I, a self-described Twitter detractor, am becoming more fascinated by its possibilities of late. Dan Miller my colleague is a big Twitter fan by contrast.
I recognize that Twitter has a number of useful functions, beyond news dissemination and entertainment, that mostly have to do with companies promoting their sales, deals or events (as reflected in this recent article showcasing Dell's use of Twitter).
Many people are caught up in the question of what will Twitter's business model become or what will Twitter ultimately be able to charge for?
I'm not so worried about Twitter being able to find a business model; I think there are probably several revenue streams in waiting.
What I'm focused on is the structuring or organizing of the content into more useful and accessible frameworks. There are more and more feeds to keep track of, whether Twitter, Facebook, FriendFeed, etc. In a way this is like the proliferation of IM platforms. But beyond simply aggregating all these feeds there is the challenge of extracting useful information from them.
AOL (MapQuest) has done something interesting in bringing all the geotagged Twitter feeds together:
As you can see there's potentially great value and local recommendations here: what to do, where to eat and drink, where to stay. But there's a fleeting and ephemeral quality about Tweets and similar feeds. How does one turn something like this into a recommendations network (a la Mosio) or make the data available for later discovery (a la Yelp) when users are really looking for the information?
Zannel, which just launched for the iPhone, tries to bring Twitter, Flickr and video together around location with its new CityWatch app:
This app is far from perfect from what I can tell, although search is going to be added to a subsequent update. That should go some distance toward what I'm suggesting. But in general CityWatch starts to point at something that can cull valuable infomation from feeds about local recommendations and improve the signal to noise ratio.