Google has published plans to build a Wikipedia rival. The key difference: There can be more than one article (a "knol") about a topic. One article is related to one author and people can approve to it by voting. Googles ranking algorithm skills shall assure that people will find high-quality articles.
Now, I think that this has great potential. For over one year, people complain about Wikipedias admin elite. The discussion pages seldom became more than a fist fight over who is right and it is never obvious if controversies exist. The problem seems to lie in the system.
We have to face it: Sometimes, there is no general consensus reachable. That is a wrong idea from the seventies. In five years, we will maybe laugh at the idea that people on the internet could agree on one way of explaining something.
As I write this, I suspect some highly-skilled Googlers are discussing how to rank those "knol" pages. Controversial terms like "facism" will have a lot of articles. When I search for "facism", what will Google show me?
What I am heading for is this: Suppose that 90% of the people agree on one type of definition, while 10% favor another (Note that I say "type" - several articles can all define the term in basically the same way). There is a controversy, and I - as someone interested in the term - am interested in learning about it. I want to see both sides of it and then decide for myself what side I am on.
A stupid, popularity-only ranking algorithm would only show me terms from the 90% bulk. But 10% is a strong minority. I say: Put both sides on the top page, Google. Use this new way of having user rankings and identify if users fall into distinct groups, upvoting distinct types of articles. And then, don't bury the minority opinion. This might be a new way to show knowledge on the internet.
P.S. Another thing comes to mind: Maybe we will have to get familiar with amazon-like greetings concerning opinions: "People who share your opinion on X, also think that Y."