Ever since inequality began rising in the U.S., in the nineteen-seventies, people have debated its causes. Some argue that rising inequality is mainly the result of specific policy choices—cuts to education, say, or tax breaks for the wealthy; others argue that it’s an expression of larger, structural forces. For the last few years, Tyler Cowen, an economist at George Mason University and a widely read blogger, has been one of the most important voices on the latter side. In 2011, in an influential book called “The Great Stagnation,” Cowen argued that the American economy had exhausted the “low-hanging fruit”—cheap land, new technology, and high marginal returns on education—that had powered its earlier growth; the real story wasn’t inequality per se, but rather a general and inevitable economic slowdown from which only a few sectors of the economy were exempt. It was not a comforting story.
“Average Is Over,” Cowen’s new book, is a sequel to, and elaboration upon, “The Great Stagnation.” In many ways, it’s even less comforting. It’s not just, Cowen writes, that the old economy, built on factory work and mid-level office jobs, has stagnated. It’s that the nature of work itself is changing, largely because of the increasing power of intelligent machines. Smart software, Cowen argues, is transforming almost everything about work, and ushering in an era of “hyper-meritocracy.” It makes workers redundant, by doing their work for them. It makes work more unforgiving, by tracking our mistakes. And it creates an entirely new class of workers: people who know how to manage and interpret computer systems, and whose work, instead of competing with the software, augments and extends it. Over the next several decades, Cowen predicts, wages for that new class of workers will grow rapidly, while the rest will be left behind. Inequality will be here to stay, and that will affect not only how we work, but where and how we live.
If we want a preview of work in the twenty-twenties and twenty-thirties, Cowen writes, we should look to the areas where computer intelligence is already making a big difference: areas like dating, medicine, and even chess. This interview with Cowen has been edited and condensed from two conversations.
If we want a preview of work in the twenty-twenties and twenty-thirties, Cowen writes, we should look to the areas where computer intelligence is already making a big difference: areas like dating, medicine, and even chess. This interview with Cowen has been edited and condensed from two conversations.
In “Average Is Over,” you argue that inequality will grow in the U.S. for the next several decades. Why?
There are three main reasons inequality is here to stay, and will likely grow. The first is just measurement of worker value. We’re doing a lot to measure what workers are contributing to businesses, and, when you do that, very often you end up paying some people less and other people more. The second is automation—especially in terms of smart software. Today’s workplaces are often more complicated than, say, a factory for General Motors was in 1962. They require higher skills. People who have those skills are very often doing extremely well, but a lot of people don’t have them, and that increases inequality. And the third point is globalization. There’s a lot more unskilled labor in the world, and that creates downward pressure on unskilled labor in the United States. On the global level, inequality is down dramatically—we shouldn’t forget that. But within each country, or almost every country, inequality is up.
You think that intelligent software, especially, will make the labor market more unequal. Why is that the case?
Because of the cognitive requirements of working with smart software. And it’s also about training. There’s a big digital divide in this country.
One of the most interesting sections of the book is about “freestyle” chess competitions, in which humans and computers play on teams together—often the computers make the moves, but sometimes the humans intervene. How has chess software changed the “labor market” in chess players?
When humans team up with computers to play chess, the humans who do best are not necessarily the strongest players. They’re the ones who are modest, and who know when to listen to the computer. Often, what the human adds is knowledge of when the computer needs to look more deeply. If you’re a really good freestyle player, you consult a bunch of different programs, which have different properties, and you analyze the game position on all of them. You try to spot, very quickly, where the programs disagree, and you tell them to look more deeply there. They may disagree along a number of lines, and then you have to make some judgments. That’s hard—but the good humans do that better than computers do. Even very strong computers don’t have that meta-rational sense of when things are ambiguous. Today, the human-plus-machine teams are better than machines by themselves. It shows how there may always be room for a human element.
You believe that, in the future, the most well-compensated workers will be something like freestyle chess players.
Think in terms of this future middle-class job: You read medical scans, and you work alongside a computer. The computer does most of the judging, but there are some special or unusual scans where you say, “Hmm, that’s not quite right—I need a doctor to look at this again and study it more carefully.” You’ll need to know something about medicine, but it won’t be the same as being a doctor. You’ll need to know something about how these programs work, but it won’t be the same as being a programmer. You’ll need to be really good at judging, and being dispassionate, and you’ll have to have a sense of what computers can and cannot do. It’s about working with the machine: knowing when to hold back, when to intervene.
Or take business negotiations. In the early stages of negotiation software, on your smartphone, there may be programs that listen to the pitch of a voice, or that test for stress. You’ll just ask the program, “Was he lying? Was he eager to do business with me?” Maybe the computer will be right sixty per cent of the time. That’s useful information, but it’s still going to be wrong a lot. And in a given negotiation, you’ll be reading off many programs, and you’ll have to decide which of those programs is more relevant.
by Joshua Rothman, New Yorker | Read more:
Image: Simone Casetta/Anzenberger/Redux
