Lawfare Daily: Grokipedia’s Deafening Silence with Renée DiResta

Hi, this is Mark Bitman from the podcast, Food with Mark Bitman. Have you ever considered how the smallest details can create real emotion? Lexus does. At Lexus, master craftsmen called Takumi are involved not just in physical craftsmanship, but in sensory details. Things like how a door closes, the click of a switch, the acoustic feedback of controls. Because a car that doesn't make you feel something is a car that stops short of amazing. Experience amazing at your Lexus dealer. Losing weight can feel like a roller coaster, full of ups, downs, and being thrown for loops, like when someone brings donuts into the office or your kids' dino nuggets start giving you that look again. That's why there's Noom. They combine GLP ones to quiet the food noise with healthy habits that help you keep it quiet, even if you get off the meds. Their program is designed to make weight loss a smooth ride, helping you understand what drives you, so you can build better habits to not just lose weight, but keep it off. Noom. Lose for Keeps. Get started at Noom.com. Not all customers will clinically qualify for medications. Individual results may vary.

can edit and GROC essentially creates and edits Grogopedia, right? And so what that means is it is interestingly... Like it would be the same for people to kind of understand. It would be the same as if Chattapedia, Chattiptiaedia, like existed. And Chad GPT would be the model that just kind of self-created it and edited it. Yes. So it was launched with about... around 850,000 pages or so. I paid attention because I was one of the first. I got a bio. And I started paying attention to it. I wrote about it for the Atlantic back at the back last year because I was very curious about it. I actually don't think it's a terrible product, right? Like my bio on it was very interesting for me because it was like, Two-thirds fantastic and one-third insane. And I was like, and this gets to the second reason why Elon Musk did it, which was basically, again, the belief that sources really matter and that Wikipedia, which has a very kind of community-designed, curated source list that they consider to be reputable sources. It's very transparent. Anyone can go look at it. But, you know, they wouldn't consider InfoWars to be a reputable source, whereas Grogapedia will and does actually cite InfoWars in some places. So this led to a lot of people. looking at what does this AI think is a good source? How does it create bios or create articles? How does it frame things based on the sort of Elonverse of reputable sources, his particular version of that term? And what that means is there is stuff in there where, you know, GROC considers it to be reality, even though most other AIs would tell you at a minimum that it is like highly disputed. Some of the pages about vaccines are a little bit wild, you know, that kind of thing. that was the, that was the big difference. That was the reason why he did it. And I followed it as more and more pages rolled out because what you started to see was that other AIs were picking it up, right? So chat GPT would occasionally return Grogoppedia as a source. Google would occasionally have a Grogoppedia link in its sources. And this is very interesting, right, because it's an AI deciding that another AI, which is, you know, kind of creating the entire site itself. is a reputable source. And so that is very interesting because then you get, again, if you go one level down, you're getting to the InfoWars aspects of Grogapedia, which are then all of a sudden being returned by chat GPT. So this leads to a lot of interesting questions around what do we consider to be reputable sources, good information, how should an AI generate encyclopedia be treated? But also, like I said, it wasn't terrible. I mean, there were a lot of entries that were quite solid. the two-thirds of my bio that was good was way deeper than anything that ever would show up on Wikipedia. And it's an interesting question, right? And what should we think about AI as a creator of reference information? Yeah, totally. And so tell me a little bit kind of about some of what, how you actually did kind of the methodology of some of this, because I actually find this super nerdy and fascinating. But it's like, I want to like kind of point out you're the first person to like go and see that rockopedia has kind of been abandoned as a project. First I want to kind of hear you say, why was this abandoned? Do you have any hypothesis? And then we'll get into kind of how you figured out all the ways that it was abandoned. So no, I don't. I actually don't. You know, Elon Musk is.

