The robots are coming, but the march of automation will displace jobs more gradually than some alarming forecasts suggest.
A measured pace is likely because what is technically possible is only one factor in determining how quickly new technology is adopted, according to a new study by the McKinsey Global Institute. Other crucial ingredients include economics, labor markets, regulations and social attitudes.
The report, which was released Thursday, breaks jobs down by work tasks — more than 2,000 activities across 800 occupations, from stock clerk to company boss. The institute, the research arm of the consulting firm McKinsey & Company, concludes that many tasks can be automated and that most jobs have activities ripe for automation. But the near-term impact, the report says, will be to transform work more than to eliminate jobs.
Globally, the McKinsey researchers calculated that 49 percent of time spent on work activities could be automated with "currently demonstrated technology" either already in the marketplace or being developed in labs. That, the report says, translates into $15.8 trillion in wages and the equivalent of 1.1 billion workers worldwide. But only 5 percent of jobs can be entirely automated.
"This is going to take decades," said James Manyika, a director of the institute and an author of the report. "How automation affects employment will not be decided simply by what is technically feasible, which is what technologists tend to focus on."
The report, a product of years of research by the McKinsey group, adds to the growing body of research on automation and jobs.
Conclusions about the relationship between the two vary widely. Examining trends in artificial intelligence, Carl Benedikt Frey and Michael A. Osborne, researchers at Oxford University, estimated in a widely cited paper published in 2013 that 47 percent of jobs in the United States were at risk from automation.
By contrast, a report published last year by the Organization for Economic Cooperation and Development concluded that across its 21-member countries, 9 percent of jobs could be automated on average.
Differing assumptions, and sheer uncertainty about the future, explain the conflicting outlooks.
Throughout history, times of rapid technological progress have stoked fears of jobs losses. More than 80 years ago, the renowned English economist John Maynard Keynes warned of a "new disease" of "technological unemployment."
Today, it is the rise of artificial intelligence in increasingly clever software and machines that is stirring concern. The standard view is that routine work in factories and offices, like bookkeeping or operating basic machinery, is most vulnerable to automation.
But A.I. software that can read and analyze text or speech — so-called natural language processing — is encroaching on the work of professionals. For example, there is a lot of legal work that is routine, said Frank Levy, a labor economist at the Massachusetts Institute of Technology. But that routine work, sifting through documents for relevant information, is wrapped in language, which had protected lawyers from the effects of automation. But no longer.
"Natural language processing opens the door to doing more and more work that was beyond automation until recently," Mr. Levy said.
The McKinsey report cites natural language processing as a key technology: The faster it develops, the more that tasks can be automated; a slower pace means less automation.
The economic cost of automation is another concern. People see advances in self-driving vehicles, and think that the jobs of America's 1.7 million truck drivers are in imminent peril, said Michael Chui, a partner at the McKinsey institute and an author of the report.
Yet replacing America's truck fleet would require a trillion-dollar investment, Mr. Chui said, adding "if you could buy a self-driving truck, which you can't."
Such uncertainties led the McKinsey researchers to calculate the pace of automation as ranges rather than precise predictions. The report's multifactor scenarios suggest that half of today's work activities could be automated by 2055. That threshold could be reached 20 years earlier or 20 years later, the report adds, depending on economic trends, labor market dynamics, regulations and social attitudes.
So while further automation is inevitable, McKinsey's research suggests that it will be a relentless advance rather than an economic tidal wave.
"We have more time than we think to adjust to the world that technology makes possible," said Matthew J. Slaughter, an economist and dean of the Tuck School of Business at Dartmouth College.
The McKinsey report also considered the other side of automation — as an engine of productivity and economic growth.
In that context, the report points to the demographic challenges ahead, as most nations age, and the need for automation-powered growth. Today, there are 46 million Americans over 65, or 15 percent of the population. By 2060, the size of the over-65 group is projected to reach 98 million people, or 24 percent of the population.
The American economy may need all the A.I.-assisted automation it can muster.
"If productivity growth continues to be anemic, we're in trouble," said Hal Varian, Google's chief economist and an emeritus professor at the University of California, Berkeley.