by Tom Wheeler
The flood of information made possible by a distributed network precipitated an unforeseen phenomenon: non-network-based centralization. The driving force behind this new recentralization is the network’s creation of a new capital asset: digital information. Previous networks carried assets to be put to work. Today’s networks create new assets by the very act of carriage. Facebook, for instance, rides on the distributed physical network to aggregate information about its users and then to perform an algorithm-based analysis of that information so as to target those users for advertisements and information.
History’s networks were “dumb”: their job was to haul the asset to a point where its value could be realized. Today’s networks comprise smart communicating computers that, by the very act of hauling information, create new information about usage and users. Every online activity leaves a digital footprint that is itself a valuable new piece of information to be bought and sold. It’s as if the Post Office kept a record of every piece of mail you sent or received, and then sold that information.
The capital asset of the nineteenth and twentieth centuries was industrial production facilitated by networks. The capital asset of the twenty-first century is information created by networks.
Our new networks are also new in their velocity. The speed of networks inherently drives the speed of change. The time between the age of steam and sparks and the age of zeros and ones was half as long as the time between the printing press and the railroad, thanks to the accelerated speed of the network that was driving change. A locomotive could deliver the benefits of Gutenberg’s product about ten times faster than when it traveled on horseback. Then the telegraph boosted that by another factor of ten. Today, tenfold increases in network speeds are puny artifacts.
Over the last few decades, network throughput has been increasing exponentially. During my tenure at the FCC, we established the threshold that defined “broadband” connectivity as 25 megabits per second (mbps)—representing an astounding eight-million-fold increase over the speed of the telegraph.4 Yet even that threshold was almost instantly outdated as new network construction at gigabit speeds (one billion bits per second) became common.
The “new new” underpinning our future is the replacement of networks functioning as a centralizing force with those that act as a decentralizing force. By pushing network activity outward, digital technology has changed the economics of networks, sped up connections (and thus the rate of change), and delivered a new collection of network-delivered challenges, including the emergence of the network-using, service-platform-based centralization of digital information.
The Capital Asset of the Twenty-First Century
The writers’ room at the Hollywood studio was spare: a long table and chairs extended from a wall covered in colorful five-by-seven cards charting the episodes for a television series’ new season. Sitting at the table, I discovered how network-enabled data had driven the decision to turn one of my favorite fictional characters into a TV series.
Author Michael Connelly has written more than twenty best-sellers featuring LAPD detective Harry Bosch. Hollywood was interested, but the show had been in the purgatory that studios call “rework” for more than a decade. That was until Connelly took the character to Amazon for its Prime Instant Video service. It was a quick sell.
The author told me how surprised he was that, after years languishing in analog Hollywood, the digital gurus at Amazon moved so quickly to bring his character to video. He discovered it was all a matter of information.
The digital tracks created by the online purchase of a Connelly book, combined with information from other online activities, indicated Bosch readers were highly qualified prospects for Amazon’s annual-fee-based Amazon Prime. In order to watch Amazon’s videos, you had to be a Prime customer. What made this significant for Amazon was that Prime customers, in part because they receive free shipping, spend three to four times more on Amazon purchases.5
Connelly told me Amazon executives decided to produce a cop show because it would help them “sell razor blades” and other consumer products.
Welcome to the digital syllogism! Information about your digital behavior has a high value because it identifies likely future behavior and allows the users of that data to target how to influence that behavior. Every time a consumer does something online, that activity becomes an asset that can be monetized.
And “online” has an increasingly broad definition. Like virtually everything in the connected world, for instance, the smart TV in your home collects information about you. When television went all-digital in 2009, a computer replaced the cathode-ray tube behind the screen. When that computer connects to the internet to retrieve a video program, it also transmits valuable information about you upstream. You thought you were being sold a television, but it turns out that television is selling you.
The job of a computer is to count, manipulate, and store digital information. And computers don’t just exist in products like your TV, PC, or smart phone. The modern digital network is a seamless collection of communicating computers. While old networks were “dumb” connections, the daisy chain of computers in IP-based networks creates analytical opportunities every step of the way. Not knowing precisely how to describe this phenomenon, conventional wisdom settled on the illustration of the internet as a fluffy cloud. The phrase joined the lexicon: information is transported, stored, and manipulated “in the cloud.”
In the physical world of the last network revolutions, the railroad and telegraph hauled information from one point to another.6 In today’s networks of integrated computer connections, the act of transporting information creates new information about the content being transmitted (for example, “Tom is looking for a hotel in Paris”), as well as the context of that information (for example, “Tom is in Columbus, Ohio, at the corner of Broad and High, and has previously searched for online French lessons”).
