Today, in 2017, the president’s top economic advisor said he had no worries about robots putting people out of work. “In terms of artificial intelligence taking over the jobs, I think we’re so far away from that that it’s not even on my radar screen,” Treasury Secretary Steve Mnuchin told an audience in Washington. “I think it’s 50 or 100 more years.”
Great! That’s a relief! President Trump can go back to horsing around on his big rig confident in the knowledge self-driving trucks won’t replace millions of drivers in a few years.
Except Mnuchin’s wrong. Like super-wrong. Artificial intelligence is not only coming for jobs, the jobs it’s coming for are the precious few left over after old-school automation already came for so many others. Technologists and economists know this. People who’ve lost jobs to robots and computers already know this. The only people who apparently don’t know it are in the White House.
“Mnuchin’s saying newfangled computers aren’t going to have any kind of big effect on the economy for the next 50 or 100 years,” says Andrew McAfee, an expert on business and technological change who co-founded MIT’s Initiative on the Digital Economy. “I don’t talk to anyone in the field who believes that.”
In a room at MIT overlooking the Charles River a few weeks ago, McAfee polled 140 of these experts in artificial intelligence on automation and employment. He asked the attendees the same question Mnuchin got this morning: When will the robots take all the jobs? Which was really another way of asking: How worried are you?
Turns out the issue is a flashing red alert on their radars. The conference included engineers and scientists, industry types from the likes of Toyota and IBM, and policymakers recently exiled from Washington by the 2016 election. They concluded, among other things, that by 2032, half the trucks on the road would not have human drivers. At the trucking industry’s current size, that transition to automation equals 1.75 million lost driver jobs over the next 15 years.
And that’s just one sector. The room was even more convinced that automation would soon sideline humans who interpret medical records for a living. The experts estimated that by 2026, most of those jobs would be done by machines. They guessed that robots would handle 95 percent of air traffic control jobs by 2028. Most factories in the US would have fewer than 20 humans working in them by 2034. Just fewer than half said they expect robots will perform most surgeries by 2036, and that Fortune 500 companies will have more robots performing managerial tasks than humans by 2034.
McAfee’s own research shows that automation is already hollowing out the middle class. This phenomenon has disproportionately hit men, since traditionally male manual labor is easier to automate than traditionally female service and care jobs. “The American middle class was built on the back of routine work. Lots of those jobs have already been automated,” McAfee says. Now, AI could come for the rest. Artificial intelligence is really, really good at the kinds of jobs that involve recognizing and matching patterns—in other words, what doctors and accountants do. Hell, a robot can even write pretty good news articles.
“Driverless cars to drive blind people around? This is amazing, science fiction stuff,” McAfee says. But policymakers and technologists need to think carefully about how to do it right. Over the course of the day at MIT, the conversation ping-ponged from bleak to impassioned. But the theme stayed consistent: Participants wanted to figure out how to ensure automation helps rather than hurts people. In other words, the exact kind of discussion Mnuchin dismissed this morning as unnecessary.
“When you are outside of Washington, this is often the most significant issue, but it’s not back in DC,” says Gene Sperling, former chief economic advisor in both the Obama and Clinton administrations.
As McAfee notes, the best players of Go and no-limit Texas Hold `Em poker are pieces of technology. Will it really be another 50 years before that same technology unleashes big changes in the global economy? Even if those changes are good for the economy in the long run, they will also hurt. And that pain should probably be on Mnuchin’s radar.