The 404 Media Podcast (Premium Feed)
from 404 Media
The Rise of AI Book Ripoffs
Episode Notes
/Transcript
YouTube version: https://youtu.be/1KfqUn0Cnj0
Hello, and welcome to the four zero four Media podcast, where we bring you unparalleled access to hidden worlds, both online and IRL. Four zero four Media is a journalist founded company and needs your support. To subscribe, go to 404media.co, as well as bonus content every single week. Subscribers also get access to additional episodes where we respond to their best comments. Gain access to that content at 404media.co.
Joseph:I'm your host, Joseph. And with me are four zero four Media cofounders, Sam Cole
Sam:Hello.
Joseph:Emmanuel Mayberg Hello. And Jason Kebler.
Jason:Hello.
Joseph:Jason and Sam, we have an event coming up that both of you are gonna be at. What's the rundown? What's happening?
Jason:Yeah. We're throwing a party at South by Southwest on Monday, March 10. We're going to have a kind of, like, live panel podcast thingamabob, about AI Slop, featuring me and Sam and, our friend Brian Merchant who runs Blood and the Machine, And then we're going to have a party after that. So to register, you can go to our website, and there will be a post by the time you read this because
Joseph:We don't have one yet.
Jason:I know, but the URL for it is really long and difficult. It's,
Sam:Put in the show notes. Show notes.
Joseph:Show notes. That's well. There too.
Jason:It's l.lu.ma/hffhi9Dot.sn.
Joseph:Yeah. Click it in the show notes and all the posts, which we will now write.
Emanuel:I love a a Live Air URL read. That's really good, Paul. Yeah. It is.
Jason:People like it.
Emanuel:Speaking of AI slop, let's get to our first story. This one is by Joe. The headline for the story is a slop publisher sold a ripoff of my book on Amazon. Joe, in case you don't know, though, I I can't believe you don't know because he's done a very good job of promoting his book, but he wrote a very successful book. And you, I think, probably were looking up the book on Amazon and other stores.
Emanuel:And what did you find while you were doing that? Yeah.
Jason:Are you buying your own book again?
Joseph:Yeah. I actually buy tens of thousands of copies, like, Trump junior style. Didn't he do that to get it to, like, the bestseller list? They just
Jason:I believe so.
Joseph:Yeah. I have not done that, to be clear. No. I haven't bought any copies of my own even though people do that to game the system. But I was, as I sometimes do embarrassingly, just checking it on Amazon, you know, to see what the ratings are and that sort of thing, it gives some indication of, I don't know, how much people like it or hate it.
Joseph:And as I was doing that, I came across this other listing on Amazon. To be clear, you don't need to buy the book on Amazon. You can do it anyway. It's just they have their metrics and their star rating, whatever. But, I I'm I'm looking for and I see this listing for a summary of Joseph Cox's Dark Wire, which is the name of my book.
Joseph:And I'm like, that's pretty strange. I wasn't expecting that. And then I go through and, immediately, I'm interested because of the crappy quality of this listing. Like, there's a really generic orange, cover. The description seems pretty generic as well.
Joseph:And just going through the other summary books uploaded by this publisher called Slingshot Books, it just seemed like a very low effort, churning out, I would say, of these book summaries. So I decided to spend $4.99 of our subscribers' money on this, piece of shit, basically.
Emanuel:So you buy it. You read it with great interest. What do you find? What what is this should I I I I have read your book, but, I mean, would I be better off just reading this instead? It sounds like possibly.
Joseph:Yeah. You just read the summary. It's fine. It gets the basic gist. I mean, on one hand, it kinda does summarize the book.
Joseph:But what literally, what it's doing is it is breaking down every single chapter, and I think there's, like, 30 or something chapters in the book. And it's just rather than them being, I don't know, twenty, twenty five pages where I can remember, all of this narrative detail, it's, like, three or four pages of just the normal plot points. But it it well, the example I give in the in the article that I wrote was that my opening of the book is about this structure I've called Owen Hansen. And there's sort of all this detail about him, and I spoke to him, and, you know, I've followed his escapades and his case for years. And you go to the summary book, and it's just two pages of Hanson did this, Hanson did that, Hanson did this.
