Ended: Nov. 21, 2016
We are morphing so fast that our ability to invent new things outpaces the rate we can civilize them. These days it takes us a decade after a technology appears to develop a social consensus on what it means and what etiquette we need to tame it. In another five years we’ll find a polite place for twittering, just as we figured out what to do with cell phones ringing everywhere. (Use silent vibrators.) Just like that, this initial response will disappear quickly and we’ll see it was neither essential nor inevitable.
Our first impulse when we confront extreme technology surging forward in this digital sphere may be to push back. To stop it, prohibit it, deny it, or at least make it hard to use. (As one example, when the internet made it easy to copy music and movies, Hollywood and the music industry did everything they could to stop the copying. To no avail. They succeeded only in making enemies of their customers.) Banning the inevitable usually backfires. Prohibition is at best temporary, and in the long counterproductive.
A vigilant, eyes-wide-open embrace works much better. My intent in this book is to uncover the roots of digital change so that we can embrace them. Once seen, we can work with their nature, rather than struggle against it. Massive copying is here to stay. Massive tracking and total surveillance is here to stay. Ownership is shifting away. Virtual reality is becoming real. We can’t stop artificial intelligences and robots from improving, creating new businesses, and taking our current jobs. It may be against our initial impulse, but we should embrace the perpetual remixing of these technologies. Only by working with these technologies, rather than trying to thwart them, can we gain the best of what they have to offer.
We need to civilize and tame new inventions in their particulars. But we can do that only with deep engagement, firsthand experience, and a vigilant acceptance. We can and should regulate Uber-like taxi services, as an example, but we can’t and shouldn’t attempt to prohibit the inevitable decentralization of services. These technologies are not going away.
Our greatest invention in the past 200 years was not a particular gadget or tool but the invention of the scientific process itself. Once we invented the scientific method, we could immediately create thousands of other amazing things we could have never discovered any other way. This methodical process of constant change and improvement was a million times better than inventing any particular product, because the process generated a million new products over the centuries since we invented it. Get the ongoing process right and it will keep generating ongoing benefits. In our new era, processes trump products. This shift
Particular technological processes will inherently favor particular outcomes. For instance, industrial processes (like steam engines, chemical plants, dams) favor temperatures and pressures outside of human comfort zones, and digital technologies (computers, internet, apps) favor cheap ubiquitous duplication. The bias toward high pressure/high temperature for industrial processes steers places of manufacturing away from humans and toward large-scale, centralized factories, regardless of culture, background, or politics. The bias toward cheap ubiquitous copies in digital technologies is independent of nationality, economic momentum, or human desire, and it steers the technology toward social ubiquity; the bias is baked into the nature of digital bits.
world without discomfort is utopia. But it is also stagnant. A world perfectly fair in some dimensions would be horribly unfair in others. A utopia has no problems to solve, but therefore no opportunities either.
However, neither dystopia nor utopia is our destination. Rather, technology is taking us to protopia. More accurately, we have already arrived in protopia. Protopia is a state of becoming, rather than a destination. It is a process. In the protopian mode, things are better today than they were yesterday, although only a little better. It is incremental improvement or mild progress. The “pro” in protopian stems from the notions of process and progress. This subtle progress is not dramatic, not exciting. It is easy to miss because a protopia generates almost as many new problems as new benefits. The problems of today were caused by yesterday’s technological successes, and the technological solutions to today’s problems will cause the problems of tomorrow. This circular expansion of both problems and solutions hides a steady accumulation of small net benefits over time. Ever since the Enlightenment and the invention of science, we’ve managed to create a tiny bit more than we’ve destroyed each year. But that few percent positive difference is compounded over decades into what we might call civilization. Its benefits never star in movies.
