Afraid of Robots Taking Your Job? You Should Be
In Rise of the Robots, a new book that predicts machines and computers will replace a massive number of human workers throughout the economy, Martin Ford sometimes supports his unsettling argument with a particularly clever technique.
He describes someone performing an intricate task, such as sorting a jumbled heap of boxes in a warehouse by size, color, shape, and markings, only to reveal that this someone is in fact a robot. The box-sorter, for instance, is a product of Industrial Perception Inc., a Silicon Valley start-up.
Later in the book he excerpts several paragraphs from a news story on a Boston Red Sox game that gives an unintended new meaning to the phrase digital journalism: The article was written by a software program designed by Narrative Sciences Inc.
I was vaguely hoping that the book would end with the shocking revelation that “Martin Ford” is actually a supercomputer, perhaps a casual acquaintance of IBM’s Watson. It would make sense for a computer to pen a book that claims the ascension of intelligent machines is overwhelmingly likely. Presumably it settled on the title Rise of the Robots by studying the names of blockbuster summer movies that many humans seem to enjoy.
Though Martin Ford shows no signs of being animatronic, he does make a persuasive and informative argument that journalism and box-sorting are only two of many jobs becoming increasingly vulnerable to automation. He cites a 2013 study by two Oxford researchers that concludes that almost 50 percent of total employment in America will be threatened by automation over the next two decades. Of course, relatively unskilled factory workers have been struggling against the specter of mechanized replacements since the Luddite protests in early 19th century England. But Ford thinks the machines are beginning to breach the walls of the citadel of skilled labor and will soon swarm over professions such as law, medicine, science, and even creative endeavors, including painting and composing music.
It would be comforting to dismiss this book as another case of a Silicon Valley entrepreneur making bombastic but silly futuristic claims. Unfortunately, at least from the standpoint of our species, Ford is not a gadget-addled lunatic; he’s a careful and thoughtful writer who relies on ample evidence, clear reasoning, and lucid economic analysis. In other words, it’s entirely possible that he’s right.
The bleakest news involves manufacturing and service jobs. He reports that Foxconn, the primary maker of Apple products, plans to deploy up to 1 million robots in its factories. A chief Nike executive, meanwhile, thinks the solution to rising wages in Indonesia is “engineering the labor out of the product.” This would not only be more affordable, it might also silence criticism about horrendous working conditions in international factories that make beloved American products. Self-checkout aisles at the grocery store, Redbox movie rental kiosks, and touchscreen ordering at restaurants are all examples of the same trend.
For those who claim that these changes simply move jobs from one sector of the economy to another, Ford points to statistics. Blockbuster once employed roughly 60,000 employees nationally. Redbox, in the entire Chicago area, has a staff of seven. A comparison between Google and General Motors is another instructive example. After adjusting for inflation, General Motors earned a profit of around $11 billion in 1979, when it employed 840,000 workers. Google, in 2012, earned almost $14 billion and employed fewer than 38,000 people. He offers many other examples suggesting that not every job lost in one area is gained in another.
The fast food industry is another of the book’s revealing case studies. McDonald’s announced in 2011 that it would transition to touchscreen ordering at 7,000 locations across Europe. And a San Francisco startup, Momentum Machines, aims to remove employees from food preparation entirely. Money saved by reduced labor costs could then be spent buying organic and sustainable ingredients.
All of this raises a fundamental question: are service and manufacturing jobs worth defending, or should we embrace the transforming economy as efficient and potentially more humane? One could argue that job satisfaction among fast-food workers is so low that the elimination of such jobs is a kindness. Efficiency and profits would rise, safety and hygiene in kitchens would improve, healthier ingredients could be used, and former employees would be free to pursue more meaningful work.
All of this might be true to a point, but the vaporization of jobs becomes self-defeating when applied on a sufficiently large scale. Wages are the primary means by which a majority of consumers gain purchasing power in the economy. If enough people are unemployed or underemployed, they will lack the ability to buy the food, products, and services that machines can offer with unparalleled speed and efficiency.
Companies could then shift their energies toward catering to high-income clientele or international markets. But for many American companies this is not a viable strategy. The super-rich can only buy so many luxury goods, and strong cultural imperatives to save money in China, the country often touted as an untapped reservoir of consumer demand, show few signs of changing. That more than half of the increase in U.S. income between 1993 and 2010 went to the top 1 percent of income distribution is morally disturbing, but it’s also an economic problem that will ultimately affect the profitability of companies that aim to reach more than a tiny sliver of the population.
The ubiquity of computer trading algorithms on Wall Street is well-known, but Ford also gives many other interesting and less-familiar examples of software encroaching on professions traditionally considered skilled. Narrative Sciences, the company that wrote the sports story on the Red Sox, was a brainchild of researchers at Northwestern and is now used by many media outlets, including Forbes. The CIA provided some investment funding for the company, presumably with the goal of using its products to expedite the analysis and processing of huge streams of intelligence from around the world.
Medical applications for IBM’s Watson, the supercomputer that defeated former champion Ken Jennings in Jeopardy! on national television, are already being tested in a variety of settings. Doctors are also training neural networks to make diagnoses, and software can already help pharmacists and doctors avoid fatal drug combinations. Specialized search engines are replacing work previously done by lawyers during the discovery phase of trials. And researchers in Spain and England, respectively, have designed software that can create intriguing, often beautiful paintings and complex musical compositions. The London Symphony Orchestra performed one piece in 2012 that was composed entirely by such a program. A Stanford professor, meanwhile, has identified two cases where genetic algorithms—programs that improve themselves through Darwinian natural selection— have created new and patentable inventions.
Examples crop up everywhere. Many Google employees would love to see self-driving cars zipping down streets and highways across America within the next decade. Because these cars could work 24-hour shifts and coordinate their efforts by sharing data on peak-demand locations and times, the pollution and inefficiency caused by countless cabs, buses, and personal cars could certainly be reduced. Of course the jobs of cab and bus drivers would also disappear.
Ford skillfully captures the complexities of these scenarios, outlining costs and benefits along multiple axes and resisting any simple narrative of technology as either Satanic or salvific. Perhaps his most interesting suggestion is for a guaranteed minimal income, possibly rebranded for political marketing purposes as a “citizen’s dividend.” This would take the place of traditional social safety nets and ensure that at least some purchasing power remains in the hands of consumers across the income spectrum as automation advances through different levels of the economy.
The amount of income would not be large enough to remove any incentive for further work; it would simply provide a basic source of funds that other jobs could supplement. This could then encourage risk-taking by entrepreneurs. Why not take a chance on a new business if you know you have some security if it fails? Proponents of such a measure have already succeeded in gathering enough signatures in Switzerland to put a proposal for a guaranteed monthly income on the national ballot. Each citizen would make around $2,800 per month if the measure passes.
How to pay for such an outlandish idea? Tax the rich, particularly the truly rich. Our current tax brackets often fail to distinguish between someone who makes, say, $500,000 and $5 million. But this difference matters, as does the difference between $5 million and $50 million. To simplify things, let the Google and Nike and Narrative Sciences executives eliminate all the jobs they please, as long as their taxes support a guaranteed minimum income.