SIO389: "Wikipedia Is the Most Accurate Form of Information Ever Created"

Transcript

This transcript was generated automatically and may contain errors — the audio is the record.

0:00:12 Thomas: And welcome to Serious Inquiries Only. This is episode 389. I'm your host, Thomas Smith, and if I sound weird, it's because I'm recording just straight into the old laptop because we are traveling right now. And I and little baby Remi had a great time at QED and since we're here, when in Rome, except not Rome, them. Figured why not also do a little bit of traveling. So we are abroad.

Actually, before we get to today's show, I wanted to mention that I think the last episode for this month going to be a bonus because we're about to meet up with one Heath Enright. I thought it would be fun. We were wanting to record a little bonus just talking about QED and some Some of our funny and fun and interesting experiences. And then I thought, wait, we're going to meet up with Heath. So maybe Lydia and I can tell Heath what has happened.

And we can get Heath's reactions to it. Which will be hilarious because there's some good stories. So that's what we're hoping to do for the last episode of the month. And that will be available on patreon.com/seriouspod. So go there. And I'm going to make this available to all patrons, any level. So get on there and pledge anything, and you'll get access to this when it comes out very soon. Today's episode. I thought it would be fitting, considering the topic and the guests, just read Dr. Amy Bruckman's Wikipedia article, for the intro, of course. Amy Bruckman, born 1965, is a professor at the Georgia Institute of Technology affiliated with the School of Interactive Computing in the GVU Center. She is best known for her pioneering research in the fields of online communities and learning science. 1999, she was selected as one of MIT Technology Review's awardees, honoring 100 remarkable innovators under the age of 35. currently directs the ELC Lab, the UROC program, and the Web Science Initiative at Georgia Tech.

Has She produced dozens of scholarship articles at Georgia Technological University and has published thousands of scholarly papers in peer reviewed journals journals, and has given presentations at high -profile academic conferences such as ICLS and CHI. Bruckman's

0:02:25 Thomas: Graduate students she advises, has dealt with topics in information ethics, game studies, social media use, and collaboration and wikis, and women in computing. And finally, her book, Should You Believe Wikipedia, was published in 2022. And that is the topic of today's show, Should You Believe Wikipedia? I I have quite positive views of that, but way I found Amy and invited her on the show was I saw her quoted in a article about Wikipedia and how AI is perhaps destroying Wikipedia. We'll talk more about that later. In it, she was quoted as saying, Wikipedia is the most accurate source of information humans have ever created, is a bold quote and I thought it would be so fun to talk to her about that so why don't we get her on the line let's go over to my interview with Dr. Amy Bruckman Amy, how are you doing today?

0:03:17 Amy: Great. It's really a pleasure to be here.

0:03:19 Thomas: Oh, thank you so much. I am excited to have you. Just reading an article the other day is how this happened. I was reading this article that was really fascinating out of, I think, New York Magazine or New York Times or both. And it was about Wikipedia and AI and the chatbots are using Wikipedia and all that kind of conflict there. And I saw your book referenced pretty early on there. And instantly I was like, I have to to speak you because it seems like such a fascinating topic at exactly the right time. And you. And you were kind enough to agree to come on pretty, short notice. So I want to thank you for that.

0:03:19 Amy: Oh, it's my pleasure. And I like the book.

0:03:57 Thomas: Absolutely. I I can't get enough. I think this is so fascinating, but I wanted to start if you don't mind with a little, the you background know, stuff. Do you mind talking a little bit about your story, how you got into the line of research you're in now and all those kinds of things, if you don't mind.

0:04:12 Amy: A student at the MIT Media Lab in the 90s, getting my PhD and was studying how to use technology to help kids to learn and doing both education research and using the internet as the medium to do education research. I created a virtual world for kids called Moose Crossing where kids learned object-oriented programming and practiced their creative writing. And then when I came to teach at Georgia Tech, over time, I kept doing lots of Internet research and less focus on education. So I started off with education on the Internet and then just said Internet on the internet over time. And it's just been very lucky timing. There's been so much going on and it's been so fun to witness it and to try to write a little bit about how we shape our world by the internet technologies that we create.

