The shiny white robot has a stooped, almost deferential stance as it approaches the Honda employees seated around a table. It turns its black faceplate to the humans, makes an open-handed gesture, and asks if they want anything to drink.
The people all speak simultaneously. What initially seems like rudeness turns out to be efficiency: ASIMO, the most advanced humanoid robot on the market, can understand multiple voices at once and uses facial recognition software to match the men with their requests. "Oolong tea, Mr. Ohara?" "Coffee, Mr. Oga?" "Milk tea, Mr. Ariizumi?" it confirms. They nod, and ASIMO heads off to fill the orders.
So far, ASIMO—at least as seen in a 2014 segment on Japanese public television—appears rather more competent than the baristas at my local Starbucks, who frequently ask me to repeat my order and haven't a clue who I am, despite my semi-regular appearances at the same location for the last six years. But as ASIMO walks away to pick up the drinks, it's apparent that there's much work ahead for Honda's engineers. The gait of the hobbit-sized machine is slow, with the knees-bent, elbows-out posture of a cautious toddler on unfamiliar turf. Honda claims that ASIMO (an acronym for Advanced Step in Innovative Mobility, not a deliberate tribute to the science fiction novelist Isaac Asimov, the company insists) comes equipped with a collision avoidance system, but that too is on par with a 2-year-old—everything is fine when nearby people are moving slowly and making allowances for the fledgling bot, but Mr. Ohara, Mr. Oga, and Mr. Ariizumi would be very thirsty indeed if they trusted ASIMO to pick up their drinks and carry them down a busy city street at rush hour.
Watching Honda's latest shuffle along creates a kind of vertigo. The robot revolution seems simultaneously upon us—look, a real robot serving coffee!—and eons away. But that dissonance is a clue that we are nestled in the elbow of an exponential curve. All around us, a Cambrian explosion of robotics is taking place, writes Peter Diamandis, chairman of the X Prize Foundation, at Singularity Hub, "with species of all sizes, shapes and modes of mobility crawling out of the muck of the lab and onto the terra firma of the marketplace, about to enter your home and your shopping experience."
Diamandis is right. Your house, neighborhood, and office are already full of the robots humanity has been waiting for with both anticipation and dread. They may be the equivalent of trilobites now, but they're multiplying and mutating rapidly. While pessimists fret that a new kind of intelligent automation will mean social, economic, and political upheaval, the fact is that the robots are already here and the humans are doing what we have always done in the face of change: anticipating and adapting where we can, muddling through where we can't, and trying to enjoy the ride.
Domo Arigato, Mr. Roomba
When it comes to prognostications about the robot revolution—and for the purposes of this article, we'll take an expansive view of what constitutes a robot, lumping together a wide variety of automated digital and mechanical deputies—Roombas are frequently asked to shoulder more than their fair share of the burden. Semi-autonomous vacuums are the most visible robots on the market, with more than 10 million sold worldwide at the end of last year. They look like the devices science fiction told us to expect: standalone machines that perform tasks on behalf of human beings, integrated into everyday life.
But if we're being honest, they're also a bit of a letdown. Anyone willing to fork over a few hundred bucks to the iRobot Corporation can have a machine zip out from under his sofa—that's where mine lives, anyway—and vacuum his house from time to time. It's oddly hypnotic to watch the device in action, as it deftly avoids falling down stairs, extricates itself from rug tassels and tight spots, and handily routes around chair legs. But it's just a vacuum cleaner, after all: a slightly smarter version of the dishwashers, washing machines, and microwaves we take for granted. And like ASIMO, the Roomba seems remarkably capable at some tasks and astonishingly inept at others, as when it accidentally bumps the door of the bathroom closed and then bounces around for hours, mindlessly cleaning the same tiny space until its battery dies.
Then there's the matter of the human maintenance required by our robot servants. The Roomba will go find its charging station when it needs more power (unless it's locked in the half bath, of course). But it requires a person to empty the reservoir when it's full of dirt and to periodically clean the moving parts. I'm terrible at taking care of my Roomba—I haven't changed the filters, well, ever—which generates a vague sense of guilt, as if I am mistreating a pet. In fact, extracting small objects from its bristles when they get caught feels surprisingly similar to the act of yanking a chicken bone from the mouth of a disobedient puppy. Small mammals love Roombas—YouTube offers an entire genre of "Roomba rodeo" videos, in which babies, cats, and small dogs glide around on the backs of the motorized discs—but the expensive machines are not meant to be used as carnival rides and are easily damaged, requiring yet more intervention. These 10 million vacuums don't exactly seem poised to gain sentience and take over the planet.
