The Inevitable

by Kevin Kelly

  This new mode of being—surfing the waves, diving down, rushing up, flitting from bit to bit, tweeting and twittering, ceaselessly dipping into newness with ease, daydreaming, questioning each and every fact—is not a bug. It is a feature. It is a proper response to the ocean of data, news, and facts flooding us. We need to be fluid and agile, flowing from idea to idea, because that fluidity reflects the turbulent informational environment surrounding us. This mode is neither a lazy failure nor an indulgent luxury. It is a necessity in order to thrive. To steer a kayak on white-water rapids you need to be paddling at least as fast as the water runs, and to hope to navigate the exabytes of information, change, disruption coming at us, you need to be flowing as fast as the frontier is flowing.

  But don’t confuse this flux for the shallows. Fluidity and interactivity also allow us to instantly divert more attention to works that are far more complex, bigger, and more complicated than ever before. Technologies that provided audiences with the ability to interact with stories and news—to time shift, play later, rewind, probe, link, save, clip, cut and paste—enabled long forms as well as short forms. Film directors started creating motion pictures that were not a series of sitcoms, but a massive sustained narrative that took years to tell. These vast epics, like Lost, Battlestar Galactica, The Sopranos, Downton Abbey, and The Wire, had multiple interweaving plotlines, multiple protagonists, and an incredible depth of characters, and these sophisticated works demanded sustained attention that was not only beyond previous TV and 90-minute movies, but would have shocked Dickens and other novelists of yore. Dickens would have marveled back then: “You mean the audience could follow all that, and then want more? Over how many years?” I would never have believed myself capable of enjoying such complicated stories, or caring about them enough to put in the time. My attention has grown. In a similar way the depth, complexity, and demands of video games can equal the demands of marathon movies or any great book. Just to become proficient in some games takes 50 hours.

  But the most important way these new technologies are changing how we think is that they have become one thing. It may appear as if you are spending endless nanoseconds on a series of tweets, and infinite microseconds surfing between web pages, or hours wandering between YouTube channels, and then hovering only mere minutes on one book snippet after another, when you finally turn back to your spreadsheet at work or flick through the screen of your phone. But in reality you are spending 10 hours a day paying attention to one intangible thing. This one machine, this one huge platform, this gigantic masterpiece is disguised as a trillion loosely connected pieces. The unity is easy to miss. The well-paid directors of websites, the hordes of commenters online, and the movie moguls reluctantly letting us stream their movies—these folks don’t believe they are mere data points in a big global show, but they are. When we enter any of the 4 billion screens lit today, we are participating in one open-ended question. We are all trying to answer: What is it?

  The computer manufacturer Cisco estimates that there will be 50 billion devices on the internet by 2020, in addition to tens of billions of screens. The electronics industry expects a billion wearable devices in five years, tracking our activities, feeding data into the stream. We can expect another 13 billion appliances, like the Nest thermostat, animating our smarthomes. There will be 3 billion devices built into connected cars. And 100 billion dumb RFID chips embedded into goods on the shelves of Walmart. This is the internet of things, the emerging dreamland of everything we manufacture that is the new platform for the improbable. It is built with data.

  Knowledge, which is related, but not identical, to information, is exploding at the same rate as information, doubling every two years. The number of scientific articles published each year has been accelerating even faster than this for decades. Over the last century the annual number of patent applications worldwide has risen in an exponential curve.

  We know vastly more about the universe than we did a century ago. This new knowledge about the physical laws of the universe has been put to practical use in such consumer goods as GPS and iPods, with a steady increase in our own lifespans. Telescopes, microscopes, fluoroscopes, oscilloscopes allowed us to see in new ways, and when we looked with new tools, we suddenly gained many new answers.

  Yet the paradox of science is that every answer breeds at least two new questions. More tools, more answers, ever more questions. Telescopes, radioscopes, cyclotrons, atom smashers expanded not only what we knew, but birthed new riddles and expanded what we didn’t know. Previous discoveries helped us to recently realize that 96 percent of all matter and energy in our universe is outside of our vision. The universe is not made of the atoms and heat we discovered last century; instead it is primarily composed of two unknown entities we label “dark”: dark energy and dark matter. “Dark” is a euphemism for ignorance. We really have no idea what the bulk of the universe is made of. We find a similar proportion of ignorance if we probe deeply into the cell, or the brain. We don’t know nothin’ relative to what could be known. Our inventions allow us to spy into our ignorance. If knowledge is growing exponentially because of scientific tools, then we should be quickly running out of puzzles. But instead we keep discovering greater unknowns.

  Thus, even though our knowledge is expanding exponentially, our questions are expanding exponentially faster. And as mathematicians will tell you, the widening gap between two exponential curves is itself an exponential curve. That gap between questions and answers is our ignorance, and it is growing exponentially. In other words, science is a method that chiefly expands our ignorance rather than our knowledge.

