Q&A: Andrew Blum on the Internet’s Physical Spaces

Andrew Blum’s new book, "Tubes: A Journey to the Center of the Internet" is an evocative trip to the heart of the Internet.

Metropolis contributing editor Andrew Blum’s first book, Tubes: A Journey to the Center of the Internet appears today in bookstores (and presumably on iPads and Kindles as well). I had the pleasure of reading it in galleys a couple of months ago. It’s an evocative trip to the heart of the Internet, a look at both the physical connections behind the web and the complex almost ad hoc infrastructure supporting it. On the eve of the pub date, I spoke with Andrew via email about the new book, his quixotic journey, and our burgeoning digital footprint.

Martin C. Pedersen: How did you get interested in the physical places behind the Internet?

Andrew Blum: I’ve always written primarily about architecture. But over the last several years, architecture as it’s traditionally understood seemed less and less to reflect the world I was living in—the world inside my screen. I realized that I couldn’t fully understand what “place” meant anymore, if I didn’t know anything about the ostensibly-virtual place I was spending my days. Then something happened: a squirrel chewed through my home Internet connection. And I realized then that there had to be something out there—something physical, perhaps even architectural. And there was.

MCP: What were you expecting to find when you set off on your journey to the heart of the Internet?


AB: A lot of wires. Some blinking lights. Rent-a-cops. Surveillance cameras. Lots of nowhereness. But it turned out to be a lot friendlier than that, and—occasionally—more beautiful.

MCP: As someone who writes about the built environment a lot, much of what you found was bluntly “un-designed.” Talk about that dichotomy, between the banal physical world of the Internet and the almost unimaginably complex system it supports.

AB: I believe that architecture expresses ideas and values, and so the fact that much of what I found was so undesigned indicated to me our broader denial about the complex systems that support our digital lives. We don’t read Terms of Service (we just click “accept”), and we don’t care where our Internet comes from. The design of these buildings explicitly said that: they are anti-monuments. As places unto themselves they declare their unimportance–even if, technically, they are essential. But this is changing. Two of the most architecturally impressive buildings are the newest: Facebook’s data center in Oregon, and a building called Telehouse West in London. They each try to express their function, in the modernist sense.

MCP: You managed to get a peek into some pretty locked down worlds. How did you manage to do that? What was the process of gaining their trust?

AB: In the case of almost all these places, their stories hadn’t been told in a decade, since the spectacular bust of the telecom world. More often than not, the gatekeepers were very happy to invite me over. Given the state of the media industry, fewer and fewer journalists were asking for an actual tour; so they weren’t exactly besieged with requests. But I was always entirely straightforward. Never once did I sneak into a place, or otherwise circumvent proper channels. The Internet is a small world; I couldn’t risk breaking anyone’s trust.

MCP: What was fascinating about the book was how “ad-hoc” (for lack of a better word) the development of the web’s infrastructure was (and is). At certain points it almost seems jerry-rigged, so that some cities are physically connected and others, invariably, aren’t. What are the urban planning and development implications here?

AB: The Internet is perhaps the greatest example ever of a human-made “emergent” system. There is no master plan. But the urban planning implications are difficult to consider. The Internet operates physically at multiple scales, which often collapse into each other: the machine, the building, the city, the region, and the globe. But it’s also an incredibly complex thing in logical and algorithm terms. That makes it difficult to draw an analogy with urban planning. Cities are certainly multivalent—physical, economic, social—but in different ways.

MCP: It seems like in the 21st century, not being fully “connected,” in the crude infrastructural sense, will be like not having a port or a public transportation system. Talk about the implications for the developing world.

AB: I think we are well past the argument that the least developed countries should focus on “essentials” before Internet access. And indeed, the places I looked more closely at—specifically Rwanda and Kenya—were making big pushes to fill-out their digital infrastructure. Interestingly, that had to happen at two scales: they needed to build robust national fiber systems; but they also needed strong international links, which came with the arrival of three new undersea cable systems down Africa’s eastern coast. Because one isn’t enough: once a country’s economy begins to shift towards telecommunications-based services, it’s too risky not to have a backup. Or two.

MCP: Is this physical infrastructure likely to change much in the future, as everything becomes digital?

AB: I don’t think the infrastructure will change. The map is quite fixed: the most important places—like Virginia, Amsterdam, and Singapore—will stay the most important, albeit growing proportionally with the second tier cities. However I do think we’ll see a significant boom in the places where the largest data centers concentrate, like Oregon, Sweden, and Iceland. As our digital footprints grow, that “deep storage” will need to grow with it.

Categories: Cities, Technology