Speaker: Jimmy Wales, Founder, Wikipedia

Moderator: Mike Masnick, Founder and Editor, Techdirt

This transcript has been lightly edited for clarity. Time codes refer to the session video.

[00:00] Mike Masnick

We are in the John Seigenthaler Center at Vanderbilt. He was a very famous free speech supporter, a strong First Amendment supporter. About five years after the founding of Wikipedia, he came out of retirement to write an article in USA TODAY in which he wasn't that happy about Wikipedia, because his profile included some information claiming that he had possibly been involved in both the RFK and JFK assassinations.

[00:30] Jimmy Wales

Basically, what had happened was somebody had put it in, intending, after he had won his joking bet with his friend, to switch it back, but he forgot. What that led to for us was a real change in Wikipedia. What we had a tendency to do then, because we were so new and young, was this: we knew everything needed to have a source, but you could write something, and then if somebody came along and said, oh, that should have a source, they would just tag it and say "no source." Now, I like to say my role in Wikipedia is like the British monarch: no power, some influence, and so on. Back then I was more Henry VIII. So I just said, look, this is the rule. And we came up with our biographies of living persons rule, which says anything negative about anyone has to have a quality source. If you see something that is negative and doesn't have a quality source, take it out immediately and discuss it on the talk page. Because what was happening is people would say, oh, that doesn't have a source,

[01:41] and they'd go to the talk page and say, I can't find a source for this. And it's like, no, you can't just leave it up while that's going on. So his concern in that editorial was that Wikipedia could actually undermine support for the First Amendment if we were reckless and irresponsible. We didn't feel like we were reckless and irresponsible, but I thought, you know what, we don't want to be, and we need to do the right thing here. Now, looking back, and looking at the state of the world, I don't suppose he had seen Twitter at that time. Speaking of poison intellects with poison pens. But for me, that was a seminal moment. Many years later, I had the good fortune to speak with him. Al Gore invited him and me to speak together—and Al sat between us. I said, yeah, he's assuming he needs to put that peace prize to use. But actually, we got on very well. What a lovely man, and an amazing giant in defense of freedom of expression.

[02:55] Mike Masnick

It was funny, too, because when that happened, it launched about a hundred op-eds asking, oh no, can you trust Wikipedia?

[03:07] Jimmy Wales

Our traffic tripled in about two months. That isn't how I wanted to grow, but it did bring us to public attention.

[03:14] Mike Masnick

Soon after that, I think just a few months later, Nature came out with a study that looked at the quality of the articles on Wikipedia and found that it was as trustworthy as the Encyclopaedia Britannica. Respond to that.

[03:31] Jimmy Wales

What was fun about that was that they had sent a certain number of Wikipedia articles to experts in the relevant area, along with the Britannica article, and asked them how many errors they could find, major and minor. We weren't quite as good as Britannica, but we were very, very close. What I've always been amused by and proud of is what happened next. After this came out, it took a few weeks, and then Britannica issued this scathing denunciation of the study. They went super defensive. On our side, the community members came to me and said, could you contact the guy and get the list of errors so we can fix them? That's the different philosophy of the Wikipedians: of course we know it isn't perfect. We know there are all kinds of things we should improve. Come and help us.

[04:29] Mike Masnick

One of the things that occurred to me in thinking about the moment we're in, and then thinking back to that time: around then, the general narrative was concern about whether you could trust Wikipedia. Anyone can edit it, anyone can do stuff, and therefore it's not trustworthy. And the feeling toward what are now the bigger tech platforms, Google and the like, was that they were beloved new services that were useful in connecting the world and wonderful. Now, 20 years later, the sense of the big tech platforms is that they're horrible and evil and a problem in the world, and Wikipedia is seen as this shining light of the wonderful side of the internet. I'm curious, from your standpoint, how that experience has been, going from a point 20 years ago when people said, oh no, you can't trust this thing that anyone can edit, to now, when people talk about Wikipedia as the one good site on the internet.

