Encyclopaedias will never be the same again.© Alamy

Inspired by the power of the web, Jimmy Wales co-founded the online encyclopaedia Wikipedia in early 2001. The reference tool takes advantage of Wiki technology, which allows Internet users to modify websites other than their own. As a result, they can change content and add entire new pages to the collaborative, multilingual project. News@nature.com 's Roxanne Khamsi caught up with Wales at the O'Reilly Emerging Technology conference in San Diego.

The idea of opening up a reference tool for people to edit is a relatively new concept. What motivated you to launch Wikipedia?

My original concept was a freely licensed encyclopaedia for every single person on the planet, in his or her own language. We want to change the world and get everybody access to fundamental information. Obviously that's important in the wealthy West, but it's really, really important in a lot of other places.

It's a really big vision and a huge job, so in order to do it we needed a lot of help. Before Wikipedia I founded Newpedia, but it had a very complicated review process: very academic and top-down, and it didn't work for volunteers. The tightly controlled, narrow model just wasn't working. Then we came across the Wiki model, which was great. When Wikipedia launched it very quickly became a success. In two weeks we got more work done than we had in two years.

There are a lot of good things you can do on the web. Why did you think encyclopaedias were the way to go?

<mediar rid='m1'><?xm-replace_text {mediar}?></mediar>Several things made me think of it. I had seen the growth of free software such as Linux and Apache and all the stuff that runs the web. I saw it work and figured there had got to be other areas in which people could collaborate. It was natural that this kind of culture started in the programming world. I thought, what other kinds of things could work?

One of the nice things about an encyclopaedia article is, if I say to a large group of people "encyclopaedia article about the Mayor of London", you know pretty much what that's got to look like: there should be some information about who is currently in the job, and its history. Or "encyclopaedia article about San Diego": it should give the population, the climate, the history and things like that. So it's easy for people to collaborate because everybody has a clear idea of what it should look like.

Do you think these flexible reference tools will overtake fixed sources of information?

Definitely. But I think it's going to vary across different domains. For example, an open Wiki-editing model, in my opinion, cannot replace peer-reviewed scientific research. If it's really open you don't have the necessary quality control from experts.

Our 'no original research' policy was developed primarily in response to physics crackpots. People would submit their new theories of magnetism, for example. A general interest encyclopaedia is not the right place for original research: it should be a synthesis of what is known.

It's one thing when experts disagree with non-experts. But in science there is sometimes disagreement among experts themselves. How do you deal with that type of situation?

For that we have our neutrality policy. There's a lot to it, but in essence we say that if there's an issue of legitimate controversy then Wikipedia doesn't take a side. We just report the controversy.

It's possible to write an article on climate change, for example, in which Wikipedia doesn't take a side. It basically just reports the data and the information. Often we have two separate articles. We might have an article called 'Climate change controversy' that describes the debate, which is mostly political. And then we might have the pure scientific article, and of course even within that there may be some legitimate dissenting opinions. Minority views need to get mentioned.

You recently launched a new tool, Wikispecies. Why did you feel so passionate about separating this from Wikipedia?

We have an encyclopaedia article on sharks, and the article discusses the movie Jaws, shark attacks, culture and so on. That's fine, because it's a general interest encyclopaedia article. But Wikispecies is supposed to be a reference work of species information for biologists, so it needs to be structured, with the scientific information and nothing else. There is some overlap and that's what was controversial, but it's a separate type of work.

Also, there are 1.8 million named species. We don't want to flood Wikipedia with 1.8 million articles that just say, "This species is a member of this genus, and that's all we have." That is more appropriate for Wikispecies: it's a starting point for dumping in data.

Would you link the concept of the wisdom of crowds to how Wikipedia works?

I think that a lot of people believe there are a million people adding a sentence each and through a giant, mass process somehow an encyclopaedia is produced. But within the community we don't see it that way. It's actually more traditional than that: a group of people who know each other communicate. It's not very different from what would go on in any publication.

What motivates people, and particularly scientists, to contribute?

A lot of different reasons. I think there are two primary ones. The first is charitable: scientists like to share their knowledge with the world. That's what they do. They learn things and they tell people about it. So they're motivated by the ability to reach a large audience and make sure the information is high quality in a project that's trying to be fully global and reach countries that are less developed. The other motivation is that it's really geeky fun.