you know, kind of notorious for starting things and then changing them or somebody actually pointed out, just I do want to clarify, when I started noticing this myself, I did go searching the internet to see if other people noticed this, right? And there are one or two people who are, and I think I mentioned this in the article, I think we put this in there. We did an analysis on like who was editing it. And there are some like real super users in there who are very, very frequent. creators of Grogapedia edits. Because even though the encyclopedia primarily writes and edits itself, humans can go. And there's a little drop down up at the kind of top right where you can say suggest an edit or suggest a change. And then you can actually submit a change request and the AI decides if it is going to incorporate your change request. I think I wrote this in the Atlantic piece because, of course, I tried to edit some of the crazier stuff in mine. Right. I was like, I didn't censor 22 million tweets. Like, that's bullshit. Let's get that changed. Right. So there was some of the. Some of the other people, though, who were in there as just diehard contributors across a range of topics, much like you see on Wikipedia, people who are just very passionate editors about a topic, were in there trying to make changes. And some of them started actually saying this on Twitter when they noticed that their edits were frozen in queue for a couple of weeks. And it was very pronounced for them and also for me because it used to accept or reject changes within a couple minutes, right? You knew maybe within... three to five minutes if your edit was amenable to the sort of AI overlord. So these other people were maybe two or three of them were saying on X and then a couple on the Grapapeedia subreddit were also talking about this, but nobody had really done any kind of systemic audit. And I think that was where Ronald and I, you know, I started it and then I reached out to him and I was like, okay, I want someone else to look at this with me, make sure I'm not wrong, you know? Totally. So you didn't have kind of a leaderboard to work with. Grogapedia kind of shows per page views counts. But there's no like kind of site wide ranking. And I thought your methodology around this was super interesting because you ended up reconstruct and basically reconstructing a leaderboard using like the search type like the search auto fill feature. Yeah, the type ahead feature across the 10,000 kind of most common words and page titles. Why go to kind of that length to do this? Like, why was that the methodology that you picked? What do you think that shows us? Why was it the right way to look at this? Yeah, so there's a couple reasons why. First, Ronald gets credit for that. He does amazing work on search quality. I'm sorry he couldn't join us for this. He had a conflict. But he is just amazing when it comes to auditing, auditing search engines and search results and such. So that was why I reached up to him. So I had noticed that controversial pages were not being updated, right? And that was just sort of a qualitative observation. I said, okay, maybe it is stepping back from certain types of pages. Maybe it's just not making certain types of changes. Let me start looking at extremely common things that should go through. And then I submitted a correction request to SpaceX saying, you know, I made an account to do this. Submitted one to SpaceX. And then it...

It just like was in queue there. And this was something where all I said was SpaceX has IPOed, just as basic and neutral and boring effect as possible on a page that Elon Musk ostensibly would care quite a bit about. And it halted in queue. And I looked at the queue because you can see that you can see about 20 of them. You can pull that from the API. And, you know, I use cloud code a lot to do these things at this point. So pulled down, I started pulling down from very, very popular pages. Things, you know. things that would naturally occur to you as likely to be popular. And then I said, let me do this a little bit more rigorously. So the way that this search token methodology works, so we started with the, there's about 5.9 million pages listed on Grogoppedia's site map. And then what Ronald did was he sort of broke those into individual words, which are called search tokens, and then found the 10,000 most frequently occurring title words. And then entered it into essentially the autocomplete, the type ahead search tool. So one of the things that happens with a lot of websites today is very common for search is that it will suggest things that you are likely to want when you type in a word. So if you type in the, you might get the Beatles, Alexander, the great, those are two of the examples that we gave. You can kind of envision how. You know, typing in a famous person's first name will likely autocomplete with the sort of second name. You'll get a small list of them as opposed to just alphabetically every single person whose name might be Alex, for example. So that... From that, we started to pull a key, started to pull the most commonly returned popular pages and then kind of combined and deduplicated them. And that got us a sample of about 300,000 pages. And again, the intent with this was to find the popular ones. Because you can use Grockapedias API to pull the page count, but it's a single shot thing. Like you can say, I want the page count for this entry, but you can't. make that leaderboard, right? So that was what we wanted to find the most commonly viewed, most commonly edited pages. The assumption being that those popular ones would be the place where we would see if edits were happening. You should expect to see edits on like Elon Musk or SpaceX, for example, Barack Obama. And that was how we sort of validated that the sort of initial sampling that I did. And I should also say the initial sampling I did was based on the Tao Center. had made a, um, they did, they wrote an article. They published it in January and they had collected edits because at grocopedia.com slash live used to show you a live feed of all the changes. And that had halted. Um, so I went to the way back machine. I started pulling way back. I scraped basically the, you know, through way back machine API, basically pulled everything I could get, um, and looked at when the, when the, you know, what change logs I was getting there. And then also we wanted to see if maybe it was just the search. kind of cue that had halted, but they were still making changes. So that was where I also did a site comparison for popular pages on their Wikipedia, on their way back machine timestamps. I know that possibly sounds very complicated. I think we explained it a little bit more clearly in the, no, no. I mean, essentially kind of, I guess what I wanted my follow-up question to this was like,

