In this episode of the NexChange Interview Series, we give you an exclusive interview with Wikipedia co-founder Dr. Larry Sanger, who shares his thoughts on his famous creation and his current blockchain-based project, Everipedia, with NexChange’s Olga Yaroshevsky. Here’s Part One:
Olga Yaroshevsky: I don’t want to start with the Wikipedia.
Dr. Larry Sanger: Good!
OY: Let’s start with the very beginning. I know that you received a Ph D degree at the very young age. How old were you?
LS: Not that young. Let’s see, I guess I was 32.
OY: Oh, 32! You are a doctor of philosophy – what was the topic of your thesis?
LS: Something called «Epistemic circularity». It’s in epistemology, the theory of knowledge. Basically, the problem is – since we have to use the ways of knowing things in order to confirm that those ways are reliable, isn’t that a kind of circularity. Can you use sense perception to confirm that sense perception is a reliable source of knowledge? That’s just an example of that kind of circularity. Actually there is a really deep problem that underlies most of the problems in epistemology.
I’ll tell you one thing – one of my professors, his name is Robert Kraut, a professor at Ohio State University. He claimed that part of the reason why I was so quickly able to latch on the idea of the back and forth that Wiki editing requires, is that I’d studied the ideas of pragmatists, which I guess is true, I did have a class on that. I don’t know if his explanation is true, but, you know, success has many parents, as the saying goes.
For one thing, there are definite ideas in economics, not once that I studied, that underlie the ideas of open-source development, and it’s like it and based on it. Robert Kraut was not wrong to say that by negotiating, and good faith anyway, about what truth is, eventually we get closer and closer to, as he puts it, illumining the true.
OY: But still, you’ve distanced yourself from the project because of ethical and moral reasons. Calling this anti-elitism, calling this non-democratic. Was that your phrase, about «the inmates got ahold of the asylum»?
LS: Yes. It continued to be actually the reason why I permanently left the WIkipedia, it is also one of the main reasons why I continue to criticise it. I noticed, as early as 2002, that Wikipedia was becoming a kind of mobocracy. That the way that people got ahead and the way that decisions were made is by treating each other in bad ways. Doing whatever you need to do to get to the head of a formerly leaderless mob. And there was a power backing, essentially, because there was no constitutional method of reaching content decisions. It’s supposed to be a consensus. Well, there isn’t a consensus positive about all kinds of issues that inform an encyclopedia. So there has to be some kind of a decision mechanism. In the absence of an editor-in-chief, or an editorial board, or some sort of a democratic decision making method, that has been regularised and can be overseen by some constitutional means – what happens? Well, some people look at whatever the local ideology is, in this case the ideology of consensus, and they say: we are the consensus. They become basically the mob leaders, it’s the way that I put it, because it’s completely unregulated. In other words, they are lawless. Of course, they can cite the alphabet soup of abbreviated regulations, but how they are applied is completely inconsistent. It depends entirely on essentially the power relations between the people in the system.
They didn’t have a solution for that, and they still don’t. I called Jimmy Wales on this privately, and said: unless you do something about this, reign in the bad actors… I also said – it’s a slightly different issue, but it’s related – do something to make experts in the project feel more welcome, because they had been driven away. All these great people that I had recruited for Nupedia, a lot of them were the original participants of Wikipedia. They were just driven away. And he didn’t seem to care. He denied to me, privately, that there was a problem. And I had made him an ultimatum. I guess it was January 2003. I said, okay, I’m done, I’m not going to have anything else to do with it. And I didn’t. So I actually became more of a critic, than a fan. Especially after they started denying that I’m the co-founder of Wikipedia.
OY: Seriously? Officially?
LS: Oh yes. For a few years, from the end of 2004 till the end of 2006 or 2007, Jimmy Wales was constantly going on the record saying: Larry Sanger is lying, he is making it up, he is not the co-founder of Wikipedia, I’m the sole founder of Wikipedia.
OY: So there was a human factor, after all?
LS: Yep, I would say so.
OY:… with your new project, Everipedia. Do you apply the decentralised of DLT technology to this consensus? Is there a consensus? Is it a democracy, or anarchy, as decentralisation maximalists often say?
LS: It’s definitely a democracy, right now. The way that Everipedia token works – you stake a small number of IQ tokens, that’s our token, if you want to make a new edit or submit a new article. You stake fifty tokens on something like that, and then the community votes on your edit. If there are more tokens upvoting than downvoting, then your edit is approved. That system itself, I think all of us would admit, is going to need refinement, it’s sort of the first draft, just to get it started on the blockchain. And we are on the blockchain, it’s live, it’s actually running. But at least it’s democratic.
In another phase of development, which should come within the next year, you will see rating mechanisms for articles. But in order to make that fair, and to actually motivate people to do the hard work of reading an article all the way through and coming up with responsible ratings, like global ratings of things like accuracy and style, and so forth – you need to persuade them that they’re not wasting their time. So it has to be one person, one vote.
We will allow people sort not just different competing articles on the same topic, but also allow the Everipedia community to choose from different competing edits. Let’s just say, for example, that there is a few different edits that are being proposed on a given really contentious article, «Donal Trump» for example. Then if there is a one person one vote system it would be much less gameable. It’s not to say that democracy can’t be gamed, but it’s gamed in a different way. It’s gamed in a mindlessly technological way. It’s more of a social gaming. I could go into that for a long time.
OY: Here’s a critical question: which flaws does this system have? Which obstacles have you faced while creating it?
OY: Be honest!
LS: I think we’ve focused a lot on pop culture topics, in the first 4 years of our development. I don’t think there is anything wrong with that, it has gotten a fair bit of traction of users. We have a younger writer base than Wikipedia does, and that’s a good thing. But going forward we definitely need to expand it with more traditional encyclopaedic topics. In order to do that we’re going to have to do another thing that we haven’t done yet. Which is constantly, like up to the second, import the latest Wikipedia edits. So Wikipedia has an API, they publish their latest edits available for anyone. So we need to grab those edits as they some online, and update our copies. So it’s a constantly forking version. This is going to be difficult, and we’ve thought about it. We need to have like a paragraph level of granularity for determining whether a part of an article should be updated or not. Like, if we’ve done a whole bunch of editing with the first two paragraphs, then the Wikipedia edits would not be copied over. But if all the rest of the article has been changed, we’re going to go ahead and continue to allow that to evolve the way that Wikipedia does. We haven’t done that.
A lot of the problems, the criticisms that somebody might have of Everipedia are basically things that we haven’t done yet. Sort of criticisms that people had of Wikipedia in the first years…
OY: And now they are applied to Everipedia. What was the hardest thing, so far? Not mentioning those things you haven’t done yet. In terms of development.
LS: That’s a question to our developers, to tell the truth. I’m not one of them, I just give a lot of advice.
We’re a little late in launching, not too late.
OY: When is the launch date?
LS: We haven’t set a launch «date», but the next few weeks, hopefully.
OY: Soon, wow. Looking forward to it.
LS: Possibly by the time this video comes out.