Joseph:And it's, I guess, the rise and fall of him. He does fall. You know, he's arrested by the FBI. But it just misses out all of the rich, detail, essentially. So to answer your question, it was just like a really bland overview.
Joseph:Now maybe that's the point of a summary book, you know, that we have cliff notes. I think we'll talk about that maybe a little bit later on in this segment as well. Like, I'm not against summary books, and I've used them when I've been studying as well. It's different when you see your work and your intellectual property, basically summarized on, like, a digital piece of toilet paper, basically. You know?
Joseph:And then I I took it and I and I showed a screenshot to you all, I think, in our group chat. Emmanuel, what did you think when I showed you the screenshot? Because I don't know whether it's AI or not. You know? Like, that's one thing we can't really determine.
Joseph:But I showed you the screenshot, I think. What did you make of that?
Emanuel:Yeah. So I've been looking and reading a lot of these, over the past couple of weeks for reasons that we'll get into soon, but, also, I believe we talked about on the pod a couple weeks ago, I wrote about AI generated books, making it into libraries, and all of these books are the summary books, so I've read them. And there's just a certain style to them, which is like you described, which is very plain and then correct, well written English, but with no stylistic flare or feeling for place and narrative and all of that. It just feels, kind of soulless. So it definitely had that style, and immediately seemed AI generated to me.
Emanuel:And while we don't have a smoking gun here, there's a lot of contextual clues that we can get into later about why, we're pretty confident that it's AI generated.
Joseph:Yeah. And then I also showed it to Jason and sorry, Jason, to put you on the spot because I didn't actually include this in the article. But I just remember you responded to it by saying, like, there's no way this is an AI generated due to, like, the sheer number of book summaries that this publisher was churning out. What do you what do you mean?
Jason:Yeah. I just mean, like, if you think about it for one second, they're publishing dozens of summary summary books, and it's some publisher you've never heard of before. There's not a human being reading and analyzing these in a formulaic way in the year 2025. It's just like Occam's razor. It's so easy to get an AI to summarize, and it's so easy to get Amazon to publish something that there's just, like, no business model where this makes any sense whatsoever for a human being to be doing this work.
Jason:And by that, I just mean, like, it's the type of AI slop that we've seen over and over and over again. And if you were to enter this market right now as a human being saying, like, oh, I'm just gonna read Joseph's book, and then I'm gonna read some completely unrelated book. And I'm gonna do this multiple times per day, publishing multiple, like, formulaic summaries that are bad. I don't know why you wouldn't have an AI do it.
Joseph:Yeah. It's, like, just way more likely to be AI, especially this year. And there's, there's actually one comment left on the article, from a subscriber, Spencer Baumann, which I'll just read out now. Quote, there's no way it isn't AI. Something about the way the chapter title is written screams chat g p t to me.
Joseph:Similar things of articles that are individual paragraphs slash bullet points with a bolded headline at the, start of each one.
Emanuel:The bullet point is a very good clue, and I just wanna, like, run through, like, if you're listening to this and you see a book on Amazon and you're not sure, like, the contextual clues that we usually use are it's the cover. Right? You'll either see that it's a clearly AI generated cover. If it's a book about someone, it's really easy to tell because they don't really look like themselves. But, also, you can see just, like, artifacts in the images.
Emanuel:It's the content of the book, which has the same style, that the commenter just mentioned. And then I think probably the best one, as Jason mentioned, is if you click on the author or you click on the author, and you see, a, do they have any online footprint? A lot of these AI publishers do not. They just exist on these storefronts. And then the other one is just the amount of books that they've published.
Emanuel:So, usually, you'll click on one of these publishers, and they've published 300 books in the last year. And, usually, it's the same author, and that is just humanly impossible for someone to do. And I think this was all true about about your book summary.
Joseph:Yeah. I'll just read a tiny bit of Amazon's statement because I reached out to Amazon for comments and they removed the that, summary of my book, but then also dozens of others. When I went to go write the article, there was only one summary left, and I don't know whether that was just created after the ban hammer or somehow it dodged it. I'm not entirely sure. But Amazon said, we limit the publication of summary books that are about other titles in our store, and the title you brought to our attention is no longer available for purchase.
