Every few years, we experience a wave of concern over the rise of robots and its effect on jobs. Automation, we hear, will rid the economy of human labor, replacing the inefficient flesh-and-blood employee with amazingly powerful computers. Yet the robot takeover has so far not occurred — human workers seem to be surviving and even thriving alongside all the machines.
Another one of these surges of concern is upon us, fueled by books such as Tyler Cowen’s Average Is Over and Martin Ford’s Rise of the Robots, as well as a spate of articles arguing that this time, the computer revolution really is different. And when we wade through the headlines, this time actually does look different.
Google’s autonomous car has already traveled nearly a million miles on California and Nevada roads. Elon Musk, the founder of the electric-car company Tesla Motors, recently predicted that autonomous cars could enter the market as soon as this year, potentially wiping out the taxi and trucking industries in one fell swoop.
Robots are getting better not only at understanding road conditions but also at “reading” their human operators. IBM’s computer system Watson, which famously defeated its human opponents in the television game show Jeopardy, has continued to rapidly improve and is now answering complex queries about such subjects as medicine. Apple’s Siri voice interface for the iPhone has also improved. Silicon Valley seems close to building a starship Enterprise–like voice-based computer, threatening hundreds of thousands of jobs in customer-support call centers.
Added to the usual worries about computers’ replacing workers is a new concern: Who will own these robots? Will a small stratum of people (capitalists, of course) control them and extract exorbitant rents from the rest of us? If you thought inequality was a problem before, the critics warn, wait until you see what happens next.
Technological change always brings out these negative voices, because we don’t have answers for many important questions. We don’t know where new jobs are going to come from or even whether there will be work to do at all in 50 years. Yet in light of the history of technological innovation, such fears are unfounded.
Far from exclusively benefiting elites, automation has allowed people of modest means to buy products that were once luxury items available only to the most deep-pocketed consumers. Robotics have caused tremendous social change and will probably continue to do so, but their long-term effect may well be to decrease inequality rather than increase it. Indeed, robotics and automation have perhaps done more to improve quality of life than has any other economic force in history. We need to keep this in mind as we assess the massive, world-changing potential of the next round of technical innovations.
Thanks to Star Wars, many of us have images of robots as humanoid figures walking around the desert, but the reality is that robots are often built into the products we use every day. Consider the Keurig coffeemaker. We place a special cup in the machine, hit a button, and the built-in computer handles the rest, leaving us a steaming hot cup of coffee, with minimal human involvement in the brewing process. The device has become a mainstay in office break rooms, and Keurig has sold millions of units around the world. Yet baristas haven’t disappeared from the work force. Despite the popularity of automated coffee machines, Starbucks continues to increase its earnings and expand to new locations, with about 1,500 new stores opening just last year.
We often think of robotics as a zero-sum economic game in which humans and machines are locked in a tug-of-war. The Keurig coffeemaker shows that the zero-sum calculus can be flat wrong. Similarly, accountants did not disappear after the arrival of Excel and QuickBooks; in fact, accounting majors have been some of the most in-demand college graduates in recent years.
Home appliances are particularly good examples of how automation increases convenience, since they are among the most common robots we use daily. Cooking is simplified by microwaves that have all kinds of automation built in, such as buttons that heat our food to the perfect temperature. Cleaning our homes takes less and less effort as well, with devices such as iRobot’s Roomba, which can automatically sweep the floors.
Perhaps no robots have had a greater effect on quality of life than the washing machine and the dishwasher. In the mid 20th century, when women no longer had to do laundry or wash dishes by hand, they suddenly had considerably more time for themselves. The devices saved hundreds of millions of hours of household labor per year. Some scholars argue that the laundry machine did more to increase female participation in the economy than any other change in the last century.
While many of these conveniences began as luxury goods, history shows us that automation tends to permeate the economy quickly. Yesterday’s computers cost millions of dollars and took up whole floors of office buildings, and printers cost tens of thousands of dollars. Today, we can carry a supercomputer in our pocket and purchase a desktop printer for less than a hundred dollars.
Those who fear that robotics will increase inequality overlook the great consumer demand for these products, and the supply-and-demand interplay and competition that force prices ever downward. Based on its autonomous-driving technology, Google could become a monopoly that owns all cars, but it’s more likely that all car manufacturers will incorporate this technology into their models.
There is little reason to think that this trend of democratization will stop, and it may even be accelerating. Soon 3-D printers will allow us to “print” millions of different objects, from mugs to the coasters they sit on. Such printers cost thousands of dollars today, but their prices have fallen dramatically over the past few years, and they will probably be in wide use by the end of the decade. Further, 3-D printers will probably increase the pace of innovation across many fields, as they make it cheaper to quickly make product prototypes and sell early models, allowing more inventors to get in the game and make their work available to the public.
