To start, generative A.I. will accelerate the path to automation. By 2030, we estimate activities that account for 30% of U.S. working hours could be automated—up from 21% before generative A.I. burst upon the scene. It could also affect a broader range of activities, including data analytics, product design, legal analysis, and research and development.
The implications of generative A.I. are complex. What is clear is that generative A.I. will fundamentally change the way many jobs are done. And we are optimistic that many of the jobs created will be highly skilled and well paid. To get there, though, the United States must invest in re-training and education to ensure that the workforce is prepared to succeed.
On the lower end of the job market—those making less than $38,200 (£32,000) a year — automation and other structural changes have already had big effects. Generative A.I. could accelerate these trends, resulting in lower wage workers being 14 times (???) more likely to need to shift occupations than high-wage workers. People without college degrees are almost twice as likely to face displacement.
But U.S. workers are resilient, as the 8.6 million pandemic-era occupational shifts proved. While it is impossible to track individual moves, many landed better-paying jobs in other fields. But we cannot simply assume resilience will continue. To adapt, more and better workforce development will be needed.
As for higher-wage jobs, generative A.I. is likely to change work activities substantially, particularly in healthcare, STEM, and professional services. In effect, it will change how these workers allocate their time, and could make these jobs more interesting. Drug researchers won’t have to do endless pre-screening of chemicals; lawyers will spend less time looking up cases; managers can pass off paperwork and instead concentrate on coaching and making improvements. These occupations will be changed by generative A.I., and all are likely to see job growth between now and 2030.
The biggest potential economic benefit of generative A.I. is that it could increase productivity significantly. Since 2005, U.S. labor productivity has grown an average of 1.4% a year. Raising that to the postwar average of 2.2% could add up to $10 trillion to U.S. GDP by 2030. Generative A.I., if coupled with the effective redeployment of the hours it saves, could increase U.S. labor productivity by 0.5 to 0.9 percentage points a year. Combined with all other automation technologies, the increase could be up to as much as 3% to 4% annual GDP growth.
For example, in manufacturing, roughly 36% of working hours could be affected by automation. The sector, however, is now short almost 700,000 workers. There will likely be fewer assemblers and machine operators, and more industrial engineers and software developers. In short, generative A.I. brings enormous potential for U.S. manufacturing in terms of higher productivity and better-paid jobs. However, to see the benefits, the sector must develop and attract a workforce with a broader set of skills.
That is the exact challenge facing the U.S. economy broadly: Workers must adapt and rise to the challenge of acquiring new skills, as so many did during the pandemic. Employers must step up too. As jobs change and labor shortages linger, companies can invest in training current and potential employees, for example through learn-as-you-earn programs. Companies can also recruit from overlooked populations, like retirees (the ‘Silver Army’ I keep banging on about), rural workers, and people with disabilities, and hire for skills and potential rather than degrees or experience. We’re applying some of these principles at McKinsey; we’ve doubled the number of schools where we recruit and are hiring more via programs like apprenticeships and coding boot camps.
If worker transitions and risks are well managed, generative A.I., combined with other automation technologies, could boost productivity and foster better jobs, contributing to genuinely sustainable and inclusive growth. This optimistic scenario is plausible—but far from guaranteed.
Almost 12 million occupational changes will need to take place between now and 2030, with over 80% of those jobs falling into four occupations: customer service, food service, production or manufacturing, and office support. Most of these workers are lower paid, and disproportionately composed of less educated workers, women, and Black and Latino Americans. The positive potential is that they move into better paid, more interesting work. But that will only happen if they have the chance to re-skill and adapt.
We do not pretend to be able to predict the future, but we don’t have to: We are in it. Generative A.I. is here to stay. The challenge is to make the best of it.