The knowledge and ownership of this computer-collected content, and its context, are at the root of the new network-driven economy. In previous network revolutions, information was static. While activity was always creating information, that information was largely inaccessible. The absence of processing power to collect and analyze the data, coupled with the inability to transmit it so that it could be combined with other data, meant the information went unused. Today, anything that touches a network becomes accessible information. And when information is put in motion to interact with other information, it creates even more new economic value.
Not surprisingly, the amount of digital information is exploding. Eric Schmidt, former executive chairman of Google’s parent company Alphabet Inc., once estimated that “from the dawn of civilization until 2003 humankind generated about five exabytes of data”—that’s approximately 5 million trillion bytes of data. He analogized that to “all the words ever spoken by humans to date.” Schmidt liked to point out in 2010, that same amount of data was being created every two days.7
Schmidt’s calculations have subsequently been surpassed. According to a 2017 study by International Data Corporation (IDC), we are creating 44 exabytes of new data on a daily basis. That’s the equivalent of 3 million Libraries of Congress being created daily!8
The creation of data is the manufacturing activity of the twenty-first century.
In the industrial era, capital assets were mined from the ground, transported to be fabricated into products, and then shipped to market. The capital asset of the new networked economy is information mined by and from connected computers and fabricated into new information products that are instantaneously available anywhere. Called “Big Data,” this information tells a story of location, network logistics, and behavior—whether of inanimate objects or of you and me.
Every day the two jet engines on a Boeing 787 Dreamliner generate one terabyte of data—approximately 1 trillion bytes of data.9 Sensors connected to microchips constantly monitor and report everything the engine does. Some of the information is networked to the plane�
�s computers to combine with other data, such as weather information, to recommend the optimal altitude for fuel efficiency. Other information is sent to a satellite to relay performance data to the ground, while different information is stored for later download and analysis.
For an airline, information that increases fuel efficiency and monitors performance is invaluable as it can mean billions of dollars in savings per year. Industrial companies such as GE—maker of jet engines, among other things—have embraced data usage to redefine industrial production beyond the nineteenth-century concept of building good products, to building the product and then subsequently managing its performance based on computing power incorporated into the product.10
Industrial activity is morphing beyond production to include high-margin services that utilize intelligent components assembled into the product to manage the product’s output. Industrial system decisions that once were based on statistical probability can now be made with near certainty. Whereas conclusions were previously extrapolated from a sample, embedded chips now measure everything, and intelligent software analyzes the data. It thus becomes possible to measure reality rather than simply forecast it. Statistical analysis just might become the new Latin—something you must learn but seldom use—as connected computing replaces statistical probability with knowledge certainty.
The applications of connected computing extend far beyond industrial activities. Information that has always existed, but was never captured and analyzed, changes everything it touches. In health care, for instance, the electronic patient records kept by 80 percent of office-based doctors are now an unprecedented research tool. If data from a jet engine can show how to fly an aircraft, data from health records can similarly change how doctors treat patients.
By correlating the data across 1,400 patients’ electronic medical records, for instance, researchers found that a heart drug called a beta blocker had the unanticipated effect of prolonging the lives of women with ovarian cancer. Similarly, data mining of health records demonstrated that a drug prescribed for heartburn increased the chance of heart attacks. Such lifesaving findings didn’t require a big laboratory project; the science was sitting there, unobserved, until the conversion of doctors’ notes into digital information allowed metadata analysis, which in turn led to new medical breakthroughs.11
For a consumer-facing company, access to such Big Data is the key to solving a long-running riddle. Retail pioneer John Wannamaker is reputed to have said that 50 percent of his advertising was useless, but he didn’t know which 50 percent. The new networks have created hundreds of billions of dollars in corporate value by using metadata to answer just such questions.
Google knows which product to sell you by estimating information about you based on your queries. Facebook knows because you told them about you, and then they augment that knowledge by monitoring your online behavior. The network that took you to Google or Facebook can watch all the unencrypted data as it passes and enrich it with real-time information such as your location, and even whether you are standing still or in motion.
And more data points are on the way. Microchips in jet engines are one example of the tens of billions of microchips being deployed in the internet of things. They will collect data and report on everything from the load in your trash can, to your driving performance, to your pharmaceutical consumption. Knowing this information, the networks will then carry back commands to the sensors to help the activity perform more efficiently, including telling you what to do.
