Paul van Gerven

22 November

The robots aren’t stealing our jobs yet, but when they do, it’s because we let them.

“It’s not obvious at all that we’ll return to the low unemployment we had before the crisis. Due to the rapid rise of robots and other technology, the future could look drastically different from the present,” Dutch Minister of Social Affairs and Employment Lodewijk Asscher warned at a conference in 2014. “Robots are fast, never sick and work 24 hours a day. They never ask for a raise, aren’t represented by unions and don’t go on strike. They’re therefore able to replace human employees in a significant part of jobs.”

Seven years later, Asscher’s warning seems a bit silly. If the robots were stealing our jobs, you’d expect them to be harder to come by. The opposite is true: unemployment is at a historical low and many companies are facing a labor shortage. And if automation was wreaking havoc on our labor markets, workforce productivity would be on the rise. Instead, after decades of steady increases, it has hardly changed since Asscher’s speech.

This doesn’t mean the ‘robocalypse’ couldn’t still happen, of course. Certainly concerns over technological unemployment haven’t faded. “Technology will take over some of our jobs. We’ll work less and have to find a meaningful use for that free time,” Eindhoven University of Technology professor Maarten Steinbuch recently echoed John Maynard Keynes, the pre-WWII economist who famously predicted that by now the main problem of industrialized countries would be boredom. Silicon Valley hotshots like Elon Musk, Mark Zuckerberg and Bill Gates have sounded the alarm as well.

Prophecy

The question, then, is: why would it be different this time? Ever since the fears of the Luddites, who smashed up textile machinery in the early 19th century, were proven wrong, the generally accepted view among economists has been that innovation – on a macro level – doesn’t harm employment: technological advances also create new tasks that need to be performed. After the big wave of automation in the latter two decades of the previous century didn’t have devastating consequences, the issue seemed to have been put to bed indefinitely.

But from 2010 onward, faced with a new impending wave of digitalization, academia once again embraced the topic. In 2013, for example, Carl Benedikt Frey and Michael Osborne of Oxford University predicted that up to 47 percent of US jobs – blue and white collar – could be fully automated within the next decade or two (interestingly, the researchers used a machine-learning algorithm to make that assessment). The next year, the Economist painted a pretty bleak picture of the future. “No government is prepared,” the weekly wrote. Policymakers, including Mr. Asscher, were alarmed.

Things seem to have calmed down again, perhaps because some recent eye-catching technology-related predictions have crashed and burned spectacularly. For example, Elon Musk whipped up excitement for self-driving taxis far too early; at one point, he promised such cars would be operating on the road by 2020. The internet of things, although perhaps not a threat to employment, so far hasn’t lived up to the hype either. Controlling coffee machines with an app and using smartphones to see who’s at the door isn’t all that revolutionary. People have become a bit wary of technological prophecy.

According to the influential analysis of MIT scientists Erik Brynjolfsson and Andrew McAfee, however, the storm is still coming. In “The second machine age” (2014), they argue that digital innovation – like steam power and electrification – is a general-purpose technology with broad applications across economies. That we don’t see much of it yet in economic or labor market statistics is because its adoption isn’t merely a matter of getting a few systems up and running. It requires a major overhaul of business processes, training of workers and other adaptations. As previously seen in the electrification of factories originally designed for steam power: these things take time.

Vigilant

So, perhaps this time, it will be different. In any case, we need to keep a close eye on the development of AI and automation. Robots may not be taking our jobs yet, but we’re already seeing examples of digital technologies being deployed in undesirable ways. Amazon warehouse workers are essentially managed by computers, enforcing a grueling pace. Scoreboards keep track of each worker’s productivity. The fastest get rewarded, the slowest fired – sometimes automatically.

The direction in which AI and automation develop isn’t preordained. Rather than replace humans, they can be deployed to create jobs and enhance human productivity. And as a society, we can make sure everyone shares in the prosperity they create. Let’s be vigilant.