What if … Canadians Henry Woodward and Mathew Evans had commercialized the light bulb? 

The entrepreneur responsible for the democratization of electric light, Thomas Edison, became one of the most famous people in the world in the 1920s and his company, Edison General Electric, was an industrial force for most of the 20th century. 

Canada could have used an anchor company like that. Alas, Woodward and Evans are protagonists in a classic Canadian so-close-yet-so-far story.

Like Edison and dozens of other inventors around the world in the later half of the 1800s, they had dreams of turning a transformational technology into practical products that would bring them fame and fortune. They secured a Canadian patent for a light bulb in 1874, and a U.S. patent two years later. But their destiny was to be forgotten. After struggling to convince investors their product had a future, Woodward sold his American patent to Edison in 1879 for $5,000, who then purchased the Canadian rights six years later. 

Canada once again finds itself in the mix to capitalize on a revolutionary technology. Some of the “godfathers” of artificial intelligence made their  breakthroughs at Canadian universities and continue to call Canada home. At the end of 2020, Canada had 20 public AI research labs, 75 AI incubators, 60 groups of AI investors and more than 850 AI startups, according to the federal industry department. That’s a sturdy foundation, but for what?

Canada’s early lead in fundamental research won’t guarantee it remains in the vanguard. Element AI, the buzzy government-backed startup that included the Université de Montréal’s Yoshua Bengio among its founders, ran out of money and was purchased by Santa Clara, Calif.-based ServiceNow in November 2020. 

Still, Bengio came closer to building a homegrown AI champion than his fellow godfathers; by then, both the University of Toronto’s Geoffrey Hinton and the University of Alberta’s Richard Sutton had joined Alphabet’s orbit. 

It amounts to one of those moments when history rhymes. 

Woodward and Evans never stood a chance against Edison, a uniquely ambitious entrepreneur who built the world’s first industrial research laboratory, compiled more than 2,000 patents in the U.S. and abroad, and was backed by J.P. Morgan and the Vanderbilt family. The dominant players in AI—including Alphabet, Meta and Microsoft-backed OpenAI—share Edison’s ambition, and are even better placed to pursue it because they have plenty of money of their own to advance the research, buy computing power and secure patents. 

So how can Canada win? Maybe by making itself impervious to losing, at home and on the world stage. That means getting the policy right.  

Any strategy starts with acknowledging the risks. Bengio and Hinton are among the luminaries who think the rapid advance of AI could take a dystopian turn. Hinton quit Alphabet’s Google in May so he could speak freely about his concerns. They worry that in the wrong hands the technology could wreak terrible havoc, and that it could become too intelligent for its creators to control. 

But acknowledging the threats doesn’t mean becoming enthralled by them. Economist Tyler Cowen ridicules the AI doomsters for failing to support their theoretical concerns with rigorous modelling. The odds of reversing climate change—an existential threat that has been rigorously modelled—are suddenly significantly better because of the speed at which AI can analyze data. The same goes for the challenge of generating enough wealth to support an aging population, and so many other daunting problems. The potential benefit from using AI to help find solutions now outweighs the theoretical threats that might emerge at some point in the future. 

The U.S. and the Soviet Union found a way to contain the threat of nuclear war, so it’s reasonable to think Washington and Beijing could agree to keep a leash on AI. Canada’s role in developing the technology could let it play honest broker, as it did in the aftermath of the 2008–09 financial crisis, when Canadian officials were influential in the overhaul of global banking regulation. Bengio has already stepped into such a role.

The struggle for AI dominance will be a feature of the most intense geopolitical rivalry since the end of the Cold War. The United States and China are the undisputed leaders in the field, and their governments will ensure it stays that way. The best a country such as Canada can hope for is to remain relevant to one or both of Washington and Beijing. 

Meanwhile, Ottawa needs to start thinking about the policies and tools it will need to make the most of a potential once-in-a-lifetime economic windfall.

In commentary the Centre for International Governance Innovation published in October, researchers Dan Ciuriak and Anna Artyushina said that Canada should resign itself to being primarily an importer of AI assets, and recognize that even if some homegrown companies break through, they will be exporting into a marketplace shaped by the U.S., China, and the European Union. 

Ciuriak and Artyushina called on Ottawa to create a federal agency devoted to AI because the technology will influence all aspects of life and an ad hoc strategy overseen by various ministers and private interests would create too many cracks. Some will cringe at the mention of more bureaucracy, but when you consider the breadth of issues that AI will force policymakers to confront, there’s merit to collecting the government’s best thinking about the technology in one place. 

That doesn’t let lawmakers off the hook. If it looks like AI is moving too fast, they could consider taxing the uses that have no obvious public good, including digital advertising. Economists Daron Acemoğlu and Simon Johnson would raise taxes on capital, and lower them on labour, to offset the incentive to replace people with machines and reduce the risk of mass technological unemployment. 

These ideas are merely a few examples of the ambitious thinking Canada will need to mitigate the risks of AI while maximizing our opportunity. There’s time, but probably not as much time as previous generations of policymakers have allowed themselves. That’s the biggest reason for worry. Technology is racing ahead at the same time that political decision making is slowing down. We need to create the conditions in which an Edison of our own can thrive. History suggests that won’t happen unless politicians rise to the occasion. 

Kevin Carmichael is The Logic’s economics columnist and editor-at-large. He has spent more than two decades covering economics, business and finance for outlets including Bloomberg News, The Globe and Mail and the Financial Post, where he also served as editor-in-chief.