This is Rick Webb's status update that made me laugh so hard at the time that I remembered it and dug it out now as quite appropriate as an illustration of the stuff that I've been thinking about for some time. Which is the problem of filtering information once the adoption of a social networking platform reaches such a number of people that there is all of the sudden more noise than signal(s). Or, the costs of using the platform become higher than its benefits. Or, network effects turn into negative externalities.
We've seen it happening with Facebook, and we've seen it happening with Twitter. When *everyone* started using these platforms, the issue moved from "being there" to how the hell do i "filter stuff out". It's like these platforms stopped being services in themselves and instead became a resource upon which a new, cooler, more useful, and more fun service has to be built.
Here comes the question what do you do with all that data, information, and connections that exist on these social platforms? Some people experiment with services like Qapture, which wants to filter Twitterers by popularity. Needless to say, the list is full of usual suspects in all three categories. I know what those people are saying even without going over to Qapture. So this made me think that popularity is really not a good filtering mechanism. It's nice to be in the loop, but I can do that just by relying on the local knowledge of people in my immediate proximity (in network terms), but that's something that Hayek discovered like 100 years ago. I just need to link to the nearest hub (a person with large number of connections), and puff, I learn about everything everyone is talking about. Something like turning the TV on.
So now, why would I want to do that on the web? There's actually data that is visible and that can allow me a much smarter and more interesting filtering mechanism: the one that introduces a high level of randomness in my network. Or, in other words, that feeds stuff that I like but would never know that I like until I came across them.
Filtering mechanism like that would be a great discovery tool (the one that allows for accidental discoveries), and it would also at the same time weed out all the information that I find uninteresting, irrelevant, or simply a noise. So the problem, how I see it, is not to fight this noise by amplifying the already strongest signals; it's by amplifying the relevant ones.
But how do I know what I will find relevant before I see it? Um. That's part of my thinking about serendipity so will write about it there :) In a bit.