Featuring

Sponsors

Tiger Data – Postgres for Developers, devices, and agents The data platform trusted by hundreds of thousands from IoT to Web3 to AI and more.

Augment Code – Developer AI that uses deep understanding of your large codebase and how you build software to deliver personalized code suggestions and insights. Augment provides relevant, contextualized code right in your IDE or Slack. It transforms scattered knowledge into code or answers, eliminating time spent searching docs or interrupting teammates.

Depot – 10x faster builds? Yes please. Build faster. Waste less time. Accelerate Docker image builds, and GitHub Actions workflows. Easily integrate with your existing CI provider and dev workflows to save hours of build time.

Framer – Design and publish in one place. Get started free at framer.com/design, code CHANGELOG for a free month of Pro.

Notes & Links

📝 Edit Notes

Chapters

1 00:00 Welcome to The Changelog 01:22
2 01:22 Sponsor: Tiger Data 01:37
3 02:59 Start the show! 01:27
4 04:26 A Wikipedian 03:37
5 08:03 Policies and guidelines 03:44
6 11:48 Wikipedia PR 04:29
7 16:17 How much is too much 01:21
8 17:38 There Is No Cabal 03:45
9 21:23 Consultipedia 02:32
10 23:55 Political leanings 04:41
11 28:36 The 6th pillar 02:45
12 31:21 Sponsor: Augment Code 01:36
13 32:57 AI reads the wiki 05:51
14 38:47 Jerod's failed edits 01:55
15 40:42 Please donate 05:41
16 46:23 The next generation 03:08
17 49:31 More than a job 00:40
18 50:11 Verifiability 07:04
19 57:15 Sponsor: Depot 02:46
20 1:00:01 Sponsor: Framer 01:51
21 1:01:52 Bill's failed edits 01:03
22 1:02:54 Adam clicks twice 03:10
23 1:06:04 Bro, do you even wiki? 02:18
24 1:08:22 Open edit requests 01:14
25 1:09:37 Edit bait 00:59
26 1:10:36 Bad actors 02:39
27 1:13:15 How a page is born 03:26
28 1:16:41 Wiki PR cabal 01:51
29 1:18:32 Wikinomics 06:20
30 1:24:52 The Notability Company 05:26
31 1:30:18 Earned media 05:17
32 1:35:35 Jerod finds a falsehood 03:32
33 1:39:07 Bill throws us an edit 02:51
34 1:41:58 Bad source beats no source 03:15
35 1:45:12 Wrapping up 01:55
36 1:47:07 Closing thoughts (join ++) 01:49

Transcript

📝 Edit Transcript

Today, we have Bill Beutler with us, a Wikipedia expert. Is that fair to say, Bill? …you are expert at Wikipedia.

It’s a yes, but as I was saying before the show, the thing is, Wikipedia is so big that even the most veteran of veterans don’t know everything that happens there. So I will do the best I can, and if I don’t know, I’ll just say “That requires more research.”

Yeah. It really is kind of like the eighth wonder of the world at this point. The fact that it exists, it’s so huge, it’s editable by anybody, it’s got all kinds of intrigue and politicking and interestingness around it… I mean, it really is a fascinating creation that exists in the digital world, isn’t it?

There’s nothing else like it. For many years – I would say that even Google has competitors, even Amazon, I guess… Even Amazon has competitors, right? Walmart… But Wikipedia, until maybe, maybe very recently, if we want to go in this direction, there’s really there’s no other substitute. And so it’s also the thing that all of the other big tech giants can agree on, that they use it to underpin information… It’s the canonical source of information on the internet.

What does it mean, practically, if you could just flex a little potentially right here at the top, what does it mean to be an expert, or a self-professed expert in Wikipedia? What does that mean?

Well, gosh, probably everybody is self-professed if they say they are…

It’s the same thing as being a Wikipedia editor in the first place. What does it mean to be a Wikipedian? Because to some extent, it is just you have an account, you make edits, you want to advance Wikipedia’s mission, and you just kind of demonstrate it day by day. And when it comes to being an expert - I mean, I guess you could ask my friends who are Wikipedia editors if I’m a Wikipedia expert, and I think they would say yes. That said, though, it is not hard – just like on Wikipedia, because it is so vast, it is diffuse, the community is very… The borders are porous. People come and go. You can just say you are, but then you have to back it up, and anyone can dispute anything. And I’m not an expert on certainly the technical side of how the squid servers work and all that, but… The community dynamics, the guidelines that underpin the content, that’s where I’m strongest.

