The future of work is changing fast. But then again, it often has.

Past generations of workers lived through changes driven by the Industrial Revolution, or the introduction of the loom, steam power, electricity, the assembly line, factory automation, personal computers, and the internet.

For 200 years or more, new technologies have transformed the workplace. And here’s some good news: Once the changes took hold, the labor market usually ended up with more jobs and newly freed-up workers to fill them. And many of those jobs would have been hard to imagine before the disruption.

“People still go to work, they still make incomes, they support their families – and that’s despite a century of tremendous technological change and an economy where work looks incredibly different than it used to,” says Daniel Gross, an associate professor of business administration at Duke University’s Fuqua School of Business, and a faculty research fellow at the National Bureau of Economic Research (NBER).

“The question people come to is, ‘Well, is this time different?’” Gross says. “People have asked that question essentially every time one of these waves has arrived of: ‘We don’t think this is like the others.’ Right? ‘This is the calamity we’ve all been waiting for’ – and that hasn’t really borne out. And given the historical pattern, I’m skeptical that it will this time, either.”

However, though some workers merely saw their jobs evolve or improve during prior disruptions, others were unable to adapt. Former factory towns fell on hard times. Careers were diverted or derailed. Workers saw fulfilling jobs carved up into unsatisfying tasks built around automation.

“Based on historical precedent, there’s no reason to think that all work will come to an end,” says Jason Resnikoff, author of “Labor’s End: How the Promise of Automation Degraded Work.” “There is reason to think that much of the (new) work won’t be good.”

So which will it be this time? Not everyone agrees, but there are common threads suggesting it will be a little bit of both – plus some new wrinkles thrown in.

White-Collar Workers in Danger, but More Jobs Coming?

For one thing, whereas many past disruptions replaced physical power with automated machines, some of today’s disruptions, such as AI, could replace brainpower. That could put more white-collar jobs in jeopardy and affect a different geography, according to a Brookings Institution report.

But overall, “we do … not see any indication in our research that there would be a long-term structural issue with the quantity of jobs,” says Till Leopold, head of work, wages and job creation at the World Economic Forum, whose Future of Jobs Report 2025 projects the creation of 170 million new jobs and the displacement of 92 million existing ones worldwide in the next five years.

“The challenge,” Leopold adds, “is the very fast and accelerating change in the types of jobs being created and destroyed, and the fact that the skills requirements in those newly emerging jobs often look very different from the ones in the jobs that are being disrupted.”

Which Jobs Face Change?

As with past disruptions, that will require retraining or upskilling workers directly in the path of change. In this case, that may be driven by forces such as artificial intelligence and emerging technologies, the green transition, and demographic shifts.

Even in existing jobs, the World Economic Forum projects 40% of core skills, on average, will change by 2030. And the McKinsey Global Institute’s latest U.S. projection anticipates that 30% of hours currently worked could be automated in that time.

That McKinsey projection has increased recently because “advancements in gen AI have just made it so much more possible to do more of the higher cognitive work … through automation,” says Anu Madgavkar, a partner at the McKinsey Global Institute focused on the future of work and related topics.

McKinsey reports its “updated modeling of the future of work finds that demand for workers in STEM-related, health care, and other high-skill professions would rise, while demand for occupations such as office workers, production workers, and customer service representatives would decline.” Broadly, McKinsey expects technological, social and emotional skills to be in higher demand. It forecasts that cognitive skills will be less sought after.

Madgavkar suggests skilled laborers such as plumbers and electricians, or patient care workers in health care could be relatively safe.

“There are a host of physical tasks which are still really hard to automate,” she notes.

AI ‘Tools Have Not Moved the Needle on the Hard Economic Outcomes’ – Yet

But how soon will most of us feel the effects?

A recent working paper examining Danish labor data found that generative AI chatbots have not yet impacted the earnings of workers in 11 exposed occupations. Those 11, measured into 2024, are accountants, customer support specialists, financial advisors, HR professionals, IT support specialists, journalists, legal professionals, marketing professionals, office clerks, software developers, and teachers.

“Despite substantial investment from both employers and workers, these tools have so far had zero impact on workers’ earnings, recorded hours, or wages,” one of the study’s co-authors, Anders Humlum, an assistant professor of economics at the University of Chicago Booth School of Business, told WorkingNation. “In the short run, our evidence shows that the tools have not moved the needle on the hard economic outcomes.”

Believe it or not, the period from 1990 to 2017 actually was less disruptive to American work than any period going back to 1880, according to an NBER working paper by a trio of Harvard University scholars, David J. Deming, Christopher Ong and Lawrence H. Summers, while their analysis found the period from the 1940s to the 1960s to be the most volatile.

But in the same paper, the authors suggest AI could be a labor market disruptor on par with steam power and electricity, and the pace of change seems to have accelerated since the pandemic.

Judging by the past, the labor economy might organically adapt. Disruption could affect some jobs gradually over generations, while others might feel the tide turn more quickly, if they’re not feeling it already. Some displaced workers will need new skills and careers right away, while others might gradually learn new skills at work as their job descriptions shift.