There was, and like as you mentioned in the article and you mentioned just now, the Tao Center had been like kind of chronicling this edit lock, right? There is this nice kind of, I mean, in terms of transparency, this was actually quite useful. But it broke. It just stopped doing that. And so kind of, and instead of just wondering, that left, leave someone with a question, right? Is the edit log broken or is no editing happening? Yes. And so those are two very different things. And you had to go to these like extreme lengths to basically answer that question, right? Yes. And like, so what is that, where does that kind of leave us with like the idea of edit log as an accountability mechanism? I guess I raised this because. I feel like transparency and accountability are things that we, are terms that we throw around all of the time as solutions. But they're, they're really just like, transparency is only as good as like the check of like, of what it can tell you to ask, right? Like that's the only thing that makes transparency super worthwhile. What it tells you to go and get receipts for, what it tells you to go and like, you know, like. ask people about letter and power about um and so i'm kind of curious like do you think that this is very useful um Well, I think the, so there are a couple of reasons why I wanted to do a million different types of checks. And that's like, you know, Elon Musk Sue's people. I didn't want to get it wrong. You don't want to get it wrong when you say something like this. So it's always been, you know, Ronald and I worked together when we were at SIO and just having like as many different types of ways to verify that what we are saying is accurate before putting it out there. So the combination of corroboration from anecdotal comments from frequent, you know, contributors. I also should say I looked at the. the edits that were kind of in Q or that had that had been in the Tao Center snapshot. And then I had Claude just kind of go and pull all of those basically through the identifiers to see what of those, again, had the, and that was how we found actually that things that had been marked accepted in Q were then retroactively changed to reject it. And it looked like there had been a major site rewrite sometime around March. So this actually made me wonder, are they halting because they're going to do another major rewrite? The transparency point, though, they didn't tell anybody about this. I think that's actually the bigger problem. So if you believe that you are... reading accurate up-to-date information, or more importantly, if other machines believe that this is a site with accurate up-to-date information, the combination of those two things is why it actually matters, that they say, hey, we've temporarily halted, the site is temporarily down for a couple months while we rebuilt the model. I actually saw Grok address this on X today, because I did go and search to see if, you know,

what the reception had been, and if anybody from XAI had responded, if not to me, then to use publicly, like, hey, we saw this article, we just want to clarify. I didn't send them a note because they sent me poop emojis back the last three times I emailed them. So I was like, no, we're not going to bother with that. If that's how they want to be, that's how they get to be, but then they can make, I'll make my statement, they'll make their statement. That's how we do this. So with that, it was, I didn't see any official responses from humans at the company, but. GROC said something. There was an agroch comment to the effect of, yes, you know, everything is halted at the moment. No edits had been accepted since April. I actually wondered if it was just ingesting what we had, you know, what we had published, what law fair had published. But it said something to the effect of as the model is, you know, is reworked or something along those lines. So maybe, again, that they're planning to. put out a new version of it. I don't know, but they did, Grock did actually confirm that this was, in fact, the case. Yeah. So I think that like let's put, I want to get back to kind of their comment on this and how it ingested the Lawfare article in like one second. Because I think it absolutely either came from that or the verge coverage of it or something else that happened. A couple of people picked it up. Yeah, it's like a couple of different things. But in the very least, it came from you. Like that was the forcing function that kind of has made an update because it hasn't released anything like this. Right. Right. Like I think that we can. We don't need, like, I mean, we can't prove the null. But, like, I do think that this is, I think that that's pretty strong evidence. But I just, before we get to that, let's put this in context. Like. Similar web has Grocoppedia at 6.7 million visits in June alone. You cite Aerof's analysis basically saying it's been cited about 356,000 times inside other AI systems, mostly CHAPT and Google's AI mode. So it's basically like this frozen in time, not updated. in a quietly inaccurate kind of reference source that is be like being fed into this model and this isn't just i mean we could i could have like we could have a whole conversation about the concept of grocopedia and whether like a self-renfrench model that builds off of like an increasingly a i populated web is ever going to be able to be accurate etc etc But setting that aside, this just hasn't been updated and is putting itself out there as a valid source and it's being ingested by the LLMs. What do you think about that? So I think GROC is actually incredibly important in the information environment today. I joke around about how I write for the LLMs now, but I really do mean it in a lot of ways in that people on X in particular really trust GROC as an authoritative source.

