Comment by hackuser | Hacker News Reader

Comment by hackuser | Hacker News Reader

Parent Full thread

Less and less do I trust the wisdom of the crowds, and Wikipedia. I want expertise, not the consensus of amateurs and the ignorant. Look how well the consensus of experts, or at least professionals, works at HN: useful, better than alternatives, but I wouldn't bank on it.

A story, possibly apocryphal, about Richard Feynman: He gave a strong negative review to a textbook that the other reviewers endorsed. When confronted he said he might not be the smartest person in the world, but was he more intelligent than the average of a hundred people? Certainly!

Or from a Car Talk brain teaser: Do two people who don't know what they're talking about know more or less than one persion who doesn't know what they're talking about?

What about 100 or 1,000 people? I'm pretty sure they know less, as they create in their echo chamber greater certainty and additional untruths.

(I am overstating the case for effect; there is value in the aggregate factual knowledge of crowds, and they are not always ignorant.)

As I have said several times before, Wikipedia doesn't follow Surowiecki's rules for "wisdom of crowds" effects.

The Infobitt model has an advantage over wikis that will enable a wisdom of crowds effect. People contribute and vote independently of each other, not by having to agree on a single version of the content. Yes, there can be comments, and that will have a biasing effect; but what people say in comments is obviously not as important as their votes, which they can exercise independently of each other and comments.

Also, since the pieces of content are short and fungible, they are therefore capable of being subject to contests. We can submit competing versions of facts, and the best can rise to the top.

Finally, every fact in Infobitt has to be sourced, and the fact is a summary of the source; and (soon) we'll be able to compete to write the best summary. So, to be sure, we'll have bad summaries occasionally, but generally I think we'll have good ones, once we've built the community up some more.

Thanks for taking the time to answer my, and everyone's questions.

One idea is to develop credibility and reputation over time by contributing useful facts. It's not raw crowd numbers but rather contributors with reputation which will yield results. Think of StackOverflow but for news instead of computer Q&A.

This is a good idea. We need to code the "like" feature as the first step to implementing this.

Wikipedia contributions are not averages of average people; they are written by those who choose to participate, not by a random sample of the earth's population. In some fields, Wikipedia is held in high esteem by many professionals.

Reply on news.ycombinator.com