And I should listen to you because...

Recently, John Dvorak wrote about trends and trend killers, and he talked about Web 2.0, the concept which comprises social computing and the blogosphere. Interestingly, he cited fragmentation as the biggest threat to this thing that is social computing. The theory goes that the more destinations there are for social networking, the more fragmented the social web will remain.

This makes a lot of sense. The value of social networking sites like linked in and del.icio.us derives not just from the "collective intelligence" of the contributors, but also from the individual intelligence and experices shared by those contributors, and, just as important, who they are. This brings us back to the identity question. There is no single authority that can vouch for a person's identity on the web. So regardless of whether a person claims to hold a Ph.D, we still have no real idea of who that person is, or how to gauge the value of the information they offer.

I am unlikely to got to multiple social websites on a daily basis; I am more likely to find one that has other people "like me" and register there. But that means that I may never connect with other people with similar interests and pursuits simply because they chose to register with a different site. So just like in the real world, there will remain numerous fragmented and duplicative networks, possibly existing without knowledge of each other.

As far as identity and collective stupidity goes, I am still trying to wrap my head around this idea of collective intelligence and online content ratings. On a music sharing website recently, I ran across an album "review". In this case, the person simply stated one word, "boo". But this word took up three lines, and was spelled with about 30 of each letter, plus an entire line of exclamation points. This was a singularly unhelpful and unenlightening review, and added nothing to the body of comments about this album. However, the review remains posted, and the 1 star out of 5 the reviewer awarded remains included in the averages. Obviously, this review adds very little discernable value to a reader, and provides no meaningful information which could be used to make a decision.

This led me to think that ratings without context, without the opportunity to comment, can lead to collective stupidity; it seems that some best practices regarding the usage of online ratings can be extrapolated. First, ratings without support or comments are only as helpful as those providing the ratings, a quality that cannot be directly observed. Second, identity adds credibility to any ratings system, or more generally, any mechanism of social networking. Third, having a way to "sanity check" ratings and comments ensures accuracy, and may control for the ulterior motives of those submitting ratings and comments. Since one of the most appealing aspects of the social web is the ability to maintain anonymity if so desired, we should probably remember the first and third practices when identity is not available.

At any rate, I have found that having the ability to see "other comments by same reviewer" is very beneficial...and perhaps in some cases, it might be helpful to have ratings for contributors. For instance, if all of your comments are marked as "unhelpful" by other site users, maybe that gives you a bad rating as a contributor to the site, or at least a mechanism to make sure that those type of reviews are ALWAYS at the bottom of the list.

Just a thought....maybe I am just contributing to the collective stupidity of the social web.