Skip to Content

Net Cerfing

He invented a key piece of what has become the Internet. The MCI vice president shares his strong ideas on where the Net should be going–and wars of the dangers of government interference.

Every time you send an e-mail, surf the Web, or buy a product online, you are using the Net that Vint built.

Well, not all by himself, as he is the first to point out. Like most technologies of significance, the Internet flowed from the minds and work of many people. “There are countless others,” he says, “who have spent their entire careers working to take the Internet where no network has gone before.” But without the innovations of Vint Cerf, the whole idea of a global mesh of computers and people exchanging messages, pictures, and software could have gotten stuck on hold indefinitely.

The Internet works because of a method of sending information called packet-switching. Although he did not invent packet-switching, Cerf, along with Robert Kahn, formulated the packet-switching recipes that have become a worldwide standard. These protocols, known collectively as tcp/ip, are the common denominator that unifies the global Internet. The Net that we have come to know-including the World Wide Web in all its graphical and interactive glory-rests on this remarkably robust foundation of tcp/ip.

Cerf is much in demand as a speaker, playing the incongruous role of the Internet guru in a spiffy three-piece suit. He has a day job, too, as senior vice president of Internet architecture and engineering of MCI Communications. There, he oversees the development of network technologies and keeps a sharp eye on the progress of the technology he helped create. This is Cerf’s second stint at MCI. During the 1980s, he helped create MCI Mail, a service that helped popularize the concept of electronic mail.

Because of his hectic travel schedule, Cerf prefers that interviews be conducted through e-mail. Technology Review Senior Editor Herb Brody held up TR’s end of the computer-mediated conversation.

TR: You are generally credited as being one of the “fathers” of the Internet. How do you feel about your child’s success?
CERF: I’ve been more than pleasantly surprised over what’s happened. I never anticipated that the experimental Internet would get larger than 128 networks-and we’re long since past that now.

TR: What do you see posing the most serious threat to the Internet?
CERF: I’m most concerned about heavy-handed intervention to attempt to regulate the Net. Although the Supreme Court struck down the Communications Decency Act, some members of Congress are trying to revive similar legislation-a development we all should be concerned about. I’ve often traveled to Capitol Hill to brief Congress on the industry, and I’m struck at just how misinformed some members are concerning the Internet. Thankfully this is changing, although not as fast as I would like.

TR: In a sense the issue seems to be who should run the Internet.
CERF: Yes, and that has come to a head most recently in the debate over the administration of domain names. More than 100 companies have endorsed a new domain naming system developed by industry organizations. But the proposal was not universally accepted, and the White House undertook to develop an alternative that it hopes will gather broader support. These two proposals have similar objectives but differ in important ways. My sense is that each has positive elements and significant commonalities that should be used to form a foundation for a broad consensus on the whole issue. The critical next step is to move from talking (and flaming, for that matter), which has occupied us over the past year. We’re running out of time, and both sides need to be ready to compromise for the greater good.

Too Much for Too Little?

TR: A few years ago, the Internet appeared to offer users an almost free lunch. Has that harmed the Net’s development?

CERF: In some ways, it has. Because of its initial sponsorship by the U.S. government, many of the costs of the Internet have been hidden. As a result, some proponents of a cyberspace utopia persist in believing that the benefits of a wired world will somehow arrive at little or no cost. The fact is that technology advances in economically self-supporting increments. As subsidies have eroded, the costs they defrayed must be assumed by someone. That seems to be a lesson that some have conveniently forgotten.

TR: How is that evident?
CERF: One example is the “all you can eat” pricing plans for Internet access. This model works only if Internet access providers can predict total usage and can price accordingly, or if they can place an upper bound on usage to achieve the same effect. It seems likely either that prices for unlimited access will gradually rise-witness America Online’s price hike earlier this year-or that people will start paying some kind of usage-based fee.

TR: What about Internet telephony-that seems like another example of how the Net can provide a kind of free lunch.
CERF: Yes. A small but dedicated cadre of boosters often hail it as a low-cost alternative to long-distance telephony. But any near-term advantage will be temporary.

TR: Why?
CERF: The cost of a long-distance call depends heavily on the access fee that the local telephone exchange has to pay. Right now, long-distance voice carriers pay substantial fees per minute for the use of the local networks. Internet service providers, on the other hand, enjoy a fixed monthly price. These differences are likely to erode with time. Still, there may well be some economies arising from the packet-switched nature of the Internet that allow Internet telephony to operate more efficiently than circuit-switched telephony. And over the long term, the technology curve favors the Internet. The cost of routers and gateways will drop faster than the cost of telephone switches. In any case, the phrase “Internet telephony” is as helpful in describing the changes we’re in for as “horseless carriage” was for foretelling the sea change engendered by the introduction of the automobile.

Tale of Two Tapestries

TR: In some ways, the Internet seems to be displacing the telephone network.
CERF: Think of the Internet and the Plain Old Telephone System, or POTS, as two tapestries hanging side by side. The Internet is dynamic and flourishing, while the phone system is venerable, ubiquitous, and reliable. Now imagine if we take these tapestries apart and reweave them together one strand at a time so the two are indistinguishable from one another. That’s the future we’re heading toward.

TR: I thought the Internet and telephone system coexisted pretty well right now. Data, voice, and video already share the same lines.
CERF: Yes, indeed, they travel over the same beloved fiber optic lines. What we don’t have in place yet, though, are technologies that get POTS and the Internet to work invisibly and in concert, delivering the integrated services that the communications revolution has always promised but has yet to deliver.

