These are some of the thoughts I’ve had in relation to the BloggerCon Overload session led by Robert Scoble. A lot of the discussion was focused at how to navigate the deluge of information that comes from being subscribed to a lot of blogs. Possible solutions raised were collaborative filtering and recommendation systems (such as the attention.xml effort), additional metadata, and filtering by intermediaries. There’s another parallel problem in how to deal with there being millions of blogs, and how do you pick the blogs to read. I think the latter problem is more general. How does the average user end up being able to include blogs in their reading habits? How do we structure a system for end users that doesn’t impose the same kind of intermediary or chanel delivery system that blogs are supposed to work around?

I didn’t really hear any comments about looking at the real end user problem for this. Reading blogs is still very much a geek activity, except the cases where people have set up sites to syndicate other content and that’s how it makes it to the users. How could we integrate a set of tools so that the knitting story that Scoble tells applies to general users? Scoble spoke about looking up some crafts activity (I think it was knitting, maybe I’m wrong there though). He did a search on Technorati and found out who the mavens in that community were based on the linking habits. That’s fantastic! But not the kind of thing we can expect the average user to do. For me the question is more how do we make a system that uses that data like Scoble did, but applies an interface suitable for the average user? Is such a system even realistic? If so is it desirable? To a degree we already have an authority system based in linking in the way that Google does page ranking. So what would be differentiating about the system based on blogs instead of pages? Is the important aspect that we can subscribe to a resource? Or that we can “join in the conversation” thanks to Technorati or trackbacks or feedster?

For me, the major advance in the habit of blogging is the recognition that there is value in loosely coupled group activity. That a group of people working together to discuss and monitor information are able to “emerge” interesting topics to the front of the content flow. That’s a reversal of what computer scientists and programmers normally look for. We tend to ignore the social issues and look for ways to use just the data itself to make new systems. The group action that arises in the blogosphere in general is a result not of one person having a great tool, but of a group of people being able to do little bits and peices. No one has authority, and yet the group can provide useful information from complex interactions. I think if we empower any one part of the system too much that system could break down. And that’s what I really think of when talking about how to make it possible to consume massive quantities of blog posts. We’re thinking about how to empower the parts that have traditionally worked, the A-list bloggers who currently get a lot of attention, so that they can do more. What we should be talking about is how to spread out the activity, not concentrate it. How do we get more people involved so that there is less filtering that needs to be done? The question in this area I think should follow the direction that attention.xml is taking. Not how do I read 15,000 posts, but how do I find the 10 other people who post exactly the 20 stories a day that I want to read? And of course, how can I continue to tune that list of 10 people so I don’t get locked into a static channel?