Canadian AI summer school shows its global ambition
The Deep Learning and Reinforcement Learning Summer School in Toronto. Photo: CIFAR & Vector Institute
Amid an intense global race to develop artificial intelligence, Canada — home to some of the field's pioneers, and among the most aggressive nations in the contest — is running a boot camp for students this week to beef up its chances to share in the AI future.
What's going on: More than 250 students and researchers from 20 countries are in Toronto, where Canada is attempting to attract as much of the world's best AI talent it can by showcasing itself as a first-rate center for research.
Canada has set a goal of recruiting 25 world class AI researchers from outside the country to chair academic programs, according to Elissa Strome of the Canadian Institute for Advanced Research. The country is specifically looking to lead in the responsible use and social impact of AI:
- "Being a moral leader is very important especially as democracy seems to be threatened on many fronts and AI could be used for good and bad," says Yoshua Bengio, a pioneer of deep and reinforcement learning and a professor at the Artificial Intelligence Institute in Montreal.
- Bengio says he is also starting a professional AI master's degree program in Montreal that he hopes will make it easier to fill the enormous demand for talent faster.
- This is all part of Canada's Artificial Intelligence Strategy, which includes three AI institutes, the AI summer school, and a number of other initiatives.
The scene: The attendees at the summer school — run for more than 10 consecutive years — are largely students and some industry and academic researchers, who come from computer science, neuroscience, quantum computing and medical fields.
- They attend lectures on topics in deep and reinforcement learning, AI methods underlying advances in gameplaying and speech and image recognition.
- "The diversity in the room is quite fascinating," Katya Kudashkina, a PhD candidate at the Vector Institute and the University of Guelph, tells Axios.
The big picture: There is high demand for AI experts as countries clamber for a top position in the global race. Training and retaining a highly skilled AI talent pool will be key to winning, according to a study by the Center for a New American Security.
- Canada isn't a big spender on the global AI stage — the national strategy is a $125 million government investment over five years versus, say, the $885 million the U.S. Department of Defense is putting behind AI initiatives.
- But researchers in Canada have been on the field's scientific edge — including Bengio, Geoffrey Hinton at the Vector Institute and Richard Sutton at the Alberta Machine Intelligence Institute.
- Last year, Toronto added more tech jobs than the San Francisco Bay area, Seattle and Washington, D.C., combined, per Bloomberg.
The even bigger picture: Annual entry-level salaries in AI at large tech companies can top $300,000, according to the NYT.
- "Hopefully as we train more people, salaries will become more normal," says Bengio. "[Right now] it makes it hard for small companies to compete with larger ones."
- Summer schools and programs are cropping up to try to hook students who are underrepresented in the field right now, and address the demand for AI experts.
- In May, Carnegie Mellon University announced an undergraduate AI program — the first at a U.S. university — beginning this fall.
What's next: Bengio says future summer schools and education in the field should include philosophers, lawyers, economists and social scientists, all of them discussing how AI should be used. And, he says, "government and funding agencies need to make sure scholars in social sciences and the humanities are involved and funded."