Does Wikipedia have things like RFCs, or specifications for how you – is there protocols that you can just like flex on and say “You know what? I know RFC 525, and on line seven it says X.”

I mean, RFC is very much a thing on Wikipedia. It’s a request or comment. Is that what your RFC is?

I think there’s a couple of different flavors of it, but pretty much. Yeah.

So an RFC would be a formal call – semi-formal. Nothing is completely formal on Wikipedia. Call for editors to give comment on a difficult issue that editors are trying to resolve. And there certainly are policies that people will cite, and they’ll do so in shorthand… Probably one of the more famous rules is neutrality; the policy that says you have to write with a neutral point of view. And so the shorthand for that is wp:npov. Or oftentimes it’s just npov. And so if you’re in a discussion with Wikipedia editors and you drop some of these terms, that’s one way to demonstrate you are speaking the language.

You’re in the know. You’re in the circle.

What are some other things? Npov makes sense… What are the other shorthand principles that all the editors know about?

I’ll tell you one that’s even more obscure. Probably many people have learned about npov over time. I’ll give you another one that is used by editors really in the know, and that is called NOTHERE. All caps, N-O-T-H-E-R-E. And this is when you’re dealing with a problem editor. Wikipedia tries to give benefit of the doubt to new editors coming in; if they mess something up, you want to assume good faith. And there’s actually a policy that says “Assume good faith.” But at a certain point, that could be abused, and you have to no longer do that. And so NOTHERE is shorthand for “Not here to build an encyclopedia.” So if two veteran editors are discussing a third editor, and one says the other is not here, that is a signal that someone’s getting ready to break out the band hammer, or put someone on suspension.

[00:08:03.00] What are the official rules? I know there’s something like no first-party sources, or like there’s specific things where if I’m just going to become an editor, I assume I go read a document that tells me “Here’s how it’s going to work.” Can you just lay the base foundation for how pages come to be, or edits come to be? Principally, not technically.

Yeah. Let me start with the superstructure of the rules, and we can get into some of the more specifics. So there’s not one document you’d be reading; it’s many, many documents. And when I got into Wikipedia, 20-some years ago, I was a much younger man then. I was single, and I had free time, and I would spend my hungover Sunday mornings just reading endless policy pages on Wikipedia.

Because the hangover wasn’t bad enough for you? You had to submit yourself to more…?

It slowed me down enough that I could just like read… They’re so long. These pages are so long.

But there are policies that are absolutely mandatory. There are guidelines which are very strongly advisory. There’s even a third level of rule, and that would be essays, community essays where they describe perspective points of view, but they don’t necessarily carry the weight of a policy or a guideline. And so the policies tend to be focused on editor behavior, such as not creating multiple accounts, no sock puppetry, things like that neutral point of view… Some of the core policies are content-focused. But once you get outside those core policies into the guidelines, then you have things like “How do you write titles of nobility? Do you put spaces in between, say, the name J.R, like J.R. Ewing? Do you put spaces in between that?” Last time I checked on this - don’t come after me if it’s no longer the case, but yes, you always put spaces in between those two letters, even if the person themselves does not.

How about the Oxford comma? Are they pro or against?

I think more pro, but it’s also not uniform.

Because the site is so vast. Again, I reserve the right to be wrong on this one, but…

That’s interesting, because the Associated Press has a similar nature. Is there any overlap between the Associated Press’s version of that and how it shows up in journalism, to Wikipedia’s?

There may be, but it would be more incidental. They’re not following Chicago style or anything like that. They’ve developed their own style over time. And by the way, another thing that they have to contend with is English comes in multiple flavors. There’s American English. There is British English. And so, for example – I’ll tell you what, some of my very earliest edits to Wikipedia back in the late aughts was changing British spellings in the articles about the Sopranos TV show back to English… Because the show had a huge following in the UK. And British editors really punch above their weight in terms of contributing to Wikipedia. But if it is an American subject, it should be written in American English, and vice versa.

