- The AI boom could change or eliminate millions of American jobs.
- Experts say many of those workers will need to be retrained for new jobs to avoid being left behind.
The US economy will face a big test in the years ahead — one it has struggled with in the past: helping workers adjust when jobs change or disappear.
Emerging generative AI technologies like ChatGPT could eliminate or change the nature of millions of jobs over the next decade. Many knowledge-worker occupations could be impacted, including those in marketing and sales, software engineering, research and development, accounting, financial advising, and writing roles.
Even workers in the jobs that remain could eventually become vulnerable. That's because AI could change their work tasks — automating some, tweaking others, and creating entirely new ones. While this landscape could bring opportunities for workers who learn to use new technologies, those who struggle to adapt could be replaced not by AI — but by fellow humans who know how to use it.
"AI won't take your job. It's somebody using AI that will take your job," economist Richard Baldwin said earlier this year.
Both scenarios point to a growing need for retraining — or reskilling — workers impacted by the coming AI boom. This includes training displaced workers for jobs less impacted by AI and giving others the skills they need to work in one of millions of new jobs that could be created due to these technologies. It also means helping workers develop AI skills in their current roles so they don't get left behind.
"Over the long term, labor disruptions could herald an unprecedented demand for reskilling displaced workers and require a significant increase in capacity for retraining," Seth Carpenter, Global Chief Economist for Morgan Stanley, said in a note in early November. Retraining and government safety net programs like unemployment insurance he added, are more likely to help displaced workers than policies like a universal basic income — which some have posed as a solution to AI job displacement.
But even if retraining is the solution, executing this transition is easier said than done. The US has seen job displacement before, and the aftermath hasn't always been pretty.
Retraining workers in certain age groups and industries can help the US avoid the mistakes of the past
In the early 2000s, expanded global trade led to cheaper products for American consumers. But this contributed to the loss of as many as five million US manufacturing jobs and the subsequent economic hardships faced by many communities — particularly smaller cities and towns in the Midwest.
When Donald Trump promised to bring back manufacturing jobs before the 2016 election, he was speaking to the Americans who had been left in the lurch.
"The mismanaged integration of the United States into the global economy has devastated US manufacturing workers and their communities," the Economic Policy Institute wrote last year, referring to the wave of globalization between 1998 and 2021.
Efforts to bring supply chains closer to home have led to a manufacturing uptick in the US in recent years. But many overseas jobs aren't likely to return anytime soon, among the reasons job retraining was — and remains — necessary for impacted workers.
If and when AI poses a new disruption to many Americans' jobs, avoiding the mistakes of the past will require significant retraining investments by US companies, schools, and governments, Michael Chui, a partner at the McKinsey Global Institute who has researched the impact of new technologies on businesses, previously told Insider.
In the "glass-half-full" scenario, Chui said, retraining could help many workers transition to higher-paying jobs. While the US has far from a perfect track record in navigating economic disruptions, he said he had confidence the US could make the transition.
"There was a time when a vast majority of the workers in the United States were involved in agriculture," he said. "And then years later, far more people were working in factories, for instance. But we didn't end up with 50% unemployment. We have historically been able to do this."
Young workers, who are more likely to have experimented with technologies like ChatGPT, may have an easier time adapting to the AI future — but extra retraining resources may be needed for older workers. When telephone operators, for instance — then a common profession for women — were automated in the late 19th century, older workers struggled most, Ethan Mollick, an associate professor of entrepreneurship and innovation at the University of Pennsylvania's Wharton School, told Insider.
"When you got rid of operators, then basically young women were able to adjust, find new jobs and were able to adapt," he said. "But older women took a lifelong hit to wages — were never able to find as good a job again."
Companies, schools, and governments have a role to play in retraining
On a practical level, retraining can happen in a few different ways: at colleges and universities, through independent programs, or within companies themselves. But it's not yet clear who will be — and should be — responsible for executing the needed training.
Some companies have already taken steps to ensure their workers keep up with AI.
The accounting and consulting giant PricewaterhouseCoopers, for instance, has plans to provide AI training for 75,000 workers in the US and Mexico. The training is designed to teach employees about new technologies and how they can safely incorporate them into their work. In an IBM survey of 300 global corporate executives conducted in May, the executives estimated that 40% of their respective workforces would need reskilling over the next three years due to the implementation of AI and automation.
It's not just about helping workers keep their jobs. Given AI's potential to help businesses increase productivity and boost profits, there's arguably plenty of reason for them to invest in AI training.
Some college professors, like UPenn's Mollick, have encouraged their students to use AI like ChatGPT in some of their studies. They say knowing how to use these tools could someday be important in the workforce. Schools and EdTech platforms have also begun offering more AI-focused courses.
Government investments are another piece of the puzzle. In addition to directing funding toward retraining programs, Congress could also adopt tax policies that encourage companies to retrain their workers — rather than replace them.
"It is true that some workers will need support and retraining as we make this transition into an AI-powered workplace," Bill Gates wrote in a July blog post. "That's a role for governments and businesses, and they'll need to manage it well so that workers aren't left behind."