The phenomenon of like at Grok is this true is a very common way that people try to fact check over on X today. These are people who are largely distrustful of what you might call mainstream fact checkers, political fact, AP, that kind of thing. I do a lot of looking at like election narratives or vaccine narratives, public health narratives. And at Grock is this true is a really big deal for people. And it is actually not that bad. I know that this really does surprise people because there are occasionally the times that you'll get the Info Wars response. But it is often actually not bad, which is why I think it's actually important to see it as. a trusted voice, so to speak, in the information environment. I think it is important what it does. I think it is very important to the fact that it is so integrated into a platform like X means that when people want an immediate response, that's what they go to. I published on Lawfare a couple months back a study of some semi- unattributed U.S. government propaganda sites, right? And when I did that work, what I found was that as these... unattributed government propaganda accounts were putting their content out. People were asking grok, is this true, right? In all sorts of languages, too, I was mostly looking at the Latin American stuff, but people were asking GROC in Spanish to essentially check the propaganda of this U.S. government site. So you really do see, I think, why it matters what information we put out there in ways that. LLMs are, you know, understand, right? So sort of curating information in ways that they can very clearly get a snapshot. In return, though, what you're seeing is this encyclopedia product is not doing that sort of rapid frequent updates. And it is, I think the problem here was really just that they didn't declare it. And yet it continued to be treated as reputable by other entities. All right, folks. Ben Witt is here, and I want to tell you the thing that if I could go back in time and change about the way I started lawfare, I would do it. I would not change anything about the way we grew it editorially. I would not change anything about the substance of it. I loved all that stuff. But, you know, when you run a small business, or in my case, a small nonprofit. Journalism outfit, you don't just do, you know, the editorial work. You also are the hiring manager. You're the payroll department. You're the benefits team. And that all sucked. But now we have Gusto, which takes a few of those things off your plate quickly and seamlessly. We didn't have it when I started Lawfare. And you know what? I wish we had. Gusto is online payroll and benefit software built for small businesses. It's all in one, remote-friendly, and incredibly easy to use so you can pay, hire, on board, and support your team from anywhere. I'm talking about...