  We have no reason to expect this to reverse in the future. The more disruptive a technology or tool is, the more disruptive the questions it will breed. We can expect future technologies such as artificial intelligence, genetic manipulation, and quantum computing (to name a few on the near horizon) to unleash a barrage of new huge questions—questions we could have never thought to ask before. In fact, it’s a safe bet that we have not asked our biggest questions yet.

  * * *

  • • •

  Every year humans ask the internet 2 trillion questions, and every year the search engines give back 2 trillion answers. Most of those answers are pretty good. Many times the answers are amazing. And they are free! In the time before instant free internet search, the majority of the 2 trillion questions could not have been answered for any reasonable cost. Of course, while the answers may be free to users, they do cost the search companies like Google, Yahoo!, Bing, and Baidu something to create. In 2007, I calculated the cost to Google to answer one query to be approximately 0.3 cents, which has probably decreased a bit since then. By my calculations Google earns about 27 cents per search/answer from the ads placed around its answers, so it can easily afford to give its answers away for free.

  We’ve always had questions. Thirty years ago the largest answering business was phone directory assistance. Before Google, there was 411. The universal “information” number 411 was dialed from phones about 6 billion times per year. The other search mechanism in the past was the yellow pages—the paper version. According to the Yellow Pages Association, 50 percent of American adults used the print yellow pages at least once a week, performing two lookups per week in the 1990s. Since the adult population in the 1990s was around 200 million, that’s 200 million searches per week, or 104 billion questions asked per year. Nothing to sneeze at. The other classic answer strategy was the library. U.S. libraries in the 1990s counted about 1 billion library visits per year. Out of those 1 billion, about 300 million were “reference transactions,” or questions.

  Despite those 100 billion–plus searches for answers per year (in the U.S. alone), no one would have believed 30 years ago that there was an $82 billion business in answering people’s questions for cheap or for free. There weren’t many MBAs dreaming of schemes to fill this need. The demand for questions/answers was latent. People didn’t know how valu
able instant answers were until they had access to them. One study conducted in 2000 determined that the average American adult sought to answer four questions per day online. If my own life is any indication, I am asking more questions every day. Google told me that in 2007 I asked it 349 questions in one month, or 10 per day (and my peak hour of inquiry was 11 a.m. on Wednesdays). I asked Google how many seconds in a year and it instantly told me: 31.5 million. I asked it how many searches all search engines do per second? It said 600,000 searches per second, or 600 kilohertz. The internet is answering questions at the buzzing frequency of radio waves.

  But while answers are provided for free, the value of those answers is huge. Three researchers at the University of Michigan performed a small experiment in 2010 to see if they could ascertain how much ordinary people might pay for search. Their method was to ask students inside a well-stocked university library to answer some questions that were asked on Google, but to find the answers only using the materials in the library. They measured how long it took the students to answer a question in the stacks. On average it took 22 minutes. That’s 15 minutes longer than the 7 minutes it took to answer the same question, on average, using Google. Figuring a national average wage of $22 per hour, this works out to a savings of $5.50 per search.

  In 2011, Hal Varian, the chief economist at Google, calculated the average value of answering a question in a different way. He revealed the surprising fact that the average user of Google (judged by returning cookies, etc.) makes only one search per day, on average. This is certainly not me. But my near constant googling is offset by, say, my mother, who may search only once every several weeks. Varian did some more math to compensate for the fact that because questions are now cheap we ask more of them. So when this effect is factored in, Varian calculated that search saves the average person 3.75 minutes per day. Using the same average hourly wage, people save 60 cents per day. We could even round that off to a dollar per day if your time is more valuable. Would most people pay a dollar per day, or $350 per year, for search if they had to? Maybe. (I absolutely would.) They might pay a dollar per search, which is another way of paying the same amount. Economist Michael Cox asked his students how much they would accept to give up the internet entirely and reported they would not give up the internet for a million dollars. And this was before smartphones became the norm.

  We are just starting to get good at giving great answers. Siri, the audio phone assistant for the iPhone, delivers spoken answers when you ask her a question in natural English. I use Siri routinely. When I want to know the weather, I just ask, “Siri, what’s the weather for tomorrow?” Android folks can audibly ask Google Now for information about their calendars. IBM’s Watson proved that for most kinds of factual reference questions, an AI can find answers fast and accurately. Part of the increasing ease in providing answers lies in the fact that past questions answered correctly increase the likelihood of another question. At the same time, past correct answers increase the ease of creating the next answer, and increase the value of the corpus of answers as a whole. Each question we ask a search engine and each answer we accept as correct refines the intelligence of the process, increasing the engine’s value for future questions. As we cognify more books and movies and the internet of things, answers become ubiquitous. We are headed to a future where we will ask several hundred questions per day. Most of these questions will concern us and our friends. “Where is Jenny? What time is the next bus? Is this kind of snack good?” The “manufacturing costs” of each answer will be nanocents. Search, as in “give me an answer,” will no longer be considered a first-world luxury. It will become an essential universal commodity.

