2006/07/28

wikipedia

I just read an article about wikipedia (New Yorker magazine, July 31, 2006. Alongside the usual praise, it captures some pretty good criticisms:

"Wikipedia has gone from a nearly perfect anarchy to an anarchy with gang rules."

"...infested with moonbats."

"...the open-source model is simply inapplicatible to an encyclopedia. For Software, there is an objective standar: either it works or it doesn't. There is no such test for truth."


Hmmmmm. Aren't we forgetting that other pillar of epistomiological virtue? Popper would say there's natural selection, the market, and ... voting.

So why not add voting to Wikipedia? Like Digg, each article would be voted on by users, and therefore have a score. Each reader would get a binary vote, thumbs up (+1) or thumbs down (-1). Uniqueness of votes can be easily enforced by IP address for each article. The article's score is then simply the sum of all the votes (+1 and -1). This score would of course help the reader adopt the appropriate degree of scepticism.

But the score could also be used in a derivative way. One of the problems on Wikipedia is situations where two contributors get into a big battle if repeteadly deleting the other's changes because say the article is on a controversial topic, or simply because whoever is wrong is stubborn or fanatical... Well each contributor could have a an editor rank based on the value they have added to articles they haved edited in the past (which would be something like the average change in score of the articles before and after their edits), and presto, in a dispute, the person with the higher rank, which should the one proven more reliable over time, wins! Truth wins, and Wikipedia lives happily ever after....

Seems simple enough... Why not? Imagime Ebay without the seller rankings... Yaiks.