Today I read a couple of things that made me think about this topic. First is this article in the NYT by Robert Shiller on the housing bubble. He talks about a study on "information cascades" and how this phenomenon could potentially cause a bubble through influence propagation across a market (in this case housing).
It reminded me of this article, also in the NYT, by Duncan Watts. In this case, they found that people's music ratings and purchases were highly influenced by those of others that had rated/bought before them (as compared to people's ratings/purchases when they could not see others' choices). So in scenarios that allowed for influence, hits became bigger and also more unpredictable.
Finally, this months Edge, has a talk/video with Nicholas Christakis where he talks about these topics and several others, all in the context of current and past research on social networks.
It is clear that social networks and markets can trigger interesting and unexpected phenomena. What has not yet been figured out is the extent of these phenomena, their magnitude and how to harness them (if even possible). This is the root of the excitement around online social networks like Facebook and why no one really knows how much they are worth. It isn't about their number of page views (even though they are huge) and their direct monetization through ads. Rather, more importantly, if they can start to effectively harness the power of influence across networks they could be worth billions. If not, they could be just fads, doomed to be replaced by the next trendy site.
Following up on this, my next post will be about Facebook monetization.