page 2 of 3
George Gilder: When Bandwidth is Free

KK : I take it then that as a believer in the human mind as sort of the ultimate repository of wealth and power, that you're not a believer in artificial intelligence as an attainable thing?

GG : Artificial intelligence is obviously attainable, it just won't be human intelligence. It's a different function.

KK : Do you think there's a downside to having everything connected to everything else?

GG : I don't see any. There must be a downside to the telephone, but I can't remember what it is.

KK : Well, when you're in the middle of supper and some solicitor is calling you, you'll probably remember what it is.

GG : Yeah, that's the downside. But that's the downside all this stuff overcomes. When you can have intelligence in your telephone it can defer the calls you don't want to voice mail and still take emergency calls. It can be adapted to your needs. So, the chief effect of these technologies is to put you in command again. The trouble with top-down centralized technologies, which the telephone and television represent, is that they're dumb equipment attached to complex switching systems and broadcasting technologies. On the other hand, the chief virtue of distributed intelligence is that the network can be dumb and the control of it can be distributed to smart users. That means that technologies are much more servants than rulers of your life.

KK : When I look at networks I see counter-intuitive behavior. Distributed networks have a remarkable ability to be slightly out of anyone's control. They possess an organic out-of-control quality. Does that concern you?

GG : I think it's good. The Internet, for instance, is an exciting kind of metaphor for spontaneous order. It shows that in order to have a very rich fabric of services you don't need a regimented system of control. When there's a lot of intelligence at the fringes everywhere, the actual network itself can be fairly simple. The future is dumb networks.

KK : Dumb networks? Why not smart networks? We put smartness in everything else.

GG : There's smartness all around the network, but the actual network should be essentially dumb glass. The fibersphere, as I call it. I think the mistake that the phone companies sometimes make is to think that they can keep up with the computer. What they call "network intelligence" will usually appear as a bottleneck to a computer industry that's rapidly rushing forward into new possibilities. So what you really want is dumb networks where all intelligence is on the fringes. You'll have intelligent devices of various sorts that are easily reachable from the network but aren't part of the actual fabric of the network.

KK : What is the fabric of the network?

GG : Photons. Electronics are not good for communications. Photons—optical computing—are. What makes photons so great for communication is they don't interfere with each other. They collide and pass on unaffected. You can send them two-way, and they are not subject to electromagnetic disruption. Many signals can flow through one fiber. But the fact that photons don't affect each other means they are cumbersome for computing, since you want interactions in computing. You need to have the charges affect one another - that's the heart of computing. The heart of the transistor function is that you can control a bigger force with a smaller force. But photons don't control each other. So for computing functions I still think that electronics will prevail; but for communications, photonics will prevail.

KK : One would think that there is wealth to be made in the interface then between photons and electrons.

GG : Yeah, there is. Opto-electronics is very important. However, opto-electronics should not be in the middle of the network, it should be on the edges of the network where it links the computing functions to the communications functions.

KK : There are some network advocates who claim that we can get a lot of what we want in fiber optic by using the existing copper wires beefed up with the ISDN communication protocol. Do you go along with the idea of implementing (and paying for) ISDN right now?

GG : Yes. The phone companies should do ISDN. We might as well get as much out of the existing copper switch system as we can. ISDN is already installed in all the new switches; it's more a matter of getting the tariffs right so they can charge some reasonable amount for its use. There's no excuse not to do ISDN today. It won't detract from the fibersphere. But while they do ISDN, all-optical networks are going to be launched all over the place by different companies. Some people have this vision that either we devote our resources to ISDN, or we devote them to creating this fabulously expensive fiber network. My belief is that fiber network is going to get rapidly cheaper, so that we're going to be able to do both it and ISDN perfectly well.

KK : What role do you think the US government should play in laying data highways?

GG : The role for the US government is to make government as efficient as possible. Government operates leviathan laboratories, hospitals, universities, bureaucracies, and post offices, and they all should be interconnected with fiber.

The government always discovers a technology after its moment is passing. If you're a winner, you don't go to the government. You're too busy. You've got too many customers. It's the people with no customers who end up besieging the government. There are all these wise-asses in Washington who really think that they can choose technologies. They think they know better. They get bowled over by every earnest representative of IBM who comes up to talk to them. Just now the US government thinks that HDTV is absolutely the future of the world because all the old farts at Zenith, and the broadcasting moguls who aren't really making it with the new computer technologies, are converging on Washington. It's always going to be that way. It's not going to change with Clinton and Gore. The dog technologies run to Washington, decked out like poodles. The politician is always the dog's best friend.

[page | 1 | 2 | 3 ]