Skip to content
Breaking News Alert It Could Soon Be Illegal For California Teachers To Tell Parents About Kids' Trans Confusion

5 Things Matt Yglesias Gets Wrong About Automation


A running theme of AMC’s “Humans” is the threat—both economic and existential—that robots pose for humanity. One of the show’s characters, Matilda Hawkins, a highly intelligent and technologically savvy teen-age girl, is beset with apathy because she believes that no matter what career she chooses, she will not be able to stave off professional obsolescence. In the show, robots are proving they can do pretty much everything humans can do, so why should Matilda care about the future?

Vox’s Matthew Yglesias has a new article calling the prospect of encroaching automation a myth. Bizarrely, he says that, rather than receive this news with relief, we should fear it. That’s just one of a number of similarly silly remarks Yglesias makes throughout the piece. I’ll focus on five.

1. ‘None of the recent problems in the American economy are due to robots.’

This is patently false. In “Average Is Over,” Tyler Cowen cites “the increasing productivity of intelligent machines” as a partial explanation for the labor market troubles young people experience today. What is more, the low-wages, persistent unemployment, and other economic maladies currently plaguing the young are all, according to Cowen, a “harbinger of the new world of work to come.”

While Cowen’s quote draws attention to the future as particularly worrisome, it also signals the problems automation poses for us today. Perhaps Yglesias is thinking that we cannot blame robots for the Great Recession. Although that is true, it doesn’t mean future economic disasters won’t be caused by automation. Neither does it mean we should ignore the real ways in which automation is having a disruptive effect on our economy today.

2. ‘Even if the pace of automation does speed up in the future, there’s no real reason to believe that it will be a problem.’

Yglesias suggests that if productivity goes up due to the acceleration of automation, we will have wealth without toil. That’s certainly a good, right? We can pursue our interests rather than waste our lives away at jobs we hate. But what if the wealth created fails to be distributed in the way he suggests? What if inequality actually widens? He offers no argument for dismissing distributional failure. After all, there’s no presumption that the new controllers of automated capital will accept an Yglesian utopia.

3. ‘The robots aren’t taking our jobs; they’re taking our leisure.’

Yglesias writes that automation is increasing our quality of entertainment, not our productivity. This has a psychological effect, too, which is to make us think automation is affecting the economy more than it actually is. Due to its ubiquity across our modes of entertainment, we get the sense that automation is on a trampling march toward conquering every other area of the economy. He cites Robert Solow’s classic assessment: “You can see the computer age everywhere but in the productivity statistics.”

But perhaps that’s because innovation isn’t always immediately revolutionary. The economist Chad Syverson saw the same sort of delayed effect in the way electricity impacted the industrial economy in the 1890s. It then took 20 years before labor productivity started to rise. This lag can be attributed to the natural gap that exists between the advent of an industry-busting technology and the discovery of the most efficient way to implement it.

4. ‘[I]t becomes clearer…over time that smartphones and the internet simply aren’t economic game changers on the same scale as air conditioning, jet planes, container ships, and televisions.’

This is at best remarkably short-sighted, and at worst hopelessly dense. In the very same article, Yglesias offers a list of ways in which his smartphone has displaced items that previously were entire industries unto themselves. The device in his pocket is a phone, music player, personal library, calculator, video-game console, alarm clock, and more. The smartphone is the epitome of an economic game changer.

Perhaps Yglesias’ reasoning is that going from no plane to a jet plane is a massive shift, but going from a clunky housephone to a small smartphone is not so drastic a change.

In their work, “The Second Machine Age,” Erik Brynjolfsson and Andrew McAfee offer two models of innovation. They write: “[T]he true work of innovation is not coming up with something big and new, but instead recombining things that already exist.” In their view, automation is this second, recombinant type of innovation “in its purest form.” Digital innovation is less like fruit, which can be “used-up” once-and-for-all, but like building blocks that can be used together to achieve ever-greater advances.

What keeps this kind of innovation from increasing productivity all at once is obvious: not all combinations are successful. Over time, recombinant technology can lead to real growth, but in any given industry there can be—and often are—a thousand failed attempts before game-changing productivity-enhancing innovation is found.

5. ‘But robots are never going to take all the jobs. The problem with trying to envision “a world without work” is that it asks us to envision an unrealistically large change.’

None of the sources Yglesias cites are projecting a world without work. This is theoretically possible in the future, but more grounded and more immediately pressing scenarios involve partial reduction in work—the elimination of targeted economic sectors in society as the jobs in those sectors become increasingly automated.

At a conference held by Oxford University’s Future of Humanity Institute, a poll was taken in which the median estimate of when there will be 50 percent chance of Human Level Machine Intelligence (HLMI) was 2050, and 90 percent chance is 2150. In other words, the kind of artificial intelligence that could displace the majority, if not the entirety, of human work altogether is not right around the corner.

What is around the corner, according to a 2013 report published by the Oxford Martin Programme on the Impacts of Future Technology, is severe economic disruption due to increasing automation. The report concludes that in the next two decades, nearly 50 percent of total U.S. employment is at risk. Even David Autor, a noted “non-alarmist” about automation’s capacity to take our jobs, says automation will increasingly lead to labor-market polarization. High-education, high-wage jobs will be available, and low-education, low-wage jobs will be available, but automation will eliminate the middle sector. This doesn’t portend a “world without work,” but I think it’s still safe to register it as a problem.