automatic payroll tax filings, simple direct deposits, health benefits. We're talking commuter benefits, workman's comp 401K, whatever you got. Gusto makes it simple and has options for nearly every budget. And simple to switch to Gusto, just transfer your existing data to get up and running fast. Plus, don't pay a cent until you run your first payroll. That's why Gusto is ranked number one on G2's highest satisfaction products list for 2026. It's trusted by more than half a million small businesses. So try Gusto today. at gusto.com slash lawfare and get three months free when you run your first payroll. That's three months of free payroll at gusto.com slash lawfare. One more time, remember it gusto.com slash lawfare. So folks, Ben Witt is here. When was the last time you got? an email or a text message that kind of almost got you from a scammer or the like. And, you know, it didn't get you. Maybe you figured it out before you clicked through and, you know, typed your Gmail password. But it almost got you. And you thought I really should be doing something to protect myself from stalkers, scammers and hackers. But you didn't really know what. So I'm going to tell you what you're going to do. You're going to go to www.com. Joindeleteme.com slash Lawfare 20. And you're going to enter the code Lawfare 20. And you'll get 20% off Delete Me. I know what you're asking. You're asking what is Delete Me. And I'm going to tell you, delete me removes your personal information that's being sold online from the internet. It has been named the top pick for data removal services by wirecutter. In the age of AI, we're all vulnerable to scammers using our personal data that's floating around on the internet and turning it against you. Google yourself. Find out that your home address is out there, your phone number, the name of a family member. And if you can't get it on the public web, you can buy it. It's unsettling, but here's the good news. Delete me can help. It has never been more affordable. Our listeners can get 20% off with that code I gave you at join deleteme.com slash lawfare 20 with code lawfare 20 and it will do the hard work. to wipe your personal information from the data broker websites. I have an active online presence. I'm not a shrinking violet, but my privacy is really important to me, and that is why I started using Delete Me.

Before they ever started advertising and I started using Delete Me because I went to a conference of pro-democracy people and I said, I really have a problem at this point. And they all said use Delete Me. So take control of your data and keep your private life private by signing up for Delete Me now at a special discount for our listeners. Get 20% off. your delete me plan when you go to join delete me.com slash lawfare 20 and use the promo code lawfare 20 at checkout. That's the only way to get 20% off is to go to www.w.w. Joindeletme.com slash lawfare 20 and enter lawfare 20 at checkout. That's www.com. Joindeletme.com slash lawfare 20 code lawfare 20. A better help ad. After my session and talking to my therapist and really feeling seeing and heard for the first time, I just felt like a weight just was lifted off of my shoulders. And I felt like it was a good match from the first time we talked. I could tell people from lived experience that they should try therapy. I could really be myself again, and I have this continued support with BetterHelp. Wherever you are, that's where BetterHelp begins. Visit BetterHelp.com slash random podcast to get started. Pay testimonials. Results may vary. Thank you for calling the Bombas Comfort Line. Bomba's make socks, slippers, teas, and underwear made with the highest quality materials. Press 1 for comfort. 2 for style. 3 for donation. You chose style. Bombas is styles for whatever you enjoy. You can run in bambas, lounge in bambas, dress them up, dress them down, but always give back in bambas. Because with every item purchased, another is donated. Bombas, comfort worth calling for. Go to bambas.com and use code audio for 20% off your first purchase. That's BOMBAS.com and use code audio. Hi, it's page from Giggly Squad and this episode is sponsored by Experian Boost. Summer glow up, check. Credit glow up, even better. Boost your credit scores instantly by getting credit for bills you're already paying. Your phone, utilities, rent, and insurance. I wish dating kind of worked like that. Connect your bank account, add those on-time payments to your Experian credit file, and your FICO score updates right away. You could instantly raise your FICO score by an average of 14 points with Experian boost. Download the Experian app for free today. Results will vary. Users who received a boost improve their FICO score 8 from Experian by an average of 14 points. See App Store or Experian.com for details. Results will vary. Not all payments are boost eligible. Users who receive a boost improve score 8 from Experian by an average of 14 points. Some may not see improved scores or approval odds. Not all lenders use Experian credit files and not all lenders use scores impacted by Experian Boost. See Experian.com for details.

