Slashdot Mirror
Digital Universe a Wikipedia Alternative
Augustine J writes "A new alternative to Wikipedia called Digital Universe is the brainchild of, USWeb founder Joe Firmage and Larry Sanger, one of Wikipedia's earliest creators. This new site differs from Wikipedia by inviting acknowledged experts in a range of subjects to review material contributed by the general public.
"The vision of the Digital Universe is to essentially provide an ad-free alternative to the likes of AOL and Yahoo on the Internet," said Firmage. "Instead of building it through Web robots, we're building it through a web of experts at hundreds of institutions throughout the world.""
Sanger said this in 2001 about Nupedia:
"The reason Nupedia is having trouble right now is that we've had trouble convincing academics that it is indeed a bona fide cathedral. If we were to convince them of that--which I think we will, eventually--you'll see just how wrong you really are (that Nupedia is a failure)."
Well, he was wrong. Experts have little time to waste on stuff like this, and Nupedia died. Will this die? Who knows.. but Sanger has been wrong before.
I suspect that this study over-estimates the accuracy of Wikipedia because it was limited to natural science topics. My impression is that Wikipedia is pretty accurate in this area because people tend to know whether they know what they are talking about or not and people who really don't know anything aren't very likely to write about something like chemistry. Where Wikipedia seems to me to have more of a problem is in areas that people who really don't know what they are talking about think that they do, which is more common outside of the natural sciences. My own field of linguistics is like this. Pseudo-experts seem to be particularly common in historical linguistics. I'd be interested to see a study like the one cited but covering areas like linguistics and psychology.
It seems to me that "acknowledged experts" is both the key to respect and the bottleneck for any on-line encyclopedia. The question is how does a online content system get acknowledge experts. One solution is to hire experts from the meat-space world -- those vetted by traditional academia, etc. Unfortunately, I'd argue that it's simply too costly to hire enough "real" experts to maintain 800,000 articles.
In contrast, wikipedia seeks to create content without this overhead to officially-hired experts. The greatest strength of wikipedia is that anyone can add to it. This encourages content generation. The greatest weakness of wikipedia is that anyone can add to it. This encourages vandals and idiots to add errors into entries.
What projects such as wikipedia need is a mechanism for creating experts and signaling expertise within the context of a corpus created by an open network. This means a better karma system and mechanism for filtering/de-editing entries. Perhaps the easiest mechanism would be a text color-coding scheme. Edits made recently by editors with no track record for stable contributions would be color coded red to caution the reader. The longer the edit lasts, the darker it becomes. Edits that we're made by those with a long history of non-edited additions would see their text quickly become normal black. Done well, such a system could even track contentious frontiers of knowledge -- showing both variants of contested facts in red until one side marshals enough evidence to induce stability.
Readers might even be able to pick which rendering of the wiki to view. They might ask to see only the content that has survived X viewings without an editorial incident (retraction or rewrite) or see only content written by contributors with some threshold level of expertise karma.
Two wrongs don't make a right, but three lefts do.
Actually, last I heard, the founder of Wikipedia is closely involved in this project. So let's not waste bandwidth on why Wikipedia is better.
Also, the people behind this Foundation have been working on other, possibly revolutionary (in a REALLY BIG way) physics research. Check it out: http://www.calphysics.org/ .
That would only work if Wikipedia also made fundamental changes to it's editing structure. Nobody is going to take Wikipedia any more seriously if a bunch of experts start contributing to it as long as any yahoo with a login can go in and change the articles they write.
Wikipedia would have to develop what is essentially a caste system so a user could only edit what has been written by people in the same or lower level as them, protecting expert knowledge from armchair scholars.
Experts already contribute to Wikipedia, and many have left because of edit wars with other users who really don't know what they are talking about. Until Wikipedia begins to show offical recognition of authoritive sources, this will continue.