By the way, if there is not a clear country of origin… Yogurt is one topic that there was an endless edit war many years ago about whether there should be an H in yogurt.

There’s no H in yogurt… Is there an H in yogurt?

Well, it depends on where you are. [laughs]

Like gas versus petrol. They say petrol over there. And I could be wrong, but the last I checked I thought it was still saying petrol. In that case, it’s just whoever created the page first, and now that was so long ago that no one remembers.

Crazy. And you actually created a company around this, because anytime you have a website as massive, that becomes kind of the standard of record. When things are on Wikipedia it doesn’t mean they’re true, but it means they’re at least vetted to a certain extent, and people accept it as true.

[00:12:03.24] And so PR is a huge aspect of that, because you are in control of your page, or a topic’s page. If you’re a public figure, you want to be painted in a good light. Maybe you want a scandal hidden etc. If you’re a business, you may want to just maintain your reputation… I’m not saying you’re doing scandalous PR stuff. I’m not sure what you’re doing, but I assume you’re helping businesses engage correctly with Wikipedia. How’d you get into that?

Yeah, I mean, the last part is definitely correct. We help our clients influence how Wikipedia writes about them, but also while following the rules of engagement for PR. So my career goes back to the early 2000s. I was a political journalist early in my career, working in Washington D.C. And around this time, the political blogosphere, which - I don’t know if that even really still exists. It’s all kind of like coalesced into – Substack has kind of made a little bit of a comeback…

…but the political blogosphere was exerting influence, and so around 2004 I was writing a column about political blogging for this insider tips sheet, sort of like a forerunner of Politico. And - well, Wikipedia kept showing up in the blog posts… It was the one thing that editors of the left and right and center all seemed to agree on, that “This is an interesting resource that can help explain concepts, so I don’t have to write them out.” And I got out of journalism and went to go work at a public affairs firm, digital public affairs firm, and I was the person in the office who talked about Wikipedia the most, just because I thought it was fascinating… Although the truth of the matter is I’d never actually edited the site myself at this point in time. What I had done was I was going to try to stand up my own wiki site. And naturally, I was going to create Blogopedia. It was going to be a directory, an encyclopedia of all the blogs out there. This thing never got off the ground, and then one day the CEO of the company asked if I would make a change to his… He had a friend who’s a member of Congress. Would I make a change to his page? And so I kind of looked into it, I was like “Alright, he’s got a point here.” And I’ll tell you what I did first, that really did set the stage for the rest of it. And that was – like, I didn’t just go make the change… I kind of like read up on the policy, and then I went to the Talk page, the discussion section for that article, and explained what I was about to do. And then I did it, and it stuck. And I got a thank you from another editor for it.

I probably wouldn’t have done it exactly the same way today, but this was like showing some foresight, showing some kind of respect for the fact that other people had views about how things were done on Wikipedia… It put me in a good headspace to kind of diplomatically work through issues. And in particular, the company I worked for at the time, it had clients like – you know, one they had back then was Domino’s Pizza. And there’s this teeny tiny - scandal’s even the wrong word, but a negative news story from way back then, where someone at a Domino’s somewhere, an employee, was doing disgusting things with the pizza, like blowing snot into it, or whatever. And it made the news. So someone had added it to Wikipedia, it was it was true, it had a source in theory… That’s all you need for Wikipedia. But I was able to make the case that these were just some random, low-level employees. This was not something that had buy-in from the corporate structure. In fact, the company responded quickly. And I got editors to agree that, “You know what? This is trivial.” It was not something that needed to be in the Domino’s page.

[00:15:56.14] And so that was probably one of the big a-ha moments, was if I make a good enough case, I can persuade editors that the Wikipedia article should say something different, in part, as you said, because it has the power… And even more so now than then, it has the power to affect the reputations of the subject it covers.

Is there a maximum length, or there’s a point where it’s like – because you could just put everything about Domino’s you could possibly gather, and at a certain point it becomes too much information.

Yeah. [unintelligible 00:16:28.28] adding too much. Not here.

How do you know when “not here” comes into play, and when just people have to make a call?

