Text Version





MARCH 4, 1997

(As prepared for delivery)

Bandwidth and Pizza

Last year, on Valentine's Day, Vice President Gore kicked off your celebration of the 50th anniversary of the invention of the electronic computer.

Now as it happens the Vice President and I were born a year after the compu- ter, 49 years ago this month; my birthday was yesterday, his comes 28 days later.

I know you each have handheld computers -- heck, you each can probably make handheld computers -- that can do that addition so you can figure out when to send him a birthday card.

But yesterday, March 3, was my birthday and incidentally also the birthday of Alexander Graham Bell.

How did you think I happened to be selected as Federal Communications Commission Chairman?

March 3 is also the birthday of the inventor of the railroad sleeping car, George Pullman.

That was a close call for me. I could have been selected as Chairman of the Interstate Commerce Commission, which was put out of business by the new Republican Congress last year.

The Federal Communications Commission they only thought about shutting down; but I'll get to that later.

In any event, as it also happens, being of similar age and growing up in the same town, it came to pass that the Vice President and I went to high school together. Now there were differences between us: he was slightly younger, had the prettiest date at the senior prom (a girl with the nickname Tipper), was captain of the football team, was a National Merit Semifinalist, and went to Harvard.

He also truly loved all science and if he hadn't gone to Harvard I believe he might well have become a computer wizard.

I didn't say that quite right. It wasn't that he went to Harvard. Bill Gates explained this to me. It was finishing Harvard that was Al's mistake.

In any event, now that I've turned 49 (slightly earlier than the Vice President), and I'm holding on for one last spring-in-my-step, irrationally exuberant year to what could at least colorably be called youth, frankly I'm delighted to remark with disdain the birthday of anyone or anything on the other side of the Great Divide of the Big Five-Oh, such as Mr. ENIAC, the electronic computer, now a dusty, decrepit, defunct 50.

Now fortunately a truth of life, vouchsafed to one only as one reaches the 40s, is that you're always as young as your ideas.

Back home in Washington the telephone and broadcast lobbyists sometimes call my ideas juvenile, as opposed to fresh and youthful, but I'll get to that later.

In my job I should have all the young ideas in the world in order to cope with the wedding of Bell's progeny and Mr. ENIAC's descendants, also called the convergence of communications and computing.

But instead I've just got questions for you.

First, your computers have made my three children -- age 8, 11, and 14 -- infinitely more knowledgeable than I was as a kid. They seem to be getting the "top sight" described by the brilliant Yale computer scientist David Gelernter -- the ability to use the machines you make to manage and manipulate their complex world.

But my kids have access to computers. What about kids that don't? Is it a problem that they do not have the opportunity to use the amazing things that you invent?

Second, all three of my kids want to use our single delightful computer that Michael Dell sent me and millions of other Americans by mail a few months ago? But I can't afford two more. So where and when can I buy a top flight $500 PC, including a monitor and a real fast modem? And maybe the answer to this question is the answer to the previous question about the have-not's of the information age.

Third, my kids want faster and more reliable access to the internet. If we give them the policies that will drive massive, network-wide bandwidth growth I think that might change the things you build. Will it?

And fourth, my kids want library-like and TV-like convenience when they get to the Internet. It's easy, entertaining and sometimes edifying to turn pages and channels. By contrast, the Internet's display of pages behind pages is not much more fun than a cluttered desk, and a long scroll seems to roll right out of those tiny little monitors that CRT technology has stuck us with for now. So I'm not complaining, but could you please get us big cheap better-shaped monitors and better GUIs and a different look to the Internet in general? If that means more of the push of TV instead of the pull of the web, that's okay -- as long as the consumer has choice of push technologies.

I know these issues are a little out of the ACM jurisdiction. But as we all tell each other, computing and communications have/are/or will be converging, so I suspect that an awful lot of the growth I'd like to see you have will come from a better Internet experience, if it were available.

Bellcore just released a study showing that the number of former Internet users approximately equals the number of current users. In other words, for everyone using the Internet, there's someone else who tried it and gave up. Mr. Bell did not have this experience with the phone. (As I recall the only person who ever told me she preferred to write me than continue a phone conversation was my date to that senior prom where Al took Tipper.)

Henry Ford did not have this sort of abandonment rate with the Model T.

I respectfully suggest we need big bandwidth, a very user friendly Internet experience, and products that fill real needs to drive computers up to the penetration rate of phones, cars and TV's -- goals all of us here want.

Network PCs and traditional PCs should be configured on the assumption that highspeed data networks can be built to every home.

Our primary goal at the FCC is to create the incentives for that sort of network.

Under the Telecommunications Act, signed in February of last year, we are pursuing more quickly than any other country by far an agenda of throwing out all the thousands of pages of state and federal pro-monopoly rules and replacing them with a play book, almost exactly as long as the rules of Major League Baseball, that will give new entrants a chance to compete in the telephone business. If we have written the right rules, and if we have the courage to finish the job over the next several months, then in time we will see as many communications firms as we have car companies. And we will have as little need to regulate prices as we do in the car industry.

Our ultimate goal at the FCC is simple: no rules.

Achieving a competitive market is a process that could easily take five to ten years, but it's worth the effort.