0:04:37 Thomas: Wow.

0:05:13 Thomas: One interview I watched of yours where the interviewer was talking about how some of this might've been a bit of a challenge to get people to take it seriously early on. Is that true? I can imagine, you know, it seems like you were pretty early on a lot of this stuff. What were some of those interactions, or was it difficult to get people to kind of care about this stuff when you first started.

0:05:34 Amy: Know, from when I first started, could see that this was going to catch on, and I think the more than anything thing has been a bit of a hangover from 1990s cyber utopianism, that we all really believed that this technology was going to be wonderful and that people are basically good and that fact would mitigate potential downsides, and of course it's more complicated than that. And I still, still am a hopeful person. I still see a lot of good in what the internet has brought to us, but clearly more complicated.

0:06:19 Thomas: If anything, I feel like maybe around 2016, well, probably earlier, it's almost like it flipped the other way, where now I wonder if there's an overcorrection. It's like, ah, no, this thing sucks. Everything about it is bad, you know, and it's kind of finding, okay, what's an actual appropriate view to take there?

0:06:34 Amy: Absolutely, there's been an overcorrection. The truth is somewhere in the middle. The technology has tremendous upsides and also has some downsides. And I'm fairly critical of AI. That AI is going to have a devastating effect on jobs and the economy. There still things that it can do? Yes, that are positive, lots of them, but we're going to have to find ways to figure out what the wise uses are and what the not wise uses are. People

0:06:38 Thomas: Hmm.

0:07:01 Thomas: Is the note that I'm using saw in this article that onto, and this is why I'm so glad to have you. Were quoted as saying, Wikipedia is literally the most accurate form of information ever created by humans. And I don't know if that's a tough sell for some people, but for me, that is not a tough sale at all. I actually, it's funny, I think I had recently been relating. I'm I'm 37 and I went to college in, so I started in 2004, 2005. So I was exactly the time where, you was there, but it was still really frowned upon. Like professors were like, don't, you can't cite Wikipedia. Like that's just the laziest, you know, it could be totally wrong there. And, you think I was pretty, quick to recognize, like, I don't know, actually, I feel like, I feel like this thing's pretty good. You know, It's a good citation. There's so many checks and balances. So that's why I really latched onto that.

Your book, Should You Believe Wikipedia, is such easy sell for me. But I'm wondering, how about for people? I mean, is that the kind of quote that, you do think the general public or the people you tend to talk to about this? Are they skeptical of that claim? How is that received?

0:08:07 Amy: Are very skeptical, but it's literally true. The reason I wrote the book is to try and teach people some epistemology. I, is the study of knowledge. I, I believe that all of us need to know epistemology, at least a little bit of social epistemography, to survive in the world today. It's not a fringe discipline. It's fundamental. Social epistemology says that the way that we know what is true is social. The more people believe something to be true, the more confident we can be that we're right. Exists.

We have only imperfect access to truth via our senses, and the way that we confirm what is true is through social consensus. That social consensus can sometimes be wrong, but the more people agree on something, the more certain we can be. And a popular Wikipedia page can be watched by literally thousands of people and updated moment by moment. About a peer-reviewed journal article.

It's reviewed by three experts who may or may not pay attention to every little detail and may or may not be skilled in every aspect of the method. That to a Wikipedia page where thousands, thousands of people can be looking at it and updating it all the time. It's always up to date and checked by tons of people.

0:08:15 Thomas: Hmm.

0:09:38 Thomas: Yeah, I always had a passion for philosophy. I almost majored in it, but not quite. But I really loved epistemology as a topic. I was wondering, yeah, into your work and how heavy, at least the kind of the introductory part is, do you have a philosophy background? Is that something you're trained in? Or was this kind of, you know, something you researched for the book? Or how has that played a part in your life and your studies?

