Tubes: A Journey to the Center of the Internet by Andrew BlumEcco (HarperCollins), 2012

In Tubes, Andrew Blum tells the story of when he “decided to visit the Internet”—and what he found there. At the start of the book Blum says that he, like many people, didn’t really think much about the physical structure of the Internet—until one day when a squirrel chewing on wires in a Brooklyn backyard made his home connection go all wonky. The fact of how “the Internet” got from the outside world to Blum’s apartment had never really mattered: but suddenly, it did. Blum started thinking about the mechanics of it, and how this idea of the physical structure of the Internet is at odds with the vague way many of us think about “the Cloud”:

What if the Internet wasn’t an invisible elsewhere, but actually a somewhere? Because this much I knew: the wire in the backyard led to another wire, and another behind that—beyond to a whole world of wires. The Internet wasn’t actually a cloud; only a willful delusion could convince anyone of that. Nor was it substantially wireless. The Internet couldn’t just be everywhere. But then where was it? If I followed the wire, where would it lead? What would that place look like? Who would I find? Why were they there? (4)

And so Blum sets off in search of Internet infrastructure and the people who create and work on it. In the first chapter, he watches the printing of TeleGeography’s Submarine Cable Map (they also have a totally fascinating (and free) interactive online version) and starts to get a sense of the physical pathways that data travels. As the book continues, he gets a better sense of those pathways: he learns about Internet exchanges (places where networks physically connect to one another to increase efficiency/cut down on travel) and visits several key ones, from California to Virginia to Frankfurt to London. Around Chapter 4, when Blum visits the Amsterdam Internet Exchange, the book got significantly more interesting to me: where in earlier chapters Blum was focused a lot on background/history and the various things he learned from various key people, the focus here shifts to what he sees. In Amsterdam, it occurs to Blum that he could/should see things in a bit of a different way from the corporate-approved tours he’s been getting. He’s found a map of data centers in the Netherlands and sees that there are plenty in Amsterdam—so he convinces a routing-table analyst to go on an 8-mile urban hike with him, to see the buildings from the outside. This results in a really pleasing section in which Blum talks about Robert Smithson’s “A Tour of the Monuments of Passaic” (which I haven’t read, but clearly should) and how it argues “that there is value in noticing what we normally ignore, that there can be a kind of artistry in the found landscape, and its unconventional beauty can tell us something important about ourselves” (151). So Blum walks and looks, and we get passages like this:

Our first data center was visible from the elevated train platform: a menacing concrete bunker the size of a small office building, with worn-out blue window trim, spreading out along a canal connecting to the Amstel River. The late-winter day was gray and damp, and there were houseboats tied up at the edge of the still water. My map indicated that the building belonged to Verizon, but a sign on the door said MFS—the vestigial initials of Metropolitan Fiber Systems, […] which Verizon had acquired years before. There was clearly no rush to keep up appearances; it seemed, rather, that its new owners preferred the building to disappear. (152)

In the next chapter, after visiting a router manufacturer in San Jose, Blum comes back to New York and spends a night watching a crew lay fiber optic cable in lower Manhattan. He talks about how, rather than being something totally new, the Internet infrastructure in a place like New York is layered on/builds off the pre-existing infrastructure from telegraph and telephone systems. This chapter was also super-pleasing to me because Blum talks about two buildings in lower Manhattan that have a telecommunications history—the old Western Union building at 60 Hudson and the old AT&T Building at 32 Avenue of the Americas—one of which is the building where I work. Both buildings are “art deco palaces” (it’s true) and when they were built, each apparently had its own “gymnasium, library, training school, even dormitories.” In 1955, the first transatlantic telephone cable went from 32 Avenue of the Americas to London. And now? As Blum puts it of 32 Avenue of the Americas, “on the twenty-fourth floor is the Internet” (176). Each of these buildings houses its own Internet exchange, and in that sense, as Blum notes, they’re not so different from the other ones he visited: except that they’re “a fact of geography,” an outgrowth of the New York of the early twentieth century, twenty-first-century spaces “built upon hundred-year-old telephone infrastructure, nestled between stock exchanges and railroad tracks” (ibid.).

Those middle chapters were the most exciting to me, but the remainder of the book is interesting, too. Blum visits Portugal to watch a new underground cable coming ashore, and visits Porthcurno in Cornwall—an important landing point for transatlantic cables. And, because the Internet isn’t just about traffic—it’s about data storage too— he also visits data centers in Oregon, getting a tour of nothing more than the parking lot and cafeteria at Google in The Dalles and getting a warm welcome at the then-just-about-finished Facebook data center in Prineville (which, naturally, has a Facebook page). Blum is a smart and engaging writer, and while the first three chapters sometimes felt like a slog, I was pleased with this book by the end, and quite glad to have read it.