Ways Change Can Benefit Some Workers

In what could be an encouraging sign for workers, Humlum says the data he’s looked at suggests high performers appear to be most avidly adopting AI chatbots. Could that be an indication that AI initially may augment high-skill work rather than displace its workers?

“If they were just the perfect substitute for their high-skilled workers, you would … expect to see the opposite path: that a highly skilled worker would have no reason to use it, but only the lowest performers would have an incentive to do so,” he says.

However, another recent paper found that generative AI has the power to level the playing field and reverse polarization by allowing less-skilled workers to rise toward the level of their more highly skilled colleagues.

“AI, if used well, can assist with restoring the middle-skill, middle-class heart of the U.S. labor market that has been hollowed out by automation and globalization,” writes David Autor, an economist at MIT who authored the paper.

Even if workforce disruption does accelerate, as some predict, history suggests a few of those in fading fields could actually see a surge in demand for their talents as incoming workers forsake disrupted occupations and labor shortages arise.

As an example, Gross cites outside research on how “anticipatory dread” created a shortage of horse-drawn freight wagon drivers as motorized trucks loomed on the horizon but didn’t immediately arrive – or how aging experts in COBOL, a coding language now rarely taught, can be highly paid to repair old devices still running COBOL because younger workers didn’t enter the field.

“Automation often takes much longer than a casual observer would expect,” Gross says. “And there are a number of reasons for that. But that’s a pattern, and I think it’s one that, to me, resonates as much today as ever.”

AI Disruption: Decades? Or Mere Years?

However, some suggest today’s changes could come faster than earlier ones, perhaps requiring faster intervention and retraining.

“The transition from a predominantly agrarian economy to an industrial one played out over decades,” says Leopold of the World Economic Forum, “whereas timeframes now are significantly more time-compressed, creating a much more immediate and less-gradual adaptation challenge, which will require a concerted effort from all parts of economy and society to manage.”

But while one study found technological disruptions have indeed affected work faster more recently, such cycles still tend to be measured in decades before full deployment of the transformative technology, rather than months or years.

A longer-than-expected transition could happen this time, says Humlum of the University of Chicago.

“In terms of thinking about productivity and the labor market and the broader economy, there’s several additional adjustments that need to take place,” he says. “Firms need to reorganize how they structure work in response to these technologies, and those adjustments take time.”

How Disruption Can Create New Jobs

Regardless of whether the coming change will be slow or rapid, history suggests there might be more plentiful jobs at the end.

In a 2018 podcast discussion available online, Richard Cooper, then an economics professor at Harvard University, and Susan Lund, at the time a McKinsey Global Institute partner who had just co-authored a McKinsey report on disruptive workplace technologies, use the example of the automobile assembly line to show how that might work. In that case, fewer workers could produce more cars – but instead of decreasing employment, the changes drove jobs.

That’s because the increased productivity allowed the price of cars to fall dramatically and drove up demand. That, in turn, required more workers to work the assembly lines and fueled whole new industries – filling stations, mechanics, tourism, and more – to support a new automotive culture.

Eventually, cars allowed longer commutes to work, bringing suburban sprawl. And they allowed people to take up leisure activities, like skiing, previously inaccessible mountain areas. Both trends drove even more jobs to serve the new realities, Cooper notes.

Of course, the assembly line also caused horse- and carriage-based industries to decline over several decades. That cost some jobs, but far fewer than were gained.

“That’s a central problem with public discussion of this issue,” Cooper says in the podcast. “The jobs that are lost are tangible. There are people in them. We can sympathize with the people who lose their jobs and so forth. There’s a human dimension. The jobs that are created or made possible have no one in them, and, in fact, we don’t even know exactly where they’re going to be until, you know, time goes on.”

New Jobs Could Be Worse

But Resnikoff – an assistant professor of contemporary history at the University of Groningen in the Netherlands and a former organizer for the United Autoworkers Union – says automation also can degrade work and lead to less-fulfilling replacement jobs.

In the case of the assembly line, Resnikoff says, “that’s a highly skilled job up until, oh, 1912, 1914, when Henry Ford brings in the assembly line. Then … you can bring in semi-skilled labor to do what before you would’ve had to be an extremely skilled mechanic to achieve.”

Some already see AI driving such degradation in computer coding jobs. The New York Times recently reported that some workers feel they must work faster to support AI on tasks they used to do themselves.

From ‘Mechanization’ to ‘Automation’

Resnikoff argues job degradation may result from disruptive technology being directed at workers by bosses. He cites the evolution after World War II from “mechanization,” a “more neutral term” for applying technology to work, to “automation.”

Before that, he argues, captains of industry, “didn’t say that every time they brought in a new machine it meant less work. And it wouldn’t have been credible if they did because, especially beginning in the late 19th century, the Industrial Revolution is producing more work, not less.

