being banished to the back rooms. He was treated as really like a pariah. Was there like a time when you thought, this just wasn't gonna work? No. And you did have some self-doubt? I mean there were many times when I thought, I'm not gonna make this work. But Geoff was consumed by this and couldn't stop. He just kept pursuing the idea that computers could learn. Until about 2006, when the world catches up to Hinton's ideas. Computers were now a lot faster. And now it's behaving like I thought it would behave in the mid '80s. It's solving everything. The arrival of superfast chips and the massive amounts of data produced on the internet gave Hinton's algorithms a magical boost. Suddenly computers could identify what was in an image, then they could recognize speech and translate from one language to another. By 2012, words like neural nets and machine learning were popping up on the front page of The New York Times. You have to go all these years and then all of a sudden, in the span of a few months, it just takes off and it finally feel like, aha, the world has finally come to my vision. It's sort of a relief that people finally came to their senses. (laughter) Next up, we have Professor Geoffrey Hinton of the University of Toronto. (applause) Thank you. (calm music) For Hinton, this is obviously a really redemptive moment. Now he's basically a technology celebrity. And for Canada, it's the country's moment as well. They have more AI researchers than just about any other place on the planet and the quest now is to see what these guys can do, starting companies and pushing the technology forward. I'm gonna set out on a journey across Canada to see the best in Canadian AI technology and to get a feel for how far the technology has come and how far it still has to go. Here is a city that gets right at the central tension of modern life and the unfolding AI revolution. (church bell ringing) It's Montreal, a place filled with beauty and old world charms that ask you to move slowly through its streets and to chill for a while, reflect, and think deep thoughts. (calm music) At the same time, it's one of the world's top AI research centers. Students flock here from all over the globe to get deep with machine learning and to take Geoff Hinton's ideas and figure out how to turn them into products we all use. To see just how successful they've been, look no further than your pocket.