Vector Institute aims to play key role in Canada’s bid to maintain AI leadership
June 28, 2017
Geoff Hinton, Yoshua Bengio and Richard Sutton may not be household names but in the world of artificial intelligence (AI) these Canadian researchers are superstars at the forefront of a field that is attracting a host of marquee tech titans angling to gain competitive advantage in the rapidly evolving field. Yet many highly accomplished AI researchers have left Canada to seek opportunity elsewhere prompting a remarkable alignment of government, academia and industry to make Canada the go-to destination for AI advancements in an ever-expanding range of industry sectors.
Enter the Vector Institute (VI), a bold play by the University of Toronto and an impressive group of AI companies to develop the technologies required to drive their adoption and commercialization.
Launched March 30, VI has already attracted nearly $100 million in public funding and close to $90 million more from 34 companies eager to get in on the ground floor and develop, attract and retain the talent needed to make Canada a global AI powerhouse in deep and machine learning. Announcements on additional corporate support are pending.
“We wanted to create a magnet for talent and create companies around it doing and using artificial intelligence … All levels of government, universities, researchers and venture capital are all aligned to seize the opportunity,” says Jordan Jacobs, Co-CEO & Co-Founder at Layer 6 AI, a Toronto-based AI start-up developing deep learning products geared towards financial institutions, e-commerce and media. “Canada has a research lead on AI and we need to capitalize on it … Everything is happening at a very fast clip.”
Jacobs launched Level 6 less than a year ago with U of T talent that trained under the British-born Hinton, regarded as the godfather of research into neural networks or deep learning. Jacobs and others recognized that the shortage of AI expertise, particularly those in niche areas, was threatening Canada’s competitive leadership in the field and were determined not to allow the opportunity to slip by. Skyrocketing salaries were favouring large, deep-pocketed firms like Google and Facebook, even luring Hinton away to Google although he splits his time between the tech giant and the U of T and is one of VI’s founding scientists.
The demand for talent is enormous with demand for up to 5,000 people but Canada only produces 500 a year. Salaries are through the roof with huge competition for it and the Vector Institute is meant to help solve that,” says Jacobs. “We’ll focus on deep and machine learning and bring in graduates and hopefully leap ahead of the others.”
In its latest Budget, the federal government jumped in with $125-million to establish a Pan-Canadian AI strategy to boost the fortunes of AI clusters in Toronto, Montreal and Edmonton. The strategy will be administered by the Canadian Institute for Advanced Research, which has made pioneering contributions to the field through its Learning in Machines and Brains program.
The Quebec and Ontario governments have also sweetened the pot, injecting $100 million and $50 million respectively into university-based AI research at the University of Montreal, McGill University and U of T. Jacobs says VI can lay claim to the idea for the federal strategy as the federal government was seeking a mechanism to support AI. Both Quebec and Ontario are also angling to be included in another federal program offering $950 million to establish a handful of so-called superclusters of which AI is a near certain recipient.
Prime minister Justin Trudeau has been vocal in his enthusiasm for AI. Speaking in Brampton ON the day VI was officially launched, Trudeau compared the importance of the emerging field with the advent of electricity and the microprocessor as he outlined the objectives of the federal AI strategy and noted the risk of not taking immediate action.
“Artificial intelligence will cut across nearly every industry in Canada. It will shape the world that our kids and grandkids grow up in and we can either be a part of that, help steer its direction and take advantage of the good middle-class jobs it will create,” said Trudeau. “Or we can watch other countries step in. They’ll happily hire our best students and hardest workers and why wouldn’t they? We are home to some of the world’s top talent when it comes to artificial intelligence and we can’t afford to lose that competitive advantage and all the good jobs that come along with it."
While major announcements have yet to emerge from Edmonton, AI investment activity is ramping up quickly. Element AI, a recent start-up co-founded by deep learning guru Yoshua Bengio — the French-born director of the University of Montreal’s Montreal Institute for Learning Algorithms — recently raised $135 million from a group of venture capital and high-tech companies including Intel Capital, Microsoft Ventures and several global sovereign wealth firms.
While VI can’t match the compensation offered by corporate heavyweights, it offers researchers and graduate students the flexibility to pursue a mix of basic and corporate research, even going so far as to establish an office of industry relations so that researchers can access government and industrial data sets and direct researchers to collaborators that match their specific areas of expertise.
“This hybrid has been successful in the people we’ve won in the talent wars with other universities, pairing machine learning researchers with coders which leads to commercial opportunities,” says Jacobs. “There’s greater money elsewhere — China is spending billions and Open AI (a San Francisco-based non-profit AI research company associated with business magnate Elon Musk) has $1 billion in commitments – but Canadian money is substantial and we punch above our weight.”