0:10:00 Amy: Am from an interdisciplinary graduate program at the MIT Media Lab, and I teach in the most interdisciplinary computer science department, probably in the world. And at all these places, we take our philosophy, our theoretical writings very seriously. We, reading groups, when I was a grad student, where we met every Thursday night and read serious theory together, and then went out to dinner and then we're not dancing. And so, a long tradition of saying, look, we can't understand what we're doing without borrowing from the social sciences, borrowing from the humanities, and synthesizing across all these disciplines.

0:10:24 Thomas: Wow, I love that. Yeah, no, I, I the reason I ask is that very thing. I think it's so important. And it's, it's really good to see. This, I feel like there's a, you know, a little bit of a negative about image some philosophy, especially, you know, more modern philosophy. It's like, it'S so specialized. Is there really any? Are they even making any progress on anything? And I think your work shows how relevant a lot of those concepts are.

0:11:08 Thomas: And you believe this thing, but if you didn't quite have, what if you accidentally believed it for justified reasons that then are wrong? And it feels a little bit, you know, like, okay, this is a little obscure. I'm not sure how relevant it is. It's really cool to see it come to life in your work and see like, no, this, this is pretty relevant. Like this is going to make a difference, uh, especially in the days of AI and, and all this information theory that you're dealing with. So that's really Cool.

0:11:30 Amy: Yeah, I mean, epistemologists have been writing about things like, for example, the problem of testimony, for I don't know how many years. Testimony is the question of when someone says something, should you believe it? What would lead you to feel like believing it is a sensible choice, as opposed to not believing it? Well, the whole internet is a problem of testimoning. It's all around us. Everywhere. All the time. I, thankful that there's some smart people who've been thinking about this for a long time. So,

0:11:54 Thomas: Yeah, for sure. Yeah. Um, I'm wondering, um, far back does your history with specifically Wikipedia go? Because I was wondering, like, that statement, you know, again, the quote I read that it's the most reliable thing. I totally agree with that. Has that always been the case? Was there a time when, you know... When it was like, all right, there's just not enough kind people working on it to Because truly, we can all imagine, and I'm sure we've all had experience with the obscure Wikipedia page that's, you know, barely anyone's looked at that, you you know, okay, I don't know how reliable that is. So talk about that a little bit. And do you think this was really true? Like your statement about it being the most reliable source of information? I'm wondering roughly, when do YOU think that really became true? Just

0:12:41 Amy: It's still the case that an Wikipedia page is not very reliable, but a popular one is highly reliable, and that's always been the case from day one. If something was edited by lots of people, the more people looked at it, the more people checked it, the more often it was updated, the more you could say it was reliable. My history of editing Wikipedia predates the edit history of Wikipedia. I have a bunch of edits that got erased when they restarted the edit -history.

0:13:05 Thomas: Hmm. Oh, I didn't know about this.

0:13:12 Amy: A couple pages that I created and it doesn't show that I've created them because the page history doesn't go back that far. Are. Was so much fun in the early days. So, in the early days, I threw up a mostly blank page for peanut allergy and milk allergy, for the seven major allergens, and I just wrote on them, you know, we have a page for food allergy, but I really think each of the seven most common food allergies deserves its own page. I put like, I don't know, a couple sentences for each one, and oh, wow, those pages now, are each of them are so developed and so thoughtful.

And I tried to edit one a or so ago, and the nice person who undid my edit very politely referred me to the journal where I would find out why the thing I was adding was now no longer current. And I thought, oh boy, This created this page, but it's way beyond my expertise now. Used to be that you could just do anything and everybody was like, hey, great, yeah, let's all help out. Now it's gotten really rough sometimes.

People want to delete everything you do and are suspicious of other people's motives. And so I do worry about the culture on Wikipedia. It's not warm and welcoming as much as it used to. Oh,

0:13:22 Thomas: Hmm, hmm.

