Bank of Canada’s Tiff Macklem warns AI could destroy more jobs than it creates

Artificial intelligence could have disruptive effects on the economy and end up destroying more jobs than it creates, Bank of Canada Governor Tiff Macklem said on Friday. He also cautioned that the impact of the technology is unknown and requires careful study.

“I would be wary of anyone who claims to know where AI will take us. There’s just too much uncertainty to be confident,” Mr. Macklem told an AI conference in Toronto organized by the National Bureau of Economic Research.

The effects of generative AI in particular have become more pressing for economists and policy makers since the release of OpenAI’s ChatGPT in 2022, which kicked off a wave of investment and hype in the technology. Generative AI products can produce text, images and other media, answer questions, summarize documents and do many other tasks, giving rise to both sunny and gloomy predictions about the impact of AI on the economy and jobs.

“As AI becomes more established in the economy and its impacts more transformative, it could end up destroying more jobs than it creates,” Mr. Macklem warned. “And the people who lose their work to automation may struggle to find new opportunities. This is a concern for all of us.”

The past 200 years of economic history show that the effects of lost employment owing to changes in technology have been outweighed by the benefits, including new jobs that are created and more demand for tasks that cannot be automated, according to Mr. Macklem.

But some contend that AI could be different, he noted, in part due to the speed at which it could be adopted and the capabilities that developers could achieve in the future. “AI could shrink the number of non-automated tasks so much that there isn’t enough work left for the displaced workers,” he said.

There is little evidence today to show that AI is displacing workers at a rate that would lead to declines in total employment. In fact, employment in computer systems design and related services, which Mr. Macklem described as a proxy for digitalization, has increased 48 per cent since the end of 2019, compared with 6 per cent for the rest of the economy.

In the short-term, AI could contribute to inflationary pressures, he added. Investment in AI is boosting demand in the economy along with equity prices, which supports greater consumption. Companies are hiring workers with AI skills, as well as building AI data centres.

Generative AI is also seen as a way to boost productivity by saving workers time as well as leading to the creation of new products and services. Mr. Macklem said that while we can expect AI to benefit productivity, the effects will take time to play out. It’s also not yet clear whether AI will significantly increase productivity or keep it rising at its historical pace, like previous innovations such as the sewing machine.

“We are still looking for the new products and services – and the new business models – that will transform efficiency and productivity,” he said.

The degree to which the technology can increase productivity is hotly debated, with economists, think tanks and consultants arriving at estimates that can vary widely.

Microsoft, which has invested billions into OpenAI, contends that generative AI could add $180-billion to Canada’s GDP annually by 2030 and boost labour productivity by “an astounding” 8 per cent.

McKinsey & Co. has estimated that generative AI could automate up to 30 per cent of business activities by 2030, and grow productivity by 0.7 per cent every year in the U.S. through to 2040.

Others are more reserved in their predictions. Daron Acemoglu, an economics professor at the Massachusetts Institute of Technology, has said that less than 5 per cent of tasks will be cost-effective to automate in the next decade, contributing to productivity growth in the U.S. of just 0.5 per cent over the same time period.

A lot depends on the sophistication of the technology and how quickly businesses adopt it. According to Microsoft, Canada lags behind its OECD peers and only 9 per cent of businesses are using generative AI.