“All of these immigrants who are coming off the boat at Ellis Island are being brought into factories. It’s a booming industrial economy. It requires a great deal of labor. And they certainly weren’t saying it was imminent that (new production machinery) was about to get rid of all the work.”

On the other hand, the term automation “was invented by Ford and the creation of their automation department in 1947 (as) … a way basically to fight their newly unionized workers on the shop floor,” he says. “This allowed Ford Motor Company to attack the job classifications of workers on the shop floor and to claim that they, themselves, weren’t doing it … (that) this is merely the result of technological progress.”

De-Skilling Labor?

He says Ford’s promotion of post-war “technological optimism,” a sense that technology would elevate workers and society beyond manual labor, soon became, and remains, the consensus mindset. And he sees it continuing.

“Many of the same functions that automation had as a way of telling a story about technology at work, that storytelling work is now being done by the term ‘AI,’” he says.

“I think that’s what a lot of the fear is about, that now … white-collar people seem more at risk,” he says. “On the other hand, what isn’t new is that this is exactly how industrial capitalism has been operating for 200 years. And people are, every generation, they’re surprised anew when employers find ways to use machines to de-skill or, let’s just say, change skilled labor. … That’s why it’s both old and new. It’s the same process, but it’s happening in a new place or a new sector.

“Whether it’s a programmer or some screenwriter in LA, these were jobs that were thought to be, you couldn’t get a machine to do them,” he says. “And now you can … (get AI to) do part of the job, if not the whole job.”

How is this a cycle that we’ve seen before?

“There would’ve been a time when you would’ve said making steel can’t be mechanized,” Resnikoff says. “It was also a very highly skilled job. But then, in fact, it was achieved – and it made some (new) unskilled jobs and it made some skilled jobs.”

Change Jolts Telephone Operators, Especially Older Workers

At least there were jobs at the end of it. Not all workers are so lucky.

Gross has studied the displacement of telephone operators over the first half of the 20^th century. Back then, callers would pick up their phone and tell an operator who they were calling. The operator, often female, would then plug a cable into a jack to connect them.

But phones started coming with dials so callers could direct their own calls. Over decades, the phone company installed local automated switchboards that could do the rest.

“There was considerable concern over what this would do, particularly to the young women who were affected,” Gross says. “This even rises to the level of congressional hearings that are more generally on automation and economic conditions, but where the mechanization of the telephone system was a point of focus.”

What happened next wasn’t good for displaced operators, particularly the older ones.

“When we observed them in the census a decade later … we find them being less likely to be employed and, conditional on being employed, more likely to be in a lower-paying occupation,” Gross says. “These effects are particularly strong for older operators as opposed to younger existing operators.

“What we think that this points to is a bit of an age gradient,” he says. “Older workers have a more difficult time adjusting to new technology and transitioning to other jobs, other sectors. And younger workers are more fluid, up to the point where those who are not yet of working age at the time that automation takes place are essentially perfectly fluid. They can prepare for careers in different sectors from the start.”

‘Dislocation Is Very Painful’

So disruption was not so good for the telephone operators. And perhaps it could derail those directly in AI’s path who don’t revise their skillsets.

“Dislocation is very painful,” says Madgavkar, the McKinsey partner. “There is, I think, no question about that.”

But for the rest of us?

“In general, over long periods of time, the quality of work people have been doing has improved,” Madgavkar says. “I do think there is less drudgery in work of all kinds today.”

Adds Gross: “It’s easy to take a view that automation is destructive to work and to workers and to therefore be reluctant to embrace it. But if we inhabit that view over the past century, there are so many things that we wouldn’t even be able to have today.”

Managing Change

He says change can be managed so it is less painful, and some commentators suggest more should be done now to mitigate the effects of AI on workers and jobs.

Gross cites the example of containerized shipping and port automation in the 1960s, when some workers gained compensation for being displaced and others got new jobs operating the automation. The changes also brought benefits to consumers and importers by making goods cheaper.

According to Resnikoff, “It’s fair to say (that) if employers are allowed to use technology however they please, and we are unable to have any rules either through the government or through collective bargaining to stop them, the history is very clear on that: Employers usually want control and they want more for less. And so there will still be jobs, but they won’t be very good.

“There’s also evidence from the past that when the state does get involved on behalf of workers to protect them, and when there are robust unions, that workers can, in fact, safeguard for themselves at least certain kinds of protections,” he says. “And this will make the job not-so-terrible, perhaps, and make the job better.”

Throughout recent history, an ever-evolving labor market has brought jobs that could not have been imagined by prior generations. Some may have arisen from technological disruptions. Others, as Resnikoff argues may be more common, could have been a natural evolution driven by independent social factors, market demand, and changing trends.

Whatever the reason, Gross says, “We’re still here, we’re still working, we are still feeding our families.

“If anything,” he adds, “on average and for most of the population, our quality of life has improved. We’re richer than we used to be. We live longer than we used to. We’re safer than we used to be. And ultimately, my perspective is to see technology as the root of progress and something that we ought to continue supporting.”