Chapter 15

THE SEARCHER

Brewster Kahle


THE WEBMASTER (Kip Parent): Brewster is one of those guys who has been successful in spite of the fact that he has never been after that kind of success. He's been pushing protocols for the benefit of humanity in order to make things run better. What he's done with Wais and protocols has been instrumental in bringing about the success of the Internet.

Brewster Kahle is inventor and founder of Wide Area Information Servers Inc., which was recently acquired by America Online, and founder of the Internet Archive.


"The Web is a lonely place," says Brewster Kahle. Brewster and I, meeting for the first time, are splashing around in a pool at a vineyard in St. Helena, California, trying to beat the 110-degree heat. We have both arrived that morning to visit our mutual friend Danny Hillis, who was spending the summer in the vineyard's mansion (home for the TV series Falcon Crest). It is not a community-building technology as it currently stands," he continued.

We are discussing the commercial possibilities of "publishing" or "broadcasting" on the World Wide Web. Brewster is dubious. "The Web was designed as a hypertext system and has been pushed in the direction of a user interface for simple services," he continues. "What is missing is discussion, hobnobbing, flirting. We get these things through email and bulletin boards, but they are stuck in ASCII-only mode. I hope that we move the community services forward in an open environment rather than having separate 3-D worlds that segment users into different proprietary systems. The Web is great in that it is a free-for-all based on open standards. Now we have to move community tools into the graphical and animated world."

Schooled at MIT, Brewster designed supercomputers in the 1980s at Thinking Machines Corporation. In 1989, he formed the Wais (Wide Area Information Servers) project to take commercial advantage of the growing Internet. He wanted to figure out if executives would turn to their computers to answer questions, rather than call someone on the phone or ask their assistant or research librarian. Would they use an online system to find the answers? What he found in 1990 was that they would. The key factor was the networks. That was when the Internet started becoming big.

So he founded a corporation in 1992, out of a joint project between Thinking Machines, Apple Computer, KPMG Peat Marwick, and Dow Jones. Wais was an Internet publishing company that created tools and services to help publishers make money by publishing on the Internet. The company made "surf" software and software for those sophisticated enough to self-publish. It also worked with publishers to create an Internet presence, through the World Wide Web and, before that, gopher. Customers included the The Wall Street Journal, The New York Times, the Government Printing Office, the House of Representatives, the Senate, and the White House, all of which were interested in making their information available on the Internet.

Brewster sold his company to America Online and then worked with AOL for a while as an Internet strategist, helping the company understand how to take advantage of the change from vertical, centralized, online services to a more horizontal Internet market.

Brewster is "The Searcher." He has started the Internet Archive. He wants to create the next great library. "We are indeed trying to archive as much public material as possible," he says. "The estimated space needed to archive the World Wide Web, NetNews, and gopher is between 1 and 10 terabytes. Some technically interesting complexities in this project seem to stem from the different systems we'd need to crawl."

"There are a bunch of legal and social issues as well. Most institutions cannot touch this because it hits every privacy, copyright, and export controversy. I feel like we've touched a raw nerve in attempting this project, since it can change the Net forever‹from an ephemeral medium to an enduring one."

THE SEARCHER (Brewster Kahle):
The Internet is causing different companies to sprout up in many different niches. America Online is a vertical-market company. It creates the publishing tools, the network, the user interface. The customer support is all in one company. There is a real need for the competencies of America Online. What America Online fundamentally does is large-scale operations, computer operations, customer service, and packaging of information for consumers, as well as being a marketing machine. It has the momentum to build an extremely large customer base. In core competencies, it is far ahead of other Internet-based services. A current idea is that AOL and the like are dinosaurs and will die off. I don't see this happening if they are smart.

America Online is not going to be able to keep up with all the technology improvements. What it must learn to do is leverage the other companies that are creating great content for the Internet. I see the Internet having a great technology stream and no revenue stream, and AOL as having a great revenue stream and no technology stream. How to get these two together is what companies are trying to figure out.

At America Online, I was helping to bring the Internet culture into a company that did not see the Internet as a big part of its operation years ago. Today, AOL sees the Internet as a major influence in what it is doing. It sees that half the email messages on AOL go over the Internet gateway. AOL is the biggest gateway to the Internet in terms of mail. Its members are demanding more and more Internet activity. AOL engineers are continually improving the capability of its Web gateway so it can handle more users.

America Online is focused on consumers, on its customers, on its "members"‹what it thinks of as its company. AOL has not concentrated as much on the information providers, and, in fact, a lot of providers are not that happy with America Online. AOL is trying to remedy that by doing things like acquiring Wais and by gatewaying into the Internet, where information providers can have their own way and use their own tools, without having to make specific deals with America Online.

The successful Internet sites are not repurposing data from other sources. They are new breeds of services. The Yahoo!s, the directories of the Internet, are a completely new kind of content. These areas are ones that are really going to grow. How long will text be the dominant form of information on the Internet? I would say it is not the dominant form now. Most people are not using the Internet and the Web for what they were designed for, which was the hypertext linking of documents to documents. They're using it as an interface toolkit for doing actual services, for interacting with customer service, finding things, doing searches. This shows that the interactive nature is really what's important to the Internet, and we are starting to see new capabilities, through Java, to make it more programmatic.

The new exciting areas are the agenting technologies and the movement toward audio and video. The network-news stations put a big dent in the newspaper business, and the same will happen to network news. The CNN site, CNN Interactive, already has textural descriptions of the hot news of the day, and you can click on short video clips. They are slow, but they show where things are heading.

