Adam Barr takes to task the insider-y side of the blogosphere’s hypocracy in this article posted to kuro5hin:
Back then you had to be able to write pure HTML to put up a site like that. With modern blogging has come authoring tools that free you from that restriction. So is this the huge breakthrough that we’ve been waiting for since 1993? Of course not. The basic facts remain unchanged: information in a blog is non-selected information, non-selected information has always been around, and it is never going to replace selected information. You can put out all the non-selected information you want, by whatever means you want, and most of it will be ignored. And if you care about sustainable business models, you sure as heck aren’t going to build one with non-selected information.
The whole article is worth reading in full. It has some excellent links and some great follow-up discussion as well (if you don’t mind gossip about Philip Greenspun).
I think Adam may be overlooking the context-building projects, from blogdex and daypop to this new metadata initiative that may at least assist in the selecting and filtering. I do think that human channels of attention will always be the most salient.
I surf blogs as I would surf the web in general, following interesting links. People who often send me on to other valuable stuff get bookmarked or blogrolled and I rely on them for part of my info stream.
So, yes. Blogs tend to be self-published (although Kaus and Alterman do it for publications) and thus “unselected” or “self-selected” but the meta-level of interpretation, through human links and machine aggregation amounts to a layer of selection that can be used the way we might rely on agents, editors, and book reviewers to indicate worthwhile books for us to read.
For example, I just noticed pb’s bookwatch, which tracks book mentions in the blogosphere.