Joseph:We have content guidelines governing which books can be listed for sale, and we have proactive and reactive methods that help us detect content that violates our guidelines, whether AI generated or not. Interesting in that the statement is it is more on we don't want summary books rather than we don't want AI books, which, I think is interesting because, Sam, you previously wrote about AI generated foraging books, I think specifically on Amazon. What happened there? Because we're I mean, this story is just about my book, and it was funny because it happened to me. But, like, the that story you wrote about could have had real world consequences.
Sam:Yeah. I mean, it's just a similar thing, but with a different topic. I've heard about this also with, knitting and crocheting tutorial books. But the mushroom one was interesting because that's actually something that, like, if you take the wrong information from, you could die from. It's like they were identifying these books were identifying mushrooms, that were potentially poisonous as safe to eat, like, good to cook with.
Sam:They're very slight, like, subtle differences in, you know, like, a death cap mushroom versus one that you can, like, actually eat, put on pizza. So, you know, if you don't know what you're looking for, you just say, oh, that's a wild mushroom. But, obviously, the consequences are more serious than that. So, yeah, it's, you know, it's people preying on beginners, people preying on people who don't want to spend more time, you know, researching something deeply. They just want the quick answer.
Sam:It's just I don't know. It sucks. It's sad to see it happen. It's sad. It was sad with the crocheters too because that's such a, like, a handmade, obviously, craft that real humans have to be involved in.
Sam:So, you know, it's, it's kind of eroding all these, like, knowledge systems, that we, I guess, we've taken for granted. But I never would have imagined that they would be under threat like this, you know, five, ten years ago.
Jason:I think when it's a summary of any other book at this point, you can pretty safely assume that it's AI generated unless it's from, like, SparkNotes. And even then, I bet that SparkNotes is is this is I have no no information to base this on, but I bet that they're utilizing AI in some way, shape, or form at this point. And it's just like I'm I'm not even sure what the utility of a book summary is at this point. Like, I believe you could probably ask Chad GPT at this point, give me a summary of Joseph's book, and it probably would give you something. Like, I don't I don't really even know.
Jason:It it's a it's a weird type of content that has, like, suddenly popped up on Amazon and all over the Internet, but I don't think that it's, like, popular or I just I don't know.
Joseph:Yeah. I mean, you're right in that you could just go to Chat GPT and get the summary that way if you have, you know, paid access or free access worth to Chat GPT. That's basically what the books are doing anyway. Right? You're just paying some publisher five bucks to basically enter the prompt for you.
Joseph:I mean, look, I don't know how this publisher or other ones got my book. Obviously, you can just pirate it. Like, I'm not gonna dance around it. That's absolutely possible, obviously. Now did they pirate it or something like that?
Joseph:I don't know. But you could probably even or Chan GPT could probably still build a summary just based on the public media coverage they had as well. So, I mean, there are different ways of going about it. And not to put it all on Amazon, as fun as that is, it is also on Apple Books. And I'm looking at it right now, and the exact same summary is on there.
Joseph:And it hasn't been taken down, not that I'm trying to go on a crusade to remove every summary of my book. It's more funny, than anything. Apple didn't respond, whereas Amazon did.
Emanuel:There's no utility to you, Jason, the reader, but the librarians I talk to are pretty confident that it's a numbers game where if they blast this summary across enough platforms, they're gonna catch some minuscule number of people who are searching for Joe's book and get the summary on accident and make enough money to make it all worthwhile. Like, that's what the scheme is in general from from what I understand. It's just like getting an errant click and and getting the the the money.
Jason:For sure. That that makes as as, like, an AI content slob factory, it makes sense. But as a, like, curated human being summarizing these books, I don't think that it it makes a lot of sense. I do know there is that, startup that, like, gives you audio summaries of books that advertise on a lot of podcasts. I don't remember the name of it, and I never tried it.
