It seems as if iTunes and Amazon, according to Anderson, have done just that. By offering everything (which iTunes does not -- BTW -- try to find obscure '80s British or Canadian pop and you're out of luck!) these companies make more money, but as both Anderson and Derekmention it is not only through the commitment to the long tail that these companies have been successful, but it is through their ability to recommend, link, and connect one song with another based on affinity that sweetens the pot. So, then iTunes and Amazon are finding ways to use power law and small world models together to produce profit.

So far, so good. But then Huberman quickly reminds us:

while economic outcomes are the result of actions by people, they do not necessarily reflect their intentions (22).

Then iTunes can only create links between Beck and Matador at 15 because customers have purchased both, not because there is any inherent link between the artist and the album. This is not shady or even a poor representation of what is going on with purchasing music, but I wonder how useful this model is for explaining how public opinion changes and circulates.

Often liberals are stunned at the inherent contradictions present in the right's agenda, as is the right with the liberals agenda. And there are voices, or clusters, on each side that are infinitely more popular than radical or even moderate sites/blogs on both sides. So if we were to use power laws and small world clusters to predict changes in political climate and belief, could it be done on a mass scale. I am skeptical that it could only show us coronations and that then both sides would come up with their stories to fill in the blanks, so that we wouldn't make much progress on either side. (Ah, the database versus the narrative rears its ugly head again.) So, maybe what this line of questioning is really showing me is that I need understand better what the use value of these models are, and pushing them into predictive and prescriptive modes is really my own need to change the world. And then that's the beauty of the power law -- it could help me devise a strategy to change public opinion, but at the same time it could be helping someone diametrically opposed to me.

And then we are back to linking, clustering, and who is reading who to see the relative success and/or failures of this change.