The future of work will be turned upside down by AI, a new study from the National Academies of Sciences, Engineering and Medicine says. Middle jobs -- those that exist between low- and high-end work -- will go away.
Jobs at risk include entry-level developers, IT support and any role that involves routine and repetitive tasks, low decision-making complexity and easily structured workflows. As this happens, the future of work will be reshaped by a widening divide between high-skill and low-skill jobs, affecting every part of the economy, according to the study.
High-skill jobs, especially those requiring problem-solving and human judgment, are less likely to be entirely replaced.
"This may be one of the biggest transformations of the economy that we've seen," said Erik Brynjolfsson, director of the Stanford Digital Economy Lab and co-chair of the effort behind the National Academies' report "Artificial Intelligence and the Future of Work."
Uneven impacts
"There's no economic law that this will automatically benefit all people," Brynjolfsson said during a webinar Thursday coinciding with the report's release. He said there might be outcomes such as job losses, wage disparities, increased inequality and uneven distribution of benefits.
There's no economic law that this will automatically benefit all people.
Erik BrynjolfssonDirector, Stanford Digital Economy Lab
For instance, software engineering only exists because of computers, but computerization upended employment for clerical workers, Brynjolfsson said. However, AI is a different technology.
One possibility is it's "so powerful that it just out-competes humans right across the board," but that's unlikely in the near future, he said.
The report also focuses on how AI could change the future of work in troubling ways, where AI tools are deployed to monitor the tone of call center workers for cheerfulness, track the computer usage of white-collar employees and enforce strict output metrics across all types of jobs.
The report noted that AI tools might be able to monitor how often white-collar workers shop online or check their email during the workday.
Education will also have to change, including instruction in secondary schools.
At one time in elementary school, it was necessary to memorize facts, said Tom Mitchell, a computer science professor at Carnegie Mellon University and a co-chair of the AI workforce committee. Now that facts are quickly available on the internet, "what's relatively more important is how to reason with those facts, how to evaluate competing versions of the fact," he said during the webinar.
Patrick Thibodeau is an editor at large for TechTarget Editorial who covers HCM and ERP technologies. He's worked for more than two decades as an enterprise IT reporter.
