OTTAWA — The roots of Canada’s flourishing modern AI ecosystem trace back four decades, to a then-new research organization’s attempt to ford the gap between the social and natural sciences. Stephen Toope, the high-profile new CEO of CIFAR, hopes to keep it “out on the forefront,” looking for interdisciplinary answers to society’s big questions.

Toronto-based CIFAR—formerly the Canadian Institute for Advanced Research— supports networks of researchers working to understand the nature of consciousness, explore beneath the planet’s crust and develop new quantum materials. It’s also responsible for much of the federal government’s $568.8-million Pan-Canadian AI Strategy, working with three hubs—Montreal’s Mila, Toronto’s Vector Institute and Edmonton’s Amii—to recruit prominent scientists to universities around the country, and to help find practical and industrial applications for their ideas.

Toope is only CIFAR’s fifth leader since the publicly funded charity was established in 1981, and takes over as it begins work on a new plan for its future. He returns to Toronto after a five-year spell as vice-chancellor of the University of Cambridge. At the U.K. institution, Toope launched new scholarships and supports for students, and expanded research programs. But sections of the British press have also criticized him over the way the university handled campus culture-war flare-ups, and for Cambridge accepting funding from the Chinese government and linked companies. 

In his first interview since starting his new job last month, Toope spoke with The Logic about his plans to expand CIFAR’s role in commercializing new discoveries, the big debates engulfing research and technology, and the importance of understanding fungi.

This interview has been edited and condensed for length and clarity.

You’ve spent the last few decades of your career running massive academic institutions.

Why take a job running a Canadian research organization?

Post-COVID, I had a really strong sense that I wanted to be back in Canada, [which] is in a relatively healthy place from a political and social standpoint. All is not perfect by any means, but when you compare it to so many other parts of the world, the mood is more optimistic. I also happen to have all my family here. 

I was really looking for a smaller organization that was much less bureaucratic than a university, as wonderful as universities are. What CIFAR is set up to achieve is genuinely unique, and hugely important, not just for Canada, but globally.

What is that?

What is unique about CIFAR is its focus on building networks that are both interdisciplinary and cross-geographic, focused on really tough questions at the edge of our ability to explore—[not] just incremental shifts within fields, but really potentially fundamental shifts that cross various fields. As an international lawyer over the last couple of decades, I personally have become more and more interdisciplinary. It just felt like that was what was needed to address some of the biggest challenges that the world faces. 

It’s really incredible to me that the first program of CIFAR was artificial intelligence, robotics and society, 40 years ago. Now, of course, the world has caught on to the importance, and there have been many iterations since then. 

You’ve said one of the reasons you took this job was that CIFAR was willing to rethink aspects of itself. What needs to be rethought?

It is not an organization that’s broken. There are three things that I’m really interested in. 

One is trying to figure out how we are more intentional about handing off the insights that come from the organization. A classic example that most people would not know about: The whole concept of social determinants of health really emerged from CIFAR programming, and that’s had a global impact. What we haven’t done as an organization as well as we might is figure out, “If these great new insights are being built in these programs, how can we hand them off to others to use for policy, for health care, for commercial advantage?” We’re not going to do those things, but there’s no reason that we couldn’t be thinking more carefully about plug-and-play opportunities.

Another is that we are a global organization, but we are really an organization of the Global North. There are some programs where we could be much more thoughtful about trying to make connections with top people in Africa, in India, in Southeast Asia, in Latin America. I really want to try to build out CIFAR’s global capacity, so that it is more connected to the Global South. 

The last one is: We are largely supported by the Canadian government. We’ve historically had provincial support. We can do even more in growing philanthropic support and looking for partnerships outside of Canada with organizations that have complementary mandates. 

The AI institutes already have structured tech transfer. CIFAR seems upstream of a lot of that. What’s its role in commercialization? And what are models that you think can be effective?

We have a program in the fungal kingdom, for example, thinking about new pandemics or how soil is actually protected [and] enhanced. These are fundamental questions with potentially huge commercial outputs. CIFAR is not going to commercialize any of these things. But could we imagine a program where we, with the cooperation and enthusiastic endorsement of fellows of our programs, introduce them to groups of people in the commercial world and in government who are interested in these issues? 

We’re not going to be creating applied-science labs. But [the National Research Council] can be doing that. I’ve actually had a chat with the president of NRC, who is keen on this sort of idea. So could top companies who are interested in whatever field these ideas are coming out of. 

You were talking about particular areas where there might be more opportunity to collaborate in the Global South. What are those?

The fungal kingdom—if you think about that set of questions, it is going to be hugely important in tropical countries. There’s no reason that Brazilians and people from Central Africa shouldn’t be thinking and participating and trying to help us understand what the real risk points and opportunities are. 

In the second phase [of the PCAIS], we’re focusing on AI and health, and AI and energy and the environment. Those are every bit as important in the Global South as they are in Canada. 

There are [also] definitely going to be applications for quantum materials that have huge implications for the developing world, across many geographies. 

If you put your mind to it, you would find that almost every program relates to the developing world in an important way. A social-science program that we have [looking] at innovation and inequality—those are hugely important questions for a country like India.

The first phase of the AI strategy was recruiting all of these research chairs.

We’ve got 119, and there will be refreshing over the next few years. 

What do you see CIFAR doing in the second phase?

Ensuring that people are continually challenged within their environments of the institutes is really important. It’s also important that we try to branch into communities that have not historically been engaged around questions of AI. So connecting AI talent in Indigenous communities; in Black communities in Toronto or Nova Scotia. Making sure that there are as many women who feel included in the opportunity that AI presents. I’ve been convinced that if you really want to measure the success of an AI strategy, it is largely around talent. 