0:14:30 Thomas: Actually, that's a, a good question. Why don't we, so it sounds like you have of significant history with Wikipedia, for those of us who, you know, maybe aren't as familiar with kind of the inner workings of it, do you mind giving us some of those basics? Like, the editors? How did people become editors nowadays? What's their, you know, methodology? Are you allowed to edit anything? Do you have to be an expert? You know, some of those basics questions for those of us who aren't aware.

0:14:51 Amy: Yeah, sure. Can edit even without being logged in. There's some pages that are protected because they're controversial. If a page is kind of a mess, then the administrators might lock it down. But most pages, anyone can edit, and you don't even need an account. Other people will look at your edits, and if you vandalize something, they'll undo it, sometimes in a matter of seconds. Page has the main page, and then has a top page. The top page you can post things where you talk about, hey guys, we could really use to fix this part of the article, it's not really up to date, or do you think about the change I made, or do you, we should do this, or is this the scope, should this be included or not? So there's a place for every single page to talk about the content and to collaborate with other editors.

0:15:39 Thomas: Is this entirely self-directed? Is there anybody, you know, saying, hey, we've noticed this kind of systemic, you, or thing that we should go through a lot of the pages, and anybody doing that kind of top-down stuff, or is it just entirely self-directed?

0:15:52 Amy: Entirely self -directed and that's why there may be more written on Wikipedia about Terry Pratchett, the fantasy author, than about some branches of science, right? Because people loved Terry Pratchet and there, know, there's always a battle about this isn't a fandom site, please keep the level of detail about the thing you're excited about down a little bit. Yeah, it's self -directed. And that leads to weird coverage biases. The things that are covered are the things that people felt like writing about. And there's good news and there's bad news. Good news is that completely naturally, the content reflects what people care about. Bad news is there's some stuff that that people probably should care about that is not covered as much as it could be.

0:16:01 Thomas: Yeah. Yeah.

0:16:40 Thomas: Yeah. It's going to reflect kind of the biases of the population that I guess is, you know. Utilizing it. Are on that topic, I mean, are there any other weaknesses or like, you that maybe this process can go wrong that are worth talking about?

0:16:41 Amy: Right, exactly. There are all kinds of ways it can go wrong. My favorite is what we call a reference loop. So So here's how this works. Journalist writes an article and gets their information from Wikipedia and doesn't really have another source. Source. Later, someone editing Wikipedia says, hey, there's no source for this statement on this particular paragraph. I'll search around and see if I can find a source. And they find the journalist's article. The appearance of a citation is created when in fact there is none, It's just circular.

0:16:59 Thomas: Yeah. Yeah. Yeah.

0:17:32 Amy: I

0:17:32 Thomas: That's funny. I, you say that in a different thing and I was so glad there was a name for it because that's something I actually noted. Do research for different shows sometimes and episodes where I'm covering stuff. Specifically around, I tend to research debunking the kind of anti-woke panics that people have. And that's a topic where this happens all the time, where somebody you'll cite like, yeah, there was this college student, you yelled over here and that was really bad or whatever it is.

And I want to follow it up. This exact thing happened to me where I was like, all this references is another reference to the same reference. And, um, I didn't know there was a name for it, but I'm so glad now I have a name to put to that. That is something that I wonder, you know, then I start to worry. Like, how much of that is there? And, you many citations are like that? What are the criteria for, you if a citation is valid?

Who does that? How do you ensure people are doing that? And I guess all those questions, plus I wonder, does something like that, or why doesn't it? Because I suppose I already know, you know, where your views are. Why doesn't that compromise Wikipedia in any serious way in your mind?

0:18:34 Amy: Don't know how often it happens. I think that's an empirical question. But people continue improving it. We are always making it better. And that is moving things in the right direction.

0:18:46 Thomas: Yeah. Okay. So So it's just the that, case like, on the topic, the more popular it is, the more important it is. The more people are going to have checked these things. And I suppose.

