User:BenKovitz

From Wikipedia, the free encyclopedia

Lover of wikis and Heuristic.

Author of the Green Light Wiki: http://greenlightwiki.com

I had a tiny role in the early history of Wikipedia.

Email: bkovitz at acm dot org.

1 The famous conversation
2 Articles started
3 Extensively edited or poured passion into
4 Want to write (or see someone else write)
5 Other

[edit] The famous conversation

There's been a fair amount of talk lately about the famous conversation between me and Larry Sanger that led to the creation of Wikipedia. Here is my memory of it.

[edit] Background

Larry had been working for a while on Nupedia, a free on-line dictionary written by credentialed experts and following a strict review and certification process to ensure that every article was high quality and, above all, factually reliable. A lot of Larry's thought since I first got to know him back in 1994 had been devoted to how sources of knowledge can be reliable and unbiased. I believe he was always looking for foundational knowledge: something solid to build on, or at least as close to that as we can practically achieve. Nothing much had come of Nupedia, though. There were hardly any articles completed.

The previous several months, I had been participating heavily on two wikis: Ward Cunningham's original wiki and one called "Why Clublet", run by Richard Drake and Keith Braithewaite. I'd become interested in the way wiki pages would grow in directions that none of the contributors had anticipated, and sometimes with writing clearer and better than any one contributor was capable of. I was especially interested in how conflicting ideas could be explored in depth by separating discussion of them to different places. No back-and-forth head-on "debate", just parallel exploration of opposing ideas. I saw it work especially well when people focused on improving the quality of the writing and poorly when people tried to make the wiki officially declare their preferred view as the correct one. I had also been experimenting with Extreme Programming at work, and had discovered amazing synergy in pair programming. I liked how these structures enabled people working together to actually be smarter than people working alone--the exact opposite of "committee"-style collaboration, and also opposite from the kind of collaboration where each person "owns" a sectioned-off piece of the whole.

[edit] That night

On January 2, 2001, Larry and I ate dinner at the taco stand at 1932 Grand in Pacific Beach, San Diego. I believe this taco stand had no name, just the words "MEXICAN FOOD" written on a window.

I suggested to Larry that he make Nupedia into a wiki. I said, instead of trying to prevent error and bias, to openly invite error and bias and make it very easy for people to correct them. I believe my exact words were to allow "any fool in the world with Internet access" to freely modify any page on the site. I told Larry to do the exact opposite of everything he'd been doing with Nupedia. No solid foundation, just endless chaos and conflict. No review process prior to publication, just go live immediately with every edit as soon as it's in the computer. I suggested that this might actually lead to better reliability and richer content than the careful, circumspect approach.

Larry was skeptical at first. Couldn't people just vandalize the site? I said yes, and other people could then repair the vandalism. Couldn't total idiots put up blatantly false or biased descriptions of things, to advance their ideological agendas? I said yes, and other idiots could delete those changes or edit them into something better.

I told Larry about some of the conventions on Ward's Wiki, like running words together to indicate a page name. I mentioned that this required me to give the Ward's Wiki page about Aristotle the name "MrAristotle." Larry cringed at that.

Within an hour, though, Larry had gotten enthusiastic about the idea. We went over to his apartment, and he tried to call his boss at Bomis, Jimbo Wales. Jimbo didn't answer, so Larry left voicemail. Larry and I talked about philosophy for a while, and roughly half an hour later Jimbo called back. They talked for ten or fifteen minutes. After the conversation, Larry had a big smile on his face. Larry said that he felt very optimistic that the idea would proceed, and that Jimbo was quite open to it.

[edit] Afterward

Within a couple weeks, WIkipedia had gone live, and Larry had posted the WikiPedia page on Ward's Wiki, inviting people to come contribute.

It's a rare thing to tell someone to do something exactly the opposite of what he's been doing, indeed the exact opposite of how he's been thinking and investing mentally for most of his life, and get a fair hearing. It almost never happens that someone actually takes the suggestion. But Larry listened to what I had to say, let his imagination engage, and ran with it.

It was extraordinarily fortunate that Larry was working for Jimbo. Jimbo had both the means and the vision to get Wikipedia moving. Jimbo had first turned me on to the power of collective knowledge years earlier in a post to the MDOP discussion list about sports betting and how it sets the "line" more reliably and accurately than any individual bettors.