The fear of commercialization was strongest among hard-core programmers who were actually building the web: the coders, Unix weenies, and selfless volunteer IT folk who kept the ad hoc network running. The techy administrators thought of their work as noble, a gift to humanity. They saw the internet as an open commons, not to be undone by greed or commercialization. It’s hard to believe now, but until 1991 commercial enterprise on the internet was strictly prohibited as an unacceptable use. There was no selling, no ads. In the eyes of the National Science Foundation (which ran the internet backbone), the internet was funded for research, not commerce. In what seems remarkable naiveté now, the rules favored public institutions and forbade “extensive use for private or personal business.” In the mid-1980s I was involved in shaping the WELL, an early text-only online system. We struggled to connect our private WELL network to the emerging internet because we were thwarted, in part, by the NSF’s “acceptable use” policy. The WELL couldn’t prove its users would not conduct commercial business on the internet, so we were not allowed to connect. We were all really blind to what was becoming.
What we all failed to see was how much of this brave new online world would be manufactured by users, not big institutions. The entirety of the content offered by Facebook, YouTube, Instagram, and Twitter is not created by their staff, but by their audience. Amazon’s rise was a surprise not because it became an “everything store” (not hard to imagine), but because Amazon’s customers (me and you) rushed to write the reviews that made the site’s long-tail selection usable. Today, most major software producers have minimal help desks; their most enthusiastic customers advise and assist other customers on the company’s support forum web pages, serving as high-quality customer support for new buyers. And in the greatest leverage of the common user, Google turns traffic and link patterns generated by 90 billion searches a month into the organizing intelligence for a new economy. This bottom-up overturning was also not in anyone’s 20-year vision.
This apparently primeval impulse for participation has upended the economy and is steadily turning the sphere of social networking—smart mobs, hive minds, and collaborative action—into the main event.
People in the future will look at their holodecks and wearable virtual reality contact lenses and downloadable avatars and AI interfaces and say, “Oh, you didn’t really have the internet”—or whatever they’ll call it—“back then.”
fruit—the equivalent of the dot-com names of 1984. Because here is the other thing the graybeards in 2050 will tell you: Can you imagine how awesome it would have been to be an innovator in 2016? It was a wide-open frontier! You could pick almost any category and add some AI to it, put it on the cloud. Few devices had more than one or two sensors in them, unlike the hundreds now. Expectations and barriers were low. It was easy to be the first. And then they would sigh. “Oh, if only we realized how possible everything was back then!”
So, the truth: Right now, today, in 2016 is the best time to start up. There has never been a better day in the whole history of the world to invent something. There has never been a better time with more opportunities, more openings, lower barriers, higher benefit/risk ratios, better returns, greater upside than now. Right now, this minute. This is the moment that folks in the future will look back at and say, “Oh, to have been alive and well back then!” The last 30 years has created a marvelous starting point, a solid platform to build truly great things. But what’s coming will be different, beyond, and other. The things we will make will be constantly, relentlessly becoming something else. And the coolest stuff of all has not been invented yet.
It is hard to imagine anything that would “change everything” as much as cheap, powerful, ubiquitous artificial intelligence. To begin with, there’s nothing as consequential as a dumb thing made smarter. Even a very tiny amount of useful intelligence embedded into an existing process boosts its effectiveness to a whole other level. The advantages gained from cognifying inert things would be hundreds of times more disruptive to our lives than the transformations gained by industrialization.
IBM provides Watson’s medical intelligence to partners like CVS, the retail pharmacy chain, helping it develop personalized health advice for customers with chronic diseases based on the data CVS collects. “I believe something like Watson will soon be the world’s best diagnostician—whether machine or human,” says Alan Greene, chief medical officer of Scanadu, a startup that is building a diagnostic device inspired by the Star Trek medical tricorder and powered by a medical AI. “At the rate AI technology is improving, a kid born today will rarely need to see a doctor to get a diagnosis by the time they are an adult.”
