On Emergent Intelligence...or Submergent Stupidity

We see group wisdom in all sorts of situations...but we also see group stupidity. The ironic thing, as Surowieki describes in his talk, is that it appears the more independent and silent the individuals in the group, the smarter the crowd. Like bees, each individual is stupid, programmed with just enough intelligence to do what they have to do, and the sum of all the small, mindless (sometime errant) moves is a complex of intelligent activity that gets a hive built. The wrong moves are averaged out and the right ones are reinforced. If the members of the group could attempt to communicate and actively try to solve a problem, they would get nowhere, because no individual is equipped to solve the problem on their own... I think this perfectly describes our Congress! They should all just shut up and vote.

On the internet, I think, it is similar. Digg, for instance, largely does give accurate information in the same way that bees build an accurate hive. If one person appreciates a news story, they presumably think it is interesting and accurate so they digg it. They may be wrong or they may be right about its accuracy, but other people will make up their own minds and digg or bury it. As the numbers add up, it becomes increasingly more likely that the sum judgment will be a correct one. The sum of a group of low-intelligence decisions is usually a very wise decision.

The problem with this is that, as Surowieki states, we don't only base our decisions on our own information, but we instinctively imitate others. So if a hundred people digg a story, its much more likely we will believe it is valuable before we even read it. In fact, its likely we will only see the stories that have been dugg; there are thousands of great stories out there that, it just so happens, were not posted. Our perspective is skewed, hence our decision is skewed, hence the group judgment will most likely be skewed. In the case of Digg, this is not such a big deal, since it is really a popularity engine, not a correctness engine (a fake story can still be popular, right?). The content being voted on is already published, and (presumably) verified news material.

Wikipedia is a different story, since articles are themselves written by the public it really is a correctness engine. Someone could write an entirely fictional entry and publish it on Wikipedia, and if it is esoteric enough (say, on the wonders of denatured alcohol) it could conceivable exist for a long time. The good thing is, as I see it, most pranksters are out for impact, so they create a fake story on popular topic (like the death of a Senator) that everyone will see, and in ten minutes the deception is revealed. Even my postulated fake denatured alcohol story would eventually be discovered and corrected by a real chemist. In this way, Wikipedia is more policed and reviewed than it may seem at first, and with time the contents do become more and more accurate. Will it be more accurate than a published encyclopedia? Not necessarily, but that's not the point. Britannica may have been painstakingly researched, but by the time the information gets out it may have changed. Wikipedia is a living and growing entity, and evolving at orders of magnitude faster than any encyclopedia can. In ten years it won't matter which is more accurate, I think Britannica will be all but dead.