David Farber

David Farber, recently named by Upside Magazine one of the most influential people of the digital age, muses on the direction that communications and next generation Internet technology are heading.

Photo by Candace diCarlo

David Farber is not plugged in, but he's nonetheless connected. On a recent trip to New York he carried a wireless phone with an ear jack, a cellular phone and a wireless computer, the latter allowing him to communicate by e-mail while riding the train. And of course he has a Web page (http://www.cis.upenn.edu/~farber).

He is the Alfred Fitler Moore Professor of Telecommunications Systems and a professor of computer information science, but he earned his full professor stripes even though he doesn't have a Ph.D. What he has is a vision of ways to communicate and ways to keep Penn on the cutting edge of research and technology.

Q. So you're wired like crazy?
Look at it another way. The business I'm in is being at the edge of communication technology.
That's what academics are good for; they live at the edge. The only way to live at the edge is to live at the edge. In the technology area, if you're not using it and understanding its limitations, you're not going to know where to go. It's very simple. So you stay at the edge and suffer at the edge.

Q. How do you suffer?
Oh, you'll get a communication device that's flaky, it's too complicated to use. You buy a personal assistant, it turns out to be useless. It's not all fun sometimes. I think I have a closet of almost-working stuff.

Q. What do you think is the next trend?
In the next five years, wireless and mobility are next - especially in the Internet world, the ability to always be connected with telephones, pagers, that kind of stuff.
The fact that you have access doesn't mean you always have to use it.
I'm potentially worldwide accessible, but when we went to our vacation spot in Hawaii, I turned it off.

Q. What are the goals of Internet II?
I am on the Presidential Advisory Committee. What we're looking for is for a general upscaling of the capabilities of corporate America, academic America and home America to have better access to the Internet information that's out there.
The network serves really two purposes. It gives you information. And it gives you the ability to communicate with others. And those two things are valuable.
What we'd like to do is get wide-scale improvements (in speed) in the infrastructure. The big question is a what do you do with it, and how do you make it widespread and get the support structure in place.
Internet II is one part of this Next Generation Internet (NGI) activity. This is a federal program. It has all the agencies involved. It has budgeted probably this year in excess of $150 million.
What we'd like to do is make it less expensive by building a customer base for it, but more importantly by understanding how it can be used.

Q. You mean it already exists?
Pieces of it have existed for a long time.

Q. Can you give me an example?
Sure. Penn, along with a bunch of others, is connected to vBNS - that's a relatively high-speed communications system that hooks together a lot of the research universities. Penn got on it about a year ago, and that basically is the zero level Internet II.

Q. And what do we do with it?
Largely a few experiments on high-speed communication because it's not well connected on the campus yet. There are proposals going in from the hospital. Anybody who wants access to it, it's there. It's not designed to be used for all your e-mail.

Q. Why would you need the high-speed communication?
If you were a hospital and you were doing experiments with letting remote doctors use your fuctional MRI system, you'd need a lot of communication, believe me.
There are two ways of sort of attacking that. One is to say, well, let's find out all the uses for it and we'll see if we have enough uses, and we'll build it. Now that's real tricky because most people don't have a use for something that's blue sky. They'll come up with Mickey Mouse uses, but they're not going to put their heart and soul into that. The experience of the communications business, especially the network business, is build it and they'll come, a sort of "Field of Dreams" approach.

Q. What else is Penn doing?
It's essentially building the local roads to the superhighway.
It's senseless to put an entrance to the superhighway if there's only one driveway on it. So you build a local road structure, you get a bunch of houses that get together to build the local road, and now they'll give you entrance to the highway.
Rutgers, Princeton, Penn and Delaware have banded together to sort of build the local roads to something called the gigapop (the gigabit point of presence), where the long haul carrier will present itself in the region.
Besides getting a more cost-effective way to access into the Internet II, there's another thing we've been trying to do for years. We're in the middle of a bunch of universities that are complementary but might have been across the country as far as our interactions are concerned. The fact that we'll build this set of local roads means that it will start being feasible to talk about doing things like remote seminars and remote symposiums, even, heaven forfend, shared classes at some point in the future.

Q. Is that the sum of Penn's role in Internet II?
No. We've been active not just as joiners but as initial pushers of the idea.
If you come down into the campus, the trouble with superhighways, if I bring a first-class road into my town and all I have is dirt roads, it's not very effective. You get bottlenecks. As we bring Internet II onto the campus, we have to build the roads onto campus to support it. It's not going to be very effective to say I want to have a joint classroom with Princeton if I can't get (the network) to my classroom.
When we built the national superhighway system, we did a lot more to the country than just let people go between cities. We changed the whole social structure of the country. Suburbia happened.
You were able to do things without having to live in the city. High-speed networking will have the same impact. We don't know how, but it's going to change it.

Originally published on February 26, 1998