I’ve been sceptical that artificial intelligence will radically remake labour markets in the short term, in part because so much hype comes from the tech industry itself. But in the last couple of months, I’ve encountered some very diverse use cases for artificial intelligence (AI) that have me thinking differently.
First, a psychiatrist in New York told me how mental health professionals are beginning to use AI to track clients’ word choices across sessions. Word choice can be an important factor in understanding mental illness and making diagnoses.
Previously, this was dependent on a therapist’s own memory and perceptions. Now, AI-driven analytics will be, in his words, a “game changer” in terms of how effectively patients can be diagnosed and treated.
Second, an investor told me about a company in the US, Axon, that is leveraging AI and proximity data. One product, called Draft One, allows police in places such as Lafayette, Indiana to download body-camera images, then push a few buttons to create a first draft of the “incident reports” that currently take up roughly 40 per cent of their time.
Maureen Dowd: At Jimmy Carter’s funeral, Donald Trump was narcissist non grata
The man gave every indication of being dead just seconds before he revived to defend himself
Donald Clarke: Megyn Kelly’s rant about Conclave being ‘most disgusting anti-Catholic film’ is big win for church
Pat Leahy: Government is getting nervous about the contentious Occupied Territories Bill
While other attempts at high-tech policing have come with unforeseen challenges (such as algo-racism), markets are keen. Axon’s stock is up 730 per cent over five years.
Finally, weary after a long business trip, I recently got a massage in Anchorage, Alaska. My masseuse, who had dropped out of high school, had managed to turn her difficult life around in dramatic ways and wanted to write a memoir to inspire others.
Discovering that I was a writer, she asked me to read her book proposal, which turned out to be as good if not better than many I’ve seen in high-end literary slush piles. Her co-author was ChatGPT.
While unemployment is low in the US today, it is notable that workers in areas ripe for disruption by AI — such as software development or middle management gigs — seem to be getting more redundancy notices.
Consider how many such anecdotes – reaching across geographies and industries – there are, alongside the fact that productivity in the US is finally on the rise after flatlining during Covid.
There may be multiple reasons for that, from low-wage immigration that allowed more skilled workers to move up the ladder into better jobs, to cyclical gains post-pandemic. But most experts agree that greater implementation of cutting-edge technologies such as AI is clearly playing some role in driving up productivity.
As a recent Apollo economic outlook note put it: “The US is experiencing a surge in corporate and research spending on the back of the AI revolution – a dynamic not seen in other developing nations or even China.”
There are three key lessons to take from this.
Could a glut of affordable EVs tempt Irish motorists to make the switch?
First, scepticism about how frothy the US market in general, and technology stocks in particular, are is entirely understandable.
But it’s hard to see how Europe and many other countries would outperform relative to the US if we are heading back to a period of technology disruption that is similar, if not exponentially more dramatic, than what we saw in the 1990s.
Back then, investors will remember, the US pulled ahead of Europe in terms of technology deployment and, as a result, in economic and market growth. AI would seem to be putting that trend on steroids today.
That leads me to lesson two: the fortunes of countries, companies and individuals often diverge in periods of technological change.
The term “jobless recovery” was coined after the 1990-91 recession, because while stock prices, corporate margins and gross domestic product (GDP) growth surged, employment growth lagged behind for longer than expected.
While unemployment is low in the US today, it is notable that workers in areas ripe for disruption by AI – such as software development or middle management gigs – seem to be getting more redundancy notices.
What’s good for markets and even GDP growth today may not be good for politics or society tomorrow — at least not without big policy shifts to help buffer the coming disruptions.
It obviously takes time for job markets and people to adapt to technological change. Yes, as the economists will tell us, new technologies eventually create new jobs.
But as historians, sociologists and social workers can attest, that doesn’t mitigate the pain of those going through such big shifts. Nor does it prevent the disruptive political convulsions that can result.
The disruptions to manufacturing jobs, which represent only a little over 10 per cent of employment in the US and many European Union (EU) countries, brought us Donald Trump and European populism. We are about to see up to 85 per cent of the labour market disrupted to some extent by AI. What will that mean?
For a start, says Nobel laureate and Massachusetts Institute of Technology (MIT) professor Daron Acemoglu, today, as in the past, “it’s likely that the short- to midterm gains from AI will be distributed unequally, and will benefit capital more than labour.”
This brings us to lesson three: people on most rungs of the socio-economic ladder are anxious about the future of work. If the AI enthusiasts are right, it’s difficult to see whose job won’t be affected by the technology. That anxiety has big consequences.
Jim Clark, founder of the recently launched New York-based Future of Employment and Income Institute, tells me that “economic anxiety changes behaviour, and that can create major political shifts”. Think about the Luddites, for instance, whose name is still given to the backlash against tech, or, on the upside, the social welfare state, which came out of the industrial revolution in Germany.
Investors, business leaders and politicians should take this to heart. What’s good for markets and even GDP growth today may not be good for politics or society tomorrow – at least not without big policy shifts to help buffer the coming disruptions. – Copyright The Financial Times Limited 2025