0:18:55 Amy: That's right. You never want to cite Wikipedia ever. What you want, well, I do once in a blue moon. My book cites a few Wikipedia pages, like, for example, I love the page on hate speech laws in Australia, trying to make the point that the definition of hate speech is different in every jurisdiction, not only in every country, but in Australia, in every province. And there's this great page, all the different hate speech loss by province in Australia, which I cite in my book just to make a point.

0:18:55 Thomas: Yeah. Hmm. Hmm.

0:19:24 Amy: But in general, you shouldn't cite Wikipedia. You should look to see what it cites and then go evaluate the credibility of that source. You

0:19:30 Thomas: Yeah. Yeah, absolutely. That's that's what I've taken to doing. It's a it's a great place to start to find sources. There interesting how often I'll come across things that are, you citation is like, well, this is now Internet Archive because it's not there anymore. You know, there's a lot of those challenges. I mean, in theory, if someone were to be directing kind of this process from the top down, you would you be hoping that people are checking those citations regularly and making sure they actually say what they say and all that kind of stuff? Is there any guidelines in Wikipedia about how to do that or how often someone should do that? Or again, is it just a question of, well, if someone wants to like kind of challenge something or check on it, it's going to happen by that sort of spontaneous group peer review process? Process.

0:20:20 Amy: Know, there are a lot of different mechanisms to maintain quality on Wikipedia. Of them is wiki projects. So a wiki project takes ownership of a particular content area and makes sure that everything is up to their quality standards and makes sure there's a certain degree of consistency. Consistency. For instance, if you have a page about a shark, WikiProjectSharks is going to say, look, the about box on the top right corner of the page should have the following information about the shark in the following order.

Then WikiProjetSharks will go and say, don't have an article about this particular species. We're going to make it our collaboration of the week. Let's all work together to write a really good article, really a shame that our hammerhead page isn't in better shape, let's all work on it this week, so maintain quality standards, and they also encourage improvements in coverage, take ownership over an entire content area, and One of the most well known is wiki projects medicine, and The most well established is wikiprojects military history, and I don't know if this is still the case, but at one moment in time, they had military ranks for the people who were regular members of military history.

0:21:14 Thomas: Well, that's delightful. Hmm.

0:21:34 Thomas: I of questions off of that, with you something like I the top example in my mind is the Amber Heard Johnny Depp trial, for example, with Something like that, where you've got a significant group group of people, you motivated to influence a page, or you may know better examples than that, where you've got, it's not necessarily like vandalism, but and it's Not necessarily like a hack or anything, but you've got a motivated group of people who perhaps have an ideology or a point they want to make. What protects Wikipedia from that kind of thing, mentioned you people somehow lock down pages. Who makes that decision? Mr. Wikipedia or who, you know, like, is is, there a hierarchy of like, who's allowed to make those decisions? How does that work?

0:22:18 Amy: Individual editors making individual decisions, and some people are quite passionate about it and follow particular pages quite carefully.

0:22:26 Thomas: But then how would anything ever be locked down then if anyone can undo it? Like,

0:22:30 Amy: There's a bunch of more advanced Wikipedia procedures, and you put put typically a special tag on a page, which triggers a social process of editors talking about, hey, should we lock this page? And there's a specific procedure that is written in policy, and you follow the procedure in the written policy about discussing whether to lock something, for example. Administrators do have extra authority. There's a huge process for becoming an administrator. And in an emergency, an administrator can take immediate action on something. But typically, nothing happens without discussion by the community. Interesting thing is that the policy pages themselves are editable by anybody.

0:23:18 Thomas: That's so, yeah. It's like, oh, it's real. It's bootstrapping. It's Wikipedia is pulling itself up by its own, you bootstraps somehow, but it works. That's the crazy thing.

0:23:18 Amy: Right?