When technology allows consumers to produce average-quality goods at home, companies must offer higher-quality products to compete. The greeting-card industry, for instance, faced extinction with the advent of desktop printing, but it started producing specialized designs that home printers cannot (yet) match. The market expanded to encompass a greater range of consumer tastes.
To be fair, patents and other intellectual-property protections ensure that the inventors of technologies are well rewarded for their efforts. Hewlett-Packard has made millions off its printer ink, much as Keurig and Whirlpool have made millions off their products. But economies of scale are no more likely to drive out competition tomorrow than they are today.
But we shouldn’t tout the benefits to the individual consumer of all these conveniences without taking a wider look at automation and its overall effect on the economy. Greater efficiency through robots allows us to produce more in less time, but these changes can force some workers to change their occupations.
The most important factor in improving quality of life is productivity growth. Productivity is simply the quantity of goods and services we can produce given limited resources, particularly our time.
If we want to improve our standard of living, there are only two options available. One way is to increase our work hours and thus the amount that we produce. But human productivity rarely grows linearly in proportion to hours worked, nor do we necessarily want to spend more time at work. The other option is to increase productivity per hour. We do this when we expand access to education and job training, increasing the productivity of individual workers. We do this also by enhancing human industry through the use of tools, which includes automation and robotics. When we use a microwave to produce a meal in five minutes rather than an hour, we have increased our food-preparation productivity more than tenfold. Multiply such improvements across all devices throughout the economy, and the massive efficiency gains we’ve made in the last hundred years look unsurprising.
Automation in the economy doesn’t strike randomly; it takes hold when market forces determine that physical capital (a robot) is cheaper than human capital (a worker). America’s entire manufacturing sector used to be heavily dependent on human labor, but today’s highly efficient factories produce more goods than ever before while employing far fewer people, because of robotics.
It’s not only a line worker in a factory or a burger flipper who might be replaced by a robot. Many white-collar workers are also at risk. The rules of the market affect everyone. Investors have poured millions of dollars, for instance, into computer startups targeting the legal industry, because lawyers read boilerplate contracts at hours billable well into the triple digits.
Creative destruction is as old as history, but the pace today is accelerating, with millions of workers potentially affected by automation in a matter of years instead of decades. Professions created just a few decades ago are now being eliminated, and entire job categories can rise and fall within a single generation.
There are no simple solutions. Increased efficiency rewards all of us with lower prices for higher-quality goods and services, but certain groups of workers could suffer deep losses. It’s possible that education itself will become more automated, which would allow more workers to take classes and improve their skills to compete in the marketplace. English-teaching robots already exist in Japan and South Korea, and more subjects may soon be offered by such automated programs. Workers must constantly improve their productivity to increase their value. This is fundamentally good for the economy, because it means that the average hourly value of a human worker is increasing over time.
We are all going to have to improve our skills to be competitive in this economy, but this transition shouldn’t distract us from the economic bounty that awaits on the other side of the revolution in robotics.
While the issue of employment garners the most attention from commentators, robotics’ socially transformative effects deserve our scrutiny as well. Perhaps no technology has more potential to improve our quality of life than the autonomous car. We will be able to relax during our commutes, reducing our stress and improving our health. Autonomous cars could almost instantaneously deliver a greater number of goods and services, such as meals, household supplies, and home-maintenance services, giving us more leisure time. Perhaps most significant, many fewer accidents would be caused by drunk driving or distraction while driving.
If autonomous cars become popular, we could greatly reduce the land space devoted to roads and parking. City governments could dedicate vast tracts of land to a variety of new uses, such as parks or housing.
Finally, and perhaps most futuristically, we will have to adapt to having more robots in nearly all aspects of our daily life. Siri and Watson are just the first steps toward fully personalized digital assistants, and future generations of these sorts of products will lead to all kinds of new social interactions and situations, affecting human relationships in ways we can’t yet predict.
There is no question that our economy will undergo vast changes in the next few years. Critics are right to warn that many jobs will be made redundant, and that automation might increase inequality, at least in the short term. But we’ve survived — and thrived — through waves of automation for centuries, and the productivity gains show that we should be championing these improvements, not hoping they stop soon. The best has been, and always will be, just around the corner. Let R2-D2 show the way.
– Mr. Crichton is a doctoral student at Harvard’s John F. Kennedy School of Government, where he researches labor economics.