On the evening of Thursday, Feb. 6, The Second Machine Age authors Andrew McAfee and Erik Brynjolffson will be at New America NYC to discuss their book with New York Times economics reporter Catherine Rampell. For more information and to RSVP, click here.
Two major publications hit bookstore shelves last week with the same overarching theme: The Second Machine Age, an important and (deservedly) much hyped new book on the future of technology and work from Andrew McAfee and Erik Brynjolfsson, and the Jan. 18 issue of the Economist, whose cover story paints a similar picture of the rapidly evolving nature of work. Yet each of these focuses primarily on one half of the story—how technology will affect the quantity of work, rather than the quality. If, as these authors argue, the key to future economic growth and prosperity is “racing with the machines,” with humans and ever-smarter computers doing what they’re best at side-by-side, is this a future you should be looking forward to, if you’re lucky enough to have a job in the future at all?
The short answer is that no one knows for sure. Predicting the future of technology will always be tough, let alone predicting how social and technological factors will interact over time. Both The Second Machine Age and the Economist are generally optimistic: McAfee and Brynjolfsson envision a world with “less need to work doing boring, repetitive tasks and more opportunity for creative and interactive work.” Similarly, the Economist takes the viewpoint that although “innovation kills some jobs, it creates new and better ones.”
But there are at least three big reasons why the work of the future won’t necessarily be fulfilling.
First, the tasks that are easy to automate aren’t necessarily the boring and repetitive ones, and the tasks that are hard to automate aren’t necessarily the fun and interesting ones. Consider, for example, the warehouses that power Amazon’s vast supply chains. As shown in a recent BBC documentary and a first-person account in the Guardian, the workers in these warehouses aren’t exactly living the dream—they are under constant pressure by their computerized overlords to meet impossible picking-and-placing targets, are physically exhausted at the end of work each day, and their working conditions may put them at increased risk of mental illness.
From a technological point of view, these warehouses are perfect examples of human-machine symbiosis in action. People have excellent dexterity and perception compared with robots, and computers can schedule workers’ movements around the warehouse efficiently, use their perfect memories to keep track of the locations of items, and set targets to motivate employees. From a subjective point of view, though, many of these workers report feeling like robots themselves.
Second, even when enjoyable and harmonious jobs are technologically and socially feasible, companies may not face strong incentives to design them like that. I don’t mean to pick on Amazon, but another one of their business units illustrates this point quite well. Amazon’s Mechanical Turk is a system for achieving “artificial artificial intelligence”—people get paid from 1 cent to a few dollars to perform small “micro-tasks” that humans can do fairly easily but computers still can’t. If you’re a company and need, say, thousands of images (or pornographic videos) labeled, or categorized, or summarized in a sentence, Mechanical Turk is a great resource.
McAfee and Brynjolfsson call this an “interesting” model and “another way for people to race with the machines, although not one with particularly high wages.” It’s certainly an “interesting” model, but “not one with particularly high wages” is quite an understatement: A 2010 study found that workers make on average about $2 per hour doing these micro-tasks, and that a significant minority (18 percent) of so-called Turkers (particularly in India) rely on the service to make ends meet. More recently, a 2013 study by some of the same researchers found that Turkers often complain about delayed payments, ignored grievances, and their work being unfairly rejected without compensation. From an economic point of view, though, why should Amazon bother with these complaints? There are hundreds of thousands of Turkers standing by to sign up for new micro-tasks, and the humans-as-a-service model (as researchers Lilly Irani and M. Six Silberman put it in the 2013 study) lends itself to anonymous, unappreciated work, not the kind of challenging, social, and meaningful experience workers desire.
Third, one person’s meaningless, repetitive labor is another person’s satisfying, hard day’s work and a big part of her identity. In a globalized capitalist economy, how do you automate one of these categories and not the other when they mean different things to different people? Even if the average job a decade from now is “better” than the average job today, this won’t stop many (often low-wage) workers from suffering as the gradual wave of ever-smarter machines makes their job, and a crucial part of their identity, redundant. Within a particular profession, too, the pleasure associated with various aspects of the job may slip away. The goal of applying IBM’s Watson to various commercial domains, as recently articulated by Jill Puleri, worldwide retail industry leader for IBM Global Business Services, is “scaling expertise.” There’s a lot of knowledge, often unarticulated, in the minds of salespeople, lawyers, doctors, and people in countless other professions that their customers need, and this white-collar knowledge is increasingly being converted into computer code and widely distributed, as Farhad Manjoo wrote in his 2011 Slate series “Robot Invasion.” This is great for IBM’s and other companies’ bottom lines and consumers who want an answer to their question as soon as possible without the small talk, but not necessarily for those who pride themselves on sharing their accumulated wisdom and experience.
Technology, then, won’t necessarily make every job more enjoyable. But it also doesn’t have to make working horrible, either. Indeed, many of the factors associated with work satisfaction are not directly related to the task itself being performed. The empirical evidence linking time spent using computers to work satisfaction is mixed, but according to organizational psychology expert Estelle Morin, some of the characteristics of meaningful work are a sense of social purpose, moral correctness, achievement-related pleasure, autonomy, recognition, and positive relationships.
The ways to bring about high-quality jobs, measured by these criteria, may not be that different in the future from what they’ve been in the past. To ensure an appropriate fit between the skills and ambitions of future workers and what jobs will be available, we should think holistically about which technologies we’re investing in and what we teach in our schools. Twenty years seems like a long time on a technological timescale, but that’s more or less when people born today will be entering the work force, so that seems like a reasonable horizon for thinking about some of these issues. Consumers may want to vote with their wallets to reward companies that provide a high-quality work environment for their employees. Unions could, as they’ve done in the past, play a role in determining technology adoption in certain industries. And governments could help address the massive imbalance between the number of jobs available and the people pursuing them. For instance, Nobel Prize–winning economist Edmund Phelps has proposed subsidizing companies to hire low-wage workers. When businesses have to compete against one another to attract employees, instead of the other way around, they may be more inclined to foster satisfying, and not merely productive, work environments.
While technology won’t, on its own, create fulfilling jobs for all, that doesn’t mean we shouldn’t try. The Second Machine Age and new articles coming out nearly every day on similar topics are calling attention to the urgent issue of the changing division of labor between humans and machines, and it’s hard to think of a more important policy issue to get right. The issue isn’t going away any time soon, but it’s not new, either: science-fiction writer Isaac Asimov was imagining ways we could get this right in 1977 when he wrote, “In a properly automated and educated world, then, machines may prove to be the true humanizing influence. It may be that machines will do the work that makes life possible and that human beings will do all the other things that make life pleasant and worthwhile.” Even as an ideal, it still isn’t clear what a properly automated and educated world would look like, and it may be a moving target. For now, though, we have a lot of work to do before we get there.
This article is part of Future Tense, a collaboration among Arizona State University, the New America Foundation, and Slate. Future Tense explores the ways emerging technologies affect society, policy, and culture. To read more, visit the Future Tense blog and the Future Tense home page. You can also follow us on Twitter.