They did not teach it how to play the games, but how to learn to play the games—a profound difference. They simply turned their cloud-based AI loose on an Atari game such as Breakout, a variant of Pong, and it learned on its own how to keep increasing its score. A video of the AI’s progress is stunning. At first, the AI plays nearly randomly, but it gradually improves. After a half hour it misses only once every four times. By its 300th game, an hour into it, it never misses. It keeps learning so fast that in the second hour it figures out a loophole in the Breakout game that none of the millions of previous human players had discovered. This hack allowed it to win by tunneling around a wall in a way that even the game’s creators had never imagined. At the end of several hours of first playing a game, with no coaching from the DeepMind creators, the algorithms, called deep reinforcement machine learning, could beat humans in half of the 49 Atari video games they mastered. AIs like this one
Three generations ago, many a tinkerer struck it rich by taking a tool and making an electric version. Take a manual pump; electrify it. Find a hand-wringer washer; electrify it. The entrepreneurs didn’t need to generate the electricity; they bought it from the grid and used it to automate the previously manual. Now everything that we formerly electrified we will cognify. There is almost nothing we can think of that cannot be made new, different, or more valuable by infusing it with some extra IQ. In fact, the business plans of the next 10,000 startups are easy to forecast: Take X and add AI. Find something that can be made better by adding online smartness to it.
The more unlikely the field, the more powerful adding AI will be. Cognified investments? Already happening with companies such as Betterment or Wealthfront. They add artificial intelligence to managed stock indexes in order to optimize tax strategies or balance holdings between portfolios. These are the kinds of things a professional money manager might do once a year, but the AI will do every day, or every hour.
But Page’s reply has always stuck with me: “Oh, we’re really making an AI.” I’ve thought a lot about that conversation over the past few years as Google has bought 13 other AI and robotics companies in addition to DeepMind. At first glance, you might think that Google is beefing up its AI portfolio to improve its search capabilities, since search constitutes 80 percent of its revenue. But I think that’s backward. Rather than use AI to make its search better, Google is using search to make its AI better. Every time you type a query, click on a search-generated link, or create a link on the web, you are training the Google AI. When you type “Easter Bunny” into the image search bar and then click on the most Easter Bunny–looking image, you are teaching the AI what an Easter Bunny looks like. Each of the 3 billion queries that Google conducts each day tutors the deep-learning AI over and over again. With another 10 years of steady improvements to its AI algorithms, plus a thousandfold more data and a hundred times more computing resources, Google will have an unrivaled AI. In a quarterly earnings conference call in the fall of 2015, Google CEO Sundar Pichai stated that AI was going to be “a core transformative way by which we are rethinking everything we are doing. . . . We are applying it across all our products, be it search, be it YouTube and Play, etc.” My prediction: By 2026, Google’s main product will not be search but AI. This is the point
Digital neural nets were invented in the 1950s, but it took decades for computer scientists to learn how to tame the astronomically huge combinatorial relationships between a million—or a hundred million—neurons. The key was to organize neural nets into stacked layers. Take the relatively simple task of recognizing that a face is a face. When a group of bits in a neural net is found to trigger a pattern—the image of an eye, for instance—that result (“It’s an eye!”) is moved up to another level in the neural net for further parsing. The next level might group two eyes together and pass that meaningful chunk on to another level of hierarchical structure that associates it with the pattern of a nose. It can take many millions of these nodes (each one producing a calculation feeding others around it), stacked up to 15 levels high, to recognize a human face. In 2006, Geoff Hinton, then at the University of Toronto, made a key tweak to this method, which he dubbed “deep learning.” He was able to mathematically optimize results from each layer so that the learning accumulated faster as it proceeded up the stack of layers. Deep-learning algorithms accelerated enormously a few years later when they were ported to GPUs. The code of deep learning alone is insufficient to generate complex logical thinking, but it is an essential component of all current AIs, including IBM’s Watson; DeepMind,
In the next 10 years, 99 percent of the artificial intelligence that you will interact with, directly or indirectly, will be nerdly narrow, supersmart specialists. In fact, robust intelligence may be a liability—especially if by “intelligence” we mean our peculiar self-awareness, all our frantic loops of introspection and messy currents of self-consciousness. We want our self-driving car to be inhumanly focused on the road, not obsessing over an argument it had with the garage.