0:23:30 Amy: I'm very proud to tell you that I made a change to an important Wikipedia policy, proposed a change that was successful. Birthday was on my Wikipedia page, and I have always been taught that you should keep your birthday private to help prevent against identity theft. And I asked them to take my birthday off of Wikipedia. Uh, asked on the policy page for policy of biographies of living people, can the subject of an article request that their birthday be removed? And the editors discussed it and they changed the policy. Uh, kind of fun, it

0:24:02 Thomas: Yeah. What does it do now? Does it just like the year or do they just take that information off? Oh, okay. Cool. Interesting. Yeah. No, that's super cool. And there's another quote I came across in, I don't know, one of your interviews that I loved, which is Wikipedia doesn't work in theory. It only works in practice. And this about that a little bit. I mean, all can imagine to some people, all this might sound like chaos. Like, how is this not just fully, sort you of a relativistic, you Know, there's no truth at all. And it's all if a group of people was so determined, they can influence Wikipedia in negative ways and all that stuff. How, how was this not that, you work in practice and maybe not in theory.

0:24:06 Amy: Just says The year Yep Yep Yeah You Know, and so you can forgive me for having being a recovering uh cyber utopian, uh, having some uh optimistic views about the potential of this technology. You look at the magic of how Wikipedia works, and it's hard not to hope that other things can work like that. Sometimes they do. Sometimes they don't. Think there's something special about Wikipedia. I don't take for granted that it will always stay that way. I Think it's something that people have to volunteer and work hard on and fight for. But it is counterintuitive that it works. I tell my my students that you can look carefully at Wikipedia and find things that are counterintuitive and learn from them. Our first hunch about how things should work is not always how they do work.

0:25:03 Thomas: Yeah. Yeah.

0:25:35 Thomas: And from this too, I think, maturing you intellectually, possibly, think part of that, you is realizing that things are not so black and white, iron clad, you know, I feel like a lot of, in studying philosophy, I feel like, a lot, of the early attempts at some of these questions, do you have this, perhaps, you, you think, therefore, I am, and all that, like, like there's going to be some solid, a hundred percent certain way, that you go from, you the stimuli, and your senses, to like, therefore we have logically proven every single belief we have, and, you know, you talk a little bit about, I think, some of that epistemological history. Don't mind doing some of that here, because I find it really fascinating. I feel like the mature kind of philosophical view is like, no, you're not going to have a 100% proof of every view you have. And knowledge doesn't really work that way.

0:26:28 Amy: Absolutely. Question So for epistemology is, how do you know what to believe. Immediate naive approach is to say you can believe fact A if it's supported by fact B and C underneath it, and then D, E, and F underneath those. And what you quickly find if you take that approach is that you can never get to ground tricks.

0:26:54 Thomas: Yeah.

0:26:54 Amy: There's always more layers down, down, down, and suddenly you're unsure how you believe anything. This is in epistemology, the problem of justification. How do you know what belief is, what you're justified in believing? Approach to justification says you can, justified in believing something if it's the result of a reliable cognitive process. So the example specifically is, let's say there are expert and novice bird watcher who see a bird and both come to the conclusion that it's a rare spotted flycatcher. You are justified in believing the expert. You are not justified in believing the novice. The novice made a lucky guess. The expert really knew that it was a spotted flycatcher. So this comes down to evaluate your sources. And that is how the internet works. Evaluate your sources, the

0:27:54 Thomas: Gosh, yeah. And you know what that makes me think of too is, I just remember that, that reminds me of like childhood arguments, where it's like someone's right, but for the wrong reasons and you're like no that but you weren't really right because you didn't actually know kind of thing and I guess the internet you gives us the ability to have access to you all of that like there's going to be tons of people who got the right answer for the wrong reasons and it even harder there's sometimes going to be like experts who got the wrong answer but for good reasons you know like that that are it's still, you would say that like, it still was a reliable process.

It just reliable doesn't mean works 100% of the time. And I feel like the internet really hits us with those challenges these days. I mean, maybe even around like COVID, for example. And like, wonder, is how do we solve that as we see an expert who's wrong and, you a non-expert who's right and then boasts about how they got it right? How do we navigate that? In this day and age, you know, do you see that as a significant problem?

