Saturday, June 18, 2005
Search, Markets, and Google vs RSS
Particletree kicks off the tags and RSS and substitutes for search meme. It's a good insight; if you've been checking out bubblegen, we've talked about it quite a bit in the last 6 months or so.
Gartenberg talks about a coming big deal in RSS land, but can't say what it is. I would think Google announcing RSS AdSense is open is a fairly good bet.
Now, more interestingly, guys like Shirky are beginning to talk about the problems with massively distributed production. Tags, RSS, and other decentralized mechanisms by themselves are not infallible, as Shirky rightly points out.
Where I disagree with Shirky (yet again) is that all of these massively distributed systems are, as I point out in my peer production presentation, essentially micromarkets.
Think about del.icio.us: it's a market, where tags are a price mechanism. When you tag something, you're letting others know how you price it (in a binary sense).
Now, what this implies is that if you get the market structure right, decentralized people driven systems (markets) will be more efficient than centralized bits driven systems (search engines).
The real issue for Search 2.0, I suspect, is going to be getting market structures right. How do you do this? Focus on the incentives different structures create.
For example, a simple way for Wikipedia to fix it's edit wars problems is to impose some kind of contribution risk. Similarly, del.icio.us is gonna get astroturfed/spammed (soon), as soon as the benefits of doing so exceed the costs, because there's no contribution risk.
There are plenty of other mechanisms to do so. The fundamental insight I'm trying to draw attention to is that the www 2.0 may resemble a vast set of interlocking markets for information, where the www 1.0 resembled a vast set of interlocking information networks (mail lists, discussion boards, etc).