Artificial intelligence is already reshaping how we work, learn, govern, and care for one another. There is pressure to use it with great enthusiasm for its many promises, and as much angst and a willingness to attempt to resist it.

For the latest episode of Inside Policy Talks, I spoke with Technology Economics Professor Joel Blit from University of Waterloo about the fourth Industrial Revolution, and assess its potential for good and ill, while focusing on what’s unique to the Canadian context.

AI is generally misunderstood. It is a general purpose technology that will transform work across numerous industries. Joel has developed a framework for understanding the impact of industrial revolutions, which he divides into phases: replace, reimagine and recombine.

I argue that those who think AI will fundamentally replace work do not understand what is at the core of markets and economic exchange. Work is ultimately a form of service, and through every technological revolution, we have simply changed how we work.

Artificial intelligence is also a misnomer – artificial, yes, but intelligent, no. It is a complex system of algorithms (language-based rule following) built on fundamental assumptions that humans give it. AI operates on the basis of its given axioms and algorithmic coefficients – even those that ‘learn’. It is running a program, which is not general intelligence. Fundamentally, intelligence involves the capacity to abstract formal features from an object of sense perception, or identifying a purely mental construct through the imagination. This requires consciousness, which AI certainly does not have, nor is consciousness reducible to functional organization, like many have tried to claim.

When it comes to the economics and policy surrounding AI, we cover many things.

Canada has real strengths in AI research and talent. But we continue to struggle with productivity, commercialization, scaling, and adoption. Only 12.2 percent of Canadian firms used AI to produce goods or deliver services in 2025, even though adoption had doubled from the previous year.

AI could raise living standards, reduce administrative burdens, and help firms become more productive. But it will also disrupt labour markets, especially entry-level white-collar work, where many routine cognitive tasks are now vulnerable to automation.

The deeper question is not simply whether AI will replace workers. It is whether it will augment human judgment, deepen expertise, and free people for higher-value work — or whether we will use it to hollow out the human relationships, tacit knowledge, and care that many professions depend on. That is especially important in fields like health care, education, elder care, therapy, and social services. AI can support nurses, teachers, doctors, and caregivers.

It can reduce paperwork and improve access. But there will be tremendous pressure – misguided at that – to attempt to replace activities that require embodiment, moral agency, consciousness, emotional experience, or the human relationship at the centre of care in its market and more important non market forms.

For Canada, the policy challenge is to pursue AI adoption while regulating concrete harms. We need to distinguish productive scale from the legitimate abuse of market power in the domain of competition policy, focus on the application layer where Canada may have real advantages and stands the most to gain in terms of productivity, and ensure any digital regulation addresses identifiable harms, like the use of chatbots, illicit image-generation, and safety and power concerns if they arise with emergent models, without importing European-style digital regulation that slows innovation without solving the real social problems.

That is the balance we need to strike.

Available on other platforms here.