When it comes to the stock market, the news cycle, and boring conversations at cocktail parties, AI is inescapable. But it hasn’t really shown up in economic data — that may be changing.
What happened: A newly published study found that, across 12,000 European firms, AI increased productivity levels by 4%. To put that in perspective, productivity in Canada increased by around 1% per year from 2000 to 2019 (and has been even lower since).
Meanwhile: Last year’s U.S. jobs numbers were revised downward by 403,000 jobs at the same time as the economy was growing quickly, a trend technology researcher Erik Brynjolfsson called a “hallmark of productivity growth.”
“We are transitioning from an era of AI experimentation to one of structural utility,” Brynjolfsson wrote.
Why it matters: You can find all manner of predictions about how AI will either have no impact on jobs or completely eliminate white-collar work, but the hard data suggests an outcome somewhere in between those extremes is more likely — at least for the time being.
Yes, but: We’re only beginning to get hard evidence as to how AI is impacting the labour market, and it’s difficult to establish causality when it comes to trends across the entire economy. We’ll need many more quarters of data before we can reach any definite conclusions.—TS