Computers Can’t Do Party TricksPosted: June 29, 2016
“The horse has lost its role in transportation but has made a strong comeback in entertainment.”
My friend Brendan can glance at satellite photos and immediately identify their locations in the world, a skill that might have qualified him as a warlock or wizard hundreds of years ago (if satellite photos had existed then). Last week, he and I were trying to figure out whether he could somehow cash in on this skill or use it to advance his career. No, we quickly decided: This is the epitome of a task that computers have evolved to do better than people, whether anyone demands that particular service or not. But seeing Brendan identify Google Earth images in person is still an impressive feat, a great party trick.
That computers can recognize images more accurately than the most adept humans, and then instantly cross-reference the results with any other set of data, requires no explanation. It’s less certain but a true enough generalization that almost anything we call work today, machines will do eventually (as we concurrently keep inventing new things to call work). Technological progress not only eliminates the need for certain tasks in their traditional domains, thereby lowering those tasks’ economic value, but it also gradually eliminates—or amputates, as McLuhan said—the faculties certain people had developed to better perform those tasks. The printing press drastically reduced the need for hand copying as a means of reproducing texts; in the much longer run, it also led to a weakening of the widespread practice of writing by hand. We’ve seen the same phenomenon time after time: technology devaluing the job and amputating the corresponding skill.
But the automations and amputations of technology are anything but the end for what they supposedly replace. Six hundred years after Gutenberg, almost everyone literate can still write by hand. And handwriting isn’t an exception. Most of what past generations did for money is still being done somewhere, probably for less money and at a smaller scale—a reality only partially explained by William Gibson’s observation that the future is never evenly distributed after it arrives.
The world is weirder and more complex than macroeconomics usually recognizes. What doesn’t contribute to GDP still exists and often still matters. Many abilities that technology debases economically, like Brendan’s knack for recognizing satellite photos, reappear in a hundred other domains: as competitive sports, hobbies, artisanal crafts, tourist attractions, historical preservation objects—the cultural substrate that underpins the productive economy where quantifiable value is added. At the very least, they often make great party tricks. McLuhan observed that the horse thrived as a form of entertainment after becoming obsolete as a mode of transportation. To say that one technology “killed” another obscures the reality that it only fragmented its predecessor and may have actually made its presence in the world more interesting.
Computers can add value but they can’t do party tricks. No one at a party gives a shit what a computer can do. Everyone at a party cares what a person can do.
Computers can’t party by definition. But computers excel at being productive and can add more value than any human. One computer can do the work of thousands of erstwhile humans. The cutting edge of technology, always pressuring mankind to justify its existence, reminds us that the productive economy isn’t the best place to affirm our own humanity.
One useful predictor of technology’s evolution, the Varian Rule, states that what the rich do today is what everyone else will be doing in the future. This has held true for nearly every major development, from the printed word to electricity to trains to cars to air travel to personal computers to iPhones. A companion to the Varian Rule might be this: What technology renders obsolete today, the rich will eventually enjoy again. The resurgence of artisanal everything among the urban consumer class is certainly a response to the dominance of industrial food production, and of course rich people are the most likely to be found riding horses in modern society today.
The horse as entertainment (source)
Cars, predicted by the Varian Rule, were the most transformative technology of the hundred years preceding the computer. They began as an expensive luxury until Henry Ford brought them to the masses. GK Chesterton described the modern transformation of so many luxuries into binding necessities—a trend exemplified by the automobile—as a degrading force (“I say the thing should be kept exceptional and felt as something breathless and bizarre”) and there is no question that driving has by now become an inescapable chore for the majority of those who have to do it, as well as a scourge on the urban landscapes we inhabit.
The car as we know it might soon get its own chance to become obsolete as a form of transportation. The long-anticipated self-driving car sits at the precise (figurative) intersection of every trope about technological progress: automation (it will eliminate jobs), social benefit (it will save us from killing ourselves behind the wheel) and massively profitable business opportunity (it will drastically reduce the cost of transportation).
If we’re lucky, autonomous vehicles will deliver the human-driven car back to the domain of entertainment and luxury from whence it came, alongside the horse. Perhaps never having to drive would restore some magic and breathlessness to the act on the special occasions when we still do it. If we’re less fortunate, however, self-driving vehicles will further solidify the car culture in which we’ve been stuck for the past century and from which we need to escape. Right now, we fulfill two roles behind the wheel, driver and passenger, and only one of those promises to go away anytime soon so the outcome will be a mixed blessing at best.