Crossing the Chasm

A chat with Robert Kahn, who has been called the “Father of the Internet.”

Amidst the tree-lined office parks of Reston, in the technology hub of Fairfax County, works the man who helped to make it all happen. Bob Kahn is often called the “Father of the Internet,” and it’s a fair description of his groundbreaking information-technology career. Kahn, who earned both a master’s and a Ph.D. from Princeton, worked at the legendary Bell Laboratories then became a professor at MIT. He left that university in 1966 to join Bolt Beranek and Newman, Inc. where he was responsible for the system design of the U.S. Department of Defense Advanced Research Projects Agency Network (ARPANET)—a seminal effort in the 1960s and 1970s that pioneered packet-switching technology (enabling multiple computers to exchange data simultaneously) and thus set the stage for the Internet and all of its subsequent manifestations.

In late 1972 Kahn moved to DARPA, the Defense Advanced Research Projects Agency (essentially ARPA, renamed) and there he and Vint Cerf created the TCP/IP protocol that enables different computer networks (and computers connected to them) to “interwork,” regardless of their specific hardware or software configurations. In essence, the two men laid the technical foundation for email and file transfers, not to mention the World Wide Web and the various social networks that enthrall us today. While at DARPA, Kahn established the U.S. Department of Defense Strategic Computing Initiative, a 10-year, $1 billion-dollar project to create high-performance computing capabilities and selected applications, as well as to grow the Internet infrastructure within the DARPA research community. In the mid-1980s, he coined the term “National Information Infrastructure,” which later became more widely known as the “Information Superhighway.” For his work, Kahn has received a score of prestigious awards including the Presidential Medal of Freedom in 2005, and in 2006, he was inducted into the National Inventor’s Hall of Fame.

But Kahn isn’t ready to rest on his laurels just yet. In 1986, after 13 years at DARPA, he founded the Corporation for National Research Initiatives (CNRI), a Reston-based non-profit organization that promotes collaborative activities in information technology among government agencies, universities and private organizations. With CNRI, Kahn hopes to foster research and development for the national information infrastructure.

What made ARPANET so integral to the development of the Internet?

It was the first real computer network. It demonstrated a technology called packet switching, which was far more efficient for computer communication than circuit switching that was the standard at that time. It had many properties that were essential to computer communications. It was really a statement about both the future of communications as well as the future of computers.

What were the greatest challenges in the Strategic Computing Initiative?

At the time, there never had been a very large or significant computer research initiative by the federal government. There were many small projects, but the idea that we would be able to effectively use large amounts of money in the computing field hadn’t really sunk in at that point in time.

We were trying to do multiple things at the time, including outfitting the research community with enough equipment and facilities so that they could make use of the then-nascent Internet. There were a lot of demands for funding beyond what was normally encountered in a typical computer research program.

Did you feel that it was an uphill battle to get funding, and to get people to realize the possibilities of the Internet?

By the mid-1980s, a significant part of the research community had realized that there was a benefit, a real significant benefit, although the world at large hadn’t yet really discovered the Internet.

I remember that the magazine InfoWorld gave the Internet its product of the year award in 1993, twenty years after we had started research on it, because that was the first time it became widely visible in industry circles. There were several reasons for that, not least of which was the passage of what I call the “Boucher Bill.” Rick Boucher, a Congressman from Virginia, allowed the National Science Foundation’s network to be connected with commercial networks and used for other than typical, research-type activities. That really opened up the Internet to the public at a time when the World Wide Web was also being deployed.

How much time do you spend on the Internet?

Well, the thing I use the most is email. I get a significant amount of email each day. And occasionally I do Web browsing, but usually for specific pieces of information because the Web has become very useful for that purpose. It’s hard to know exactly how much time I spend; it’s easy to become too dependent on it. I’m not prone to want to sit in front of a terminal for eight hours a day.

I spend most of my time interacting with people here actually, since my role is a management job. We’re not having this interview on the Internet. Most of my interactions are person to person.

Do you have a blog? Do you ever respond to blog posts or stories on the Internet?

I do not maintain a blog and usually refrain from responding to material on the Internet, other than personal email, unless it’s really important that I do so.

The Internet has truly manifested the term “vox populi,” in that anyone with a PC can toss out a thought or opinion, expert or ignorant, on any topic for all the world to read. Is that a good thing?

I’m a strong supporter of free speech, and if one cannot communicate effectively, it’s hard for that speech to be conveyed adequately. The bigger challenge to each citizen is to then sort things out. When virtually everything is made available through formal professional channels, a certain degree of careful pre-selection and professionalism in expression takes place before it ever gets published, which improves certain things in many ways. On the other hand, it’s not the same as complete individual freedom to express yourself in the same marketplace.

Is there a downside to the Internet—having such a vast trove of information at our fingertips?

Well, I suppose, if you focus on the information that is harmful or wrong, or the information that is inflammatory or inciteful, then, yeah, that’s a downside. I mean, it’s true of information of any form, whether it’s on the net or on paper.

