Internet Father Contemplates State of His Creation
Robert Kahn, the man who figured out how to turn networks into inter-networks, says we haven't seen anything yet
P> In a recent interview with Washington Technology, Robert Kahn reflected on creating the ARPAnet, which spawned the Internet, for the Defense Department's Advanced Research Projects Agency. At one point in the conversation, he ran out of a conference room to grab a photo of the dozen or so "founding fathers" of the Internet standing under a multicolored map of the world. Kahn gave a brief description of each person, and his relevance to the project. Kahn said he himself was the techie on the job. After creating the Internet, what comes next? Some say the revolution is over; others say it's hardly begun. As head of the Reston, Va.-based Corporation for National Research Initiatives, Kahn works on managing the content -- particularly through the use of digital libraries -- of the Internet beast he helped create.
WT: When you began work on what would eventually be the Internet, did you picture what we have today?
KAHN: When I started this work in the early 1970s, the real question in my mind was how we would link together different packet-switched networks. It was a technical challenge. It was not a question of developing a multinational infrastructure for the world. I didn't know exactly how this would play out. It was a focused effort to understand the technical issues in linking different nets and computers on different nets. In that time frame, there were only a handful of computer networks around.
WT: What other people were instrumental in the development of the Internet?
KAHN: It's always hard to give credit when something is a success. If something fails, it's easy to find someone to blame, but when something succeeds, it's got a lot of fathers. It never serves anyone well to try to garner credit for things like that. You wouldn't have an Internet with five or 10 or 30 million users if a lot of people weren't involved over the years. Various people have contributed to parts of this. Some have been on the pure R&D side, some on the deployment side, some on the commercialization side. They've all played a role in the process. However, I would specifically cite Vint Cerf, who collaborated with me on the TCP/IP protocols. Also, Dave Clark of MIT, who applied the protocols to PCs and LANs, was an important contributor. Finally, I would cite the people -- myself included -- who were involved in the ARPAnet project and in the development of packet radio and satellite technology. These formed a basis for the early Internet experiments. When you trace the history, there are two main developments worth citing. One is the notion of packet switching. This was a new innovation that underlied the whole development of computer networking. No. 2 is the Internet as the preeminent example of an open architecture network. For the first time, it demonstrated that you can build a network out of pieces that are independently contributed.
WT: What are you working on today that will have the impact that your work 10 or 15 years ago has now?
KAHN: The whole field of managing content in a network environment and the use of remote network resources is something I am working on. When we were first building networks, we had somewhat of a traditional phone company mentality. This was apparent only in retrospect. You know, here are these bits, and how do you get them someplace else. The post office has the ability to move packages from A to B, but without regard to their content. Now by analogy we're trying to figure out how to access stored content out of warehouses and how to make this value-added service more useful. I really think of it as the next step up in networking where you're dealing with the ability to manage content. That's going to have a profound effect in the next 10 years or so.
The next thing will enable people to use this powerful combination of content and resource to form virtual communities to carry out collective functions.
WT: What are the best uses of the Internet so far?
KAHN: On a scale of one to 10, we haven't seen anything yet. Despite the fact that you can do a lot on the Internet today, it's still a tiny, tiny subset of all the things you might like to do on the Internet. Either the speed isn't there, the delay is too long, or something hasn't been invented yet. Electronic mail is useful, although you can get overloaded, as I have. That's a positive and a negative simultaneously. The fact that you can't move large high-resolution images quick enough or motion picture video around easily on the Internet is a disadvantage, but if you could it would be both an advantage and a bigger disadvantage because you might clog the Internet with a lot of stuff you didn't need. Percentage-wise, it's very hard to find things of real value on the Internet. The amount of nonsubstantial material on the Internet is growing at a faster rate than the substantial stuff. If people had tools to ferret out what was substantive, that would be a real plus. We still don't know much about how to search and find information. If somebody says, 'read the paper I wrote last year about such and such,' and you happen to know enough to find it, that's fine. But what do you do if you can't remember the title, and you can't remember the author or any distinguishing words that don't hit 2 billion other titles?
