So, Washington Post today had a story on how wikipedia has absorbed the news of Kenneth Lay's death, and (do I even need to say it) it's absorbed rumor & conspiracy theory as well.
I find this entertaining. Not just the wikipedia stabbing (I enjoy that a bit), but that people think it's worth reporting about John Gabriels Theory. I hate the burst bubbles, but the wikipedia has not magically unlocked doors to greater human intelligence/fairness/beauty/truth-iness. It's just an accelerated version of what happens with all kinds of things already studied in different types of group intelligence and decision making. What I'm curious about today is What factors can be measured in different group intelligences, to see how well they work?
My theory of possible factors to measuring Group Intelligence:
- Accessibility: how easy is it to participate in the group for a person?
- Moderation: How much moderation before the input is accepted?
- Correction: How much correction once the input is accepted?
- Investment: what is on the line for the individual if their input is bad/wrong/not liked?
- Measurability: Can the input from individuals be measured against an abstract number (how many jelly beans in a jar?) or is it ascetic (how pretty is this kitten?)
I'm thinking these items could be measured on a scale, and I theorize (off the seat of my pants) that as Accessibility goes up, and Investment and Moderation goes down, that the group intelligence will be less accurate. I think these factors can be used to consider why The Jellybean intelligence works but why the Twin Towers memorial drove people to nearly riot in trying to decide which design to build.
Most importantly (to me ) is the question of how do these factors interact with projects like our own RocWiki?
Also posted at IndustrialSomething.org