0:28:57 Amy: Right way to navigate it is through social consensus and finding ways to, if you understand how reliable knowledge is constructed, then you can design the online ecosystem to provide more opportunities for reliability to happen naturally. I

0:29:14 Thomas: Yeah. That article was about AI, you known chatbots and AI kind of utilizing Wikipedia. I'm just curious, is that something that, is that also a part of your area of interest as well? Or is that something like where they kind of just cited you for the, the purpose, but that, that isn't something that you've studied? Cause I, I'm very curious about those questions, but I don't want to pose them to you if it's not something that you've really looked into, like

0:29:38 Amy: Not worried about the impact of chatbots on Wikipedia at all. No. You know the author of that article had a opinion that that was something that's worth worrying about. I respectfully disagree. Maybe I'm naive, but I think people have have to know the difference between the kind of stuff that a chatbot spits out and the more reliable information in a curated information source.

0:29:42 Thomas: Oh, really? Okay.

0:30:05 Thomas: Well, if you don't mind, could you set up some of those arguments for audience? I are some of the reasons that author and other people are really worried about chatbots and AI kind of ruining or being a threat to Wikipedia.

0:30:17 Amy: Think as chatbots' generative AI becomes readily accessible, it's fairly easy to ask the chatbot to questions for you. And the question is, are people going to rely on that too much? I think they will rely on it too much. Hopefully we can teach people not to do that. The thing that makes this another level complicated is the chatbots of today are highly unreliable in a dozen different ways. They will spew absolute nonsense at you.

I don't know if the chatbot of 10 years from now will have made significant advances in reliability. And in that case, we're in a totally different ballgame of where we will be and where we should be. Right now, the reliability is low enough that it doesn't make any sense to really rely on them. Just sometimes make things up Out thin air. So you've got to be really careful. Mean, I think that I don't think this is a threat to Wikipedia. I do think that it's a threat to how we understand what is true, and even more so, it's a threat to jobs. So, for example, there are already marketing departments where instead of having five staff working in marketing, they have one person who is going to use generative AI to write a lot of their copy. Is a for our economy.

Think there are a lot of people who, when robots took over assembly line jobs, said glibly, well, those were boring jobs anyway, and now we can move those people to more interesting jobs. Now that it's people taking away white -collar jobs, are you still saying the same thing? Who's laughing now? Not just, you union member who's been forced out of a high -paying job. It's the marketing department, too.

Our textbook that for, I teach a class called a computer society and professionalism are required undergraduate ethics class. Our textbook for that class says one argument is that while we're getting rid of jobs, we're going to be creating new kinds of jobs. Those jobs will be more interesting. And not a lot of evidence that the jobs are for the same people and that the number of jobs is the same number of of jobs.

Think more about, I've always been a fan of the science fiction TV show, The Expanse, and in the future world imagined in this books and TV show a significant portion of the population is stuck on universal basic income with no job because there's nothing for them to do. Don't think that's an unreasonable thing to worry about. Wow.

0:31:14 Thomas: Yeah.

0:33:09 Thomas: Mostly because I don't think we'll even get that good of a situation because we won't get a universal basic. I'm just kidding. I'm a little pessimistic about the outlook in America. But yeah, that reminds me, man, I'm blanking on it, somebody, but somebody, I don't know if it was a tweet or something, but somebody did point out kind of the irony that, you the belief know, maybe at one time was, all right, the robot's going to do all the manual labor and then that way we can focus on stuff like, you know, writing poetry and reading and painting and doing all that stuff. And right now it feels like somehow the inverse of that, like we are stuck doing a lot of, you delivery, warehouse job, like a lot of that labor that's really tough for not enough money. And meanwhile, now the robots are the ones that are taking over illustrating stuff for us and knowledge work. And that feels just so dystopian to me. Is that something that, I you worry about that? Or do you think that's just kind of a temporary, condition you of how things have developed?

0:33:15 Amy: Well, I think it's a problem. Sorry.

0:34:10 Thomas: That, yeah, that doesn't give me any hope. You've been so techno -optimist. No, I'm just teasing. Seem like there is an optimism throughout your work still? Or am I incorrect in reading that? And if so, where does that come from?

