March 14, 2000

May not be reproduced commercially without permission of the author

I've always maintained that a metaphor for the Internet is rock music, with trends and stars emerging in unexpected ways. Last week, I attended an event called "Live Cluetrain," which reinforced this metaphor in many respects.

The event featured, in concert as it were, two of the authors (three, if you count the guy who wrote the forward) of the essays in a book called "The Cluetrain Manifesto." It is one of the books I have reviewed in the Short Book Reviews which are linked on my web page.

So here are the authors of these random essays, coming out with a few more random comments (my personal favorite being Chris Locke's offhand remark that education and government will catch on about one hundred years after business). And how many people are in attendance?

Would you have guessed over 1000?

David Weinberger, one of the authors, asked what appeared to be a rhetorical question. "What's the Web for?" The complete question, which I copied from the transcript available on Netpreneur, is this:

"A telephone is for calling people. Highways are for going places. What is the Web for? I haven't heard an answer. Why does the growth curve look like a hockey stick with the fastest vertical acceptance line since the discovery of fire? Why are we rushing to this thing that we have no idea what it's for? Something deep is going on."

At the time, Weinberger seemed to me to be saying that we do not know the answer to the question. In fact, he appeared to imply that anyone who claims to have an answer thereby reveals himself to be a fool. Foolhardy though I am, I was sufficiently intimidated that I did not raise my hand.

But here is the answer:

The Web is for decentralized collaboration in the dissemination and organization of information.

This answer helps to align the Web with the Internet itself. The Internet Protocols facilitate decentralized communication among computers, and the Web facilitates decentralized communication of information using those computers. The Web's hyperlinks are analogous to the Internet's routers.

I think of the essence of the Web as hyperlinks. Whenever I create a hyperlink, I am engaging in the dissemination and organization of information. Because I need no one's permission or authority to create hyperlinks, the Web is decentralized. (Actually, I believe that a few years ago Ticketmaster threatened to sue if you linked to them. If there were a meter to measure cluelessness, they broke it.)

What makes the Web tricky is the "collaboration" part. How is decentralized collaboration supposed to work? The people who are trying to solve this problem are the ones who are aligned with the technology.

In contrast, there are still many people who are fighting the technology.

--Companies focus on capturing information (the data mining fad) rather than on disseminating it.

--Web sites that are poorly designed from a user's standpoint, thereby frustrating the organization of information.

--Portals, including AOL, that try to "control" the user's path, thereby frustrating the decentralization of information.

Off the top of my head, here are some examples of decentralized collaboration:

--the original Yahoo hotlist, in which users created the links and the sub-categories. It could be that the Netscape "open directory" project will re-invigorate the original Yahoo approach.

--Amazon's affiliate program, in which individuals who mention books can link to Amazon's page. Amazon's on-site book reviews facilitate some decentralization in the dissemination of information, but the fact that Amazon deletes any links out from the reviews is an inhibiting factor.

--Many of the individual web sites listed on my favorite links page. See particularly "Tomalak's Realm" and "Arts and Letters Daily."

Going forward, I believe that the important improvements in the Web will come from:

--XML, which facilitates decentralized descriptions of information. In addition, I continue to believe that decentralized, overlapping directories (overlapping in the sense that an Internet Business based in Silver Spring could appear in both a directory of Internet Businesses and a directory of businesses based in Silver Spring) will emerge to overcome the weaknesses of the central portals.

--tools that allow people to rate products, services, web pages, and so forth. Some people rave about epinions.com, but I personally do not yet feel confident about the information I obtain there. I am not sure that there is a single process can be used to rate everything effectively. In some cases, the "mass" opinion should get the most weight. In other cases, expert or idiosyncratic opinion might be the most valuable. My guess is that ultimately the rating function needs to be decentralized and specialized, just as I believe that the directory function needs to be decentralized and specialized.

--web sites that are less complex and more narrowly focused than the current sites. My guess is that Amazon's attempt to branch out into other areas of commerce ultimately will prove to be a huge mistake from the standpoint both of their internal management and the user's experience.

--"groupware" tools that reside on the Net. Many people now find that their desktop PC is a "ball and chain." The next step in decentralization is to enable people to organize and disseminate information from anywhere.

Overall, I believe that my proposal to say that "the web is for decentralized collaboration in the dissemination and organization of information" is a good working answer. My prediction is that companies that align themselves with this model will do much better than companies that try to fight it.