I’ll tell you what, I don’t think they come into play there, because if you’re adding too much, you’re like here to build too much of an encyclopedia. And that’s –

It’s definitely such thing as too much. So there is answer to your question, and I believe it is around a hundred thousand bytes. There’s rarely hard and fast rules… This one says as an article gets closer to a hundred thousand bytes, you should consider splitting it. And so this is why you have articles that have like – there’s a parent article, and there will be a child article. There’ll be Microsoft is the article, then there is Criticism of Microsoft, which is - it could be nearly as long as the actual Microsoft article.

Well, they’ve been around so long… I mean, at a certain point there’s just a lot. There’s a lot of history for a lot of these - either whether it’s an organization, or a topic, or… I mean, it’s overwhelming sometimes to think about how you would even manage the information architecture of such a thing. And it sounds like it’s somewhat organic, and has grown over time…

How many editors are there? So we had a friend of ours, Andrew Nesbitt, on the show a couple of weeks ago, talking about ecosystems in open source, and how he found that when it came to certain popular open source packages, there are about 15,000 people who are kind of underpinning the entire software world, so to speak.

There’s the old trope from the XKCD comic, like it’s one person in Nebraska who’s like running all the dependencies… Well, it turns out 15,000 around the world is pretty close to that. And I think with Wikipedia, I assume there’s a – I don’t want to call them a cabal, but there’s got to be a core group of N editors who are kind of in charge. Do you know that number? Are you one of that group? Do you know that group?

I am not personally in that group, but I know plenty of people in that group. It’s funny… Did you pull the word cabal out of thin air, or did you pull that from doing some reading?

I just thought of it because it seemed like appropriate at the time…

It is. Well, so there is even an essay on Wikipedia called “There is No Cabal.”

And the reason why that essay exists is because – the cynical take on it is yeah, there’s not one cabal. There’s many cabals. And that meaning there are kind of groups of editors who will talk together offline. And offline coordination is frowned upon… But let’s not pretend it doesn’t happen. It can happen in good ways, it can happen in bad ways. I think it’s kind of value neutral.

But to your question about like – because it’s similar to the open source communities, for sure… Which it is - it’s an open to open knowledge community. Open source is certainly a component of it. I know it follows a power law, where there is a small number of highly active editors up at the very top, and a long tail of contributors who might contribute here and there.

So the numbers I’m going to give you are going to be very approximate, and they’re just kind of offered for demonstrative purposes, not to be quoted as accurate… And these numbers are very findable. But let’s say that there’s 3000 editors who are editing every single day, often hours a day. And they are really like the core group that keeps Wikipedia going. And a lot of them are not always writing the articles that you read. They are kind of arguing over policy, banning problem editors, and kind of working out kind of the structure of the behind the scenes; all the behind the scenes stuff, janitorial stuff at best. I don’t know about at best… But you know, Wikipedia has an arbitration committee that is sometimes considered to be Wikipedia’s Supreme Court… They’re in that 3000.

[00:20:18.25] And you’ve got maybe 30,000 editors who are editing in a given month. I’m in that group. I don’t edit as often as I used to in the early years… And there’s also definitely a lifespan. There’s… I’m forgetting the term I’m looking for here, but like a course of events, where…

Lifecycle. That is what I’m looking for. A lifecycle to being an editor, where you kind of get in, you get real excited, and you add a lot of material… And then you kind of like drift away, as you’ve already –

Burn out is totally one thing. Shared all the knowledge you wanted to share is another… There’s a lot of editors who really don’t create content. They just kind of shuffle around categories, and poke around here and there… And then to finish out the down at the long tail end, there’s about 100,000 editors who are - and this one I’m a little more certain about… There are 100,000 editors who make one edit, minimum one edit per month. And so that’s great, that it is really the people at the very top who are doing the most of it.

It’s kind of crazy. So going back to your work and how that works, how your work works, is like - when a company engages with you, do you then advise them how to engage with Wikipedia? Or is it like “Hey, we really want Bill’s edits, and he has clout…”? Because there’s like a clout system that’s just part of humanity, that’s built into this, as you edit, to a certain extent… Like “This guy’s reliable because he’s been not just doing this one company or this one topic… There’s clearly people that just come with their agenda, and then there’s people who are just there to edit and then they may also have it… There’s so many different angles.

How do you tell brands to engage?