WT: What is the research area that interests you the most?
KAHN: I'm quite interested in the applications of networks to the management of content, what I call the digital library area, which is really closer to a marketplace for content in my view. I've been spending a lot of time on the issue of copyright management. I'm very interested in the use of networks to aid in manufacturing productivity, getting designs for things on-line.
I'm also very interested in the idea of a national knowledge bank where you create codified knowledge and make it available to people and programs. That's really the same as content. But there's no infrastructure for supporting this. Knowledge, especially knowledge of the technical sort, comes out of scientific and engineering disciplines. You've got a healthy dialogue going when people publish ideas and criticize them. The same thing needs to happen in knowledge structures.
WT: Who should govern the Internet?
KAHN: The Internet is an exchange medium for access to information and other resources. The first order of responsibility for governing the Internet ought to be modeled after what happens in other analogous situations. If you asked what's similar, I'd say the ether that we deal with is similar for things like speech. Who governs what goes through the ether? The answer is the Bill of Rights. Everybody can say what they want, but there are some exceptions. You cannot libel people in public without expecting some legal recourse. Just like you have certain rights or expectations of privacy, people can't arbitrarily come and disturb your space. There's a lot we can learn from what goes on in the area of individual rights in the rest of society.
WT: There are so many companies getting in the Internet business -- cable companies, telecom companies, Internet access providers. Is there room in the long term for all of them?
KAHN: It's almost always the case when you have a new and burgeoning area that you get a lot of initial diversity. It was true in the automobile industry; it is still true in the software industry today. Some entities will move at faster rates than others, better products will do better in the marketplace, more efficiency will usually win out over less efficiency. You get certain economies of scale as things grow. The larger companies buy out the smaller ones because people are willing to let go of what they've developed for bigger returns; they can move along to work on ideas they want to pursue. The process of evolution tends to introduce a need for structure as time goes on. The only difference from other fields is this one tends to move at a faster rate. I think a shakeout is now happening continuously.
WT: Do you think enough money is being spent on R&D?
KAHN: First of all, I think there is a key continuing role for government to sponsor advanced research and development -- to fund the things that wouldn't happen in the normal marketplace. Ideas that aren't tightly tied to near-term product development. Leveraging opportunities for which the risks are too high for industry to act on them themselves. That's how society makes progress. You never want to turn off a mind. You never want to shut down productive intellectual endeavor. I think we can make productive use of more R&D funds. But if we can't afford it, it's a very different issue, and sometimes you have to bite the bullet.
During the Cold War, there was a solid agreement on support between the government and the scientific community. I hope that will continue into the future, but I think it needs to be renegotiated.
WT: How do you personally use the Internet?
KAHN: I have my hands full managing the organization. A typical day for me is meeting with people. If I could manage to keep my Internet use to an hour a day I would be very grateful. When I do spend time on the Internet, it's generally to respond to electronic mail. My assistant scans it and flags the things that are important. I used to do that myself, but then I started getting 200 or 300 messages every single day. About a year ago I was off the Internet for several months. When I got back, there were so many messages waiting in the system that knew I was never going to catch up. So, in effect, I hit the delete button. I use the Web from time to time to pull down something that someone suggests I look at. Unfortunately, I cannot afford to spend hours going through e-mail or looking for material on the Internet, though there is a slew of interesting stuff out there.
WT: What changes are coming up in the next year for the Internet?
KAHN: What's going on is the inexorable process of commercialization and internationalization. I think the Internet will become an accepted part of the infrastructure of the world, unless it doesn't. Various countries might take steps that could be hostile to its evolution or even existence. In Germany, they banned certain content. If they could ban that, they might ban other things. People might worry about interfaces functioning across borders. Things could become fragile if we don't fight for them. I'm hopeful that won't happen. People need to be vigilant and fight to defend the openness that exists in the network environment. If that exists we will have an Internet as a piece of the infrastructure for a long time.