[05:33] Jimmy Wales

Going back to Seigenthaler, the core point I took away from that is that if you are irresponsible in how you're exercising your rights—even though you have a right to do that; you can be as irresponsible as you like with your freedom of expression—you put it into play politically. And I actually think a lot of the social media companies, not all of them, have something to answer for here. I'll tell you a great story. Before Elon bought X—I can't call it X, I'll never be able to—there was an incident of a kind that was quite common on Twitter. Some random person started calling me a pedophile and making very specific claims about an incident in 1997 and so on. Absolute lies, complete nonsense. I figured out later it was somebody who'd been banned from Wikipedia. This was at the height of the #MeToo allegations, so it was the sort of thing you'd be concerned about. At first I did what you're supposed to do:

[06:47] I reported it, because under the terms and conditions you're not supposed to do this anyway. Within minutes I got an automated email back saying, great, we're looking into it. A few hours later, I got an email: we've looked into this and we don't see any problem here. So I emailed Jack Dorsey and a couple of board members I knew, and I'm like, Jack, come on. And he's like, oh, Jimmy, don't worry, we'll take care of it, we'll take it down. And I'm like, you're missing my point, Jack. I'm Jimmy Wales. I can email Jack Dorsey because I've met you many times. What about the teenage kid who's being abused on the platform? This isn't okay. You need to be more responsible about how you run things. And now Elon has fired what few people they had working on that. So as much as I support the rights of these companies to run their platforms this way, I'm like, it ain't a good idea. It's like I support the right of a restaurant to serve really bad food. I ain't going to eat there.

[07:47] Mike Masnick

We're having that debate again now, or maybe still, 20 years later, about things like Section 230 and intermediary liability, and whether you can moderate a site how you want, which goes to the point you were just making, and about anonymity online. Have those debates changed in the past 20 years, or is this the same thing all over?

[08:10] Jimmy Wales

I think it's broadly the same debate, because the issues are fundamental and human and will always be with us. Anonymous speech, I think, is incredibly important. We have many Wikipedians we know—we're not a particularly secretive bunch of people—but people do use a username, and we don't know who everybody is. What we do is look at the quality of the work, and you can get banned if you're not doing the right things. But we have people editing Wikipedia from very difficult and dangerous places. If we have a Wikipedian in Iran editing—and we've had situations; we've got Wikipedians in jail right now in Saudi Arabia and Belarus. When we review their edit history, it's very clear there is nothing bad about what they're doing. They are being good Wikipedians. They're just speaking facts. These two didn't take advantage of being anonymous. They didn't necessarily use a VPN and all the things they should probably be doing in a dangerous place. But anonymity is incredibly important to us. I'll give another example, and this is not about being in a difficult place. It's about the importance of being able to freely work on something of great interest.

[09:33] We have people in the community who are very vigilant. One of the things we always have to look out for is pedophilia advocates—people who are trying to write in favor of pedophilia, trying to tone things down. They are not interested in a neutral point of view. They are campaigners for a view. And there are Wikipedians who have a second account simply to edit in that area, because they have a professional job and they don't want anybody looking at their edit history and saying, oh, this is weird. Bob, I thought he was a nice attorney, but all he edits is about pedophilia. How strange. That's a serious thing. We want these people editing, but maybe you don't want to do that publicly. You want to fight the good fight, and you need a bit of anonymity for the rest of your life. Even if you've got no legal risk, it's still important. So we are very, very firm about user privacy. We've had different kinds of incidents in different places around the world, and it gets very complicated sometimes.

[10:46] Mike Masnick

To that extent—you talked about the large platforms and the responsibility they have—what responsibility does Wikipedia have to its volunteer editors in those situations? Obviously those are extreme situations, but when things happen.