Yeah, so that's one part of it. But I kind of want to actually also loop us back around to why in the first place we're covering stuff like this at Lawfare, which has kind of a national security bent. It has a rule of law focus. I think that one of the main things that I see something like this as is that these are highly exploitable systems. And so, and they, and they change people's information ecosystems entirely. And they're doing it before our very eyes. And they're getting incredibly sophisticated. And the cat and mouse game is getting harder and harder. And so I'm just kind of curious what you think about this, how this story, how this. particular thing that you're researching, how that kind of doubles down on that thesis? Yeah. So I am very interested in what Grockapedia considers to be a reputable source, what it considers to be a legitimate edit, right? That was sort of what got me paying attention to. I've been paying attention to it since it launched, which I guess is, I think it's maybe right around a year now. Maybe I'm trying to remember if it was August or October. But the... The thing that, because per your point, there's a term that gets tossed around a little bit, but like data poisoning or model poisoning, right, where you are actively intentionally trying to manipulate an AI system by creating a perception of a reputable source with, you know, accurate information. And one thing that we see is that they do, in fact, ingest and regurgitate. material from the open web that they think is accurate, right? And so that question of how much of this stuff makes it in is actually why the sort of source wars really do matter. We see Russia doing this with content farms. We see Iran doing it with content farms. I'm sure China is doing it, though their content farms are usually not as good. So you just have this phenomenon of... In the days of just plain old SEO, it was called a data void. Search engine optimization. Sorry. Yes. No, no, no. Like I'm just like, so like in the days where you kind of would like write an article and fill it full of keywords so that it would be picked up by Google search. So you were writing for pickup by the search engine. Yeah. Now we're in this age of like AI. AI optimization. I guess like basically like AI. Yeah. Or AI optimization. AIO. Sorry. Yeah. I think people call it like. GEO generative engine optimization. I use AIO answer, sorry, AEO, answer engine optimization is the one that I went with, but, you know, it'll become a term at some point. One of them will win. But the point about if you can make a machine think that you have created something accurate, or if you do create something accurate and you get it out there as fast as possible, optimized for them, then that is going to influence how they see the world, what they synthesize, you know, what a model synthesizes as reputable information, and that in turn is what is returned to someone who is searching for it. And I see this all the time because there are, the thing that gets me about the data void problem and a lot of the time is that

There are, and a data void is just a thin search term. Basically, when you look for it, there's not a lot there. Or there's a lot of stale stuff and not a lot of recent stuff. So if a court decision comes down and media doesn't report on it, they actually don't really pick it up for a while because they're not just there scraping PDFs out of databases, right? And so you have this moment where the machine doesn't realize that the world has changed. And so that provides really a prime opportunity to put out a whole pile of press releases and shape public opinion about that event. And that's where I'm like, okay, that really irritates me. And I do think that there are actually information integrity issues with that. And so I'm very curious. From a research standpoint and, you know, just as a person who follows information integrity, you know, for the last, I guess, almost 10 or 12 years now, what is essentially the... playing field that is created and how do people engage with it? Yeah, and so one of the things that I'm curious about that you've kind of mentioned a few times is like the staticness at the page or they're not going into a database and pulling out a PDF. They don't know that the world has changed around it. One of the things that kind of, I know that you did at the Stanford Internet Observatory and other types of things. And I know this is also just how a lot of platforms do things like spam removal or bot removal or cybersecurity kinds of things is actually behavior based, right? Is you're tracking certain types of behavior. So one of the things that I'm interested in, and I wonder if you've like seen any of this or if Gwacoppedia is like exemplary of this. Is how much is did that, do we have any idea if the stativeness of Grogapedia tanked actually? It's, it's like decreased its role in the LLMs. Like maybe. Actually, yeah. Somebody, yeah. Somebody posted about that. There's a guy on X who's, I unfortunately didn't like jot his name down in notes or anything, but I feel like I started with a G. He's been posting about this. He actually posted about it. two days ago, just as we were getting ready to publish, he said, look at the tanking of Grockapedia in search results. That was something that he put out. And the speculation, because I saw him comment on the Lawfare article today, and the speculation was that it had realized that this had essentially stalled, that nothing knew was happening. And there is a, and this again, this is sort of people speaking in hypotheticals because, you know, it's not like. Bing has come out and set it, and it's not like XAI has come out and said anything. But a lot of times you will hear that your performance in search ranking is highly dependent upon your site looking fresh. This is true in social media recommender systems too. They are looking for, you know, if you don't post a lot and then all of a sudden you post once, your post isn't going to get seen by very many people. It's the same phenomenon with if you have a website and you update it very, very sporadically, maybe it's not going to get indexed as quickly. Maybe it's not going to show up. Maybe it's, you know, you're going to be playing a little bit of a different, um,