We’re also trying to focus energies around two distinct themes. One is AI and health; the other, AI and energy and the environment. These are fundamental challenges for Canadians: How do we improve the health system? How do we improve health outcomes? How do we shift to new energy sources over time without bankrupting the country? 

There’s an ongoing debate in Canadian innovation policy about who benefits from publicly funded R&D. The three AI institutes’ chief scientists have corporate roles—Geoffrey Hinton with Google, Yoshua Bengio with ServiceNow and Richard Sutton with DeepMind—which have led to criticism that we’re funding research that benefits these foreign tech companies. Is there something to be done, and if so, whose job is it?

I don’t think there’s a problem. Every country on Earth is trying to create incentives for these large corporations to set up research labs. Cambridge has exactly the same group of people—the Googles, the Amazons, the Ubers. It was a huge advantage to have them present. They have technical challenges that need to be resolved. Connectivity to academic institutions is important for them, but it’s also important for researchers to understand what the issues are out there in the world that they might wish to address. 

It’s also a way of creating further anchors for talent. We don’t want the Canadian AI strategy to bring in a bunch of professors, PhDs and postdocs, [only to] have them all leave because there’s nowhere for them to work. We want some of them to be working at great Canadian startups and scale-ups. But the sad truth is there are precious few—in any sector—Canadian anchor companies.

We may need that direct foreign investment and connectivity for the ecosystem to continue to grow. A number of these AI startups have already been acquired by Canadian corporations, including the banks. They now are serving as anchors within the financial-services industry for other startups around tech and finance. 

It would be, frankly, an own goal to be trying to put too many barriers around large companies that are actually at the forefront of technology and can theoretically help Canada grow. Now, we have to have a strategy that doesn’t just leave us in that position forever. 

On a lot of technology issues, we’re seeing a bifurcation of policy and industries between the U.S. and China. Where do you think Canada plays in this? 

The world has changed pretty dramatically in the last 10 years, through the regime of Xi Jinping. There are domestic sources of nativist politics, populism and nationalism that have crept into a range of countries. That’s not just about China; [it’s] broader. It seems to me that Canada really has to be primarily allied with the U.S., its European partners, and Australia and Japan. 

It looked for a while like Canada might be able to—and the United Kingdom thought it was going to—balance itself between China and the U.S., always being closer to the U.S. but with strong links to China. That’s looking harder and harder. Xi Jinping and his regime are talking about self-reliance; they’re not talking so much about partnership. In terms of values, security and IP, the pressure will be to align more with what you might think of as our traditional partners.

What does that mean for a research organization?

We have to make sure that we are strongly connected to those places. CIFAR is very well connected in the U.S., pretty well connected in the U.K. and parts of Europe. But if there is this bifurcation that’s emerging globally, we also have to be aware of what’s happening in places like India, Japan, [South] Korea and South Africa. We don’t want to just abandon the playing field to other forces.

You got a lot of attention for a speech you gave at Peking University [in Beijing] in March 2019. You said “collaboration is not optional” in dealing with global challenges, and you talked particularly about the importance of universities working together.

And I still would stand by that. My views haven’t changed, but I think the practical application has to change. 

This is a global phenomenon. It started in Australia, morphed into the U.S. [and] hit Canada and the U.K. And then of course, our European partners in the Netherlands and Denmark. This was amongst university presidents in Europe, the single most common topic. I can tell you that everyone’s thinking about it. The vast majority—if not almost unanimity—would say that we simply can’t afford not to connect with China around climate change, around some security issues, around the future of food. Over the last number of years, they have invested heavily in their scientific progress. They’re important scientific players, as well as political players and social players. 

You can’t work with China on everything. But there are some things, frankly, where you just have to work with China. Canada in its new Indo-Pacific Strategy has just said essentially the same thing. I’m not out on much of a limb on this one. We’re co-hosting with China the Biodiversity COP, intriguingly, given the pressures that have been felt recently. 

You have to cooperate with your eyes wide open. At Cambridge, for example, we established a set of international cooperation principles—not targeting China, because this is relevant for working with Iran or Russia or any number [of other countries]. It said to our scholars: You have to think about a range of issues if you’re really going to try to enter into collaboration—your own safety and security; IP; national-security tech-transfer issues; the pressures on academic freedom, and whether Chinese students studying at western universities are being pressured. We created a series of tools, mostly online, to really get people to think through these issues. 

Has the experience of being a public figure in the U.K. where your leadership was being called into question changed anything about the way you’re going to do this job?

No. 

In England, anything that happens at Oxford and Cambridge is front-page news. It’s a bizarre phenomenon. There’s just a fixation on the two institutions.

Within the university, I always had incredible support. If you actually look at most of those articles, they quote the same three people. There are literally three people, two of whom were emeritus professors—firmly entitled to their view, but not representative of the wider community. 

I was brought to Cambridge, in part, to address the issues that I addressed. That was very much a part of the hiring process: around greater openness to people from less privileged backgrounds; greater connectivity with the wider world; less self-absorption than has historically been true in the institution. Those were things on which I wanted to deliver. And ensuring that students had better support, both in terms of their finances and in terms of mental health and well-being.

AI was the thing that CIFAR made happen for Canada 40 years ago. What is CIFAR’s next AI?

The easy answer is quantum. But the next really exciting set of questions may be in the social sciences, of innovation, inequality, and social cohesion, addressing that set of challenges.

Disclosure: Irfhan Rawji, chair of The Logic’s board of directors, is also on CIFAR’s board.

Correction: This story has been updated to clarify details about CIFAR’s creation.