Asterisk: You're famous for the Depths of Wikipedia account, where you share factoids from some of the most arcane, interesting, and surprising pages on Wikipedia. But you're also now a part of the broader Wikipedia community. How did you first get interested in the site, and how has your involvement changed over time?
Annie Rauwerda: I started back in high school editing typos and adding things that I noticed were missing — like items to lists. But I had never done anything more than that because I was afraid of it because there are so many rules. Like, I'd seen the talk pages. And many of Wikipedia’s policies and guidelines and essays are very wordy.
Then I started the account — even though I felt a little like a phony. But I remember the first time I felt really excited about the Wikipedia community was when I got on a call with the president of Wikimedia, New York City, back in 2020. And she had told me about a guy named Jim who retired from working at the phone company. He worked in that big AT&T building that doesn't have windows. I don't know exactly what he did in there — but cables and stuff. Anyway, he's retired now, and he spends all day biking around New York City and taking photos of infrastructure for Wikipedia, because Google Maps photos — and so many other photos — aren't freely licensed. And I was like, that's amazing.
So I kept hearing about more and more individual people and their shticks. There are a lot of generalists who edit Wikipedia, but there's something so endearing about the people who have just one thing. That's the first time I got excited. That was when I was brave enough to start making bolder changes and writing my own articles.
A: Wikipedia's tagline is the “free encyclopedia, created and edited by volunteers around the world,” which makes it sound like a cohesive, happy little family. But as you alluded to, there's a lot of rules. It’s intimidating to write articles. Is it actually the case that there is a Wikipedia community, or is it more accurate to talk about communities within Wikipedia?
AR: The answer is both, of course, but when people talk about Wikipedia as a decision making entity, usually they're talking about 300 people — the people that weigh in to the very serious and (in my opinion) rather arcane, boring, arduous discussions. There's not that many of them.
There are also a lot of islands. There is one woman who mostly edits about hamsters, and always on her phone. She has never interacted with anyone else. Who is she? She's not part of any community that we can tell.
But then there are hundreds of thousands of editors on English Wikipedia. And within that there are very specific communities that are really interesting. There's the military history WikiProject. Maybe this makes sense because of the whole military thing, but they are very hierarchical. They have a lot of rules. They're very efficient in reviewing articles. Also Wikipedia has a pretty outdated rating system for articles — one of which just got deprecated when “Featured Articles” and “Good Articles” became a thing — except for in military history, because that community was like, well, we need to have every level.
The same thing is true of the tropical cyclones community. They also do a lot of reviewing, and they also tend to skew very young. It's a lot of teenage boys in tropical cyclones. There's also a very strong anti-vandal community, who similarly skew very young.
A: Before I had Wikipedia friends, I did not really understand how many different silos there were.
AR: So many. There are acronyms that I hear that I do not know. The anti-vandal people especially are really off in their own world. One of the most effective, consistent, long serving anti-vandalism patrollers is not a teenager. He's a PhD-level material scientist who used to write very high quality articles about chemicals. And then he just got too upset about vandalism and decided “I have to devote hours of my life — even though I have a family and a really demanding job — to this.”
A: What is your sense of the people who do get drawn in and commit so much time. We hear about these people who've edited over three million Wikipedia articles. What's driving them?
AR: Well — two things. The three million articles guy, that's Stephen Pruitt. He's wonderful. He kind of became the face of super editors. And it's because the edit count on Wikipedia is an imperfect proxy. Stephen uses bot-assisted editing — which might be something like doing categorization — and so he can do a lot of edits at one time. And so like, yeah, he's made an edit to a bunch of articles, but it's not the same as writing three million articles. But then again, he also writes a lot of articles.
As to what makes people get sucked in? I think that one shared quality of every single person I've met who has stuck around Wikipedia for a long time is that they have very little hesitation to work hard, and they put a low value on their own time. Maybe they are even willing to waste time. Like, people are very willing to raise concerns and discuss and accept critiques. I also think people tend to care a lot about precise language and about being correct. Some of that comes from the culture of Wikipedia, but I think it's also what predisposes them to edit it in the first place.
A: Editing can seem incredibly inefficient sometimes, given that Wikipedia is so democratic and anyone can enter into any discussion. Take the page for yogurt. The debate over whether to call it “yogurt” or “yoghurt” lasted seven years and totaled over 140,000 words. What happened there specifically, and what makes the process so inefficient in general?
AR: I have not read all 140,000 words. That's multiple Shakespeare plays — which I would rather read.
My impression is that for the debates that last for so many years, people will tap in for a couple of months, and then things will kind of settle, and then they'll tap out and let other people argue. I don't really know what's going on there. I wonder if maybe it's just because people feel so invested because they've already sunk so much time and energy in and think they need to “win.”