So we’ll do both, by the way, myself or members of our strategy team. We have a six-person strategy team. Or seven, I suppose… But folks who are working on Wikipedia all day, every day. And we will, using our disclosed business accounts that say “I’m Bill and I work at Beutler Inc, and my client is such and such, and I’d like to propose a change on this page.” We’ll do that. We also will coach our clients through leading outreach themselves. For large companies, where they have a corporate comms division, and they have more resources, they more – I don’t know if they have time for it, but they… Like, if a big company shows interest in Wikipedia and cares enough to put an employee on it, and share information, and talk with editors - that can be a good look. And so that is certainly a thing we do for certain, especially like Fortune 50 clients. But also, we will represent clients ourselves.

If you’re having us do it, then we’re going to decide what we’re going to be willing to ask for or not. If the company is the one who’s out there with the disclosure, it’s all on them. If they want to push a little harder on something that’s in the gray area, we might be willing to do that… But the fact is - yes, there is an aspect of reputation where we’re likely to get faster replies, myself or my colleagues, because editors will be more likely to recognize us. On the other hand though, there’s thousands of editors out there, and they don’t all know us. So there are projects where we work on the Talk page, communicating for a company, and we’ll be talking to editors we’ve never run across before. So there’s no – Wikipedia is all shades of gray. There’s never any one right way to do everything.

[00:23:55.02] So a lot of sites or a lot of tech companies, because of their location and their employee base and stuff, they tend to lean to the left side of the political spectrum here in the United States. And there’s lots of claims that Wikipedia also is captive to that. And I’m curious your perspective from a “in the trenches” kind of guy who’s getting changes done… Do you see overwhelming political leanings, generally speaking, in Wikipedia? Do you think that that’s bonk? What are your thoughts on that?

So the point about companies and their lean - I mean, they certainly did, I would say, in the Obama administration, Obama era… But especially now, in Trump too, I think that they kind of blow which way the winds are going.

They’re very malleable. They’ll go where the dollars tell them they need to.

Right. They have shareholder value to worry about.

A hundred percent. Wikipedia, I would say - and just for the record, I’m pretty centrist myself. I was a little more on the right when I was in my 20s, I’m a little centrist, not quite the left, in my 40s… But I’ve kind of seen the full gamut. And Wikipedia definitely has a center left, even maybe a little bit left bias. And I would say that that has something to do with the fact that a lot of the contributors, many have an academic background, and that obviously leans heavily left… They have journalistic values, which sometimes do have, often have at least, even if not, I would say, a left wing bias, they are at least parallel, there’s a parallel perspective there… That said, right now there’s probably more right wing criticism of Wikipedia than there has been in ever, really. And it’s been kind of rising over the last few years… And it’s not to say that there aren’t some decent criticisms of the handling of certain topics… But I think a lot of them kind of willfully misunderstand how Wikipedia uses sources.

So one of the big complaints would be “Well, we can’t use Breitbart, or The New York Post, or The Daily Mail to cite sources.” And kind of blow right past the fact that these are all publications that are not known for their journalistic scrutiny, integrity… No one’s going to flag you for using The Telegraph, or using the Wall Street Journal. But the fringier sources… And now there are more fringy sources on the right. I think there’s a whole interesting thing to be written, or many things to be written about why that’s the case. Why the right does not have kind of journalistic values as part of its animating values, and the left does more. That’s a huge topic, and it could be the rest of the show, if you want it to be. And I’ve written about this…

I don’t want it to be. I just wanted to tiptoe into it and then move on, but…

…this is an interesting perspective from you.

I think the way to maybe approach that without going left, right or center is to talk about how power is distributed. Because you mentioned there’s no cabal, right? So if there’s no cabal, how is power distributed? How do you acquire power? How do editors squash other editors? How are there clicks? …things like that. I think that’s kind of like maybe a version, Jerod, of what you’re talking about… Because it’s not quite that, but it’s how do you leverage your own bias and your own power.