Jason:But that, that was popular for a while because it was like, you can read these self help books in seven minutes and, you know, learn what to learn what the four hour work week, says. But I've never tried it, and I I don't know how popular it is. Also, you can just ask chat g p t for a summary of Joseph's book. I've I've just learned.
Joseph:Oh, good. Can can you can you read us a tiny bit?
Jason:It says it's shit. Says it's great. So I asked for a a summary of Dark Wire by by Joseph Cox, and then it just said, like, you know, a three paragraph summary. And then I said, can I have a chapter by chapter summary? And then it said and now it reasons because DeepSeek reasons, so you can, like, see what it was thinking.
Jason:And it first said, am I allowed to do this? And then it said, summaries are permissible, so I'll proceed without including direct content. So I guess it's not quote quoting anything from you, but it says, chapter one, the encrypted underworld focus, introduces the reader to the ecosystem of encrypted phone networks, discusses how certain companies market specialized devices to criminal clientele emphasizing privacy features and anonymity, and then it, like, kind of goes on and on. And it never says
Joseph:That's wrong. That's not what happened that's not what happens in chapter one. That's interesting.
Jason:Yeah. It never produces the word annom, which is kind of interesting.
Joseph:Which is the main the main company name.
Jason:Book is about. Yeah. Right. Yeah. I don't know.
Jason:Was chapter five called The Covert Takeover? Do you remember?
Joseph:No. It's not called that.
Jason:Okay. Well.
Joseph:Dude, I'm I'm on the slot publisher side now. At least they, like, got it right. Yeah. That changed. It's not
Jason:a very good summary.
Sam:The summary of mine is pretty spot on,
Joseph:unfortunately. Go on.
Sam:Yeah. Yeah. I mean, it's just it's they're pulling from a lot of the sources are, online, like, media coverage of the book. So it's pretty accurate even without having access to the book probably. But, yeah, it's pretty it's correct.
Sam:It's very specific. You know? It's like you could get a lot of summary out of the title, which is how sex changes Internet, Internet change sex. It could bullshit pretty easily off of that, but it's very specific about, like talks about, like, post board systems and OnlyFans and
Joseph:And and sources, like, are they mentioned? The people you spoke to?
Sam:This the main source is next big idea club.com, which is, like, a blog, that I gave an excerpt to. And what it's what it's pulling from is, like, the the subheadings from that blog, not necessarily the chapters. But then again, it's like my book is on LibGen free to download that people torrented. So it's like it could it could access the whole thing if it wanted to, probably, but I don't know if it has that capability.
Joseph:I'm sure mine is as well.
Sam:That's how you know you
Jason:made it. It reads, like, you know, like, fifth grader did a book report.
Joseph:Fast chat at GPT, man. Emmanuel, you you touched on some of your library stuff, but to close the loop on that close the loop, the Hoopla? No. Doesn't matter. Briefly remind us what the library story was and then what has happened since?
Joseph:Because we'd so we did talk about it on the podcast, but there's been some developments.
Emanuel:Yeah. So Hoopla is a service that provides public libraries with access to ebooks. So if you wanna get an ebook from your public library, you're probably going through Hoopla or this other service called Overdrive To illustrate how common the summary AI generated summary book issue is, Hoopla responded to the story eventually. They emailed a bunch of librarians and addressed a bunch of the issues that I mentioned in the story. They are kind of working on new policies on how they curate and filter out AI generated books.
Emanuel:They removed a bunch of AI generated books, but I thought it was very telling that Hoopla said that it is just removing all book summaries from its platform as a rule. Like, no more book summaries with the exception of HMH Books, which is the publisher of CliffNotes, and maybe SparkNotes as well, but they haven't mentioned that. But they did call out CliffNotes as being the one publisher that's allowed to to to have summaries, but all all of them are gone. So, for example, I, mentioned this one publisher called IRB Media that had hundreds of summaries up on Hoopla. Those are all gone.
Joseph:So and I mentioned CliffNotes briefly in my one as well because I think it's the one that everybody knows. But is the to speculate slightly, is the reason that they're kind of exempt in CliffNotes is just because well, they're legit. You You know what I mean? And, like, there's there's more there's more to it. There's study guides rather than
Emanuel:subways. Exactly. Yeah. I was gonna say, I have I've never used SparkNotes, so I don't really know what those look like. But I have used CliffNotes in my life, and what I remember is, like, I don't know.
