August 1998, 6.08

Stupid Is Smart

By Steve G. Steinberg

David Isenberg has emerged from the belly of the beast. Last year, as a member of AT&T's technical staff, this self-described telco nerd sparked an influential debate with his essay "Rise of the Stupid Network." In that essay, Isenberg argued that the "intelligent" architecture of the telephone networks has retarded innovation, while the Internet, by offering "stupid bandwidth," has become a breeding ground for smart apps. Not surprisingly, the relationship between Isenberg and AT&T turned cool. Today, he runs his own shop, calling himself a "prosultant" ("con is negative," he sniffs) on next-generation networking. Here's the iconoclast on what's next for networks.

Meet the new Net, same as the old Net. "We're already seeing intelligence start to rear its head in the Internet. New protocols, like RSVP (which tries to guarantee users a specified amount of bandwidth), require a smart network. We need to find stupid solutions to these problems to avoid the same kind of big monopolies we have in the telephone industry."

Voice isn't the killer app. "The real win for voice-over-IP will be allowing people to mix real-time voice with data, images, and video. In the meantime, the circuit-switch guys will have to reduce prices to stay competitive - and still offer better quality."

The customer doesn't know best. "If you're listening to your customer it's almost preordained that you'll miss the new market. And when the new market expands to encompass the old market ... that's when companies can become obsolete."

Why AT&T is doomed. "Today, Internet telephony has crummy voice quality. The average telephone customer couldn't stand it. But suppose that Net people start using voice to, say, supplement online gaming. Because it starts off looking more like a Nethead game than a phone call, AT&T might not realize how cross-elastic it is with telephony until it is too late."

Date last modified: 9 Oct 98