Our intelligence is a society of intelligences, and this suite occupies only a small corner of the many types of intelligences and consciousnesses that are possible in the universe. We like to call our human intelligence “general purpose,” because compared with other kinds of minds we have met, it can solve more types of problems, but as we build more and more synthetic minds we’ll come to realize that human thinking is not general at all. It is only one species of thinking.
Humans are for inventing new kinds of intelligences that biology could not evolve. Our job is to make machines that think different—to create alien intelligences. We should really call AIs “AAs,” for “artificial aliens.”
Imagine that seven out of ten working Americans got fired tomorrow. What would they all do? It’s hard to believe you’d have an economy at all if you gave pink slips to more than half the labor force. But that—in slow motion—is what the industrial revolution did to the workforce of the early 19th century. Two hundred years ago, 70 percent of American workers lived on the farm. Today automation has eliminated all but 1 percent of their jobs, replacing them (and their work animals) with machines. But the displaced workers did not sit idle. Instead, automation created hundreds of millions of jobs in entirely new fields.
By 2050 most truck drivers won’t be human. Since truck driving is currently the most common occupation in the U.S., this is a big deal.
While the displacement of formerly human jobs gets all the headlines, the greatest benefits bestowed by robots and automation come from their occupation of jobs we are unable to do. We don’t have the attention span to inspect every square millimeter of every CAT scan looking for cancer cells. We don’t have the millisecond reflexes needed to inflate molten glass into the shape of a bottle. We don’t have an infallible memory to keep track of every pitch in Major League baseball and calculate the probability of the next pitch in real time. We aren’t giving “good jobs” to robots. Most of the time we are giving them jobs we could never do. Without them, these jobs would remain undone.
Jobs We Didn’t Know We Wanted Done This is the greatest genius of the robot takeover: With the assistance of robots and computerized intelligence, we already can do things we never imagined doing 150 years ago. We can today remove a tumor in our gut through our navel, make a talking-picture video of our wedding, drive a cart on Mars, print a pattern on fabric that a friend mailed to us as a message through the air. We are doing, and are sometimes paid for doing, a million new activities that would have dazzled and shocked the farmers of 1800. These new accomplishments are not merely chores that were difficult before. Rather they are dreams created chiefly by the capabilities of the machines that can do them. They are jobs the machines make up.
The initial age of computing borrowed from the industrial age. As Marshall McLuhan observed, the first version of a new medium imitates the medium it replaces. The first commercial computers employed the metaphor of the office. Our screens had a “desktop” and “folders” and “files.” They were hierarchically ordered, like much of the industrial age that the computer was overthrowing.
The second digital age overturned the office metaphor and brought us the organizing principle of the web. The basic unit was no longer files but “pages.” Pages were not organized into folders, but were arranged into a networked web. The web was a billion hyperlinked pages which contained everything, both stored information and active knowledge. The desktop interface was replaced by a “browser,” a uniform window that looked into any and all pages. This web of links was flat.
Now we are transitioning into the third age of computation. Pages and browsers are far less important. Today the prime units are flows and streams. We constantly monitor Twitter streams and the flows of posts on our Facebook wall. We stream photos, movies, and music. News banners stream across the bottom of TVs. We subscribe to YouTube streams, called channels. And RSS feeds from blogs. We are bathed in streams of notifications and updates. Our apps improve in a flow of upgrades. Tags have replaced links. We tag and “like” and “favorite” moments in the streams. Some streams, like Snapchat, WeChat, and WhatsApp, operate totally in the present, with no past or future. They just flow past. If you see something, fine. Then it is gone.
A universal law of economics says the moment something becomes free and ubiquitous, its position in the economic equation suddenly inverts. When nighttime electrical lighting was new and scarce, it was the poor who burned common candles. Later, when electricity became easily accessible and practically free, our preference flipped and candles at dinner became a sign of luxury.
Deep down, avid audiences and fans want to pay creators. Fans love to reward artists, musicians, authors, actors, and other creators with the tokens of their appreciation, because it allows them to connect with people they admire. But they will pay only under four conditions that are not often met: 1) It must be extremely easy to do; 2) The amount must be reasonable; 3) There’s clear benefit to them for paying; and 4) It’s clear the money will directly benefit the creators.
