When WebLogs started spreading, many people felt that this would make everybody a publisher, and everybody a reader. And this has in fact happened, more or less. An unforeseen development, however, is that human nature seems to compel us to read the same weblogs everybody else is reading. The result is that a very small number of weblogs is read by a very large number of people. Everybody publishes, but almost everybody reads what a few big-shots have to say.
Attention is concentrated on a very small number of individuals.
It seems hopeless to change human nature; but perhaps this tendency can be channeled using technology.
The thing to avoid is GroupThink -- social dynamics that reduce the number of options and channel individual opinions into well-trodden paths. On the one hand, this kind of harmony prevents society from being torn apart by diverging opinions, but on the other hand it also reduces criticism. It effectively prevents an OpenSociety where people can make informed decisions about their government. So the negative aspect of Attention Concentration becomes visible when the number of sources for a particular topic is smaller than possible. If there are only 10 independent sources for a certain conflict, for example (3 major news networks, 3 involved parties, 2 NGOs, and 2 famous bloggers, for example), then decisions will be affected by a very small number of people. The important thing is that this small number is smaller than the number of good sources.
All this says is that if a small number of corporate news channels (media concentration) is bad, then this applies to blogs just as to any other news source. If we substitute 5 corporate news channels reaching 90% of the adult population with 5 weblogs reaching 90% of the adult population, then the situation is as bad as before.
Potential ways out:
As I understand GroupThink, it's primarily a small group phenomenon. Classic examples include corporate boards and government task forces, where people who disagree can face direct and immediate punishment. It's very difficult to enforce at the level of an entire society. Even if you resort to police state tactics, you'll only drive dissent underground. In an OpenSociety, someone is always going to point out that the Emperor has no clothes.
Addressing your example of AttentionConcentration directly, anyone who can read the two famous bloggers can also read dozens of independent news reports from all over the world. If GlennReynolds? thinks Instapundit is more influential than Reuters, he's deluding himself.
The difference between corporate news channels and blogs is that the 5 weblogs have no ability to drown out the thousands of others except through quality and/or quantity of content. MediaConcentration? happens because the big guys put the little guys out of business. AttentionConcentration happens because no one has more than 24 hours in every day, so everyone is going to read a handful of trusted sources. One is source-driven, the other is audience-driven. -- KD
It's a question of Balance. There's a natural tendency for human beings to want to be connected to one another. We're monkeys, social animals. But there's also a creeping sense of wrong when everyone looks at the same thing. Evolutionary dynamics tend to take care of this in a rough and tumble way: when everyone's doing the same thing the system is very vulnerable to the unexpected and some sort of failure is inevitable. As noted above, a few bankruptcies are almost certainly in the mail. Weblogs (and related technology) seem to be reversing the over-concentration brought on by TV. People are starting to pay attention to more real things that really connect them. #joiito probably has a lot more to do with the participant's lives than an episode of Friends. That's probably a good thing
The AlertBox ([Diversity is Power for Specialized Sites]) by Jakob Nielson explores this a bit more, by refuting some arguments presented in a recent New York Times op-ed piece, [More News, Less Diversity], by Matthew Hindman and Kenneth Neil Cukier. They claim attention concentration on the web. Nielsen, on the other hand, doesn't want to examine whether some sites dominate a single topic, but whether these same sites also dominate other topics. And that is not the case; there is no overlap between sites, and specialized queries lead to different sites (niche within a niche). Nielsen finishes with More proof that sites might be big in some contexts, but are rarely big everywhere.
TechnoRati had a section listing the top 50 interesting recent blogs; it showed blogs with a significant number of links (40-50) that just gained a lot more links. It thus filtered out blogs with less readership, and also filtered out the famous blogs.
TechnoRati allows you to find the blogs that link to you -- your "cosmos"; TrackBack works in a similar way. This kind of technology allows you to engage in a dialogue with other people, even though everybody only has his own WebLog, because you get notified on your own weblog when somebody links to you.
The problem is that you might only get notified of the start of the conversation. If you want to continue the conversation, you have to constantly refer to the postings of your partner, in order to notify him of your recent addition to your own blog.
Personal InternetRelayChat (IRC) channels is something JoiIto tried ([Attention Concentration and becoming a place]). The negative aspect is that this reinforces attention concentration - people will join famous people's channels. The positive aspect is the personal touch. Joining personal IRC channels is like visiting other people's homes.
Since publishing costs are so low, even those that cannot gather a large audience can publish, and therefore at the lower end of popularity, the individuality of sites grows in importance. And that is a good thing.
Similar things happen in music. Unpopular experimental bands have small audiences. If producing music is cheap, however, even these bands can get published.
I don't know whether the history of mp3.com shows that it works or that it does not.
Wikis are rooted in their communities; the individual control of the founder is weaker than on a WebLog or on a personal wiki. This has the benefit of acting like a public space, encouraging collaboration, without abandoning the idea of responsibility (see AbsentLeader and related problems).