[11:06] Jimmy Wales

They're important. One of the things we're very clear about is that access to knowledge is a fundamental human right. In India, a newspaper is suing us—it's a long story, but it's ridiculous—and we got a demand to reveal the identity of some of the editors. What made it super complicated is that we're based in the US. So we could just say, sorry, ain't going to happen; bring it to a court in the US and force us, and good luck with that. But if we did that, we would lose the ability to defend the case. We would lose the ability to go to court at all in India. No lawyer could represent us if we openly defied a court order. So that wasn't a practical strategy, but it was coming down to: they're going to have to block us, because we're not going to do this. We did manage to work with the judge and with the volunteers to find a compromise solution that involved the volunteers filing through their attorney. We managed to get through it, but it was a very complicated situation. And the other thing is, we are quite confident we will win in court.

[12:25] So losing the ability to fight a case we're going to win, which could be an important landmark case—we didn't want to lose that. But we're not going to compromise the safety of the volunteers either. It's really hard.

As another example, Turkey. Turkey blocked Wikipedia for three years. When they first did it, I'm convinced, because of things they told me, that they were absolutely shocked we didn't cave in immediately, because they have experience with this. They blocked Twitter, even after Elon owned it, for a little while, and Twitter caved in and took down the tweets. In our case, the articles in question were very well sourced and very well written. They were sourced to The New York Times, the BBC, something the European Union had put out. So we said, we're not going to take this down. And they said, oh, it's fine, you don't need to take it down. Just block it from being seen in Turkey. And we said no. They were really surprised, because a lot of the companies do cave in immediately. I think this is one of our great advantages.

[13:38] We're a charity, a nonprofit organization. Small donations are a huge part of the support for Wikipedia, and so our business model is quite consistent with being principled on these kinds of things. Our donations largely come from the US—not all, but a majority of the money. There's a much bigger tradition of charitable giving in the US than in Europe. The Europeans all think the government should pay for it, and I think they think that a little bit less now that Europe has become less free-speech friendly, but that's a whole other complicated story. I always say that if we compromise somewhere like Turkey, our donors are going to be very disappointed in us. We have a standing open offer. We're blocked in China. We could open in China anytime we want. I've been told very clearly that all we have to do to be available in China is partner with a Chinese university that will manage the project and make sure all the content is legal in China. And I'm like, yeah, no. I met once with a professor—I was in China, and they dragged this professor out of bed in the middle of the night to come and see me. He said, you have to understand, Jimmy, the government is very patient.

[14:52] They're not going to back down. And I said, I will wait a thousand years. This is not going to happen. I just turned 60, so I've been joking that my life is one-third over, but for the next 120 years, I'm going to be here saying absolutely not.

Where was I? I'm getting off on my stories. Anyway, we fought in the courts in Turkey, all the way to the Supreme Court, and we won. We're really proud of that. We're now unblocked in Turkey, and it's a landmark case: the president's office can't just decree that a website is blocked. They have to go through a legal process. I'm not saying they've got the First Amendment in Turkey, but it made a difference. And one of the reasons I think this is incredibly important is that I know with absolute certainty that if we cave in on one of these things, they're going to come out of the woodwork everywhere.

So we've got a lot of problems. I live in the UK, so I get more involved in our situation there, because I'm there and I know all the relevant government ministers.

[16:01] I've made it very clear to them. You could tell us to age-gate Wikipedia. And here's what's going to happen: we're not going to do it under any circumstances. So then it's going to become not just a regulatory issue between us and Ofcom, the UK regulator; it's going to be political, because if you block Wikipedia, the whole country is going to go insane. That's absolutely ridiculous. So I don't think you're going to do it. My team is a little more reserved than I am. I came out of a meeting once and one of our lawyers said, I can't believe you just told the minister that you're just going to fucking not do it if you're told to. I'm like, well, we're not. They should be aware of that in advance, before they screw up.

That brings us to something I want to say. One of the things I've always counted on is that we're based in the United States and we have the First Amendment, and it's incredibly important. People ask, did the government ever pressure you? There's this trope on the right that the Biden administration was pressuring social media companies to censor. I have no opinion about that.

