The robots are coming
The Washington Post, January 9, 2013
"The Robots Take Over!" cries this month's cover story in Wired Magazine. "They're coming for your job—and you'll be glad they did." The piece, by Kevin Kelly, chronicles the amazing roles robotic technology is poised to take on—from warehouse worker and waitress to artist, musician, therapist and even comedian. It trumpets the news with the breezy optimism that characterizes most talk of technology's impact on society.
And why not be breezy? One of the core tenets of economics is that technological advance is the wellspring of human betterment. Yes, new technologies disrupt old arrangements and devastate industries and workers they displace. But, over time, such innovation spawns new industries and jobs whose scale vastly exceeds the losses suffered by technology's "losers."
It's the nature of economic change. The Luddites, as we learned in school, were wrong. Or, to put it more precisely, no one could blame them for fighting the machines that eliminated their skilled textile jobs. But the broader notion—that technological advance could destroy more jobs than it creates or cause widespread economic harm—is a fallacy.
Well, what if the Luddite Fallacy was a fallacy only for the first 250 years of modern capitalism's existence? What if we're entering an era of geometrically accelerating technological advance in which artificial intelligence, robotics and nanotechnology will together pose much more profound threats to jobs, wages and social stability than has commonly been imagined? And what if it's not just the "unskilled" who are at risk, but most of us?
That's the unsettling scenario sketched by computer engineer and software entrepreneur Martin Ford in his cautionary book, "The Lights in The Tunnel: Automation, Accelerating Technology, and The Economy of the Future." It's the most provocative 2009 book you'll read in 2013.
"Tunnel" had been on my list (well, on my Kindle) for ages, but I finally got to it over the holidays—doesn't everyone like a little dystopia with their egg nog? And be warned: You need to press past Ford's offputting "tunnel" metaphor to get to the guts of his unconventional analysis.
But if (like me) you're fascinated by futurist Ray Kurzweil's arguments that accelerating technology makes this unfolding era truly different, Ford's logic, and fears, will haunt you—and seem impossible to rule out. Is mass replacement of human work without the simultaneous creation of enough decently paid new work going to happen? If so, is the troubling inflection point 75 years away? Or two or three decades? If it's the latter, what should we be doing about it?
"The central thesis of this book," Ford writes, "is that, as technology accelerates, machine automation may ultimately penetrate the economy to the extent that wages no longer provide the bulk of consumers with adequate discretionary income and confidence in the future. If this issue is not addressed, the result will be a downward economic spiral."
In essence, Ford is hypothesizing that Marx may just turn out to have been a little ahead of his time when he talked about capitalism's "contradictions." Eventually capital will concentrate in fewer and fewer hands (in tomorrow's case, the robot owners'), and surging unemployment will combine with sagging wages to undermine the mass markets capitalism requires in order to function.
To be sure, every generation has seen such dark forecasts come to naught—but only this generation is living through the geometrical technological acceleration that the digital revolution (via Moore's law) has ushered in.
"The alarm has been raised, but so far the wolf has not shown up," Ford writes. "Does that really mean the wolf is only an illusion?"
In a phone interview Tuesday, Ford emphasized that the pace of technological change is a much bigger force for disruption than globalization—yet it's the latter that generates the ink and the fears. It's wrong to think that only less-skilled workers are at risk, Ford adds; it's much easier to automate a radiologist's job than a housekeeper's. For this reason, the idea that the policy answer is "education and training" strikes him as self-evidently flawed. Yet less-skilled workers will have no haven either, as Foxconn's recent order of 1 million robots for its low-wage Chinese factories proves.
In the end, Ford says, if something like the scenario he sketches comes to pass, capitalism's salvation will require that mass consumption not depend on income from work. What does that mean exactly? It means government redistribution on a scale that today seems as unthinkable as does the economy-wide automation of jobs he foresees.
Is Ford's scenario worth losing sleep over? Clearly it's not imminent, as Lawrence Summers, former economic adviser to President Obama, pointed out in an interview, because if we were in the situation Ford describes, productivity (i.e., output per human worker) would be rising dramatically, and it's not. Economist Erik Brynjolfsson at MIT, co-author of "Race Against The Machine," is looking at the same trends with a more hopeful sense of our prospects. Yet Jaron Lanier, who helped create virtual reality in the 1980s, and who authored the culture of technology critique, "You Are Not A Gadget," has an important book coming out this spring that will raise deep questions about technology's threat to the middle class. The New York Times' Paul Krugman, meanwhile, has tiptoed toward these questions (and the distributional issues they raise) in a few recent commentaries.
As for me? When I hear Kurzweil talk about the Singularity, I think, boy, if he's right—and we're going to essentially merge with machines and transcend human biology a few decades from now—what do I tell my 15-year-old daughter about how to live?
I feel the same way about Ford's vision. It's important to consider. And unsettling. I'm not sure what the odds are. I see signs out there that make me worry we're heading toward something like what he sees. And I'd like more top technologists and economists to do some hard thinking together about it now. Over to you, Paul Krugman.