Once again, networks are driving a new economic model. This time, however, it is for a different kind of product. The network revolutions of the nineteenth century enabled industrial production at scale. The networks of the twenty-first century are enabling the use of information at scale.
But What about My Privacy?
But what happens to my privacy when the digitally networked world knows everything about me?
Mark Zuckerberg once observed that privacy is no longer a “social norm.” “People have really gotten comfortable not only with sharing more information and different kinds, but more openly and with more people. That social norm is just something that has evolved over time.”12
Privacy has always been a relative concept. As America expanded westward in the eighteenth and nineteenth centuries, homesteads had the privacy of isolation. Yet in small frontier towns everybody knew about everyone else. Privacy was a casualty as the Industrial Revolution drew people into jam-packed tenements with thin walls and open windows. Then, after World War II the movement to suburbia began a return to geographic privacy protection, while in the cities the commercialization of air conditioning allowed privacy behind shut windows.
That privacy is an evolving concept relative to technological innovation can be traced back to—of all things—the box camera. In the late nineteenth century, George Eastman began selling a portable camera. It created a furor about personal privacy because anyone could record anyone else’s images and actions without the other person’s knowledge or permission.
In a seminal article, “The Right to Privacy,” written in response to the camera’s intrusion, Samuel Warren and Louis Brandeis argued, “Recent inventions and business methods call attention to the next step which must be taken for the protection of the person, and for securing for the individual … the right ‘to be left alone.’ ” They went on to warn that “numerous mechanical devices threaten to make good the prediction that ‘what is whispered in the closet shall be proclaimed from the house-tops.’ ”13
Louis Brandeis subsequently became one of the great Supreme Court justices. His concerns seem quaint in an era in which smart phones generate billions of pictures daily. But replace his term “mechanical devices” with “digital devices” and “whispering in the closet” with the digital utterance in an email or website visit, and it delivers us to our own question about the “next step which must be taken for the protection of the person.”
How we respond to the new networks’ ability to know everything about us will be one of the great cultural challenges of the twenty-first century. This response is made more complicated by the “privacy paradox”—the difference between expressed concern about private information and actual behavior. One study even suggests that society has moved beyond the paradox, that a majority of Americans are resigned to the fact they have already lost control of their information.14
Factors as diverse as age and nationality affect privacy perceptions. Millennials appear more accepting of the idea that information is known about them than do baby boomers. In the European Union, privacy is specifically identified as a civil right guaranteed to all, while in the Bill of Rights and the U.S. Constitution, the word “privacy” never appears.15
Yet networks have always posed a privacy threat. The founding fathers feared their revolutionary correspondence was being read. Messages sent via telegraph during the Civil War were often sent in code in case the enemy gained access to the line. Telephone calls were first held on party lines available to nosy neighbors; subsequent private lines were vulnerable to wiretaps.
The new networks add a compounding dimension to privacy. Historically, network privacy has been based on protecting the content of a telegram or a phone call, as well as protecting the information necessary for the network to operate, such as routing instructions and account information. Laws and regulations protected both the content and the context of communications. The extra effort required to access this information also served to protect it.
But what happens when the degree of difficulty to access your private information disappears, when the network is a connected chain of computers with the capacity to analyze and store digital information, and that information is a capital asset essential to a large segment of the economy?
The business models of early internet companies did not permit waiting around for gradual growth through subscriptions. As a result free distribution supported by advertising became the primary source of revenue. In an effort to solve J
ohn Wannamaker’s problem, the early internet companies introduced cookies and other user identifiers: from that point forward, the ability of computers to count, calculate, and store information, coupled with the networks’ ability to deliver that information with increasing granularity, has meant that tracking individuals is the default business model of the internet.
If you are not paying for an online product, then you are the product as information about you is collected and monetized. Even when you are paying for something like online connectivity, you are being monitored and monetized.
The smart phone has deepened the development of a revealing portrait about each of us. That wonderfully helpful technology is an unseen witness that constantly reports what we are doing. Your phone knows the basic information about who you are, where you live, and what you are doing online, but it also knows a great deal more about each of us.
When a smart phone communicates with a home Wi-Fi network, it also can report on how many people are in a room and even where they are sitting. When you are on the move, your smart device knows where you are, whether you are stationary, walking, running, or driving—and from where and in which direction. Because of a built-in altimeter, it knows the floor of the building you are on. It knows your likes, dislikes, and preferences; the books and news you read, the music you enjoy, and the TV and movies you watch. And, in the ultimate intrusion, your mobile can be turned on remotely without your knowledge to allow someone to eavesdrop and listen for key words, much like some email services collect key words in what you write.
