Mr. Carney comes to Canada to talk about artificial intelligence

TORONTO—It’s not often that Bank of England Governor Mark Carney makes it back home to Canada to deliver a public speech. But he did that just the other day in Toronto on the subject of artificial intelligence, outlining its benefits but also pointing to needed new policies. As Carney quipped to an audience of several hundred enthusiasts gathered for the fourth annual AI summit at the University of Toronto, “you know artificial intelligence is growing when you invite a central banker.” The audience included David Dodge, a former fellow governor of the Bank of Canada, along with Ed Clark, former CEO of TD Bank Group and a driving force behind the creation of the Vector Institute, a leading centre for AI research at the university. While Toronto has emerged as one of the talent and research hot spots in the world of AI, it is also where there is a growing and much-needed attention to the economic impact and social challenges of AI. This was much of the focus at the Toronto summit, organized by the university’s Rotman School of Management. It is noteworthy that the U.S.-based National Bureau of Economic Research a year ago chose Toronto to hold a symposium to set out its research agenda for the coming economic impact of AI, followed by a second such gathering in Toronto last month. The NBER is the leading U.S. institute for economic research and disseminating its results. There are many challenges as the advancement of AI research and its adoption continues, not only by business but for its use in healthcare, smart city management, law enforcement, infrastructure, education and many other areas. AI is a general purpose technology—a technology, like electricity, the internal combustion engine or the internet, which has the potential to affect all aspects of economic and social life and which is disruptive as well as transformative. The development of self-driving vehicles is just one example of the spin-out effect of AI as a general-purpose technology. But our biggest gaps in knowledge are less in science and technology, where research is moving rapidly, and more in the social sciences, as speakers at the Toronto summit indicated. Concerns around the future of work; inequality if wage gains go largely to the most skilled while mid-level jobs disappear; the identification of future skills needs; how AI impacts productivity and innovation; the level of competition and the threat of big data monopolies that engage in monopoly pricing and thwart the entry of new players; issues of privacy and data ownership; built-in AI biases on gender and race; the growing gap between firms at the frontier of AI and the rest; the challenges for a country like Canada in building globally competitive AI businesses rather than serving as a research centre to develop intellectual property for multinationals based elsewhere; and how to accelerate technology diffusion so that the potential benefits are more widely shared by small and midsize companies as well as the biggest. This is why we need to pay much more attention to the need for greater understanding of the economic and social consequences of AI, while of course continuing to invest in our talent-based technology research capabilities. The Canadian questions are very much about how can we build AI companies so that some of the coming products and processes will be developed in Canada, and how can we accelerate the adoption of AI technologies, especially in an economy which has a higher proportion of small and midsize businesses and a much smaller population of large-scale global businesses, so that our businesses achieve levels of innovation and productivity that underpin their future growth and generate the wealth for successful society. As Ontario’s Institute for Competitiveness and Prosperity warned in an AI report earlier this year, if we fail to retain our skilled talent here for lack of opportunity in commercialized applications, our economy will “eventually stagnate, as it shifts to being an importer (purchaser) of technology rather than an exporter .” Losing our costly investment in talent to better opportunities elsewhere at a time when they are at their prime innovation creation potential “is an unsustainable business model.” Competition policy, commercialization, technology diffusion and balancing privacy against the need for big data are among the major challenges. Linda Hasenfratz, CEO of auto parts maker Linamar Corp. with global revenues in the $6-billion range, outlined how her company—with an ambitious 100-year vision—is looking to AI and machine learning as a core part of the company’s future innovation. Yet a recent report on the success rate of business successfully adopting A.I. in10 countries, Canadian businesses ranked at the bottom of the list. Another report, from Deloitte Consulting, showed that for early adopters of AI, investments are paying off significantly. However, the adoption rate for emerging technologies by the largest corporations are 10 times higher than for small and midsize businesses, with the largest businesses having information technology departments and resources to test, deploy and manage emerging technologies while smaller companies can’t afford the budgets and lack the people to pursue emerging technologies such as AI. The big picture challenge of AI as a general-purpose technology was set out more than a decade ago by a leading thinker on innovation, Carlotta Perez. She stressed that each major technological shift “breaks the existing organizational habits in technology, the economy, management and social institutions.” The present generations, she wrote, “are living through a period requiring intense social and institutional creativity” which will bring social conflict, confrontation, negotiations, agreements, and compromises on policies and institutions at all levels. This is not trivial change. Politically, it is hard to see how one can be an effective MP without understanding what is at stake in the AI revolution for the kind of country we have in the future. David Crane can be reached at crane@interlog.com. The Hill Times