Yeah. I mean, you can definitely discuss it in terms that are not explicitly left/right political. There’s kind of a truism that if there is no visible hierarchy in an organization or a community, then it becomes an invisible hierarchy. And so that is 100 percent true at Wikipedia. And to my knowledge, nobody has ever, say, mapped out exactly what that looks like…

[00:27:54.27] If you were to look at Wikipedia’s own maintained list of the most active editors in Wikipedia, within that top 50 or so, you’re going to find some of those editors who are, just by virtue of being there all the time, the ones who they know the most policy, they have the most connections… And in certain topic areas - there are topic areas that do get captured by the contributors. Most of those are not necessarily the most active pages. A lot of times that’s in a little out of the way places where most editors don’t care to spend any time there… And so - you know what? When I talked about the rules of Wikipedia, I mentioned the policies and the guidelines. I didn’t mention that there are five pillars of Wikipedia, which are like Wikipedia’s encyclopedia, Wikipedia has no firm rules, etc. etc, things like that. There’s a sixth pillar that people sometimes refer to, and that is “The person who cares the most ends up getting their way.” [laughs]

That can be life sometimes, too. Yeah. Just like “Well, you care more than the rest of us, so go ahead, man… Just go ahead.” [laughs] That can be shaky when it comes to what’s true or not. Like, the person that cares the most gets their edit… It’s like “Well, how about the most truthful thing?” But when we all disagree about what’s true - I mean, those are really hard problems, aren’t they?

It is. I mean, you remember Stephen Colbert back in the early days of his Comedy Central Show coined a word… It was wikiality. And wikiality was – it’s the truth that we all agree on, because we just agreed with the case. And so it was couched as a criticism of Wikipedia, but also was more broadly intended to comment on shared delusions… In his case, he was definitely criticizing the American right… But certainly, shared delusions have no political home. They could be all over the place. And so there’s a bad version of that, and there’s a good version of that.

And one of the things about Wikipedia that I find really to be fascinating, probably one of the reasons why I got drawn to the project in the first place, was life does not have a black and white answers for many, many things. One plus one equals two, yes… But the right form of government is something that can change over time, even. As democracy works in the United States, it has had a harder time in other countries… It’s not because democracy itself is bad or wrong in some other place, it can be about what the culture around it is. And so there’s just no way that you can set it and forget it with Wikipedia, or with all the topics it covers. There’s no substitute for the daily task of rolling up your sleeves, getting in there, and making sure that things stay correct. And that’s why Wikipedia’s job will never be done. It’s an ideological informational battle that rages on around us, and never will go away. If it goes away, then there’s a problem. If Wikipedia calms down, that is when I get worried. As long as people are arguing about it, you know it’s healthy.

So one of the things I’ve been telling people who don’t understand how all these new AI chatbots are doing what they do - just as a mental model, I know it’s not actually true. I just say “Just imagine that they’re kind of reading Wikipedia to you.” Like, they basically read the page so you don’t have to, and then they summarize it. I mean, obviously that’s a simplification and not always true, but certainly, these things have all read Wikipedia and continue to, right?

That is probably the most succinct way of putting it, that they have read it, and continue to.

Well, I’m trying to think of like why it might go away, is because maybe there’s a demand for the information, but not for the website, so to speak.

And then there becomes either a perverse or inverse incentive to even edit, if there’s no demand for the website. I mean, I feel like a lot of our internet underpinnings are kind of up in the air right now, of like “Where will they be in 10 years?” And I’m not so sure about Wikipedia. Because it’s certainly – it’s tantamount to these things right now, and new things they need. But why edit it if there’s no there there?

It’s a really fascinating question, because it’s the conversation that Wikipedia editors are having themselves. And the foundation that runs Wikipedia also is very much focused on this issue. I am a little more bullish. I’ve been around for Wikipedia about 20 years, and I’ve seen many predictions of its demise.

I remember I was even interviewed for an Economist story back in, I want to say like 2012… It was called WikiPeaks… You know, a play on WikiLeaks, but WikiPeaks. It was like a “Has it reached its zenith?” thing.

Oh, that’s like a Twin Peaks thing. Okay. They peak of Wiki. Gotcha.

Right. Or like Peak Oil, you know? That concept.

And so I’ve seen – rumors of its death have been greatly exaggerated in the past. And I will say that I do think that the core community doesn’t do it for the clicks. They do it because it’s a hobby. It is something that – I oftentimes will liken editing Wikipedia, especially the people who are just recategorizing things, and kind of going about their way, just working through multiple articles and making the same kind of change over and over… It’s like knitting. It’s a repetitive, soothing…

A hundred percent. I have friends who will sit down after a hard day’s work and just kind of click through the edit lists and make the same change, over, over, over. And there’s a weirdly satisfying thing in doing that. And that’s one mode of motivation for Wikipedia editors. Of course, there are many. And as long as Wikipedia itself continues to underpin what Google, what ChatGPT is putting out there, its influence is going to be continued.