Emanuel:If you if you get the cliff notes on Moby Dick, it's not just like, here's what happens in every chapter. It's like, oh, like, it is commonly interpreted that the whale signifies yada yada yada. And it gives you, like, historical context. You know?
Joseph:It's just like it
Emanuel:tells you something. There's some analysis being done and presented to you about the book, so you know what to tell your teacher when you pretended to read it. You know?
Joseph:Cliff notes, the original cheating with chat GPT. I mean, I don't know. I never used them. That makes sense. Alright.
Joseph:Let's leave that there. When we come back after the break, we're gonna be talking about another AI startup that is dehumanizing factory workers. This time, it's a lot more explicit, I would say. We'll be right back after this. Alright.
Joseph:And we are back. This is one that Sam wrote. Y Combinator supports AI startup dehumanizing factory workers. We'll play the video in a second, for our listeners. But, Sam, maybe just give us an overview of what is this video exactly and and where did it come from?
Sam:Yeah. So this is something that was going around on Twitter last night and this morning that I caught there. And you know something is pretty bad when it's getting panned on Twitter even though it's a y combinator startup about AI. Twitter at this point is very much like the tech bro boosterism world, and even those guys were upset about this. So, you know, it's bad.
Sam:So, yeah, it's a startup called Optifye dot a I, spelled super dumb as these startups are, o p t I f y e. Optifye.
Joseph:That by the way, that is the worst, like, best Fi I've ever seen. Spot Spotify is one thing. This is crazy.
Sam:The Just adding letters,
Joseph:but we
Speaker 5:don't need them.
Joseph:Y e, I, like, could not understand it. My eyes were, like, going all cross eyed when I was trying to read the title.
Sam:It's bad, and I keep calling it the wrong name in my head and forgetting what the actual name is. I'm like, Syncdify. Like, I don't know. It's Productify. I don't know.
Sam:But okay. So it's a startup that, quote, unquote, it's for AI performance monitoring system for factory workers is their proposal here. In a very small nutshell, they are using machine vision to track workers in a factory and see what their productivity levels are, how many, like, per items they're making, what their output is, their efficiency, quotas, things like that. It's like this whole, like, dashboard that they're advertising, like, bosses can see, and then it shows you exactly who on the line is going too slow or isn't being efficient. And that's the idea.
Sam:And the video, in it, the founders of this startup are kind of doing, like, a skit. And one of them's a boss, one of them's a worker. And we'll just we'll play the play the clip. But
Joseph:Yeah. We'll we'll we'll we'll play that now, and then, listeners will have some understanding what we're talking about, and then we'll come back.
Speaker 5:Yeah. Oh, it's Workspace 17. Workspace 17 is the bottleneck, the worst performing workspace here. Hey, Number 17. What's going on, man?
Speaker 5:You're in dread. I've been working all day. You've been working all day? You haven't hit your hourly output even once today. And you're at 11.4% efficiency.
Speaker 5:This is really bad. It's just been a rough day. Rough day? More like a rough month.
Sam:Yeah. So, basically, if as you can tell, this is productizing bosses being assholes to workers is the real kind of innovation here as Emmanuel put it when he was reading and editing the story. It's not that this is like a new technology, that surveillance on the line is something that has never happened before. We see this happen with Amazon. We see it happen, all over the place with different warehouses and also just with, you know, like, computer jobs and remote work.
Sam:But, the shocking thing about this isn't the technology. It's the blatant, normalization of turning around, calling one of your workers by a number, and mocking them about having, you know, a rough day, a rough rough month, and being, an ass to to your workers and, you know, being this cruel overlord while they're working on your production line.
Joseph:Yeah. And we'll bring the y combinator stuff in in a second. But just while we're talking about the video itself, it like, is this video a bit? Or, like, what what do we make of that?
Sam:I don't think it's a bit. I think if it were a bit, it would be more overt. I think good satire is often toes that line, but it's not if it's satire, it's not good satire because it just is not overt enough for people to say this is this is a send up of this type of AI. And I don't think it's a bit because Y Combinator is celebrating it, congratulating them online. Their website says they're backed by Y Combinator.