But how can you be against having the ability to access more information? Knowledge is something we seek to create, develop, understand. I think the more opportunity you have to access information, the better informed you can be—if you use it wisely.

You could have a debate about who or what to trust as an information source. We always have a responsibility to form our own judgments. We can’t be everywhere in the world at once, so we have to rely on sources to keep informed about what is going on and make our own judgments about the validity of the source.

Where do you see Internet applications and personal technology going in the next five to ten years?

Innovation is something that persists in the human spirit. People will always have interesting or new ideas. Given the Internet, begin to think of all the new things you can do with it. In the early stages, we weren’t trying to figure out all the possible things people could do with it; we were just trying to figure out how to make it happen. We were trying to cross the chasm. The challenge was how to get to the other side. That’s very different than trying to figure out what you’ll do once you get to the other side. Once the bridge is there, then you can begin to think about the opportunities for building on that.

Most of what is going to happen, I think, is going to be a big surprise to many of us—things none of us thought about. Or maybe we did, [but] just didn’t think had any potential. Take the example of some of the new services today where you get short messages, and you subscribe to them. Well, the military did that years ago, for communications on the battlefield. The fact that that could be viable as a commercial service probably didn’t occur to most people. But there it is. Social networking is something that may or may not continue to evolve the way people envision it.

Do you think more restrictions are inevitable? Will the current period become known as the wild west of the Internet?

The wild west [connotes] an awful lot of lawlessness. I don’t know that what we have today is lawlessness… My guess is we probably will have more regulations. I don’t know if that is necessarily all bad; it depends on what the rules and regulations are trying to deal with.

It’s remarkable to me that people would try to misuse the Internet as much as they do. I would not have predicted that. But then, in hindsight, I guess I should have expected it. And when people do misuse the Net you have to take steps to prevent that where you can, though maybe not through direct intervention [such as] laws. It could be charging for things that are now free. Cyber security is one area of concern, and spam is one of the biggest issues that we now have. I think most of the spam would go away if the spammers had to pay significant fees to send the spam. On the other hand, if we charge for email, that might inhibit many other valid communications. So it’s a balancing act.

You’ve expressed your dislike of Net neutrality, the idea that all Internet traffic should be treated equally no matter who uploads or downloads data or what kind of data is involved…

Actually, I have not. What I have said is that I have never taken a public position on Net neutrality, since I viewed it largely as a business matter. I think that there are things that we need to focus on very strongly. One is to enable research to take place whether it’s internal to the communication network providers or external to those networks; and another is not to take steps that will cause the Net to fragment.

The problem with Net neutrality is that it is still primarily a slogan. If you ask somebody if they are for or against Net neutrality, it would be like asking, are you for or against political neutrality. You might wonder what that means. Parties with different opinions on Net neutrality might both say they are for or against Net neutrality because they interpret the term differently.

Now if you were to ask me for specific views, I’m against discrimination against customers as a general rule. If one party’s traffic is not permitted to go through at all, or is seriously delayed at the expense of another’s, and both have reason to expect similar service, that’s normally a bad thing. I think you’d have to look at the specifics.

You have been described as the “Father of the Internet.” Is that an accurate description of your role in its creation—and if so, how do you react to that phrase when you hear it?

I don’t like the term “Father of the Internet” since it does not speak to what I actually did. Indeed, with the success of the Internet, many others whose work had little to do with what I think of as the Internet often use that same descriptor for their work. Of course, I’m proud of the work that I did—much of it jointly with Vint Cerf—in helping the Internet to become reality, but some of the most important work we did involved developing and managing the Internet standards process and the associated social structures that enabled the Internet to grow and flourish in a commercial setting and around the world.

What does CNRI provide that is missing from traditional funding sources?

I don’t think we’re really a funding source. We try to formulate and then enable and carry out projects that are called research initiatives, which address national infrastructure capabilities for the future. A highway system is an infrastructure for transportation; the Internet is an infrastructure for computer applications. Those are the types of projects we formulate and carry out. We do it in a non-profit context simply because we want to bring all the relevant, and usually diverse, parties together.

If you have different groups investing in things they find interesting, when they’re all done you have a number of programs that are not interoperable. What we’re trying to do is recognize those issues upfront and deal with them and raise the funds from wherever we can.

We may even find that this is true with, say, electronic health records. If one organization builds one system, and another builds a different system, how will they interoperate? Agreeing on standard record formats is just one possible first step toward interoperability.

You have received a lot of impressive IT, engineering and other awards. Which mean the most to you?

I’m proud to be a member of the National Academy of Engineering and a Fellow of several professional organizations in my field. I could cite many awards, but let me limit my choices to the two national medals [the 1997 National Medal of Technology and the 2005 Presidential Freedom Medal], which are particularly important to me, as is the Turing award that is perhaps the highest award from the computer science community, and, of course some of the international recognition such as the Japan Prize.