The previous generatives resided within creative works. Discoverability, however, is an asset that applies to an aggregate of many works. No matter what its price, a work has no value unless it is seen. Unfound masterpieces are worthless. When there are millions of books, millions of songs, millions of films, millions of applications, millions of everything requesting our attention—and most of it free—being found is valuable.
Digital display manufacturers will crank out 3.8 billion new additional screens per year. That’s nearly one new screen each year for every human on earth. We
But to everyone’s surprise, the cool, interconnected, ultrathin screens on monitors, the new TVs, and tablets at the beginning of the 21st century launched an epidemic of writing that continues to swell. The amount of time people spend reading has almost tripled since 1980. By 2015 more than 60 trillion pages have been added to the World Wide Web, and that total grows by several billion a day.
Most of its 34 million pages are crammed with words underlined in blue, indicating those words are hyperlinked to concepts elsewhere in the encyclopedia. This tangle of relationships is precisely what gives Wikipedia—and the web—its immense force. Wikipedia is the first networked book. In the goodness of time, each Wikipedia page will become saturated with blue links as every statement is cross-referenced. In the goodness of time, as all books become fully digital, every one of them will accumulate the equivalent of blue underlined passages as each literary reference is networked within that book out to all other books. Each page in a book will discover other pages and other books. Thus books will seep out of their bindings and weave themselves together into one large metabook, the universal library. The resulting collective intelligence of this synaptically connected library allows us to see things we can’t see in a single isolated book.
The link and the tag may be two of the most important inventions of the last 50 years. You are anonymously marking up the web, making it smarter, when you link or tag something. These bits of interest are gathered and analyzed by search engines and AIs in order to strengthen the relationship between the end points of every link and the connections suggested by each tag. This type of intelligence has been indigenous to the web since its birth, but was previously foreign to the world of books. The link and the tag now make screening the universal library possible, and powerful.
Science is on a long-term campaign to bring all knowledge in the world into one vast, interconnected, footnoted, peer-reviewed web of facts. Independent facts, even those that make sense in their own world, are of little value to science. (The pseudo- and parasciences are nothing less, in fact, than small pools of knowledge that are not connected to the large network of science. They are valid only in their own network.) In this way, every new observation or bit of data brought into the web of science enhances the value of all other data points.
Over the next three decades, scholars and fans, aided by computational algorithms, will knit together the books of the world into a single networked literature. A reader will be able to generate a social graph of an idea, or a timeline of a concept, or a networked map of influence for any notion in the library. We’ll come to understand that no work, no idea stands alone, but that all good, true, and beautiful things are ecosystems of intertwined parts and related entities, past and present.
Libraries (as well as many individuals) aren’t eager to relinquish old-fashioned ink-on-paper editions, because the printed book is by far the most durable and reliable long-term storage technology we have.
Possession is not as important as it once was. Accessing is more important than ever.
On average most modern products have undergone dematerialization. Since the 1970s, the weight of the average automobile has fallen by 25 percent. Appliances tend to weigh less per function. Of course, communication technology shows the clearest dematerialization. Huge PC monitors shrunk to thin flat screens (but the width of our TVs expanded!), while clunky phones on the table become pocketable. Sometimes our products gain many new benefits without losing mass, but the general trend is toward products that use fewer atoms. We might not notice this because, while individual items use less material, we use more items as the economy expands and we thus accumulate more stuff in total. However, the total amount of material we use per GDP dollar is going down, which means we use less material for greater value. The ratio of mass needed to generate a unit of GDP has been falling for 150 years, declining even faster in the last two decades. In 1870 it took 4 kilograms of stuff to generate one unit of the U.S.’s GDP. In 1930 it took only one kilogram. Recently the value of GDP per kilogram of inputs rose from $1.64 in 1977 to $3.58 in 2000—a doubling of dematerialization in 23 years.