Sam:Y Combinator is, like, it's a venture capital firm, that, you know, creates you know, makes companies huge. You know, it's, some of the biggest companies in the world have come out of Y Combinator. So I don't think it's fake. I think it's real. I think these guys are just, like, really out of touch, for various reasons that, you know, we don't have to get into, but, like, it's just it's a bad look for everyone involved, basically.
Jason:Well, I think also it's not just a venture capital firm. It's a startup accelerator that picks ideas that it thinks are not represented in the market yet. And so they end up being like, the market doesn't have this yet, so we are going to help you incubate this idea and, like, turn it into a reality. And so I don't know, it would be very odd for Y Combinator to to, like it's just it's not satire. Like, it's it's just not.
Jason:It's it's like a promotional video for their idea.
Joseph:Well, speaking about Y Combinator's support slash backing, they deleted some posts. Right? By they, I mean, Y Combinator. What happened there exactly? Like, what did they delete?
Sam:So Y Combinator had posted on Twitter I mean, I was because of screenshots and also you can see it in, like, just you can find the URL through deleted posts. People were applying to Y Combinator. But, they were saying they were introducing Optify dot AI as building AI performance monitoring for factory workers, And then they link to, a link about their launch, and then they say congrats on the launch, and then they tag the founders on Twitter. And on LinkedIn, they also made a post, which I found the LinkedIn post through just like googling Y Combinator and OptiFi. And it comes up in the cache, which you can still see what the what the LinkedIn post part like, the first sentence of it was or whatever.
Sam:And it was a similar thing. It was like, you know, congratulations on launching, to Optify, and they're gone now. They're deleted now. I think And the idea
Joseph:when they deleted because the because the the promo video, which was in very, very poor taste, that was going viral Yeah. Yesterday. Did did they delete it, like, after that, or do we not really know? They deleted it at some point.
Sam:I I'm pretty sure it was after that because that's what is all over Twitter. Like, the replies to that tweet that's now deleted are people being like, are you fucking kidding me? Like, this is Right. Disgusting. So I think they got a lot of backlash, and they were like, oh, like, instead of, you know, like, hiding replies or closing comments or whatever, they were just like, just nuke the post, and we can pretend we did not see it.
Joseph:Or you can pretend that we never supported this company, which is dehumanizing people by calling them a number. Yeah.
Sam:Yeah. I mean, we'll see. I mean, I like, I reached out to, Y Combinator, obviously, to ask, you know, what's what's going on with this? Why did you delete the post? Are you backing this company still?
Sam:What's going on there? And I haven't heard back yet, but, we'll see. It's also all over the Optifys site that they're backed, you know, quote, unquote, backed by Y Combinator, which is obviously a big, like, stamp of approval for the industry. It's like, oh, that's a big deal. Like Jason said, it's like that's, you know, that's kind of a mark of, like, oh, you're serious.
Sam:Serious people think you're serious. So Yeah. You know, this is worth looking at.
Joseph:It it legitimizes startups and the eyes of other startup founders and investors and all that sort of thing. Emmanuel, while I I think you added to this, correct me if I'm wrong, but while we were talking about it in our group chat, I think you did bring up that there was a video from a few years ago or something where somebody did do a bit about this. Right? Like, again, I'm not saying but we know this one's real. It's bad by y combinator.
Joseph:It's just funny that this is you can't really tell exactly. And then there was a bit earlier. What was the deal with the the earlier one? Was it somebody just mocking this sort of thing?
Emanuel:It was, I don't think it was I think it was it was either a joke or it was, one of our favorite, you know, local dipshit, you know, like local artist makes a statement videos. But it was a video of a coffee shop, and it was using machine vision that looks almost exactly, like this actual startup to kind of track the productivity of the baristas in there. And that went extremely viral because it freaked people out. It seemed so dystopian. And it's one of those videos that kept going viral even though it turned out that it wasn't a real technology from a real company.
