A new worry is rippling across the stock market lately: entire businesses, not just their employees, may be thrown out of work. While most economists say fears of an AI job apocalypse are overblown, seismic shifts have happened in the past after big tech breakthroughs.
The IT revolution of the 1990s led to a surge in productivity that sped up the US economy for several years. It also rendered companies or even industries largely redundant – from travel agents and stockbrokers to classified advertising and newspapers, or video rental stores.
Economists expect artificial intelligence will deliver higher productivity, which is key to raising growth rates in the long run. But investors are growing nervous about what damage might be done on the way, in capital markets as well as labour markets – especially because AI threatens disruptions on a broader scale than the Internet boom.
“Is this time bigger? Yes,” says Anton Korinek, an AI expert at the University of Virginia – perhaps by a factor of 10. “The key difference from the 1990s is that the Internet only disrupted information distribution,” says Korinek. “AI disrupts cognitive production at large. That’s a much bigger economic surface area.”
Ultimate promise
To be sure, all of this is early-days speculation over a fast-changing and largely untested technology, whose ultimate promise is to make workers more productive.
Productivity is essentially a measure of how much output workers can deliver with the available tools, so it tends to surge upward when someone invents important new ones like the Internet or AI.
US labour productivity rose 2.8 per cent last quarter, according to data out on Thursday (Mar 5). Economists typically don’t read too much into one quarter, since the numbers tend to jump around. Still, the trend has been ticking up. After big swings in the pandemic period, productivity has grown at an average pace of 2.8 per cent since the start of 2023. That’s more than double the average for the decade through 2019.
There’s intense debate over how much of this acceleration is due to AI. But even analysts who reckon the new technology isn’t yet making a big contribution mostly expect it will do so before long.
A more productive workforce drives the kind of efficiency gains that can allow both corporations and their employees to boost earnings, without triggering inflation. Historically, economies adapt to big tech breakthroughs – creating new industries and professions that nobody could’ve envisioned before – and living standards rise.
‘How it has to be’
That’s the long-run view – which smooths out lots of bumps on the road.
“Having some boom-bust in a sector is normal,” says Simon Johnson, the Nobel prizewinning economist at MIT. “Maybe even how it has to be.” But as companies go under, he says, it can create wider risks – especially if the failed businesses borrowed lots of money. “What you don’t want is to infect the credit, and you definitely don’t want to get inside the banking system.”
As of now, on US capital markets, what’s become known as the “AI scare trade” is barely a blip.
The S&P 500 is up by about two-thirds since the release of ChatGPT in November 2022. A big chunk of those gains has been driven by the surging value of AI companies themselves and their suppliers – giants like Meta Platforms and Nvidia Corp. – which creates one set of risks if their technology disappoints.
But there’s another set – the one behind recent market wobbles – which is different. It stems from the possibility that AI does deliver the promised productivity leap, and then some. That idea, captured in a research note by the little-known firm Citrini, sent the S&P into a brief nosedive at the start of last week.
Citrini’s scenario of massive white-collar layoffs driven by AI was basically sci-fi, set in 2028. There’s no sign of anything like that now, with US unemployment at historically low levels.
Still, Daniel Keum at Columbia Business School, who’s studied how automation technologies like AI change the balance within firms, sees signs of a shift. His research, based on comments in earnings calls and annual reports, found that bosses become more likely to refer to their employees as costs, among other evidence of a tilt in power away from workers.
Also, even if companies aren’t yet cutting jobs or wages, they are trimming in areas like health care, remote work and even free snacks, Keum says. “These side benefits is what the companies go after first, before they go after reducing your paycheck.”
‘Essence of capitalism’
When businesses can cut payroll costs because technology enables them to do more with less, that’s often good news for their profits and shareholders. Take Block, the fintech firm run by Twitter founder Jack Dorsey, which said on Feb 26 it’s slashing almost half its staff in a bet on AI productivity. The shares are up more than 15 per cent since then.
But last week also offered an example of how productivity gains can have a downside for investors – involving the storied International Business Machines (IBM). The startup Anthropic said its AI tool can do something that once needed “armies of consultants”: modernise Cobol, a dated programming language run on IBM computers. IBM shares plunged the most in a quarter-century, before recouping most of the losses.
Past technology booms have seen household-name businesses fall by the wayside – like camera-maker Kodak and video-rental chain Blockbuster, left behind by the Internet. It’s all part of what the economist Joseph Schumpeter called the “creative destruction” which leads to progress.
Federal Reserve Bank of Richmond president Tom Barkin referenced that phrase last week when asked if the Fed should be trying to counter AI disruption to businesses and the labour market. “That’s been happening for hundreds of years in this country,” he said. “It’s part of the essence of capitalism.”
That’s not necessarily reassuring for industries – and their investors and employees – confronting the short-term risk. Korinek at UVA lists some of them, including back-office services, content production, customer support, legal and financial analysis, and coding.
“Eventually, the disruption will extend to any firm whose competitive advantage lies in human expertise that AI can replicate,” he says. “The transition period may involve stranded assets, debt overhang and the potential for sharp market corrections.”