Who Controls the Internet?

Sam Peet In 1986 Stephen Wolff took an obscure job with the U.S. government: division director for networking at the National Science Foundation. This meant he was effectively in charge of the internet—at the time, a conduit for academic messages and file sharing. But over the next few years Wolff came to realize that if he did his job well, he would put himself out of one. He reckoned that eventually the network he was overseeing could, should, and would be open to everyone, and thus too big and complex for any single person or agency to manage. Wolff was right. As Shane Greenstein, of Harvard Business School, documents in his richly detailed history How the Internet Became Commercial, the decentralization of control over this resource resulted in one of the most significant periods of technological innovation and economic transformation in history. Consider the Mosaic browser—created at the University of Illinois at Urbana-Champaign with an NSF grant, building on technology that had been developed at a physics lab in Switzerland (CERN), commercialized by a start-up in Silicon Valley (Netscape), and eventually licensed to a 20-year-old tech company (Microsoft). This is what Greenstein calls innovation “seemingly from the edges of the network”—not led by a “single, hierarchical, and deliberate organization” (as sci-fi writers such as Arthur C. Clarke, the inventor of HAL in 2001: A Space Odyssey, had predicted future technology would) but arising from dispersed, cumulative, unpredictable work, enabled by forward-thinking laws, norms, and institutional decisions. Whereas Greenstein analyzes the past, the authors of several other recently released books are looking forward, imagining what’s next for the internet as it becomes ubiquitous in our lives. The Inevitable, by Kevin Kelly, a cocreator of Wired, offers a futurist’s perspective; The Third Wave, by Steve Case, a cofounder of AOL, gives an entrepreneur’s; and The Industries of the Future, by Alec Ross, formerly of the State Department, presents a statesman’s. These three books have a lot in common. They all describe technologies that are poised to change society: machine learning, robotics, virtual reality, and connected products, among others. All cite the challenges raised by those developments, while expressing a sincere hope that their impact will be broadly positive. And all predict that innovation will continue to come “from the edges,” though each author offers his own spin on what that means. Kelly, whose book is the vaguest and most speculative, sees decentralization as inherent in a networked world. Case envisions the internet extending to new industries and enabling “the rise of the rest,” in which cities beyond Silicon Valley come to dominate specific fields. And Ross predicts similar diversification by industry and geography with an even more global view—the acceleration of mobile banking in Africa, for instance, and of robots in Japan. But what if the decentralized model that Greenstein outlines is going away? After all, today’s internet is clearly dominated by what Farhad Manjoo, of the New York Times, calls the Frightful Five: Amazon, Apple, Facebook, Google (now a unit of Alphabet), and Microsoft. Might these incumbents be more inclined and better positioned than past giants to defend their turf and maintain centralized control? For one thing, they own much of the technology deemed critical to the next wave. Facebook has acquired one of the leading virtual reality firms, Oculus. Alphabet has absorbed seven separate robotics companies along with DeepMind, the artificial intelligence outfit that recently established a milestone by beating a world champion at the game Go. Amazon and Alphabet are both experimenting with drone delivery, and, of course, Alphabet is developing a self-driving car. Data is another advantage. The Frightful Five have vast amounts, which is exactly what you need to build machine-learning applications. Because they’ve been in business (and collecting information on their many customers) for years, they have a dramatic, perhaps insurmountable, head start in this area. As Ross asks, “Will big data serve to centralize businesses, pulling more industries into the gravitational field of Silicon Valley?” Case places his faith in entrepreneurs and their record of displacing incumbents. After all, AOL survived Microsoft’s attacks during his tenure; perhaps today’s start-ups can do the same against the Frightful Five. But, as he admits, AOL’s disastrous merger with Time Warner was precipitated by public policy decisions that denied “open access,” allowing cable companies to discriminate against content providers. “If we couldn’t partner with a cable company, the thinking went, maybe we needed to buy one,” Case writes. That’s just one example of how easily the scales can be tipped away from decentralized control. Greenstein cites another—antitrust law—noting that although the breakup of AT&T furthered technological development, today’s rules are narrowly focused on how competition affects consumer prices rather than on the need for multiple perspectives in the innovation process. It would be a stretch to call Jeff Bezos, Tim Cook, Mark Zuckerberg, Larry Page, or Satya Nadella the “manager of the internet.” But the companies these people lead do represent a centralization of power, and each of them will have a disproportionate say in the network’s future. Although the potential for innovation from the edges still exists, it’s not inevitable.

Who Controls the Internet?

Partner Center