I would say that because of the rise of AI search and the widespread awareness that AI has read, and continues to read - I love that formulation. I’m going to steal that.

[00:36:00.19] Yeah. That it’s more important than it was two years ago. Or people have realized that, and so it’s more front of mind. So not to get too much into my business specifics, but this has been a really interesting year. It was kind of a rocky first half of the year, and I think that had something to do with the election, and brands kind of sitting on budgets, waiting to see… But then by the middle of the year, the AI – like, this is really the year, in multiple ways, where AI became part of business workflows. People had to figure out how to make use of it, even outside of technical circles. And so the second half of the year has been crazy for us. We really have not seen this kind of interest in a long time… Because all of a sudden there’s now – it’s not just Google as a driver. But that was historically the driver of clients to us, because Google relies on Wikipedia, it shows up at the top of some of these search results… Now there’s a whole second driver, and that’s ChatGPT.

It’s interesting to think about that though, because you’ve got these non-interfaces interfaces to Wikipedia… Like you said, the website not being there, Jerod. I think we still need – I think what you’re saying is we need this source of truth. I think so long as we have that centralized and societally elected source of truth, then Wikipedia still has a pretty good underpinning in terms of its foothold to being just that. I can’t imagine that the LLMs extracting that, and being trained on that supplants that, because you still need – those are distributed. You’ve got various frontier models. They’re not all controlled by one… And so they’re disparate sources of truth, if you want to call them that. And they’re really just copies of the truth, not the actual source.

I still value, personally - this is where I’m personally at… I still value the original database, which is Wikipedia. So I’m bullish as well.

What if you’re extracted from that for 5 or 10 years? Are you still going to value that? Or are your kids going to value that?

Well, your kids might not even know Wikipedia is back there, behind the scenes, doing its thing. Now, to your point - and I agree with you guys, that doesn’t mean that getting your information in there isn’t still important, because eventually it makes its way out of the GPTs, maybe. But they can also opt to not use your – so you’re kind of a degree of separation away from your edit actually being useful, because they may or may not use it in the final output. Whereas anybody who goes to the webpage, they’re going to read your sentence, as long as your sentence stays. Right?

So there is like a decreased value of an editor for those who are not editing because it relaxes them, but they’re editing because they want some sort of –

I have no problem with people editing for self interest. The only edits I’ve ever made on Wikipedia was adding links to our podcast…

For self-interest… [laughs]

No, it’s fine. Because we are a source of information. There’s a public figure on our show, they said something interesting, you go put it in the page, you get a link back to your podcast… Like, that’s self interest, but it also is adding value to the wiki. And so it’s totally fine. I’ve got no problem with that.

Did those edits stay? Have they remained?

I don’t know. Well, the problem was… I think I realized – I thought maybe they were going to be… They’re all no follow links, aren’t they? Like, every link is a no follow…

They are no follow, but it’s still value being there.

Yeah. I think there’s still value being there. It was kind of like an ROI question of like how worth it is it for me to go be doing this, and the answer was “Well, maybe I should call Bill Beutler Inc. and have them do it for me… Because okay, my personal use - I never look at the sources very often. And when I do, I even less often click through to a source. So if it was a follow link, I would at least be getting Wikipedia juice, and I think that was worth it… But without that, I was kind of like “Meh…” And then I think I got one denied, and I was like “Well, this sucks.”

I mean, this is why they no-follow it. Exactly.

[00:39:53.25] Yeah. I understand why they no-follow it. It would be so many more edits if they didn’t. But yeah, I just kind of gave up. I think I did two or three, and I was like “This is not worth the squeeze.”

Yeah. And this is why we work with SEO firms a lot less than we work with PR firms.

PR firms trying to tell a story about a client, and SEO trying to push a website up. I will say, though, that’s starting to change just a little bit in this AEO, GEO era. The links between – actually, less adding links, and I’m thinking more of like creating links between pages to strengthen the relationship between, say, a client’s page and a concept they want to be associated with. I’m not saying we’re seeing a lot of that, but it’s just a conversation that comes up more often than it did for a long time, which is it almost never came up.

