Exoflood an Exoyawn


The exaflood was the catchy name wrapped around a prediction that exabyte-scale demands for data would destroy the Internet, making it unusably slow and erratic. Year after year, Internet doomsayers make the same tired prognostication. Karl Bode of DSLreports.com notes that mainstream media is finally starting to get the message. The two leading prognosticators of doom, Nemertes Research and the Discovery Institute, seem to be driven by an interest in battering the concept of network neutrality and broadband regulation. I wonder why? (You may best know the Discovery Institute for its support of schools teaching sloppy magical thinking.)

Bob Metcalfe, Ethernet's inventor, famously and literally ate his words in 1997 because of a promise he made of a gigalapse in 1996 that failed to come to pass. No word yet from the current Cassandras Bhagwan Shree Rajneeshes on swallowing their pride and their continuously inaccurate doomsaying.

The Internet turns out to be resilient, not brittle, partly because money funds growth, and companies are dying to take our money. While broadband providers may try to spend the least amount to bring us passably usable service, the Internet's backbone is driven by service-level agreements, steely-eyed technologists, and filthy lucre. We may put up with "up to 15 Mbps* (*as little as 768 Kbps)" connections, but Comcast, Verizon, et al., don't play that game with their network interchanges.

The analysts who make these predictions also fail to account for dynamic feedback. Once you start engaging in behavior on the Internet that fails, you stop. When I'm watching Hulu or Netflix, and the video becomes choppy and unwatchable, I stop watching. What a concept.

Photo by yours truly.