0:34:15 Amy: Oh, Absolutely. This technology has done and will do things that are wonderful beyond our wildest dreams, and things that are horrible, all mixed together. Still believe there are lots of places where technology has tremendous positive potential. Potential well.

0:34:41 Thomas: Interesting that the chatbots are having that issue. I mean, I know, again, AI is not like your specialty or anything, but I would think that with all the knowledge available, like the chat bots have all the access to Wikipedia, right? Or at least they did. And it's weird that they can't take that plus everything else they have and come up up with something even more reliable than Wikipedia. It's kind of weird that, you know, they're just, they seem to be doing these hallucinations. And

0:35:05 Amy: Garbage in, garbage out. I read one report, I don't know if it's true, that the training data used by OpenAI included content posted on Kiwi Farms. Kiwi

0:35:15 Thomas: Don't even know what that is.

0:35:16 Amy: Firms is a hate site, uh, Don't want to know. It's One of the worst. So, I don't know whether the report that said that they trained on Kiwi Farms data is true or not, but, but the point is, they trained on a lot of data that is not as carefully edited as Wikipedia. So I am a very active Redditor. I love Reddit.

0:35:18 Thomas: It is? Oh, Oh, wow.

0:35:36 Amy: I am loyal to Reddit. But you know that when you read Reddit, there are some things that you believe and some things thatyou do not believe.

0:35:36 Thomas: Me too.

0:35:45 Thomas: Yeah.

0:35:46 Amy: And as far as I know, the AI systems just ingested all of it. You know, so I'm a mod on our science. They ingested our science, thank you. But they also ingested r slash conspiracy, right? And that's kind of a problem. I

0:35:52 Thomas: Yeah, yeah. That's interesting, because on one, just thinking philosophically, would you think that, but again, because we've talked about how this knowledge works and how this reliability works and it's counterintuitive and it is chaotic and it it is messy, but it works and it just feels, I don't know, something philosophically doesn't feel right there. You would think that a chatbot having access to all that would be able to kind of reproduce that messy reliability. I wonder what the missing ingredient is, if you have any thoughts on that.

0:36:36 Amy: Think what you're talking about is one of the most important research problems in AI today. And I know there are people working on it. So hopefully they'll make progress. Us.

0:36:46 Thomas: Yeah, Gotcha. Yeah. Fair enough. Yeah. You're not going to solve that one here today for everybody. That's a tough one. Yeah, I I hadn't heard about the Kiwi Farms thing, but boy, that sure would explain some of the stories that we've heard about these these AI chatbots where they they seem to go from one to Nazi pretty quick, you know, like out of nowhere pretty quickly. And it's like, that seems suspicious.

0:37:09 Thomas: No kidding. Yeah. I sort of coming to a close, would what you like people to get out of this book? And what are you hoping, you you had your magic wand in the work you're doing, what are You hoping? How are you hoping to influence the conversation and maybe even influence, you society at large around these issues.

0:37:08 Amy: Garbage in, garbage out, right?

0:37:28 Amy: Would like us to teach kids in school a little bit of theory of knowledge. And I think if we understand what truth is and what knowledge is, that we can move forwards as a society in new ways.

0:37:43 Thomas: To that. I always love philosophy being taught anywhere. I'm a fan of that. So I'm an easy audience for that pitch. But well, Dr. Amy Bruckman, thank you so much for coming on. I highly recommend the book. I'll throw it out, of course, one more time, Should You Believe Wikipedia. Fascinating. Thank you so Much for being here and giving us some of your your time.

0:38:03 Amy: A pleasure. Thanks for having me.

0:38:04 Thomas: Is there anything else you'd like to throw out as a plug for people to visit or go check out or anything while you're here?

0:38:09 Amy: That's good. That's great. And glad to be able to spend some time with you.

0:38:14 Thomas: Fantastic. Thank you so much.

0:38:16 Amy: A lot.