Om Malik has a wise commentary today on how peer-to-peer services (p2p) is the killer app for broadband.
He offers a Cachelogic chart showing how p2p services (but more specifically eDonkey) are driving total Internet traffic. In fact, more than half the total Internet traffic monitored by Cachelogic, according to the chart, is eDonkey traffic. (The illustration was copied from Malik's blog, but credit should go to Cachelogic.)
Then Malik makes some really key points (boldfacing is mine):
What's troubling to Malik is how distribution costs are being shifted wholesale to "pipe owners," the Internet backbone owners. With major studios and other content producers looking at p2p as a distribution method, this trend will only accelerate.
So let me add a question.
At what point does all this force the construction of new Internet backbone capacity? At what point do we overwhelm Moore's Law of Fiber, in other words, and require new builds? And, if those builds are based on Moore's Law of Fiber, how do we make sure those new builds aren't built with an eye toward monopoly?
Let's assume for a moment (for the sake of argument) that a fiber line installed today will have 1,000 times the capacity of one installed five years ago. This makes it difficult, economically, have more than a few suppliers, once the capacity of the present system is reached.
The best solution would seem to be for the backbone to move to a utility model, where build-out investment can be controlled and prices can be as well. But given the recent history of regulated utilities, this sounds like a disaster.
Once current backbone capacities are stretched, in other words, the fight for control of the Internet begins in earnest.
One more point. Isn't it possible that backbone pipe owners are encouraging this growth of p2p to hasten that day, and increase their control of the market?