[17:16] I just know what happened with us. We did talk to the Biden administration, and there was never any suggestion of censoring Wikipedia. It was only us explaining how Wikipedia works. This was during COVID. They said, we want to understand the information space. How do you keep misinformation out? And we said, here's how it works. It was all very polite, and they didn't pressure us at all. I don't know if they did in other cases; I'm not privy to that. Only more recently, we got this absolutely unhinged letter from Ed Martin, who was at that time the interim DC prosecutor. He didn't get the job, so then they forgot about it. But it was a letter hinting about our nonprofit status because of content issues in Wikipedia. After the fact, I thought, it's really a good thing that we now have serious, thoughtful, professional staff, because they answered the letter in a very matter-of-fact, thoughtful way. Whereas I would have just scrawled "go fuck yourself" and put a footnote to the First Amendment. That's just not okay.

[18:25] Mike Masnick

You've walked through all of my questions.

[18:28] Jimmy Wales

This is great.

[18:29] Mike Masnick

One question answered them all. But I did want to talk more about age verification, because this is an issue that has really caught on in the past two years, globally. Certainly in the US, but Australia, the UK, the EU, all over the world. Suddenly people are saying the internet is unsafe for kids, and we have to do something about it. The widely agreed-upon position among many is that the best way to do that is to age-gate: identify how old everyone is and say, if you're under a certain age—it's different in different places, and it doesn't seem clear how they pick it—we're going to gate it and not allow access. How do you, and how does Wikipedia, view this rush to put up age gates around the globe?

[19:17] Jimmy Wales

It's terrible. Not just from our perspective. For example, in Australia, for reasons that are entirely unclear, we're not included in the ban. If you read the letter of the law, we absolutely would be. The same in Mississippi, I think, which has a rule where, in the definition of what counts as social media, we tick all the boxes as far as I can tell. But they just don't dare do it, because it would expose the ridiculousness of the whole thing. Everybody knows kids who are 12 years old—I picked 12 because you're probably old enough that you're looking at Wikipedia to do your homework, but you're well under any age limit. Is this the problem you're concerned about? Kids reading an encyclopedia? Really?

One of the things we've started doing—and I'm comfortable with it, but I have to tell my team I want to be really careful here—is that we say to legislators, whatever you're planning to do, I want you to think about the Wikipedia test. If what you're about to pass would damage Wikipedia, you should think again. Implicit in that, also, is that we're not going to do it anyway. The reason I'm slightly uncomfortable with it is that I don't want to just defend Wikipedia. I think defending all free speech and all free expression is important, so I don't want to say we need an exemption. I actually got really irritated once. In the EU, there's the copyright directive, which is complicated and problematic, and there's a clause in it that exempts any nonprofit encyclopedias. I'm like, yeah, there's only one. That's us. My team was quite rightly proud that they had gotten that in, and I said, yes, but realistically, I'm worried about the rest of the internet as well.

[21:17] Mike Masnick

You talked about the value of Wikipedia being based in the US and protected by the First Amendment, but also about attacks on free speech and the First Amendment in the US from the federal government and from the states. At the same time, with the laws we were just discussing around the world, we're in a moment where it feels like the internet is fragmenting because of global regulation. Section 230, which we talked about before, exists only in the US. How does Wikipedia remain a global property in a world where you have different laws everywhere?

[21:52] Jimmy Wales

We've been very careful, from before it even mattered that much, because I always thought it was a good idea. We exist in the United States. We're a nonprofit organization in the US, and we do have a few staff members in other countries, so that raises a bit of exposure in certain places. But broadly, we don't exist separately as an organization anywhere else. Then we have local chapters. Wikimedia Deutschland, for example, is the biggest local chapter, but it's an independent nonprofit organization. We fund them, they do their own fundraising, they do community meetups and partnerships with galleries, libraries, archives, and museums, and some software development. But they don't control the website. So if you sue in Germany, you have a really hard time suing the local chapter, because they can say, we can't control the website. We're a club of users here.