If I look at the classic magazine publishing model, there are three different sources of revenue: advertising, subscriptions, and mailing list sales. All three revenue streams are necessary to make most magazines go. What do we have currently on the Internet? We have a start at advertising-based revenue, but we don't have sales of subscription or mailing lists. The subscription-based revenue stops at services like PSINet, America Online, and UUNet, where people pay for their Internet access. None of that money is paid to the people creating the content, or the Web sites, or whatever software that's making it interesting. A model that took some of the revenue from subscription payments and had a royalty structure paying money back to the content creators who made the Internet interesting would make for a more robust Internet. It would also enable providers with only a small niche to have an easy mechanism to make a little bit of money. We're getting there so far on the advertising model. We have to get the subscription model going.

Mailing list sales seem Big Brotherish to most people, because they give other people lots of information about what you want and what you do in your spare time. This kind of information in the computer world will get even more dangerous as more and more of our interactions can be watched online. For some people, dangerous is another word for opportunity, so this area has to be watched carefully and crafted well with appropriate protective laws. Mailing-list sales will also be used to help find the information that you do want, so there will be good aspects, as well as bad, to having this information out there.

My current project is to launch an Internet Archive that gathers, stores, and allows access to all public information on the Net. The Internet Archive has a project going with the Smithsonian Institution to archive the 1996 presidential election on the Internet, to see how the medium affects the political landscape. In addition, we're collecting Web sites over time and acquired several that have already gone dark. We are contributing those to the Smithsonian, where they will be put into an exhibition that will be on view between the election and the inauguration.

That project is a starting point for something much broader: archiving the whole Internet in order to understand this Internet phenomenon in the future. The interesting thing about the Internet is how fast it's evolving, and the goal of a running archive is to track the changes in the Internet over time. It's important for historians and scholars to be able to look back and understand the Web at different times.

The Internet Archive is not just for historians. It can be an active component of the Internet infrastructure itself by becoming, over time, as critical to the workings of the Internet as the domain-name service or directories. The archive can be a backup for dead sites when Webmasters graduate or when Web sites just go away.

The size of the Internet currently could be somewhere between 1 and 10 terabytes. In fact, no one really knows how large the system is. In ten years, systems of that size will be commonplace and we'll have the old source material. If you have a full copy of the Internet in one place, you can do clustering studies to understand the evolution of communities and their overlaps, as we move from being a global village where everybody is chanting the same theme song from a popular sitcom to having lots and lots of different communities out there. We can track demographic shifts and even experiment with new indexing technologies. A centralized resource may not be the correct long-term solution, but it's a way to get started quickly.

The idea is to do the archive as a commercial-nonprofit combination. That has some precedent, but we will be creating a new model. The commercial entity is extremely important in fueling the technology for the archive, but the nonprofit entity will actually hold the bits. A nonprofit entity garners a certain amount of trust because it's going to be there for the public good and over a long period of time. I am endowing the Internet Archive with enough funding so that any bits donated will be kept alive forever.

I see this in a broader context than just making a time capsule. I'm not proposing that I know how we can build the ultimate digital library, but at least we can start the collection for those libraries that in a few years will become an integral part of our information ecology.

Marshall McLuhan had it wrong when he talked about a global village. With the Internet we're not constructing a global village. We're constructing a globe of villages where every group has its own separate culture. The villages are overlapping and geographically distributed; people can be part of several different villages that span the globe. We've moved beyond the brain-dead nature of the mass media.

THE GENIUS (W. Daniel Hillis): Brewster was one of those people who saw the potential of the Net very early. He realized that the hard thing was going to be making connections between people with content and people looking for it. So he invented the idea of Wais, which is the first Net-searching engine that I was aware of.

THE SCOUT (Stewart Brand): When Brewster shows up, I feel an elf has just arrived. A shockingly effective elf.

THE STATESMAN (Steve Case): One of the early pioneers in the Internet. I first met Brewster when he was still at Thinking Machines and coming up with the Wais protocol, pursuing a better way to search through a growing sea of information. He was one of the early pioneers of building what is now the Internet.

THE EVANGELIST (Lew Tucker): Brewster and I both think best when we're walking, and we spent many hours walking around the Charles River discussing Wais when I was sponsoring it internally at Thinking Machines.

THE CATALYST (Linda Stone): Brewster is one of those very persistent people who had an idea and kept digging and digging and drilling and drilling until he started to get somewhere with it.

THE IDEALIST (Denise Caruso): Brewster is one of the nicest and smartest people I've met in this business. I remember that during one of his many cross-country driving trips‹it might have been when he moved from Cambridge to Menlo Park‹we exchanged quotable quotes all along the way. He also has vision galore. Not only did he foresee the importance of search engines for the Internet years before anyone else, now he's taking on the equally important task of archiving the Net.

THE MARKETER (Ted Leonsis): Brewster's a great guy who's off of his next journey, trying to catalog the entire Web. He's a deep thinker, someone who firmly believes in the merits of Web technology and packaging. Brewster will always be a generation ahead of where the commercial application is. If you listen to Brewster and you have patience, you'll eventually get to the right spot with him. Brewster founded Wais, the first company that understood that publishers would want access to the Web and would need databases and services to post and bring up their content. We needed that kind of expertise and personnel. We are very happy with the acquisition. We bundled it into AOL's operations. Now Brewster's off doing his next big thing on the Web.

THE SEER (David Bunnell): Brewster is brilliant. He might have the best ideas of anyone I've talked to. However, he is also very introspective and somewhat unsure of himself. If he had Bill Gates's confidence, he would change the world.


Back to Contents

Excerpted from Digerati: Encounters with the Cyber Elite by John Brockman (HardWired Books, 1996) . Copyright © 1996 by John Brockman. All rights reserved.