But also — to the yogurt example specifically — when it comes to British versus American English, or Australian or Nigerian or Indian English, basically, the policy is kind of a free-for-all, and that's how it's been for a long time. It usually comes down to whether the topic has to deal with Nigeria or India. If it's about George Bush, then, like, please write it in American English. If not, the rule is to stick with whatever the original writer used. And so that's not always very satisfying. And then there's yogurt. That's not tied to any one country.
A: You had a TikTok about another internal Wikipedia debate. Until 2018, there were no freely licensed photos of Kim Jong Un. No one wanted to violate Korean copyright law (though there was a debate about that). And no media outlets freely licensed any photos. And this never got resolved, and so we had hand-drawn images of Kim Jong Un instead.
All of which leads to my larger question, which is — is the writing and editing process working?
AR: I mean, the answer is no. There are people who just kind of wait for their opponent to get carpal tunnel — and that's not productive for anyone. I would call those sorts of debates tendentious. You're not trying to find consensus.
I think when it comes to Wikipedia rules and interpreting the rules, it's impossible. You can't follow every single rule. And I'm using the word “rule,” but, on Wikipedia there are “policies,” which are like the Bible. But they change all the time because people work on them. And if you're going to change a policy, then it has to be discussed. And that happens all the time.
But then there are essays which are just written by a person. And they look a lot like policies! They just have a little thing on the top that says: “This is an essay.” And then there are guidelines! They basically look like rules, and people kind of treat them like rules. And so you can't really follow all of them because there's stuff that contradicts other stuff. It's a Catch-22.
Back in the day, there were five — call them guidelines, decrees, commandments — sent down from on high from the less good co-founder, Larry Sanger. And the last one was: “Ignore all rules.” And ha! That's still there.
But within that, there’s a guideline: “Be bold.” And people still talk about both of those all the time. But there are rules where, if you don't follow them, and you don't have good standing or a good reputation, you're gonna get immediately blocked. And so “Ignore all the rules” and “Be bold” aren't actually true.
And that's the unfortunate thing about growing and becoming a massive encyclopedia that's more respected than it was in 2004. It's a lot harder to truly be this free-for-all, libertarian utopia that the founders thought it might be.
A: You said there's, like, 300 people who are really making a lot of the decisions about Wikipedia editing. Who are they?
AR: I threw 300 out there because — if there's a big, big vote — it's generally about 300 people who show up. About two years ago, they changed the Wikipedia skin, and so the table of contents appeared differently, and there was more white space, which people care a lot about. More than 300 people showed up for those votes — and that's about what you can expect for big votes. Usually those people are either administrators, of which, on English Wikipedia, there are fewer than a thousand active. And sometimes they're just people who care a lot about policies.
Who are they? I think that I can generalize because I go to a lot of meetups. There's a bimodal age distribution. There's a lot of students, very bright college students, grad students, and sometimes a precocious high schooler. (They're the best — I love them.) And then, there's a bunch of retirees. I think common professions are software and computer engineering.
A lot of people work blue-collar jobs. Wikipedia is like their intellectual outlet. And there are some people who are severely disabled in some way, so they don't leave the house for a traditional job — and that gives them more time to edit Wikipedia. It's more male than female.
A: You’ve said that the Jimmy Wales, aughts-era goal of “building the sum of all human knowledge” was perhaps not the right goal. Why?
AR: There are seven million articles on the English Wikipedia — which is, I think, the biggest Wikipedia. So many of them are getting dusty. They are outdated. They are suffering from, uh, a not very ethical PR person going in there. Basically, it's a lot to maintain. Do you remember Aaron Swartz?
A: Sure.
AR: Okay. So he was a Wikipedian. Back in 2006 he put out a blog post that made a splash because it wasn't what people expected. The thinking back then was — okay, we have these few thousand people that are quite active on Wikipedia — they’re the ones doing all the work, right? And he was like, no. The bulk of Wikipedia's work is very much the people in the background that just see when something's wrong and fix it. Like what I did when I was 18.
So if the article is cottage cheese, and someone comes in and replaces it with “poop,” those will get fixed right away because first of all, there's anti-vandalism people watching, but also, even if they miss it, people are looking up cottage cheese a lot.
But if you're vandalizing or if you're doing something sketchy in an article about some really obscure person that doesn't totally meet notability standards and who isn't getting any page views, it's just impossible to maintain that stuff.
A: What's the process for something getting deleted?
