Monday, May 18, 2015

Soon They’ll Be Driving It, Too

Soon They’ll Be Driving It, Too

Intelligent machines are ousting low-skilled workers now. Next they’ll start encroaching on white-collar livelihoods.

[We Americans can be pretty good workers, too]

By Sumit Paul-Choudhury in the Wall Street Journal

Should you be worried by the emergence of intelligent machines? To some the answer is clear. “Full artificial intelligence could spell the end of the human race,” Stephen Hawking warned recently. Martin Ford’s “Rise of the Robots” offers a more prosaic reason for concern: Partially intelligent machines might render humans not so much extinct as redundant. “No one doubts that technology has the power to devastate entire industries and upend specific sectors of the economy and job market,” writes Mr. Ford, a Silicon Valley software developer turned futurist. Will machine intelligence, tackling tasks once thought of as humanity’s exclusive preserve, “disrupt our entire system to the point where a fundamental restructuring may be required if prosperity is to continue?”
Mr. Ford invokes Norbert Wiener, who in 1949 prophesied an “industrial revolution of unmitigated cruelty” in which machines would outstrip humans in routine work “at any price.” In Mr. Ford’s view, just such a revolution is under way in blue-collar work. Robots are ousting low-skilled workers everywhere, from fast-food joints to factory floors—a trend that Mr. Ford argues is central to the puzzling “jobless recovery” of the past decade as well as to other anomalous trends in pay and employment.
Now the machines are encroaching on white-collar livelihoods, which is why the intelligentsia have begun to wake up to their advance. To date, most automation has been of routine tasks that are relatively easy to describe in terms of simple instructions. But the combination of ever faster processors, ever smarter algorithms and ever bigger data is yielding supercomputers that are ever more capable of tackling complex challenges. IBM ’s Watson, having triumphed over human champion Ken Jennings at “Jeopardy!,” is now turning to medicine and cookery. Other machines are proving their mettle in fields ranging from scientific research to the stock market. Creativity no longer seems an insurmountable obstacle: Computers are starting to compose music or create paintings that could pass for the work of humans.
We are still a long way from all-round human intelligence—smart machines are becoming more flexible but still tend to excel in only a specific area—but Mr. Ford lucidly sets out myriad examples of how focused applications of versatile machines (coupled with human helpers where necessary) could displace or de-skill many jobs. If you are of the professional classes, you will likely read with mounting dismay Mr. Ford’s compelling explanation of how tools that encapsulate “analytic intelligence and institutional knowledge” will enable less qualified rivals to carry out your job proficiently, quite possibly from another country. An intelligent system might mine huge corporate data sets to distill years of experience into simple instructions for an overseas worker—who can then use translation and telepresence to overcome linguistic and geographical barriers. When the tools systems have become smart enough, those offshore workers may in turn be deemed surplus: In a particularly dastardly move, computers may even acquire those smarts by spying on their human users.
The author is persuasive in his discussion of the business logic that makes this process seem all but inevitable. Machines may be less accomplished than humans, but they are often cheaper, more dependable and more docile. While you might worry about their growing abilities, it is the economic incentives that seem truly problematic. Mr. Ford worries that if this trend runs away it will prove bad for all but the ultra-wealthy capitalists who own the machines. Because workers are consumers too, a declining workforce translates into declining demand, and that threatens the entire edifice of modern capitalism. Continue as we are, he suggests, and we may return to feudalism.
Will we? Why should this time be any different from previous waves of automation, in which displaced workers have moved, after some initial disorientation, to satisfactory new jobs? Machine intelligence, says Mr. Ford, is a general-purpose technology with broad applications: There will be few untouched fields to which workers can turn in their search for employment. Still, his copious examples, striking though they are, add up to no more than strong circumstantial evidence for that claim.
We should always be skeptical about the difficulty of transferring polished theories into unruly reality. And for the moment, there will remain bastions of human exceptionalism. One recent analysis suggests that “highly creative” work (including architecture, design and entertainment), which accounts for around a fifth of U.S. jobs, will prove intransigent. Mr. Ford also dedicates chapters to the ways in which the health-care and educational sectors have resisted automation.
Could we find new jobs in these areas for those put out of work by automation? The author’s short answer is that we can’t. Those at the bottom of the labor pyramid aren’t capable of doing jobs higher up it, and there wouldn’t be enough of those jobs anyway. Rather surprisingly, he gives only passing treatment to the potential deployment of intelligent machines to up-skill workers. “For the majority of people who lose middle-class jobs, access to a smart phone may offer little beyond the ability to play Angry Birds while waiting in the unemployment line,” he writes. Today’s smartphones, yes; but tomorrow’s smarter phones may enhance their owners’ reach and abilities in more productive ways.
The author’s apparent reluctance to engage with technological solutions to a technological problem perhaps reveals where his true object lies. His answer to a sharp decline in employment is a guaranteed basic income, a safety net that he suggests would both cushion the effect on the newly unemployable and encourage entrepreneurship among those creative enough to make a new way for themselves. This is a drastic prescription for the ills of modern industrialization—ills whose severity and very existence are hotly contested. “Rise of the Robots” provides a compelling case that they are real, even if its more dire predictions are harder to accept.

— Mr. Paul-Choudhury is the editor of New Scientist.

No comments: