How To Guess If Your Job Will Exist in Five Years
Will my job even exist in five years? Following the rise of Claude Code and ChatGPT, pretty much every white-collar worker I know has been asking themselves that question. AI can code like an engineer, write a business plan like a consultant, decorate like an interior designer, and answer medical questions better than a doctor. It can make up a shockingly catchy and shockingly filthy country tune, and croon it in a voice drenched in Tennessee whiskey. The realization that America might not need so many engineers, consultants, interior designers, doctors, and country singers in the future naturally follows. Searches for the phrase job apocalypse are spiking. Polls show that voters are beginning to freak out.
But there’s a better question for white-collar workers to ask themselves: Am I coal, or am I a horse?
Horses and mules have had a rough go of it in the labor market, to say nothing of hinnies. American farms employed 26,493,000 equines in 1915. One hundred years later, the number of such animals on the payroll had collapsed to 700,000. (To be fair, the data aren’t great.) Farmers needed horses until tractors and trucks did their work better, so farmers hired millions of them instead. (Again, not great data. The government inexplicably stopped keeping a tally of farm trucks in 2013, though it still counts the number of tractors.)
The problem with horses—one of the problems with horses, I say as a former horse girl—is that they are as stubborn as mules. When the combine rolled onto the alfalfa field, horses did not see the writing on the barn wall and start applying for factory jobs. They didn’t learn to code or attend community college. They stood there and ate carrots.
Humans have managed the tides of change far better. More than half of the American labor force worked in agriculture in 1880, compared with 2 percent today. But farmers didn’t become obsolete. They became sewing-machine operators whose children became steamfitters, whose children became teachers, whose children became contestants on reality shows. Many jobs common 100 years ago—drayman, telephone operator, woodchopper, hoistman—are niche or nonexistent today.
[From the March 2026 issue: America isn’t ready for what AI will do to jobs]
I’m smoothing things over, granted. The transition from agricultural employment to factory employment involved wrenching mass migration, the utter misery of the Great Depression (as well as other brutal recessions, now faded from collective memory), and the painful dealmaking of the New Deal. The shift from blue-collar work to white-collar work has decimated communities such as Dayton and Youngstown and Muncie, and contributed to the rise of extreme inequality. We’re still experiencing the social, political, and public-health consequences. When technology leads to huge numbers of jobs disappearing quickly, bad things happen. Still, humans are transformers. We adapt, we change, we prosper, and sometimes we thrive, even as a robot revolutionizes our job—which is where coal comes in.
In 1865, a brilliant English economist named William Stanley Jevons published a book, The Coal Question: An Inquiry Concerning the Progress of the Nation, and the Probable Exhaustion of Our Coal-Mines. Coal “is the material energy of the country—the universal aid—the factor in everything we do,” he wrote. “With coal almost any feat is possible or easy; without it we are thrown back into the laborious poverty of early times.” The text is rhapsodic, romantic even. (Find someone who looks at you the way Jevons looked at a furnace.) But Jevons wasn’t overdoing it. England was one of a handful of places where coal sat on or near the surface of the Earth, along with Appalachia and the Ruhr, the economic historian Brad DeLong told me when I called him to chat about Jevons. You could dig it up and heat a home, boil water, forge a metal, or power an engine. Coal was why England was the seat of the Industrial Revolution. Coal made England rich.
Yet the country had only so much of it. (Problem.) Efficiency improvements in steam-powered engines would ease demand, economists who weren’t Jevons argued. (Solution?) But the same improvements would drive down the price of consumer goods, push up corporate profits, and extend the penetration of coal into the economy; efficiency would lead to the demand for coal increasing, not decreasing. That insight became known as the Jevons paradox within the world of academia. England would need more coal, or it would decline as an economic power.
You’ll be hearing a lot about this obscure Victorian notion in the coming years. Silicon Valley types can’t stop talking about it. Searches for the phrase Jevons paradox are looking a lot like searches for the phrase job apocalypse. Contemporary examples of the paradox in action are plenty: LEDs, heat pumps, and front-loading washer-dryers use less electricity than incandescent bulbs, furnaces, and top-loaders. But the United States uses more electricity now than it did 20 or 50 years ago because we have so many more electrical gadgets. Broadband, mobile data, and semiconductors are Jevons-paradoxical too. Better phones and faster networks have led to people watching short-form videos every waking moment, meaning we need more bandwidth. Advanced chips that turn everything into a tiny computer means that someone can hack your coffee maker and demand a ransom, meaning we need more chips.
The paradox occurs in the labor market as well, with humans in many jobs standing in for coal. In 2016, the Nobel laureate Geoffrey Hinton argued that we “should stop training radiologists” because software would soon render them obsolete. But improvements in medical imaging unlocked new use cases for CTs and MRIs; patients demanded, and doctors ordered, more tests, and radiologists were the doctors administering and interpreting them. Technology acted as a complement to human work rather than a substitute for it. Ditto with radiology and AI. At least for now, artificial intelligence is changing how doctors do their job, not eating their lunch.
Perhaps the most relevant example is software engineers. Earlier this month, the fintech company Block fired half of its employees. “We’re already seeing that the intelligence tools we’re creating and using, paired with smaller and flatter teams, are enabling a new way of working,” Jack Dorsey, the company’s CEO, explained on X. “I had two options: cut gradually over months or years as this shift plays out, or be honest about where we are and act on it now.”
To translate that from corporate-speak: To Dorsey, people are horses. Innovation is driving them out of existence. But people are coal—or, to be more precise, coders seem to be coal at the moment. Businesses employ 6 percent more software engineers now than they did a year ago, in part because corporate executives are desperate for workers to figure out how to develop or implement AI products. (Dorsey might have had other reasons for axing those 4,000 Block employees, by the way. A former employee speculated in The New York Times that the cuts were the result of “standard prioritization and cost management, not an A.I.-driven reinvention.”)
[Kwame Anthony Appiah: The age of de-skilling]
Of course, people can be horses and coal and a thousand other things, because AI will have different effects on different workers in different industries in different places and in different times. The government, too, could affect the job market: It could raise taxes on corporate profits, or pass AI-constraining or labor-protecting regulations. Part of the reason Hinton’s prediction about radiologists never came true is that medical-imaging software needs to go through the notoriously challenging FDA review process, slowing down its deployment and raising costs. Given the friction out there, AI’s labor-market and productivity effects could be far more muted than many people (me) fear. And some sectors of the economy won’t be affected much, if at all. The most common job in the Bay Area isn’t AI-systems architect. It’s home health aide.
As AI changes everything or nothing or both, coal is finally headed the way of the horse, in England at least. Jevons correctly predicted that technological advances would increase demand for coal, yet he profoundly underestimated the ability of British mines and the global market to meet that demand. The country’s production volumes increased until the early 1950s. And while Jevons was writing his book, a Pennsylvanian called Colonel Drake (not actually a colonel) was figuring out how to drill deep into the Earth and pump out petroleum. Soon after, oil and gas supplanted coal, as coal had supplanted biomass, and as solar will someday—perhaps—supplant oil and gas. All of England’s industrial-scale coal mines have closed. Its last coal-fired power plant closed in 2024. Today, the country uses as much coal as it used in 1666, when the most common profession was peasant.