A recommender system is an information organization tool which extracts knowledge of individual users of a specific (online) resource based on their activity within that domain, and uses this knowledge to generate for them individual recommendations. These recommendations are made based on the broad assumption that people who have agreed on some things in the past will likely agree on things in the future.  Because these systems classify content based on how it is engaged with by previous users, they have emerged as an effective (and profitable) way to organize content on the web, the web itself being resistant, almost by its very nature, to the imposition of top-down, ontological classificatory control.  This paper interrogates some of the assumptions and biases involved in the personalized organization and presentation of digital media content, in an effort to devise a more critical analysis of the economic, technical, and social forces that contribute to the growing popularity of the personalized recommender systems. It pays particular attention to how personalized content filtering finds expression in the recommendation engines of taste-coordinating websites like Amazon and Pandora, in the self-organization of information through social classification sites like LibraryThing and Delicious, in the adaptive capabilities of next generation search engines (what Michael Zimmer refers to as "Search 2.0"), and finally within locational media software like FourSquare. Discussion of these recommender systems will refer to their mechanics as well as their accompanying rhetoric, which often associates the personalized delivery of content with a more empowered individual user. Promises of individual empowerment attached to emerging expressions of personalized media are generally made in opposition to the broadcast media model of the mass market. The concluding section of this paper will examine the rhetoric of these promises in a discussion that considers such novel adaptations of content production and delivery as a market response to current media configurations. These adaptations, it will be argued, serve in large part to define differences to be "commercially exploited."  The nurturing of differentiated markets represents a potential challenge to some of the very characteristics that have traditionally been associated with the internet's social and political potential.