in a different league as far as your search rankings. So that was the speculation about this was that that was what had happened. And I should say I did also reach out to the way back machine folks. And I was like, hey, you know, I'm about to put this out. If you see something different, like this is what I did. I used your API. I pulled this. I did that. If you see something different, please let me know. But, but they sort of saw the same thing I did. So, you know, that question of just no updates. What does some of this mean for the weaponization of these systems? So, and so I'm going to specifically, you kind of, you've said one thing, which is like maybe keeping your sight fresh, however that is my, I mean, you could just put gobbledygook onto the page and white text and that would be the same fresh, right? Like, you don't need to actually be like, you know, doesn't need to be human eye legible, at least not initially. But also one of the ways you could gamify this is occurring to me is you were kind of talking about how people use GROC is this true is like I wonder how much XAI is training GROC on the material, like the volume of material and the like how often people ask X is this true and whether or not. that type of information that gets lots of questions around that gets a like gets a disproportionate amount of uptake into into grok and so it probably does right and so like what's stopping now to kind of like to play this out what's stopping the Russian, Chinese, like kind of interference from hiring a bunch of bots to spam X and say, Grock, is this true? 30,000 times every time like something comes out. This is something where I'm going to make one transparency comment here, which is just that like we used to have research access to Twitter and we used to be able to actually look at stuff like that. And now if I want to know. you know, who's saying what with at Grock is this true? You know, you're either like trying to scrape from somewhere or trying to like cobble it together or trying to find somebody who has access to the, you know, is paying for the fire hose or whatever. So it's actually very hard to really feel like you have a comprehensive sense of what's happening. And this was because, you know, Elon Musk also switched it to much more of a logged in experience because he didn't, as X became training content for Grock. And. I will say another thing that I use GROC for in terminal quite often is I'll have it do fact checks of recency biased things. So I wrote a survey of a whole bunch of different startups in a particular space. And my co-author and I started that, you know, three months before we finally published it. So it took a while to do the research. And the very, very last thing I did was basically say, like, hey, Grock, can you, can you fact check this? Specifically because I knew that it would actually go pull the numbers. Like, it is just way better at finding the most recent numbers because I think it is ingesting from X and you see companies will drop their latest stats on X. They'll put a press release on X. And so some of the other models don't fact check very, very, very recent stuff quite as well. And it's for like very niche things like startup stats and stuff. But.

But it is, it is an interesting way that he has designed this particular AI to have that in there. Obviously, there's like significant other challenges that come with being very heavily trained on Twitter at this point. But it is, you know, there's different ways that I think. My hope is that there's somebody in there who's thinking in adversarial abuse terms, but you could either be flooding the zone with, you know, a million different bots, talking about a particular topic in a particular way, saying something happened when it didn't. I'm trying to remember, I wrote this on my substack. There was this very dramatic thing that happened with Spencer Pratt during the primary in Los Angeles, Los Angeles mayoral primary. And somebody had posted a screenshot. of a ballot dump, sort of as the AP was updating its numbers. And in this one second of this screenshot, it looked like Spencer Pratt had gotten no new ballots in this drop of, I think it was 24,000, if I'm recalling correctly, ballots. Now, that is statistically very weird. That is also not what happened, right? It updated his number like instantly. But that screenshot went viral. Like Elon Musk posted about it, I mean, all of the big right-wing influencers posted about it. And if you asked at Grock, is this true? it would answer yes. For a period of time, it kept answering yes. Because what it is saying is that screenshot is accurate. That screenshot is real. It's not forged. And that's where you see these like interesting ways in which. It eventually did update, right? As AP and others put out statements and fact checks, then it eventually did understand that no, it was not true. But for a while, the screenshot of GROC saying, yes, this is true, also went viral because it was people saying, look, they cheated and Grock knows it. And that was a very interesting thing to watch happen. And this is where you can see the kind of gaps in the machine. And I think that's an interesting space to be looking at. Yeah, I think that that's exactly right. Like you were saying that it's really valuable for checking things like numbers that a company puts out its press releases or its quarterly kind of earnings or whatever. I mean, yes, but a company could always put that out or something could always, someone can always be putting that kind of information into the system. And that never means that it's necessarily true. It is just true in that it exists in the system. And maybe it's even true. that the company released it. We don't know. A lot of these unverified accounts or whatever else is like it's hard to kind of know, although Grock does kind of answer for that type of thing by having paid for verified accounts. But yeah, I know, which has its own problems. Exactly. And so that's exactly right. And so we're just kind of in this, in this really strange, this really strange epistemological. loop, I think essentially. And it's, there's a delay. And the delay, as you kind of put it, is never is like, you know, the, you know, the lie can get around the world before the truth can put its boots on. And, you know, and like I see that. What you're describing makes that idiom kind of feel real and on a speed run. Like it feels like it's kind of just happening every day in all of these small ways. Yeah.