The reason even solid physical goods—like a soda can—can deliver more benefits while inhabiting less material is because their heavy atoms are substituted by weightless bits. The tangible is replaced by intangibles—intangibles like better design, innovative processes, smart chips, and eventually online connectivity—that do the work that more aluminum atoms used to do. Soft things, like intelligence, are thus embedded into hard things, like aluminum, that make hard things behave more like software. Material goods infused with bits increasingly act as if they were intangible services. Nouns morph to verbs. Hardware behaves like software. In Silicon Valley they say it like this: “Software eats everything.”
TV, phones, and software as service are just the beginning. In the last few years we’ve gotten hotels as service (Airbnb), tools as service (TechShop), clothes as service (Stitch Fix, Bombfell), and toys as service (Nerd Block, Sparkbox). Just ahead are several hundred new startups trying to figure how to do food as service (FaS). Each has its own approach to giving you a subscription to food, instead of purchases. For example, in one scheme you might not buy specific food products; instead, you get access to the benefits of food you need or want—say, certain levels and qualities of protein, nutrition, cuisine, flavors. Other possible new service realms: Furniture as service; Health as service; Shelter as service; Vacation as service; School as service. Of course, in all these you still pay; the difference is the deeper relationship that services encourage and require between the customer and the provider.
enough. You usually wait too long for one, including the ones you call. And the cumbersome payment procedure at the end is a hassle. Oh, and they should be cheaper. Uber, the on-demand taxi service, has disrupted the transportation business because it shifts the time equation. When you order a ride, you don’t need to tell Uber where you are; your phone does that. You don’t have to settle payment at the end; your phone does that. Uber uses the phones of the drivers to locate precisely where they are within inches, so Uber can match a driver closest to you. You can track their arrival to the minute. Anyone who wants to earn some money can drive, so there are often more Uber drivers than taxis, especially during peak demand times. And to make it vastly cheaper (in normal use), if you are willing to share a ride, Uber will match two or three riders going to approximately the same place at the same time to split the fare. These UberPool shared-ride fares might be one quarter the cost of a taxi. Relying on Uber (or its competitors, like Lyft) is a no-brainer.
In the past few years thousands of entrepreneurs seeking funding have pitched venture capitalists for an “Uber for X,” where X is any business where customers still have to wait. Examples of X include: three different Uber for flowers (Florist Now, ProFlowers, BloomThat), three Uber for laundry, two Uber for lawn mowing (Mowdo, Lawnly), an Uber for tech support (Geekatoo), an Uber for doctor house calls, and three Uber for legal marijuana delivery (Eaze, Canary, Meadow), plus a hundred more.
The Uber-like companies can promise this because, instead of owning a building full of employees, they own some software. All the work is outsourced and performed by freelancers (prosumers) ready to work. The job for Uber for X is to coordinate this decentralized work and make it happen in real time. Even Amazon has gotten into the business of matching pros with joes who need home services (Amazon Home Services), from cleaning or setting up equipment to access to goat grazing for lawns.
number of startups and venture capitalists are dreaming up ways to use blockchain technology as a general purpose trust mechanism beyond money. For transactions that require a high degree of trust between strangers, such as real estate escrows and mortgage contracts, this validation was previously provided by a professional broker. But instead of paying a traditional title company a lot of money to verify a complex transaction such as a house sale, an online peer-to-peer blockchain system can execute the exchange for much less cost, or maybe for free. Some blockchain enthusiasts propose creating tools that perform a complicated cascade of transactions that depend on verification (like an import/export deal) using only decentralized automated blockchain technology, thereby disrupting many industries that rely on brokers. Whether Bitcoin itself succeeds, its blockchain innovation, which can generate extremely high levels of trust among strangers, will further decentralize institutions and industries.
The wealthiest and most disruptive organizations today are almost all multisided platforms—Apple, Microsoft, Google, and Facebook. All these giants employ third-party vendors to increase the value of their platform. All employ APIs extensively that facilitate and encourage others to play with it. Uber, Alibaba, Airbnb, PayPal, Square, WeChat, Android are the newer wildly successful multiside markets, run by a firm, that enable robust ecosystems of derivative yet interdependent products and services.