Competition gives us many different kinds of cars, and opens the roads to new innovations every year.

Competition too can give us many different kinds of interconnected phone networks -- wireless, hybrid fiber coax, SONET rings, switched and unswitched ISDN, ADSL, and so on.

Bandwidth ought to be like pizza. You should be able to get whatever you want -- small, medium, or large -- delivered piping hot to your door, with any combination of voice, video, or data toppings.

Nobody wants to have the status quo shaken up so that you can order bandwidth like pizza. You know why? There's a lot of risk involved. But risk doesn't stop companies in competitive markets like computer hardware or software. If you want bandwidth delivered like pizza, you're going to need a lot of bandwidth companies.

I often find computer companies just want the phone networks to work cheaply and well, with high bandwidth on demand.

Many of today's telephone companies will be able to meet those demands. But they are much, much more likely to do so if we have competition in the local market.

The 1996 Telecom Act can be a charter for choice among computer companies: if you don't think the local telco is giving you or your customers the bandwidth and the price you want, shop for the alternative from the new entrants.

Competition will also give us bandwidth. There are T3 backbones, but we are stuck having to access them with 33.6 kilobit modems. If our real highways worked that way, you could drive 55 miles per hour on the interstates, but would be forced to go less than one mile--per day--on the on-ramps.

Tens of kilobits a second is fine for E-mail, but not for graphics and the World Wide Web. And Real Audio and Real Video are great, but at 33.6, they are not very "real."

Last year, Lehman Brothers released a report stating that "'demand elasticity for bandwidth' is a key communications theme, as the unending appetite for bandwidth drives industry growth." The huge pent-up demand for more bandwidth means that as competition pushes prices down and makes more services available, usage will skyrocket.

On January 23, the FCC held a Forum on Bandwidth. You can read the complete transcript on our Web site, WWW.FCC.GOV. At the forum, Les Vadasz from Intel described how he would solve the bandwidth problem. As he pointed out, you can turn on your tap and get water, you can put a plug in a socket and get electricity, you turn on your TV and get cable. Why can't you just connect your PC to an outlet and get data?

He laid out seven requirements for this data service. Number one, of course, is bandwidth. Number two, is instant access--no dialing in. Number three, plug and play service, just like cable. Number four, multimedia capability--voice, video, data. Number five, store-and-forward capability, so that voice mail and E-mail are easy and reliable. Number six, security. And, number seven, affordability.

I don't think the right approach to achieve these goals is for local phone companies to impose access charges on Internet service providers. The Internet shouldn't contribute to the subsidies that access charges represent. But the Internet also shouldn't be subsidized in the form of below-cost second lines. If someone wants to buy a second line to access the Internet, the phone company should be entitled to charge them what that line costs.

The need with bandwidth also creates great opportunities for other kinds of providers that don't use the local phone network to connect to their customers. Both wireless and cable have the potential to break the bottleneck of local exchange carriers in providing high-bandwidth data connectivity to the home.

Unfortunately, rate regulation provides disincentives for companies interested in deploying such a service. The sooner we get to a fully competitive market for telecommunications services, the sooner we can throw out the old regulations, and the sooner companies will deploy new technologies--rather than more lobbyists and lawyers--in order to beat their competitors.

Our job should be to create efficient incentives for companies to deploy these technologies. That doesn't mean subsidizing industries we like. I it means allowing the market to work without distortion. Companies that can provide higher bandwidth more cheaply won't need subsidies to be successful as long as government or monopolies don't try to distort the market.

In April, the FCC will be starting a proceeding on innovation. I hope the computer industry will follow that proceeding and submit comments. We're looking at questions about how regulation can help or hinder technology development and deployment, and your input will be extremely valuable.

As we think about bandwidth, we also need to keep in mind where the bandwidth is going. Businesses are already developing high-capacity corporate intranets that let them take full advantage of the potential of the Internet. But what about American residents? American consumers? Today we have fine companies, but we don't have big bandwidth networks. This highway needs to be public, it needs to be widespread, it needs to be ubiquitous.

The universal service provisions of the Telecommunications Act concern the essential public benefits of the Act. Under the Act, we are working with the states to completely restructure the way we ensure that people in poor areas or remote, rural areas can afford telephone service.

But Congress saw that doing the job also requires a national commitment, so it included the Snowe-Rockefeller provision in the Telecommunications Act of 1996. In November 1996, a bipartisan, state-federal Joint Board handed the FCC a set of recommendations including:

The FCC is now reviewing the public comments on the recommendation and will issue final rules on May 8.

There are those people who see the $2.25 billion per year recommended by the Joint Board and say, "I don't want to pay." But we're talking about a tiny fraction of the revenues of the industry. With this modest amount of revenue, our children can become the beneficiaries of a big bandwidth road.

We spend $88 billion a year on roads. Don't you think we could spend $2 billion on this particular road?

Late in April, President Clinton, former President Bush, Colin Powell, governors, mayors, CEOs, and leaders of non-profit groups are meeting to determine how volunteerism can help us address this country's social problems. The computer community has shown that volunteerism can work, especially in our schools. Net Day is just one example.

It is not too late for your companies and universities to make commitments to volunteer to help wire our schools and to help train teachers and students on how to use computers and the Internet.

Thank you.