Social networks like MySpace & Facebook have activity feeds of all the network interactions in a users social graph, coupled with photos from sites like Flickr, videos from sites like YouTube and the morass of data from sites like Twitter....

Aggregating all of this creates so much "noise" that the user is overhwhelmed. Their mailboxes and feed aggregators are currently not able to provide a semantic filter on all of this data....

How do we find ways to meaningfully aggregate this data going forward, filtering out what the user ACTUALLY wants to see (through some form of AI) and how can the semantic web possibly help.

Where is this noise overload going, and will users simply switch it off when it becomes too much?