Still, having a Roomba means that I spend less time cleaning up crushed snack-food items—or less money employing someone else to perform that task. Does the fact that a machine instead of a person is lowering the Cheerio-load in my carpet mean it's time to start freaking out about the future of employment?
The Automation Jobless
When we talk about robots taking jobs, strong hydraulic arms looming over factory assembly lines is what comes most readily to mind. The International Federation of Robotics put the population of industrial robots at more than 1.1 million in 2013, making robots a well-established component of U.S. manufacturing.
But the more interesting (and less well understood) phenomenon is the advent of robot replacements for jobs long considered immune from mechanization, particularly the service functions that make up a significant part of our day-to-day interactions.
Speaking at the conservative American Enterprise Institute in March, Bill Gates hinted that a little freaking out might be in order: "Software substitution, whether it's for drivers or waiters or nurses, [is] progressing…Twenty years from now, labor demand for lots of skill sets will be substantially lower. I don't think people have that in their mental model." A September 2013 study from Oxford University looked at 702 occupations and found that 47 percent of total U.S. employment faces the risk of being eliminated in favor of computerization.
But this isn't the stuff of a misty, menacing future. It's already underway. The Botlr robot, deployed in some properties of the Starwood hotel chain, delivers extra towels and forgotten toiletries to hotel guests. Having a robot show up with your missing items sounds much better than awkwardly answering the door with your bare legs sticking out of the bottom of a hotel robe with a couple of crumpled dollar bills awkwardly clutched in your hand.
Singapore's Timbre restaurant group signed a deal in November to bring flying Infinium-Serve robot waiters to their five locations in the labor-crunched country. The robots would deliver food and drink—acting as propellered busboys, but not fully replacing waiters and bartenders, who would continue to be tasked with "higher-value tasks such as getting feedback from customers," CEO Woon Joonyang said in a press release. The Consumer Electronics Association predicts that commercial sales of unmanned aerial vehicles will reach $130 million in revenue in 2015, up 55 percent from last year, putting 400,000 units into the skies.
While unemployment rates have fallen to 5.6 percent and financial markets have largely recovered from the recession, ordinary people share the intuition that technology may be to blame for some unpleasant economic undercurrents, including high joblessness rates among young people, record numbers of Americans who say they have stopped looking for work, and expanded disability rolls. A December New York Times/CBS News/Kaiser Family Foundation poll of unemployed 25- to 54-year-olds found that 37 percent of those who said they wanted a job believed technology was a reason they did not have one.
In his 2012 book Coming Apart, political scientist Charles Murray charts a widening gulf between the white upper and lower classes between 1960 and 2010. Murray sees some of the few occupations left bridging the gap—low-skilled white-collar jobs, for instance—disappearing thanks to automation. Jobs like phone operators, once upon a time, and tax preparers or travel agents more recently.
Holding back automation is impossible, says Murray: "This is not something where you can artificially subsidize people to become buggywhip makers." But thanks to our unimpressive education system, he argues, Americans are less well-equipped to flexibly handle change than they once were-and anyway, the latest round of automation isn't creating new jobs the way previous advances in industrialization once did. The upshot: an even faster social and political bifurcation.
Fretting about the impact of automation on employment is a time-honored tradition. In 1961, after 10 months of recession, Time published a story on "The Automation Jobless." As economist Timothy Taylor points out on his Conversable Economist blog, the text could have been plucked from this week's issue of the magazine. "While no one has yet sorted out the jobs lost because of the overall drop in business from those lost through automation and other technological changes, many a labor expert tends to put much of the blame on automation," the 1961 essay intones. "Throughout industry, the trend has been to bigger production with a smaller work force…Many of the losses in factory jobs have been countered by an increase in the service industries or in office jobs. But automation is beginning to move in and eliminate office jobs too…In the past, new industries hired far more people than those they put out of business. But this is not true of many of today's new industries."
Politicians took up the refrain then, just as they do now. In the famous speech where he vowed to put a man on the Moon, President John F. Kennedy delivered a line that could have been dropped into Obama's State of the Union this year verbatim: "I am therefore transmitting to the Congress a new Manpower and Training Development program to train or retrain several hundred thousand workers particularly in those areas where we have seen chronic unemployment as a result of technological factors and new occupational skills over a four-year period, in order to replace those skills made obsolete by automation and industrial change with the new skills which the new processes demand."
Yet the legacy of the 1960s is not one of apocalyptic unemployment and social breakdown. As Taylor notes: "The U.S. unemployment rate had declined back to the range of 5.0 percent by August 1964, but concerns over how the U.S. economy might adapt to technology and automation remained serious enough that President Lyndon Johnson signed into law a National Commission on Technology, Automation, and Economic Progress. The Commission eventually released its report in February 1966. When the unemployment rate had fallen to 3.8 percent."