  Very soon now we’ll live in a world where we can ask the cloud, in conversational tones, any question at all. And if that question has a known answer, the machine will explain it to us. Who won the Rookie of the Year Award in 1974? Why is the sky blue? Will the universe keep expanding forever? Over time the cloud, or Cloud, the machine, or AI, will learn to articulate what is known and not known. At first it may need to engage us in a dialog to clarify ambiguities (as we humans do when answering questions), but, unlike us, the answer machine will not hesitate to provide deep, obscure, complex factual knowledge on any subject—if it exists.

  But the chief consequence of reliable instant answers is not a harmony of satisfaction. Abundant answers simply generate more questions! In my experience, the easier it is to ask a question and the more useful the reply, the more questions I have. While the answer machine can expand answers infinitely, our time to form the next question is very limited. There is an asymmetry in the work needed to generate a good question versus the work needed to absorb an answer. Answers become cheap and questions become valuable—the inverse of the situation now. Pablo Picasso brilliantly anticipated this inversion in 1964 when he told the writer William Fifield, “Computers are useless. They only give you answers.”

  So at the end of the day, a world of supersmart ubiquitous answers encourages a quest for the perfect question. What makes a perfect question? Ironically, the best questions are not questions that lead to answers, because answers are on their way to becoming cheap and plentiful. A good question is worth a million good answers.

  A good question is like the one Albert Einstein asked himself as a small boy—“What would you see if you were traveling on a beam of light?” That question launched the theory of relativity, E=MC2, and the atomic age.

  A good question is not concerned with a correct answer.

  A good question cannot be answered immediately.

  A good question challenges existing answers.

  A good question is one you badly want answered once you hear it, but had no inkling you cared before it was asked.

  A good question creates new territory of thinking.

  A good question reframes its own answers.

  A good question is the seed of innovation in science, technology, art, politics, and business.

  A good question is a probe, a what-if scenario.

  A good question skirts on the edge of what is known and not known, neither silly nor obvious.

  A good question cannot be predicted.

  A good question will be the sign of an educated mind.

  A good question is one that generates many other good questions.

  A good question may be the last job a machine will learn to do.

  A good question is what humans are for.

  * * *

  • • •

  What is it that we are making with our question-and-answer machine?

  Our society is moving away from the rigid order of hierarchy toward the fluidity of decentralization. It is moving from nouns to verbs, from tangible products to intangible becomings. From fixed media to messy remixed media. From stores to flows. And the value engine is moving from the certainties of answers to the uncertainties of questions. Facts, order, and answers will always be needed and useful. They are not going away, and in fact, like microbial life and concrete materials, facts will continue to underpin the bulk of our civilization. But the most precious aspects, the most dynamic, most valuable, and most productive facets of our lives and new technology will lie in the frontiers, in the edges where uncertainty, chaos, fluidity, and questions dwell. The technologies of generating answers will continue to be essential, so much that answers will become omnipresent, instant, reliable, and just about free. But the technologies that help generate questions will be valued more. Question makers will be seen, properly, as the engines that generate the new fields, new industries, new brands, new possibilities, new continents that our restless species can explore. Questioning is simply more powerful than answering.

  12

  BEGINNING

  Thousands of years from now, when historians review the past, our ancient time here at the beginning of the third millennium will be seen as an amazing moment. This is the time when inhabitants of this planet first linked t
hemselves together into one very large thing. Later the very large thing would become even larger, but you and I are alive at that moment when it first awoke. Future people will envy us, wishing they could have witnessed the birth we saw. It was in these years that humans began animating inert objects with tiny bits of intelligence, weaving them into a cloud of machine intelligences and then linking billions of their own minds into this single supermind. This convergence will be recognized as the largest, most complex, and most surprising event on the planet up until this time. Braiding nerves out of glass, copper, and airy radio waves, our species began wiring up all regions, all processes, all people, all artifacts, all sensors, all facts and notions into a grand network of hitherto unimagined complexity. From this embryonic net was born a collaborative interface for our civilization, a sensing, cognitive apparatus with power that exceeded any previous invention. This megainvention, this organism, this machine—if we want to call it that—subsumes all the other machines made, so that in effect there is only one thing that permeates our lives to such a degree that it becomes essential to our identity. This very large thing provides a new way of thinking (perfect search, total recall, planetary scope) and a new mind for an old species. It is the Beginning.

  The Beginning is a century-long process, and its muddling forward is mundane. Its big databases and extensive communications are boring. Aspects of this dawning real-time global mind are either dismissed as nonsense or feared. There is indeed a lot to be legitimately worried about because there is not a single aspect of human culture—or nature—that is left untouched by this syncopated pulse. Yet because we are the parts of something that has begun operating at a level above us, the outline of this emerging very large thing is obscured. All we know is that from its very beginning, it is upsetting the old order. Fierce pushback is to be expected.