Interested in more careers-related content? Check out our new weekly Work Life newsletter. Sent every Monday afternoon.
The impact of artificial intelligence on the workforce is the topic of the year, and the estimates of job losses range from meaningful to terrifying. Still, as we weigh the potential damage, we might be missing an important upside.
If the end game with any new technology is to increase productivity, then a technology as powerful as AI can surely be expected to cause a surge of economic activity and wealth. As that happens, there is a case to be made that demand for workers will actually rise although it might take some time to get to that point.
The idea that AI and productivity will rise together is obvious. You can certainly see it in the financial markets, where optimism about profitability has lifted valuations. Although some worry that surging stock prices are indicative of a bubble, the prevailing view is that they are rational given the potential of AI. That is, if all goes well with AI, then the productivity of each worker employed will go exponentially higher, hence the need for fewer workers.
Economic forecasters are also upbeat on what might come about as AI is broadly deployed, with the International Monetary Fund (IMF) in February saying they saw it as likely to raise global growth substantially.
The idea that workers will share in the riches is harder to believe. That is no surprise given that companies are already shedding workers, saying they will need fewer as they deploy AI.
Since the beginning of 2026, the list of companies culling staff because of AI has been growing and now includes Block, TD, Scotiabank, Air Canada, Rogers, Meta, Amazon and Microsoft. In some cases, companies have said directly that they can get AI to do what workers used to do, while in others they have cited putting more money toward AI investment as the reason they need to pare back payrolls. Either way, there is an inverse relationship between employment and spending on AI.
Making the case that AI will cause hiring starts with the idea that it will eventually be good for the broader economy. A new study by Boston Consulting Group (BCG) makes this case, albeit with the caveat that it will take some time.
Looking at the U.S. labour force, they conjecture that over the next two to three years 50 to 55 per cent of U.S. jobs will be ‘reshaped’ by AI. In some cases, that will mean an augmentation of current roles. But, over time, they also see the productivity gains from AI triggering higher demand and then a need for new human roles.
That some jobs will be lost is not a question. BCG estimates that in the U.S. about 10 to 15 per cent of jobs will be eliminated over the next five years. The flip side is the companies that use AI aggressively will be able to deliver their products more efficiently, which will trigger declines in prices. This, after all, is what we have seen with any number of products that came from new technologies from flat screen TVs to laptop computers to cars. That might cause a rise in demand that, in turn, means a need for more workers.
Economists call this relationship, where efficiency improvements trigger higher demand, the Jevons Paradox – Jevons being the 19th-century economist who figured out that more efficient ways of using coal would trigger more demand for it, not less.
In the case of coal, policymakers at the time were worried they were running out of the resource, but if appliances could run on less of it then coal would last longer. The premise that Mr. Jevons laid out in an 1865 book and which came to pass, was that as coal could be used more efficiently, businesses would prosper and use more of it.
In the case of AI, the question is whether its widespread use does boost worker productivity and whether the resulting fall in prices causes demand to surge and industries to expand hiring.
You can extend this further and look at whether some occupations might get a Jevons Paradox surge. This is already happening with software developers and coders, with demand for their services rising even though AI tools have raised industry productivity and threatened to put many out of a job.
Another oft cited example is radiologists, whose work has been transformed by new technology which has increased accuracy and decreased cost. As a result, instead of radiologists having less work, doctors are ordering more scans which has increased demand.
At this point, it is early to know which industries and which occupations might get a boost from widespread AI adoption. In the short run, we will continue to see organizations freezing hiring or paring back headcount as they revise their needs and invest in AI.
Over time they will reassess, perhaps investing in training of workers so they can make the most of the new technologies. The headlines might continue to seem bleak for workers but if AI technologies live up to their promises, we might eventually need more workers, not fewer.