AR: There's this place. We all solemnly march down to it. Anyone who wants to show up can, and then we just vote. It's called articles for deletion. Let's see what's up right now. The most recent thing I deleted was a page that had barely been changed from the 1911 encyclopedia, which was totally plundered back in the day when they were just trying to fill Wikipedia.
A: Because it was public domain.
AR: Right. And it was a term for a mountain that hadn't been in use since the ‘30s. It was what colonists called it. So it was a duplicate article, essentially.
Anyway, you nominate something for deletion, and people vote. Your vote really doesn't count very much if you don't add a reason. But then, once there's been a little bit of discussion, like, I don't know, usually if five people weigh in, that's sufficient. If it's a weird name of an old mountain, people don't really care. But if it's a big deletion, they get really huge.
I think more than half of the things that get nominated end up getting deleted but I need to double check that. And then, if people don't participate very much for some reason, which usually happens when it's a really boring topic that no one cares about, or it's boring and it's not clear whether or not it should be deleted — then it gets re-listed.
The standards change sometimes. Once in a blue moon, you'll have a huge deletion event, like — I don't know — we're deleting all the Olympians that don't have sources apart from this one register. So goodbye curlers from 1930! We actually do not know who you are.
A: Speaking of sources, we are both editors. We've had the debate, sometimes, over whether or not we should cite Wikipedia. Can you answer this for us?
AR: If you're putting anything out into the world, please do not cite Wikipedia! Do not do that. I don't trust it. Ever. That's why you check all the sources. Ideally, every single thing, every single piece of information besides, like, “the sky is blue”-type stuff should be tied to a source. I would say, just cite those because people interpret things weirdly sometimes, and you never know.
A: I'm actually surprised to hear you say that though maybe I shouldn’t be. I've definitely had conversations with people that assume that there's been so much work put into Wikipedia at this point — especially on really well sourced articles — that it's trustworthy enough.
AR: If it's a featured article, I trust it a little bit more. But I still wouldn't ever do MLA formatting with Wikipedia. But I love to trust Wikipedia when I'm just doing cursory reading about things I am interested in.
A: This gets to areas where Wikipedia is strong. The Napoleon page, for instance, has over 500 sources. And you've touched on Wikipedia article grades. Where is Wikipedia strong and where does it need work?
AR: Okay, strong suits. Military ships of history. So strong. Tropical storms, strong. US military history, British military history. I would say, generally, it's pretty good on US politics — even local politics. There are articles about pretty much all the recent elections in random counties for the past, like, 75 years.
A: This fits with my theory that these are the same people who are going to city council meetings.
AR Yeah. Things that are bad? I don't really talk about this as much as I should, but as an editor, I am very passionate about the articles that have a lot of page views but get very little attention from editors. Those often are things that are low-brow pop culture topics that are not naturally interesting to the median Wikipedia editor. For me, the one that I always notice is influencers. Influencers often have horrendous articles. You can tell that somebody that didn't really get the norms of Wikipedia went in there and added — I don't know — six paragraphs about the most minor scandal, all of which cited tabloids.
Also business articles. Or pages on CEOs. Sometimes those articles are really unbalanced, poorly written, weird, or outdated. That's not good. If they're a powerful person, we should know more about them.
And then of course people talk about the gender gap. There's been this huge push to write articles about women, and that's been pretty successful. One thing that doesn't really get at, though, is articles about hobbies or topics that are traditionally more popular among females.
A: I have some friends who are very into this, and they're going back and fixing the textile history articles.
AR: Yes, textiles! It's really amazing how much hasn't changed since 2003, when one Australian woman basically did it all. And then there haven't been a lot of contributions since, which is a huge bummer.
A: Are there any internal mechanisms in Wikipedia for the allocation of attention beyond just the natural interests of the editors? A clearinghouse for saying: “This thing needs help. Go fix that”?
AR: There's the Signpost newspaper, which is Wikipedia's internal newspaper. That always has an “article for improvement” or whatever. There's daily tips that you can sign up for. And then there are big campaigns that come from on high from The Foundation. And those are almost always something like “photograph a monument” because people love to do that and it's a good gateway activity. Or “write a biography about a woman,” which is good. Sometimes there are photograph campaigns for culture or heritage. But you can only nominate so many topics, and they'll only get attention for a week.
A: It sounds like Wikipedia needs to bring in newer, different demographics of writers and editors — but you've also alluded to the fact it's very intimidating to start up. Is there a process for acculturating new writers?
AR: Yeah, definitely. I care a lot about holding newbies’ hands. A lot of the old timers that came up in 2004 — when there were people who actually believed the “Ignore all rules” thing — they never had to deal with all of the norms that exist now. Even if you're a Nobel Prize winner, just getting all the norms takes a minute. You have to learn it. There are some people who are truly so welcoming and kind that it makes me feel very, very excited and hopeful about the future. And then there are others that are total assholes. And that's how it goes, I guess.
