How Canada Pushed The AI Agenda in 2017

Artificial intelligence was reignited in Canada this year, creating new interest in the complex field with new dollars dedicated towards the growing industry.

Universities, the Big Five, and even Prime Minister Justin Trudeau became vocal supporters of AI this year, not only advocating that Canada becomes a global leader in the burgeoning field but backing that sentiment with hundreds of millions of dollars in funding.

But Canada’s history with AI isn’t new. For the past two decades, universities across the country have been hotbeds for research and development, allowing scientists and researchers to lay the groundwork for breakthroughs in AI.

Geoffrey Hinton, from the University of Toronto, has been dubbed the godfather of deep learning; Yoshua Bengio, from the Université de Montréal, is a machine learning pioneer, and Richard Sutton, from the University of Alberta, is a leader in the subfield of reinforcement learning.

Canada Invests Big in AI to Retain Talent

With a renewed focus on research, the Canadian government launched a $125-million national AI initiative called the Pan-Canadian Artificial Intelligence Strategy. It is being led by the Canadian Institute for Advanced Research with the committed funding split between three academic institutions: the Montreal Institute for Learning Algorithms (MILA), the Alberta Machine Intelligence Institute in Edmonton, and the newly formed Vector Institute for Artificial Intelligence—centres that are being led by Bengio, Sutton, and Hinton, respectively.

The Pan-Canadian AI Strategy is focused on some noble causes, including increasing the number of AI researchers and skilled graduates—something the country needs to keep up with innovation in the field.

The Vector Institute was officially launched in March at Toronto’s MaRS Discovery District with the goal of advancing AI research, supporting scale-up firms and producing more deep learning graduates than any other institute globally. Since launch, they’ve added 10 new people to the institute.

The Quebec government allocated $100 million to the AI community in Montreal, while Ontario committed $50 million to Vector, both investments to keep academics and research in the country.

Despite the multi-million dollar investments, the fear of losing Canadian talent to technology juggernauts in the United States continues to be top of mind, as companies south of the border can fork out huge salaries and opportunities for growth.

“We need to create the right set of incentives to get people to continue to work in Canada,” said Jodie Wallis, AI lead at Accenture Canada.

“That means we need to pay people at the same level as other global institutions, and we need to give our researchers the ability to work on projects of their choosing and participate in the industry, if that is what they want to do.”

This past year saw more venture investment in AI and technology startups than Canada has ever seen. Venture capital investments in AI set a record with $191-million USD committed to 22 AI-focused firms and startups around the country in the first three quarters of the year. That included the enormous $102-million USD Series A secured by Montreal’s Element AI in June.

When Techvibes spoke with’s CEO Steve Irvine in the summer, he talked about how Canada needs to firmly command their position in the AI space to truly become recognized as a global leader.

“One of the things I’d love for us to not talk about is how we can be the next Silicon Valley… Let’s highlight our strengths as Canada. Let’s focus on where we have the advantage,” he said. Irvine launched in January with a $5 million investment from Georgian Partners to develop an AI-powered enterprise software platform.

The Call of the North: Canada’s Research Advantage

Other countries are also beginning to take note of Canada’s booming AI sector.

The United Kingdom-based and Google-owned DeepMind established its first international research lab in Edmonton, Alberta. Opened in July, the new lab is being led by Sutton and two University of Alberta computer science professors.

“It’s becoming more widely recognized that Canada does have leadership position. One of the points of evidence is how global technology companies are setting labs up here,” said Wallis.

American companies have tapped Canadian talent to head new AI arms, pulling university researchers away from academics and into the industry through the creation of development labs—specifically ones in Montreal.

Facebook set up a AI research lab in Quebec’s largest city and committed $7 million in new partnerships. Joelle Pineau, the co-director of McGill University’s Reasoning and Learning Lab, was selected to lead the new lab while continuing her work at the university.

Microsoft and Samsung also pegged Montreal as the place to be for AI research, rather than Silicon Valley where the technology giants are often headquartered. France’s Thales created an AI research centre in the city but will collaborate with both MILA and Toronto’s Vector. Locally-based Stradigi also opened up an AI-focused technology incubator in Montreal.

Canada’s own RBC got in on the action too, announcing plans to open a Borealis AI lab next year. The lab will work with MILA and Bengio, who is also one of the frontmen of Element AI, part incubator, part AI development hub.

Meanwhile, many academics and entrepreneurs have pitched Toronto as Canada’s true AI hub.

“I actually assumed at the beginning that I would start Integrate in Silicon Valley… I started to realize that Toronto really is the founding city of modern AI,” said Irvine. “We have this unbelievable research advantage here that I don’t think a lot of people know about.”

Toronto’s position as an epicentre has been supported with the launch of Google Brain, as well as Uber establishing a driverless-car project in Toronto, bringing in University of Toronto’s Raquel Urtas to lead the lab.

AI-driven startups have also cropped up in the city this year, while early-stage firms secured new funding.

Deep Genomics raised a $13-million Series A this fall. Headed by Brendan Fray, Deep Genomics is building data-driven AI platforms for geneticists, molecular biologists and chemists with the lofty goal of creating life-saving genetic therapies.

With a handful of AI-driven health and medtech startups working out of MaRS, the innovation hub is living up to its founding vision of commercializing breakthrough medical discoveries.

Toronto scientists Frank Rudzicz and Laim Kaufman founded WinterLight Labs to build a speech-based platform capable of detecting cognitive diseases. Armed with $500,000 from Novatio Ventures, the startup is set to commercialize their product—that’s something Wallis said the country needs to focus on in 2018.

Shifting Focus from Research to Commercializing

While funding for research and development is critical to get ideas off the ground, Wallis wants the commercialization of AI to become the centerpiece of Canada’s AI agenda next year.

“We haven’t quite solved the problem of commercialization; the connective tissue that links the researchers to private industry, government entities, and small and medium-sized enterprises,” said Wallis.

Research is the first part in a development lifecycle that spans from idea to economic value.

“The injections of funding from government is tremendous, but they have all gone into research. Commercialization is where we need Canadian startups to play a larger role. It requires talent, it requires funding, but it also requires these startups to get customers,” she said.

Wallis presented three recommendations for how Canadian-grown startups can be supported to commercialize AI-based products in 2018.

  1. Advance Canada’s data policies: “Canada is definitely progressive when it comes to privacy by design, but we need to advance our data policies to the next level to allow Canadian companies to take advantage of data in a meaningful way.”
  2. Tackle the scale-up challenge: “We’re moving in right direction on funding and on talent but we need to move in the right direction to motivate Canadian companies and government entities to buy from Canadian AI startups.”
  3. Sector-focused solutions: “Early AI solutions were around AI technology: natural language processing and building out machine learning algorithm capabilities… But now we should start looking at solutions in sectors like mining, retail, and agriculture.”

Next year, Wallis hopes to see Canadian AI startups moving away from creating generic AI-driven solutions to building specific industry solutions, “that’s where we can get real value.”