When it arrived in December, the White House’s report, “Artificial Intelligence, Automation, and the Economy” felt like both a parting shot and warning to the incoming administration. That it was perfectly dovetailed with heated election rhetoric around domestic job loss was simply a bit of serendipity, however.
The subject matter had been at the forefront of the paper’s authors for some time. In fact, it was a sequel of sorts to study released by the White House the month prior to the election. AI and automation were topics co-author Jason Furman says the Obama White House discussed regularly, between meetings about autonomous vehicles and factory safety.
“It’s one that showed up in a lot of different places in our policy making,” Furman, now a Senior Fellow at the Peterson Institute for International Economics told TechCrunch in an interview this week. “We would have a meeting about where the jobs were going. AI was popping up in so many separate policy processes and issues that the White House chose to do a pair of reports, the first focusing on the issue and the second focused on the economic aspects of it.”
Between campaign trail promises and post-election photo ops, the thread seems to have been lost on the subject. Economic discussions in recent months seem focused specifically on trade and immigration – longtime talking points and frequent scapegoats that are far easier to distill on the stump than cutting edge technologies.
In March, Treasury Secretary Steve Mnuchin even went so far as brushing off such questions by characterizing questions about AI-driven job loss as if it were something out of some dystopian sci-fi novel, telling Axios that it wasn’t “even on my radar screen. Far enough that it’s 50 or 100 more years.”
Now, the January report wasn’t exactly a call to take arms against Skynet, but it does address some important truths about automation and AI that have already begun to have a very real impact on both the economy and domestic jobs. For all the growth potential these technologies offer to the U.S. economy, it’s an unavoidable fact that there has been and likely will continue to be short-term gross domestic job loss.
Though Furman says he believes that, as with other economic factors, technology too often serves as a scapegoat.
“I think automation is too easy an out and lets policy makers off the hook,” he explains. “France has a much higher fraction of prime age workers in its work force than the US does. That’s not because France has less automation. It’s because they have labor market institutions that, while still very problematic, do a better job of helping people find work than they do here in the United States.”
Those who are particularly bullish about the future of automation point to past technology breakthroughs like the industrial revolution, which didn’t so much kill jobs as shift the economy away from things like agriculture. What that means, ultimately, is that short term job loss can give way to long term job growth, as new and potentially better jobs are created. And certainly one thing automation and robots have the potential to do is help eliminate what the industry has handily deemed the three Ds – dull, dirty and dangerous jobs.
But among these polarizing conversations around the growth of technology in the industrial setting, an important point often gets lost: if left unchecked, it can serve to expand the already vast economic gulf between what have been deemed “skilled” and “unskilled” workforces. MIT economist David Autor summed the fear up well in a conversation we had earlier this year.
“The labor market for college-educated workers is very, very strong,” he told me. “And those people continue to get paid better. The set available to people who just have a college education or less has dramatically contracted as a function as well as trade, but automation has been the bigger fact. It’s an important part of the growth of inequality we’ve seen over the past several years, the decline of earnings and fortunes of people without a college degree.”
The solution, it turns out, could be simple. Investments in education and job training could help stem the bleeding and bridge the gulf.
“More education would absolutely help,” says Furman. “I think more of what economists call active labor market policies – training, job search assistance and subsidies for jobs. Not all of those programs work, but they’ve gotten a worse reputation than they deserve. We basically don’t try very hard. We spend 0.2-percent of our GDP helping people find jobs and prepare them for jobs. That’s lower than any other OECD country except Mexico and Chile.”
Of course, all that requires financial investments which may not prove a popular solution in a political climate more invested in building walls and hampering trade with other countries. But AI’s impact on jobs is not a conversation to put off for 50 or 100 years in the future.
“I am not sure whether what we want is hard skills like STEM or soft skills like judgment and figuring out how to be nice to the person next to you, which robots aren’t nearly as good at,” says Furman. “My guess is it’s some combination of both. It would be good to get the answer exactly right, but I think getting the answer even partly right would be a good improvement.”