Well, here’s a sustainability question, because we’re in the current time of year when Wikipedia grants us the awesome “Please donate” button with a picture…

…and the sob story… Which, for my money, I’d rather you just charge for the website and not do that, but I understand. People have their opinions on it. I like the fact that it’s free for everybody, so I understand why they go the route they go… But you know, what Reddit does is they let everybody contribute to Reddit, and then they take all the information and they sell it to Open AI, to be trained on.

And Reddit has advertising. It’s a for profit company, right?

Whereas Wikipedia is not. Yeah.

Yeah, exactly. But they’re sustainable in that way. Like, they’re not just sustainable, they’re actually like profit-generating. Do you think that Wikipedia could potentially just – I mean, they are the world’s source of truth on many topics… Couldn’t they just charge the model trainers for that, and not have to – like, wouldn’t humanity better off if Open AI and Microsoft and Google were paying Wikipedia, and then they wouldn’t have to take donations?

Well, they would have to change their license, their creative commons license, that - you can use it. Anybody is free to use anything on Wikipedia, remix it, even use it for commercial purposes… And all you have to do is say you got it from Wikipedia, and you’re covered. And that’s a very kind of liberal license that they have chosen. Copyleft, as opposed to copyright.

Very permissive. It’s kind of like Jimmy Wales. Jimmy Wales was profiled in the New York Times magazine about a decade ago with the unfortunate kind of demeaning headline “Jimmy Wales is not an internet billionaire.” Among the people who’ve founded top 10 global websites, those guys are all billionaires many times over, with the exception of Jimmy Wales. And the problem is - and it’s not a problem, honestly. Jimmy Wales’ wisdom - his smart decision at the time was that if he had tried to add advertising and make it a for profit site and then limit it that way, it would not have become, not have grown to be what it is. You know, contributors really were drawn to the altruistic nature of it, in the sense that they were all kind of collecting all the world’s knowledge. If they were trying to if they were trying to monetize it, then the whole thing would collapse like a house of cards.

So this is still true today… If they were to start charging for it, they could – and then you know what? I should be careful. There is a one part of the Wikimedia Foundation that is charging for some parts of it. So the Wikimedia Enterprise is a for profits company inside the Wikimedia Foundation. And I guess I should say they were a one-time client of ours a few years ago. Not now. But I still have friends there. And they do offer up a professional version of the Wikimedia API, which is much more reliable, it has guaranteed uptimes, and an SLA… Which the regular Wikipedia API, which my firm does use for monitoring software that we built and maintain - it’s an unwieldy API.

[00:44:00.16] As Wikipedia’s knowledge is useful, but untamed, so too is their free API… And so they’re never, however, going to charge, say, the AI companies for the main unvarnished product. But what they did do a few months ago was they did – like, the Wikimedia Foundation was annoyed with the big AI companies for putting strain on their servers by crawling the pages, using up a lot of their server time. And so what they did is they put out a couple of different cleaned up versions of it that they were like “Hi, Open AI. Hi, Anthropic. Please, please, please crawl this one. Don’t crawl our main site. We’ll give this one away for free. You can crawl it.” But they can’t, due to their licensing, force them to use it. They’re not charging for it. Does anybody use it? I don’t know the answer to that, and I’m not even sure the foundation knows the answer.

So there’s a lot of interesting questions like this that we won’t know the answer to, I suppose, for a long time. And to your question earlier about “Will people contribute if they don’t see the page show up in search results?”, so there’s not the same glory in their work being in the limelight of the top of Google search results. I will say that for the current generation of editors, for the people who built Wikipedia, that does not matter. They do it because they love it.

The real question - and this has been a conversation I’ve heard at a few Wikipedia conferences in the last few years, as Wikipedia approaches its 25th birthday anniversary in January… It’ll be celebrating 25. You know, the people who started Wikipedia - they are all 25 years older now. And obituaries for Wikipedia editors have been – I don’t know if it feels like it’s been getting more common. It would make sense, as they all get older…

I know a handful of people who’ve been editors, who passed away, from even old age… And so the real question is, will there be the next generation of editors? Will they come on and continue the work? That is a source of anxiety, for sure.

Yeah, I d