Jimmy Wales on the Failure of Nupedia
I knew there was a precursor to Wikipedia, but until I read this blog post, I didn’t know how bad the state of affairs really was:
Before Wikipedia there was Nupedia, where the articles had to be written by experts and peer-reviewed. After three years, Nupedia had produced a grand total of 24 articles. Then a tiny, experimental adjunct to Nupedia—a wiki-based peanut gallery where anyone could contribute—exploded into the flawed, chaotic, greatest encyclopedia in the history of the world that we all know today.
In this video, Jimmy Wales talks about the failure of Nupedia, and he says:
When I first had the idea for an encyclopedia, there was a prior project called Nupedia, and I’d hired this PhD in philosophy to help me organize it, and he organized a very top-down system, very academic. There was a seven-stage review process and you had to apply and get permission before you could write an article. He was actually having people fax in their credentials — their degrees and things — to prove they were who they said they were. And it failed. It failed because it wasn’t any fun. For anybody.
That’s really interesting. It’s also a bit ironic, since there is in fact a peer-reviewed free encyclopedia out there which is doing excellently. It’s the Stanford Encyclopedia of Philosophy. I think it’s safe to say its system is designed by PhDs in philosophy: the footer even says “Copyright © 2008 by The Metaphysics Research Lab”. The content is truly excellent for the topics it covers — there really aren’t any sites out there that can match it.
And yet, this didn’t work at all for Nupedia. It was a total failure. Wikipedia is an extraordinary success, despite its flaws. What is it that makes SEP a success, despite using more or less the same model that doomed Nupedia to failure? Is it simply the subject matter?
If you compare Wikipedia to SEP, when both have an article on a topic, SEP’s almost always comes out far ahead of Wikipedia. But Wikipedia has a whole other breadth, both within the subject (philosophy) and outside it (the rest of knowledge). Clearly both models can work, and I’d say that under ideal conditions, the peer-reviewed process produces superior results (which is to be expected), but ideal conditions are hard to produce, as Nupedia shows.
Later in the clip, Wales talks about his decision to keep Wikipedia open, after seeing Nupedia fail:
In the early days I thought, I’m just gonna keep everything as open as possible for as long as possible, but I assumed that that was gonna be about two weeks. I actually had a hard time sleeping at night, because I figured someone was gonna come and wreck the wiki over night. That never happened. And if people did come in over night when there weren’t many people around and they did some vandalism, we just cleaned it up in the morning and that was fine.
Wales says this isn’t surprising if you think about it, because most of us are really decent people, and for most people, the moral ground rules of Wikipedia resonate pretty strongly. The people who are out to purposefully vandalize are really quite few. Maybe that’s so, but I think he’s wrong to attribute Wikipedia’s relatively low rate of vandalism to people’s common decency. After all, we know from experience with other sites that trolls are abundant. Even if internet trolls are a very small minority of internet users, they’re very effective. Their small numbers don’t stop them from spreading their shit far and wide. I think the reason Wikipedia is relatively clean and free of this kind of crap is a bit more complex than simply, “People are basically decent.”