This happens in a lot of organizations. It's not just Wikipedia. There are people who are very good at what they do, but they're also unpleasant and gruff. They get called “unblockable” because they don't violate the rules enough. But — and I don't know if this is really a trend — from my vantage point it seems like people are less tolerant of the old school, laddish internet type trolling. And that is exciting too.
A: Wikipedia is unique in that it still feels very web 1.0 in its design and aesthetics. But it has attracted, like you said, lots of younger editors. You are Gen Z and you are both internet native and also Wikipedia native. How intergenerational is it?
AR: The stats on Wikipedia editors are so bad that they're almost useless because people tend to be quite private. And if it's a survey that's coming from the Foundation, there's a whole mistrust of the authority thing. People don't really like to fill those out. But that's the best, objective guess at how multi-generational it is. And that is showing that people are aging. The median, and probably average, Wikipedia editor appears to be older now than ten years ago.
A: Okay, you've alluded to the Foundation a couple times, and the tone of it has made it sound like the Capitol in the Hunger Games.
AR: It's not that. So there's a lot of overlap. If you love Wikipedia, working at the Foundation is probably your dream job. It's amazing. But people butt heads with the Foundation a lot. It's fun to complain about The Man. That's definitely part of it.
A major factor is that, maybe four years ago, there was a big flare-up called Forum Gate, which people haven't quite recovered from. Wikipedia has this arbitration committee, and the executives at the Foundation permanently banned a guy who — while he is in the category of gruff, not friendly, not the most empathetic — was a very effective editor on the site. And tons of people either stepped back for a while or resigned in protest.
Also the Foundation is big now. And I think that some people, even if they don't admit it, hold some jealousy and resentment that there's this fancy office in San Francisco. For a nonprofit job, it's a good one. Whereas the people who actually make the product? They don't make any money. It doesn't always feel like we're trusted by the Foundation.
But I like the Foundation. There's a natural animosity that comes up sometimes — but overall, we work together. And more than half of the Foundation staff works on the tech — servers, MediaWiki — because you can't get enough volunteers to do that.
A: How much authority does the Foundation have over the site?
AR: The Foundation doesn't really know what it's there to do. That's kind of what people have been saying for, like, 20 years. Some roles are very clear. If someone is threatening violence, give them to the trust and safety division. If there's a major tech problem, that's the Foundation. The volunteers just don't have the bandwidth for all that. But the patrolling, the determining of the rules, the participating in discussions — the Foundation is very hands off for all of that.
A: The Did You Know section on Wikipedia features a trove of random facts every day, and that's run by a bureaucracy of volunteers. But it's not neutrally curated. It's an area for editors to show off their special interests, which has resulted in some topics being, I think, restricted. And some of those are ones you would expect, like Taylor Swift. But apparently, also, such topics as Gibraltar and German hymns?
AR: So it's not like German hymns is “banned,” and neither is Taylor Swift, or even Gibraltar — although that one was a bit more firm. It's mostly just like — hey, can we cool it? Can we cool it with the German hymns?
The idea behind the Did You Know section is to be interesting for readers, sort of. But really it's there because people want to show off their brand new articles, and we want to let people feel the pride of showing off their new articles. It's also a way to get a mini-review process — to make sure everything in there is cited — and it will get copy-edited for you. It's a good way to get people used to getting edited and reviewed.
But it's mostly the same squad that does the reviewing. So there are some people who never submit to the Did You Know section, and there are a lot of people that submit everything. So if you pay attention to it consistently you'll be, like, oh, another New Zealand coin. (There's a girl who's been doing lots of those, and she rocks.) Or another article about a time zone! It makes me feel a little tender. It's kind of cute.
A: People are going to ChatGPT now as their first place to learn about a topic, compared to Wikipedia. I'm curious as to your thoughts on how LLMs are going to shift what Wikipedia feels like, and what you see as the future there.
AR: The traffic to Wikipedia has not taken a dramatic hit. Maybe that will change in the future. The Foundation talks about coming opportunities, or the threat of LLMs. With my friends that edit a lot, it hasn't really come up a ton because I don't think they care. It doesn't affect us. We're doing the same thing. Like if all the large language models eat up the stuff we wrote and make it easier for people to get information — great. We made it easier for people to get information.
And if LLMs end up training on blogs made by AI slop and having as their basis this ouroboros of generated text, then it's possible that a Wikipedia-type thing — written and curated by a human — could become even more valuable.
