Book Reviews

Too Much Information

When flawed algorithms meet naïve expectations

By Howard P. Segal | August 8, 2018
Flickr/BASF

The Efficiency Paradox: What Big Data Can’t Do by Edward Tenner; Knopf, 282 pp., $27.95

The systematic quest for greater efficiency, both at work and at home, began in the mid-18th century with the industrial revolution in Great Britain and has since become the central preoccupation of the modern world. Many thinkers have hailed efficiency’s benefits, including lower production costs and ease of distribution. The comparatively few to have offered serious critiques, such as Karl Marx and 20th-century American economist Eric E. Lampard, have pointed to the frequent exploitation of workers, both physically and psychologically, in efficiency’s name. To be sure, the blessings of the industrial revolution have been mixed—prosperity for many and misery for many more—but nuance does not often figure into the debate, long dominated by one-sided arguments, for or against efficiency.

Edward Tenner’s Efficiency Paradox acknowledges this complex legacy even as it reckons with a more recent era: the high-tech transformation of much of the world during the past five decades. He focuses not on the endless stream of mass-produced material goods that line the shelves of every big-box store but on the practice of “matchmaking,” made possible by artificial intelligence, in which digital algorithms use consumers’ past purchases to suggest new ones. Amazon.com exemplifies this approach. But the relentless pursuit of Big Data and of supposedly flawless algorithms has not, alas, fulfilled Silicon Valley’s vision of a technological utopia, or what Bill Gates and others in 1995 called a “friction-free” world. Instead, flawed algorithms, naïve expectations for what data might reveal about the rhythm of daily life, and obliviousness to the nontechnological conditions that often undermine supposedly scientific prophecies have posed obstacles to “progress.” As Tenner puts it, “Efficiency makes the world more predictable. But if everything is as direct as possible, we are also deprived of the benefits of occasional randomization and of productive mistakes.”

Tenner is hardly the first critic to analyze the consequences of our contemporary zeal for efficiency. Indeed, his two previous books—Why Things Bite Back: New Technology and the Revenge of Unintended Consequences (1996) and Our Own Devices: The Past and Future of Body Technology (2003)—covered similar terrain and deserve to be reread. The Efficiency Paradox, however, is the foremost study to date of efficiency’s unexpected outcomes. Other popular works, such as Ray Kurzweil’s The Singularity is Near (2005) and Kevin Kelly’s What Technology Wants (2010), make embarrassing ahistorical or pseudohistorical generalizations—above all, that little of importance occurred before 1970. A lone standout is historian Jennifer Karns Alexander’s The Mantra of Efficiency: From Waterwheel to Social Control (2008), which complements Tenner’s work. She demonstrates how the concept of efficiency changed, from a concept that applied only to machines to being a means to control human behavior.

Tenner is no neo-Luddite. He pays tribute to high-tech advances that enabled him to research and write his books, and embraces efficiency insofar as it produces the least possible waste while expending a minimum of resources—a rationale that harkens back to 19th- and early 20th-century obsessions with avoiding waste. Today, though, efficiency has come to mean accomplishing a task with the least possible human intervention—a goal that often turns out to be self-defeating, particularly when efficiency becomes almost an end in itself. Recall Thomas Edison’s famous line that genius is one percent inspiration and 99 percent perspiration and contrast it with our contemporary enthusiasm for machine-like efficiency. Reconciling these two seemingly incompatible ideas is Tenner’s task. As he notes, the children of many of Silicon Valley’s top engineers, scientists, and executives attend schools in which laptops, smartphones, and coding are either absent or are marginalized in favor of books, pens and pencils, and even knitting needles. Their parents, on the frontlines of technological innovation, presumably know better than anyone the costs of the increased efficiency they themselves have made possible. They could hardly be called neglectful.

As recent controversies over the misuse of websites like Facebook (by the now-defunct British firm Cambridge Analytica) have shown, manipulation of public opinion, above all during political campaigns, poses a threat to American and other democracies—a threat that transcends particular political parties, candidates, and officeholders. Yet the company’s employment of “psychographic microtargeting” to try to pinpoint the views and personalities of millions in order to influence elections is the kind of enterprise that Tenner finds almost delusional.

Here, as in his first two books, Tenner catalogs some of the unexpected consequences of blind faith in technology. No simple-minded “technological determinism” for him. His examples in this book include overdependence on GPS devices weakening traditional navigational skills for hikers and climbers; notetaking on mobile devices undermining traditional use of pen and pencil; and overdependence by pilots—and soon by ordinary car drivers—on computerized programs resulting in sometimes fatal “skill erosion.” Tenner, in his book and in subsequent interviews, also considers other unexpected outcomes, such as how consumer preference for ride-sharing services like Uber and Lyft over public transportation is leading to more traffic congestion, not less; how the biases (despite disclaimers) of search engines favor certain restaurants, hotels, or books; how overreliance on sleeping pills worsens insomnia; and how mobile devices, while reducing time spent in the office, allow for unprecedented monitoring and tethering of employees’ daily lives. Not to mention ever-longer professional baseball games: managers, coaches, and increasingly general managers rely on data to make time-consuming decisions about lineup changes during games, decisions once made instantaneously. Meanwhile new technologies increasingly used in cricket, tennis, and soccer to correct allegedly bad calls themselves make mistakes and undermine the traditional faith in experienced referees and umpires.

Tenner advocates what psychologists call “desirable difficulty”: working harder rather than relying completely on devices in order to assimilate information that, in the long run, may make us more efficient. He seeks to combine Big Data and algorithms with intuition and experience to produce the “right blend.” And somehow, despite his skepticism about technological panaceas, he remains optimistic.

With current controversies over unwanted access to our personal information through Facebook or breaches of corporate or government websites and databases, The Efficiency Paradox is a timely and welcome exploration of a desperately important subject. May there be more to follow.

Comments powered by Disqus