I'm not leaving my kids with the Roomba when I go out, no matter how much they love it. But I'm perfectly happy to deputize a robot sitter from time to time. Like so many overanxious yuppie parents, I keep a small video camera in my (very young) children's bedroom. The Dropcam is WiFi enabled, so I can check it from anywhere, including my phone. I can also set it to alert me if there is unusual noise or movement in the room. What that means is that I can stretch the boundaries of being "at home" with the kids to include dinner at the next-door neighbor's house or even the Italian restaurant on the corner—anywhere I can (a) see my house to make sure it's not on fire and (b) get home quickly. This $99 Internet-enabled, infrared, motion-sensitive digital eyeball has put a sitter out of a job on more than one occasion, when I have happily deployed a machine to keep watch on my kids in marginal circumstances when a human being would previously have been necessary.
People with single-level houses and a little more cash to spare can do more than just watch their kids sleep. They can actually follow them around the house and nag them to do their homework or eat their peas using one of several telepresence robots now on the market. The general phenotype of these machines is something like an iPad mounted on a Segway. Products like the Double, Beam, and Kubi—all of which are currently available for purchase—let a person who is not in the room act like he's there. That means less work not just for babysitters but also for airline pilots, as telepresence becomes increasingly common in offices as well.
Of course, the act of flying a plane is itself heavily automated. The pilot and first mate are increasingly there just for show, and they may soon vanish as driverless cars acclimate the population to the idea of vehicles without humans at the helm.
Bot, You Can Drive My Car
Driverless cars are often cited as the Typhoid Mary of the coming robot plague. But for now, Uber and other car service apps are great examples of technological change generating more jobs, while simultaneously creating a consumer surplus as customers buy superior goods and services for a lower price. Data released by the company in January showed that Uber drivers were earning more than their professional taxi driving counterparts—with take home pay as high as $17 an hour in Washington and Los Angeles, $23 in San Francisco, and $30 in New York.
All of these services may someday be fully automated. Uber CEO Travis Kalanick has made no secret of the fact that he plans to replace human drivers with self-driving cars as soon as possible. But for now, new tech is generating new kinds of human jobs that are arguably better than the similar ones they replaced, even as the supposedly menacing robots crowd in around us.
All 2015 models of the Tesla S come equipped with enough features to constitute an autopilot mode: Thanks to radar, ultrasonic sonar, a camera with image recognition, GPS, and more, the car boasts adaptive cruise control that adjusts to the speed of traffic, the ability to read speed limit signs and stay in its lane, self-parking (both parallel and garage), and self-stopping if a crash is about to occur. Many of these features are already standard in other luxury car brands as well. A significant percentage of cars on the road could pilot themselves much of the time if we let them—and increasingly we are letting them—which makes the handwringing about self-driving cars seem both premature and a case of too little, too late.
Transformers: Robots in Disguise?
Marc Andreessen, who invented the Web browser and is now a leading venture capitalist, has taken to Twitter and his blog to decry automation alarmism. He writes that the fear "robots are going to eat all of the jobs" is a prime example of the "'lump of labor' fallacy—the idea that there is a fixed amount of work to be done." With the rise of smartphones and broadband, he says, an unprecedented number of people have access to the means of production. And it's crazy to think they won't do new, impressive, resourceful, economically stimulating things with those tools.
The "this time is different" argument, Andreessen continues, contains the subtext "there won't be new ideas, fields, industries, businesses, and jobs. In arguing this with an economist friend, his response was, 'But most people are like horses; they have only their manual labor to offer' I don't believe that, and I don't want to live in a world in which that's the case. I think people everywhere have far more potential." Andreessen isn't alone. In February, the University of Chicago asked economists if they thought that automation had historically decreased employment. Some 76 percent agreed that it had not.
But Tesla's Elon Musk and others have pushed back, arguing that the economic threat is compounded by a more serious existential threat from artificial intelligence (A.I.). The Future of Life Institute released an open letter in 2014, with an impressive list of signatories including Musk, physicist Stephen Hawking, and actors Morgan Freeman and Alan Alda, expressing concerns about the rise of A.I. "The potential benefits are huge, since everything that civilization has to offer is a product of human intelligence; we cannot predict what we might achieve when this intelligence is magnified by the tools AI may provide, but the eradication of disease and poverty are not unfathomable. Because of the great potential of AI, it is important to research how to reap its benefits while avoiding potential pitfalls." The letter itself is vague, but Musk has called A.I. a "demon" that is "potentially more dangerous than nuclear weapons." And other Future of Life Institute documents fret about how to ensure that A.I.s use weapons systems or surveillance cameras appropriately.
Laser Eyeballs and Hamburgers
Back in 2002, when LASIK was still in its infancy, I went under the laser to get my atrocious vision corrected—it was a graduation present from my parents. Even then, the surgery itself was almost entirely automated. I realized the doctor wasn't doing anything even remotely related to the actual procedure when she started chatting with me about the best place to get a burger in New Haven while the smell of burning eyeball filled the air. More than 20 million people have had the same experience, though for many of them the recall of the experience may be blurred slightly by Ativan or other anti-anxiety meds, typically administered to people who are nervous about letting a robot shoot lasers into their eyes.
Medical robots have gotten a whole lot smarter since then. Watson, which you most likely know as the IBM Jeopardy champ, has turned its attention to human anatomy. To kick off its medical education, Watson "read" all of PubMed and Medline, two enormous databases of medical journals. In March 2012, Memorial Sloan Kettering agreed to allow Watson to consume tens of thousands of cancer patient's records. Forbes reported in 2013 that Watson had analyzed 605,000 pieces of medical evidence, 2 million pages of text, and 25,000 training cases and had the assistance of 14,700 clinician hours fine-tuning its decision accuracy. Watson's skills as a diagnostician are already outdoing human doctors in some areas, including detecting lung cancer, where Watson's 90 percent success rate is much better than humans' 50 percent.
After Watson diagnoses a problem, using natural language inputs from human clinicians as well as diagnostic images and the patient's medical history, its client could be turned over to one of the more than 3,000 da Vinci surgical robots in hospitals worldwide. The robots are controlled by a human surgeon who is typically in the same room as the patient, but have the precision and extrasensory capacity to make surgeries—particular hysterectomies and prostate removals, where they are most commonly used—less invasive and more accurate.
Working Their Way Up From Getting Coffee
In January, Persado Inc. raised $21 million in venture capital. The company has created software that replaces copywriters—when Verizon wants to get a customer to renew their contract, for instance, Persado helps craft an email that is calibrated to maximize the chances of success, strategically deploying key words and creating appealing financial deals. But the emails also play on emotion, choosing whether to threaten a customer with a lost opportunity or gratefully thank them for their continued business. And it seems to work: Citi, a Persado customer, told The Wall Street Journal that the tool has increased the rate at which emails are opened by 70 percent. The clickrate inside the emails has gone up by 114 percent over human-crafted missives.
Should I fear for my job? After all, it's just a hop, skip, and jump from heart-tugging ad copy to readable magazine features, right? Maybe. But buried in the coverage of the significant investment was this little tidbit: CEO Alex Vratskides says the new V.C. money will be used to expand Persado's salesforce. Which consists of humans who win over other humans as customers by showing them how much better computers can be at the jobs where they are currently employing humans.
As some jobs fall by the wayside, others are created. Another company, Journatic, offers a different kind of computer generated copy: hyperlocal news. Bots extract information from publicly available data sources, such as real estate transaction records and press releases, and recombine the information into the form of a traditional news article, which can then be reproduced in local broadsheets, neighborhood supplements, and websites. But the company also employs human copy editors to clean up the text, bring it into conformity with Associated Press style, and generally check the computers' work. In 2012, the company got in trouble for putting fake bylines on its content, something it quickly agreed not to do again. We might like to think a person wrote the story we're reading, but when it comes down to it, some newspapers—and readers—are already willing to let that illusion go for the sake of the bottom line.
Welcoming Our Robot Brethren
The robot-ridden future may sound vaguely terrifying, but it's unlikely to be terribly different from the robot-ridden present. You are already the commander of a tiny but powerful robot army. In lieu of hiring human beings or doing the work yourself, your bots do your banking, cleaning, babysitting, letter writing, and more. Perhaps your job will disappear, but a new one—one you probably can't imagine any more than an 18th century farmer could imagine an I.T. support tech—will emerge.
Soon we will find it jarring to discover a flesh-and-blood person doing tasks that were once "impossible to automate." Is it ever good news when you need to talk to a bank teller in person or "speak to a representative" about something that's not on an automated phone menu? Getting and giving directions has ceased to be a point of tension or confusion; ubiquitous, traffic-savvy GPS has it covered. It's not that people will interact less; it's that we will be forced to transact less. As our machines take care of more business, we will be free to pursue other things. It may be unnerving to talk to someone about all-beef patties while they oversee a surgical procedure on an important body part, but I'd still rather have that procedure done by a competent, consistent machine than by a person. The day is fast approaching when you will sigh with relief to see ASIMO or a drone busboy—not a messy, fallible, inattentive human being—headed toward your table in a restaurant.