The Vector Institute launched in March to help Canada become a leader in artificial intelligence. The Toronto-based, independent, nonprofit research institution is focused on three goals: to be a world-leading center for AI research, to help universities graduate more master’s and Ph.D. students in deep learning and machine learning, and to grow the economy by implementing AI technology in finance, retail, transportation, and other fields.
The institute was founded with assistance from the University of Toronto and funding from the Canadian and Ontario governments and private companies. To learn more about it, The Institute interviewed its research director, Richard Zemel, a professor of computer science at the University of Toronto. Here is what he had to say about making the Vector Institute a leader in AI.
What are your plans to make the institute a success?
My first priority is to bring in faculty members and research scientists who will be the drivers of deep learning and machine learning education and research at the institute. We have a deep talent pool of expertise in these areas in Toronto, and our goal is to build on and expand them. We’re preparing to make several announcements in the next few months on our new hires.
Another priority is to create a flexible operating model—which means our researchers won’t have to decide between working in academia or industry. Vector researchers will be able to work on projects in the private sector as well. This will enable companies in Canada to adopt some of the most advanced AI technology.
Why is the Vector Institute needed in Canada?
The institute is a response to both a challenge and an opportunity. Graduates of Canadian universities often pursue careers in machine learning, and then leave the country to work at some of the world’s leading technology companies. The challenge Canada faces is that while our graduates are in high demand, the job market here hasn’t offered them enough opportunities, so they’ve tended to leave. Moreover, companies need to do more when it comes to investing in and innovating with AI technology. The Vector Institute is working to create meaningful opportunities for the world’s top talent to pursue careers here in Canada.
Since the institute’s launch, more opportunities have become available here. Uber announced it will open a new branch of its Advanced Technologies Group to improve the company’s AI systems. Google launched its Google Brain Toronto, dedicated to AI research. And DeepMind, a leader in AI research, announced the opening of its first international AI research office in collaboration with the University of Alberta, in Edmonton.
Anecdotally, I’ve heard from my own students that they are deciding to stay in Canada because of the emerging AI ecosystem.
Can you say more about the support the institute receives?
The institute launched with an unprecedented level of support from both government and industry. The Canadian government committed US $100 million, and the province of Ontario committed $40 million. Federal funding will be provided through the Canadian Institute for Advanced Research (CIFAR) to support core activities, and the University of Toronto helped establish the institute with administrative, operational, and policy support. (You can read more about the founding team on the Vector Institute website.)
In addition, more than 30 companies have committed more than $85 million over 10 years. These commitments reflect an acknowledgement that AI will be transformative. We, along with other institutions, will participate in the Canadian government’s Pan-Canadian Artificial Intelligence Strategy. This plan will help attract and retain top academic talent and researchers and promote collaboration in Canada.
We also plan to work with AI startups to help them grow, and to drive the adoption and commercialization of their technologies.
How will you share the technologies that come out of the Vector Institute?
Our board is undertaking a process to develop an intellectual property policy. Our end goal is to encourage commercialization by Canadian entities and to help launch and grow companies that develop their IP domestically and commercialize it globally. The IP policy will be developed with expert advice and input from researchers as well as a broad range of Canadian industries. The results of our research will be published in conference and academic journals, as will research results from labs in the field of machine learning.