Lunchtime Musings: too much bandwidth, or too little imagination?


BusinessWeek, May 29, 2007

But once you have 100Mbps or more available at home, what the heck are you going to do with all that bandwidth? For the average consumer, 6Mbps should more than suffice for today’s typical needs, whether it’s downloading music, watching the occasional video, or even running a home network that lets two or three computers do the same all at once. Does anyone really care whether that song download from iTunes (AAPL) takes 10 seconds or 2 seconds?

GigaOM, December 20, 2005

“Can a consumer tell any difference between 1.5 megabit per second and four megabit per second service?” asks Bruce Leichtman, president of Leichtman Research Group, Inc. “The answer is no.

And if I had the time, I’m sure I could dig through Usenet archives and find an article from 1995 noting that for most users—other than those W4r3Z guys, of course—the difference between 14.4 and 28.8 really wasn’t noticeable.

Thinking about broadband, in the US at least, is fundamentally warped because we’re stuck in the mental rut of “a fatter pipe means we get things faster,” rather than thinking about the fact that a fatter pipe means that we can move a lot more data around. Both quotes above display the same plodding failure of imagination: they’re thinking about how long it will take to do what we can already do, and largely ignoring the fact that a fatter pipe also opens up new possibilities.

If we stipulate that “more data moving around” is the really interesting part of fatter pipes into the home, that does underline one big issue: we need to correct the current dramatic asymmetry between download speeds and upload speeds. If I’m 30Mbit down but still stuck at an absurd 768k up, very little that’s interesting can happen. Bump the upload speed a bit and things start getting fun.

You’ve got P2P/Bittorrent/change-how-distribution-works scenarios, both commercial and non-commercial. The “local computer becomes dumb local storage, buy bulk storage, processing power, and software as a service” idea starts looking more feasible. You can flip that idea on its head and create a VRM style “I am my own silo of identity, transaction, and attention data stored securely and locally, and I can quickly and at will provide a subset of that data to a vendor when initiating a relationship or transaction” deal.

Sure, the ideas above are varying degrees of half-baked and/or wild-eyed, but if you stop thinking about “fast” broadband and start thinking about “can move a lot of data” broadband, I’ll bet that you’ll all start coming up with better ideas of your own real quick. You’re welcome.