There's a phrase, I have to credit Google with this one. It was their assertive provenance paper, which was about image verification. But I like the phrase a lot. It was the difference between is this real or is this true, right? And is this real? Like that screenshot is real, right? That's a great, you know, I think this is a really great example of that. Is this real? Yes, that image exists. That is actually the AP page that did go up that happened at that time. Is this true? No, Spencer Pratt did get ballots in that drop. And that's the gap. And the question of how do you handle this? Now, this is where I think community notes comes into play, right? Or that is supposed to be the system that adjudicates reality with more human oversight because people are still voting on the notes. But interestingly, what you see there is it's very slow and increasingly human, there's so much polarization and humans deciding something is true that those notes actually aren't clearing. is why the combination of GROC being faster, and honestly, GROC actually being willing to, in this particular case, once that AP fact check came out, GROC changed its answer. It updated with, you know, it treated AP as a reputable source, even as community notes never got to the point of actually finding enough agreement between diversion publics on X to make that note show up. So it is like, you know, all these different things kind of feed into each other. And the question is what matters in the very short term when people are forming opinions about things or on the flip side, is this entering into training data or search engine data? And that's, I think, the grocopedia problem that halt. I think that that's a great way of thinking about it. And I guess we're going to leave it there. But I do really, really appreciate this article. I hope it gets fed into the ecosystem. As we have evidence of from GROC itself, it already has been. And things are kind of at least being updated on the XAI universe. If they're not fixing Grapapedia, then maybe they will. I don't know whether that's good or bad. And so hopefully we'll have you back on to kind of give us at some point, yeah, an update about whether or not Grogapedia is just going to be this dinosaur that stops existing or whether it will come back on. I really do wonder if there will be a moment in which it becomes politically expedient. for for Elon Musk at some point and thus he like kind of like forces it to go into reboot and puts the energy and money and again to it or not but no I think I think that he will like I actually really do expect it to kind of come back in some capacity but you know I think that AI assistance in the encyclopedia space we had Jimmy Wales on on Lawfare on the pod and that was a really great interview that I did that one with him I hosted that one when this book came out but this question of How can you use it potentially as a tool where there is still more active human involvement so that if there is a halt like this, there are still people who are doing something? It's the difference between the fragility of machines, like the sort of breakage there, versus the innate biases that all humans have. Yeah, totally. Well, Renee, thanks for coming on. Bye, guys.

The Lawfare podcast is produced by the Lawfare Institute. If you want to support the show and listen ad-free, you can become a Lawfare material supporter at lawfaremedia.org slash support. Supporters also get access to special events and other bonus content we don't share anywhere else. If you enjoyed the podcast, please rate and review us wherever you listen. It really does help. And be sure to check out our other shows. Scaling laws, rational security, allies, the aftermath, and escalation. Our latest Lawfare Presents podcast series about the war in Ukraine. You can also find all of our written work at lawfaremedia.org. The podcast is edited by Jan Patya. Our theme song is from Alibi Music. And as always, thanks for listening. This message comes from Jackson. Taxes aren't something you can only think about once a year. With investments, planning for tax days year-round. Fortunately, Jackson offers tax-efficient products. Visit jackson.com for more information on how our products can make your tax bill a little bit less painful. Jackson is short for Jackson Financial Incorporated, Jackson National Life Insurance Company Lansing, Michigan, and Jackson National Life Insurance Company of New York, Purchased New York.