That's been really important. In Germany, there was a really odd case. A man murdered a TV star—a very famous case in Germany. He was convicted and went to jail for 14 years, which I think is the maximum you can get in Germany.

[23:20] Then he got out. And under German law—if anybody's a German legal expert, I might be getting the details a little wrong, but I'm pretty sure I've got the basics right—after that, it's not legal for a newspaper to call that person a murderer, even though it's a complete fact and a famous case. The tradition in German newspapers is just to use the person's initials after that, which is odd. He came out and started suing, mainly about a book that was going to come out, but he sued Wikipedia as well, and got an order that we should change his name to his initials. Our basic response was: that's interesting. You can bring your case. There's a procedure to bring your judgment to the US and show it to a US judge, and they're going to tell you, no, First Amendment, that's not okay here. But here's what happened that was really odd. The German community voluntarily decided that it's within German cultural traditions, so they changed his name to just the initials. I don't agree with it, but I'm not going to stop them from doing that. It seems odd to me, though.

[24:31] Mike Masnick

Everyone's talking about AI, and I know there have been concerns about the impact of AI on Wikipedia, both good and bad. How does Wikipedia view everything happening in the AI space right now?

[24:44] Jimmy Wales

All the major AI models train extensively on Wikipedia. The only thing we really ask for is attribution, and the AI models aren't very good at attribution. The different language communities around the world have sometimes quite different ideas about the use of AI in Wikipedia. Broadly, I would say English Wikipedia is one of the more anti-AI communities, although there's a diversity of opinion. I understand that Polish Wikipedia is quite excited about AI and how they might use it. That's part of the democratic process in Wikipedia, the discourse and dialogue. People will try an experiment here and an experiment there and see what works and what doesn't. We do have a problem with people using AI to write a Wikipedia entry. If you say, write an encyclopedia entry about Taylor Swift in the style of Wikipedia, it does a pretty good job, and you look at it and go, well, not great, but it's okay.

I know we're out of time, but I've got to tell this story. It's too funny. What I always do, to test every new AI model I come across, is ask: Who is Kate Garvey? My wife.

[26:01] My wife is not a famous person, but she worked for Tony Blair. She worked in Downing Street under Tony for ten years, so she's in all the books about the Blair administration, and she's had some news coverage. She currently promotes the Sustainable Development Goals, the Global Goals. She works with the UN, funded by the Gates Foundation and Google and so forth. So she's a little bit known, which is perfect for my purposes, because the AI sort of knows who she is but isn't quite sure. It always responds, but it's always wrong. One time it said that she set up a nonprofit to promote women's empowerment in the workplace, which sounds like Kate, and that she did this with Miriam González, who is Nick Clegg's wife. Nick Clegg was the former deputy prime minister who went on to work for Meta. Of course I know Nick, and we know Miriam, and our kids go to the same school, because London is a small town. I showed this to Kate, and she said, oh, that could have happened.

[27:00] We could have been sitting together at a dinner party before she started Project Everyone. Actually, that's plausible. But then I asked, and who did she marry? This is fantastic, because it's never me. It's usually somebody we know, because, as I always joke, I married into the Labour Party, so it's UK politicians or political journalists. That's always very amusing. But my favorite, and this one is a little less plausible—I can tell this story here in the US; it's very well known in the UK, but I think you'll know it. I asked who she married, and it said Peter Mandelson, who was the ambassador, and the Epstein scandal and all that. So I said to the AI—this was before all that came out—isn't Peter Mandelson quite famously gay? And then it got very woke with me. It said it's not appropriate to speculate about people's personal sex lives. So the point is, you cannot use AI to write a Wikipedia entry. It'll make up a lot of weird stuff.

[28:15] Mike Masnick

On that note, everybody, please. Thank you.

[28:18] Jimmy Wales

Thank you. Fantastic. Lovely.