Death Knell for Associations?

Back at ASAE's Great Ideas conference in February, there was a lot of talk about social media--and a lot of talk about how social media could eventually spell the death of associations. Since I work for an association and don't relish the idea of being out of a job, I didn't give it much thought at the time.

Fast forward about a month. I get an email telling me that radRounds is following me on Twitter. Never having heard of radRounds, I clicked the link to their website. And there it was--what could well be the future of associations.

I preface this by saying that I could absolutely be wrong about radRounds--it could very well be a product of one or more of the many radiology associations out there. But as far as I can tell, it appears that the radRounds is not affiliated with an association; this is from the "About Us" section of the site:
radRounds was founded on the eve of January 2008 by a group of radiologists who saw the need for a collaborative, open, and easy-to-use web tool for networking, learning, and collaborating. We are completely managed and created by both informatics experts and physicians who specialize in the wonderful field of radiology and medical imaging.

Networking, learning and collaborating? That's pretty much what associations are for, right? Think about it. Compare your association's website to radRounds: Industry news? Check. Forums? Check. Career center? Check. CME course listings? Check. Groups, case reviews, file sharing, podcasts? Check, check, check...you get the point.

RadRounds seems to offer a LOT of what traditional associations offer. Granted, a lot of what they offer--guidelines, for example--may be created by associations--but really, is an association truly a necessary middleman for stuff like guidelines? With social networks like RadRounds, where professionals can connect and collaborate--what's to stop them from eventually creating their own guidelines?

Again, I am NOT an association executive and I know there are MANY aspects of associations that, conceivably, at least, independent groups couldn't easily re-create on their own. That said, exactly how many functions is that? Enough so that associations have nothing to fear? Or few enough that a dedicated bunch of members could pretty much replicate most functions of an association, thereby eliminating the need to pay dues and support staff and overhead?