Primal’s Primal Pages, announced at the 2010 Semantic Technology Conference, tries to create a custom webpage that contains the exact information you’re looking for.

It uses Primal’s “knowledge model” to figure out which content from sites like Wikipedia, Flickr, and Amazon would give you the answers that you’re looking for. Then you can tweak your results by looking at “related pages” or “deeper pages” about narrower topics.