While I'm livetweeting the Techonomy 2012 conference (www.techonomy.com/#techonomy12/Causeit's list of related accounts), I'm inundated by the hashtags of thought leaders (#bigdata, #p2p, #socbiz, #innovation). For the uninitiated, this little # symbol has begun to be the web's low-tech way to have high-power sorting of conversation. It's a way for users to toss a bright little flag in the rapid-moving waters of web discussions, serving as shorthand, its own poetry form, and a tool for tagging.
There's a great article recently published by NY Times Magazine that talks about the cultural usage of hashtags. I even used hashtags almost exclusively to pare down an otherwise text-heavy presentation recently.
There's another element of hashtags which is extremely important to companies and organizations (or even committed individuals) interested in listening to their ecosystems. It's called semantic analysis—the reading of conversations for meaning. Hashtags allow machine systems a shortcut to sentiment analysis, the analysis of moods in conversations. Linguistic tools look for subtle indicators, and are complemented by users' conscious decision to emphasize a particular statement, concept, or feeling through a hashtag.
You can read a lot more about the wide world of social media for large organizations, and what changes it demands in their capabilities and structures in Accenture's rather dense, spot-on Social Media Management Handbook, which has a great microsite for those looking to dip their toes in.