Emanuel:It's not something that people are actually using. So it just it's it's funny that I mean, it's funny slash horrible that this thing, that was not real and everybody thought was horrific is now actually a real company. And I don't know. To me, sort of seems inevitable. As Sam said, this type of thing exists in many workplaces, whether you're white collar and people are tracking what you're doing on your computer and your mouse movements and your keystrokes, or you're at an Amazon warehouse and people track how many packages you're handling in an hour and how long your trips to the bathrooms are, etcetera.
Emanuel:It just the idea that you can use machine vision to track whether you're tracking it well or not, the idea that you can track any workers' productivity with just putting a camera in there. I don't know if this company will survive. I don't know if this will become a real thing, but I feel like the idea is just so attractive to bosses that there's no way that we're not gonna see something like this eventually.
Jason:Yeah. I mean, it's already here like both of you have already said. I mean, it it kind of has its origins in this idea called scientific management, which, Edgar Wang Waiso used to write about a lot for us at Motherboard. But it's, like, based on management practices from, like, the eighteen eighties and eighteen nineties, that has largely become, like, algorithmic management on Uber, DoorDash, etcetera. And then, yeah, it's just like, this this it's just a difference of degree, not of kind, really.
Jason:But I think that the video here was so dehumanizing and so fucked up that it went viral. But as you both said, like, a lot of people work under conditions that are like this to some degree where they're being managed by a computer software or their output is being like, their their worth as a worker is entirely based on numbers and whether a computer thinks they're doing a good job or not.
Joseph:Yeah. I I guess just, last question, Sam, to wrap it up. Like, what does this show us? Is it just this one company sort of said what all the other startups are thinking and probably want to say and they just, like, fucking went for it and broke through? Or is it, you know, a mask off moment for labor and AI?
Joseph:I know it's it seems like to me that a lot of companies will want to do this, but they would not all be as stupid to make this video.
Sam:Right. Yeah. I again, these are these are I don't wanna, like, shit on these guys too much, but, like, they're they're computer science students at Duke University. They're in an extreme bubble as far as interacting with normal people, I'm sure. They have little life experience to speak of.
Sam:Their families come from, you know, they own they've said they say in their bios, their families own manufacturing plants. Which yeah. I mean, yeah, it's like and that's a it's a whole complicated conversation, but that we're definitely not gonna get into here. But, like, it's it's something that we see over and over again in, in the AI startup world that this sort of thing, is very much applauded. Like, efficiency numbers, you know, it's like timing your bathroom breaks.
Sam:And even even if you're the boss, people impose that on themselves. Like, they say, you know, I did I slept three hours. I slept on the floor. You know, I I slept at the office. I worked my efficiency rate, you know, my heart rate, whatever.
Sam:It's quantifying of your work and your worth, actually has really damaging effects on workers, whether that's yourself or your employees, and your factory workers, honestly. It's like, this is not gonna go well, no matter what. It doesn't go in a in a good direction. And a few people have pointed out just, like, on social media, and I think I had the same thought. It's like, it would be way more efficient just to talk to that worker and be like instead of being like, hey.
Sam:You know, number 17, what's your deal? What's your problem? And then he says, I'm working I'm working all day, and you just say, well, it looks like you've had a bad a rough month or whatever. It's like just talk to people and say, hey. What's what are the inefficiencies on your section?
Sam:What are the inefficiencies that you're seeing, without threatening them with being fired probably is a really efficient way to cost effective way to get to the bottom of what's going on at your own company. But instead, people wanna distance themselves from, from their own workforce, from the actual labor being performed so that they don't have to think about it. They can just look at a graph instead of a person making sweatpants, you know, below them, how they view it. So yeah. I mean, it's it's just and, yeah, like you said, it's that they're saying it out loud.
Sam:It's like, oh, like, he admit it. It's just very blatant in this case because they are so young and they come from that society where it's just like they've they've seen it play out their whole lives. So they don't have that kind of filter to know, oh, it's not cool to just put this out there. You have to put it behind a few layers of, like, politeness or whatever, pretend it's productivity focused when really you're just like, how hard can I abuse people before they break sort of mindset? I don't know.