TR: I still don’t understand what exactly is lacking in the present system. From my desk at MIT I can make a phone call, send a fax, check my e-mail, and surf the Web-all pretty much simultaneously.
CERF: Yes, but integration will provide more seamless interaction. You could have a universal, multimedia “in-box” that can accept e-mail, voice mail, video mail, and faxes. That’s not going to happen until the Internet and the phone system are blended better. By the year 2010, more than half of all voice traffic will be traveling over packet-switched networks. But for the great majority of users, placing a voice call over the Internet will be no different than placing a phone call today.

TR: What will we be able to do that we can’t do now?
CERF: Say you’re catalog shopping on the Web and you’d like to ask a specific question about a product you’d like to buy. Already, we’re seeing applications that would let you make that call using the same Internet connection you are using to view the page. It might also allow integrated messaging, so that your pager alerts you when a particularly important e-mail arrives in your in-box. Services like these are being introduced for the corporate marketplace now. What will be really exciting is when they become available for widespread consumer use.

Multicasting into the Millennium

TR: What technologies do you think will change the way the Internet works?
CERF: One key development is something called multicasting, which vastly improves the ability to send real-time audio and video signals through the Net. The way things work now, a Web server has to send a new copy of the audio or video data to each individual who requests it. With today’s technology, a high-end personal computer “unicasting” on the Internet serves a maximum of several hundred people simultaneously.

TR: What does multicasting do differently?
CERF: Imagine you’d like to share today’s Dilbert strip with some of your colleagues working in a different building. You would not fax 10 separate copies-instead, you might fax one and ask a friend to make copies. Multicasting works along a similar model. A server sends only one copy of information, which can then be reproduced and distributed throughout the network.

TR: How might this be beneficial?
CERF: First, it reduces the unnecessary repetition of traffic on the network for material that needs to get to a large number of recipients at the same time. This applies, for example, to large-scale software distribution, database updates, and securities transactions information.

TR: How close is this notion to a reality?
CERF: Multicasting is already here. A number of Internet service providers, including uunet, are offering multicast services. Last August, MCI and Real Networks announced the establishment of the Real Broadcast Network-an architecture that pushes information content to the periphery of the network and thus multiplies the potential audience dramatically. Before, a live concert that was broadcast on the Internet could reach an audience of only about 10,000 simultaneous users. The Real Broadcast Network makes it possible to fill a much larger “virtual stadium.”

TR: Multicasting seems like a way to turn the Internet from a “many-to-many” network into a broadcast system. Won’t that change the Net’s character?
CERF: Yes and no. While we’ll begin to see many broadcasting applications ported to the Net, we’re not going to lose the many-to-many character that you and so many others appreciate. Also, it’s not as if the Internet has never changed radically before. For many years, the Internet was almost entirely a text-based medium. The Web has certainly altered the character of the Net, but I’m not sure you hear much complaining-although we do need to pay a lot more attention to the presentation of Web content to people with limited sight and hearing.

The Ever-Evolving Net

TR: How can the Internet be changed to make it a better tool for business and commerce?
CERF: Well, there is nothing in the present structure of the Internet that isn’t friendly to commercial transactions. Dell Computer reports sales of $6 million a day on their site, while Cisco Systems reports $2 billion per year on their own. Also, corporations are spending billions of dollars to create private intranets as a way to increase efficiency. Mass consumer spending on the Web seems to be in the early stages of lift-off. One very visible example is bookseller, which recently reported quarterly revenues of $66 million. That company grew to 700 employees in less than a year.

TR: What have you found most gratifying about the way the Internet has evolved over the years?
CERF: When I began my work, I was happy just to have packets travel successfully between points on the network. We didn’t realize that we were developing a blank canvas that future technologists would paint upon with such broad and bold strokes. For instance, while we did do tests in the 1970s with transmitting audio and video, I never dreamed what we’ve seen develop over just the past few years. Today, in fact, we’re seeing many types of media-most notably radio-being reborn on the Internet.

TR: Do you see lessons from innovations past to guide those who strive to refine the Internet today?
CERF: Look at the telephone. The underlying technology has advanced radically over the decades, but in one crucial aspect-the user interface-it has barely changed at all since the 19th century. So I think the most important technological innovations we will see with the Internet will be the ones that devise new interfaces to older services that are richer and easier to use than the simple but constrained interface of the telephone today. Of course, if we are successful with speaker-independent voice recognition, we may find a kind of renaissance in the use of voice as an interface to complex services.

TR: Put yourself 10 years in the future. What is it about the Internet in 1998 that will seem the most quaint?
CERF: The idea of having to dial a phone to make a temporary Internet connection-which is the way many people use the Net today-will seem pretty odd. We’re going to evolve to the point where network access is provided like electricity-in other words, always on. Today, we use a circuit switch (telephone) to get to a packet switch (the Internet). In the future, we’ll use a packet switch to get to a circuit switch-if we need the circuit switch at all.

TR: The Internet has gone from obscurity to near ubiquity in less than five years. That pace can’t be sustained, can it?
CERF: Well, we are a long way from ubiquity. Only about 40 percent of all U.S. households have Internet access. In other parts of the world, the penetration is considerably less. Besides, nothing happens overnight-it only seems to. The Internet today is a result of many thousands of people devoting their entire careers to moving network technology forward, step by incremental step. When the Internet burst into the public consciousness, it was the culmination of a process that began in the 1960s. Looked at that way, the rate of change isn’t necessarily unsustainable.

Keep Reading

Most Popular

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

It’s time to retire the term “user”

The proliferation of AI means we need a new word.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Sam Altman says helpful agents are poised to become AI’s killer function

Open AI’s CEO says we won’t need new hardware or lots more training data to get there.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.