[Wikipedia-l] A parallel World Wide Web?

Wed Jun 2 21:00:24 UTC 2004

> create an encyclopedia or database containing all the human knowledge.
> Of course this aim can never be reached, but it should be approximated

There is a difference between knowledge and information. Wikipedia contains 
too much information and too little knowledge. This results in attracting 
more and more people who think wikipedia is another good place to share 
information instead of knowledge, people who believe that their information 
is knowledge. In most cases it isn't.

To say it bluntly, the information contained in a "List of porn actors who 
died wearing a hobbit costume" isn't knowledge for 99.999999999% of the 
Wikipedia users. I agree that it might be knowledge for 0.000000001% if they 
are looking for that particular kind of information. However, if Wikipedia 
does contain that information to fulfill the needs of that little amount of 
users, that specific information is noise for the vast majority. Since there 
are no retrieval methods to sort out relevant from irrelevant information, 
the noise is there. (No, google does not work.)

Knowledge is information that rests if everything unneeded *is sorted out*. 
Forgetting information is a very important thing in creating knowledge. 
That's the point Wikipedia is missing today. That's why some people see a 
need to talk about "Wikipedia 1.0" and "peer review" and that sort of things 
every now and then.

If we were aware during the regular article editing process that - given that 
our mission is to collect knowledge - our strategy should be to *sort out 
information*, not to *collect information*, that 1.0 stuff wouldn't have to 
be discussed.

Uli