As much as one-third of the United States workforce could be out of a job by 2030 thanks to automation, according to new research from McKinsey. The consulting firm now estimates that between 400 million and 800 million individuals globally could be displaced by automation and need to find new work.

“Even if there is enough work to ensure full employment by 2030, major transitions lie ahead that could match or even exceed the scale of historical shifts out of agriculture and manufacturing,” according to a report by the McKinsey Global Institute published this month. “Even as it causes declines in some occupations, automation will change many more – 60 percent of occupations have at least 30 percent of constituent work activities that could be automated.”

Professions most susceptible to automation include physical ones in “predictable environments,” including operating machinery and preparing fast food, according to the research. Alternatively, automation will have a more muted effect on jobs that involve managing people, expertise, and those that require frequent social interactions.

“Unpredictable” jobs such as gardeners, plumbers, or providers of child- and elder-care are also less likely to see automation over the next decade as they remain challenging to automate and don’t usually earn high wages, according to McKinsey.

This selective effect on the workforce has many worried that income inequality could continue to worsen in the United States.

“Income polarization could continue in the United States and other advanced economies,” noted the research. “If reemployment is slow, frictional unemployment will likely rise in the short-term and wages could face downward pressure.”

To be sure, McKinsey isn’t the only group studying the effects of artificial intelligence on wages and economic growth.

A recent working paper penned by economists from Northwestern, Stanford, and the College de France explored what would happen to economic growth if artificial intelligence starts generating original thought. A rapid uptick in the rate of innovation – and in new ideas – have led some to speculate about economic hypergrowth and ever-increasing GDP gains.

But to help transition to a future with increased automation, businesses and policymakers will “need to act” to keep people employed, suggests the McKinsey research.

McKinsey notes that governments will have to develop and provide extensive job retraining to help displaced workers as well as providing more generous income supplements.

“Beyond retraining, a range of policies can help, including unemployment insurance, public assistance in finding work, and portable benefits that follow workers between jobs,” the researchers wrote. “Possible solutions to supplement incomes, such as more comprehensive minimum wage policies, universal basic income, or wage gains tied to productivity.”

The research cited the U.S. High School Movement at the turn of the last century and the GI Bill as key examples of how developed countries can cope with the disruptive effects of a transforming economy.

“In many decades hence, the value of this labor may be diminished if we reach a state in which machines can do a large share of the work,” concluded the report. “For workers around the world, policy makers, and business leaders—and not just social scientists who specialize in socio-economic paradigms—that should give pause for thought, and be a spur for action.”