Sam:It's a huge it's a huge complicated thing, stemming out of a very, goofy startup, frankly. So we'll see. I don't know. I hope it doesn't have legs, honestly, but like, we've established it's something that's already happening. So Yeah.
Sam:Being out there in the open is is powerful.
Joseph:For sure. Alright. We'll leave that there. If you're listening to the free version of the podcast, it'll now play us out. But if you are a paying for a full media subscriber, we're gonna talk about a, particular AI clip involving Musk, Trump, and Toews.
Joseph:It's the perfect four zero four media story, AI, politics, hacking, Toews. You can subscribe and gain access to that content at 404media.co. We'll be right back after this. Alright. And we are back in the subscribers only section.
Joseph:Sam, you wrote this one as well. AI video of Trump sucking Musk's toes blasted on government office TVs. Like, I didn't know whether I wanted to read out that headline in the first section.
Sam:That headline, Joseph.
Joseph:Oh, yeah. But I don't but
Sam:I don't understand. Let's say blasted.
Joseph:Yeah. Yeah. I mean, that was the impression I got, and people will understand why in a second. But, like, in the free section that we just transitioned from, I didn't know where I wanted to read out the headline because it's like, are people gonna subscribe because they wanna hear that? Or are they gonna be like, I absolutely do not wanna hear that and I'm not going to subscribe?
Joseph:I don't know. It really could have gone either way. But when people walked into the offices of the Department of Housing and Urban Development, a US government department, this week, what did they see on, it seems, multiple TVs, in the DC office. Right?
Sam:Yeah. Yeah. Imagine you're like, it's a Monday morning. You got your coffee coming out of the break room or whatever, going to your desk, And then all over every television, every screen in in your office is this AI generated video of Musk getting his feet licked and sucked by Donald Trump and with an overlaid word, like, sentence, big capital words that says, long live the real king, and it's just on every TV. And no one can figure out how to turn it off.
Sam:Just a beautiful way to wake up, in my opinion. I love love that image so much. Yeah.
Jason:Should we should we discuss the Venn diagram at play here of our of our interests?
Joseph:Go ahead.
Jason:Okay. So you have, like, hacktivism, hacking, like, more or less. I believe I believe we can consider this hacking. So that's your interest, Joseph. Has AI swap.
Jason:I would say that's my interest or maybe Emmanuel's interest. And then it has Elon Musk, which I would say I've been writing a lot about lately, so I'll put that my interest. And then it has feet. Sam, you get feet. Or Emmanuel gets feet.
Jason:You guys can fight fight over it. Okay. Okay. But you put that you put it in the middle, and it's a story, and it's all It's for Fort Meade. All of us.
Jason:Yeah.
Joseph:Someone someone made a Twitter bot years ago that would generate headlines for motherboard. Right? The Uh-huh. Tech vertical advice we used to work at. And it would just, I presume, scrape real headlines from motherboard and then gobble them up and then output a a new one.
Joseph:This is this sounds like the result if somebody did that to four zero four media, basically.
Sam:Yeah. If someone knows how to make that, like, blue sky bot, please
Joseph:Yeah. Please. Yeah. If anybody's listening, make a blue sky bot where you scrape our headlines and then you make more noise. Headlines.
Joseph:As long as they're all nice, and, otherwise, please don't.
Sam:Anyway We only write nice stories.
Jason:It's definitely like r slash not the onion, but, like, r slash not four zero four media, but but is. It is.
Joseph:I will say, the video is graphic, and I was not expecting that.
Sam:And Emmanuel's like, yeah. It's not.
Joseph:No. No. Graphic for me. Okay. And it's,
Sam:Some people did say content warning. Sure. Content warning.
Joseph:And Trump is, being, he's sucking those toes enthusiastically. I'll I'll say that. As far as a I know it's a stupid question. But as far as AI videos go, is it any good?
Emanuel:Yeah. It's not bad. I mean, as far as AI videos go, it's like the likenesses are there and, like, the movement is there. It's pretty good. Correct number of toes on the feet, I believe.
Joseph:I did see some people claiming, and maybe I just need to watch the video again now, but maybe there were two left feet. Did did anyone spot that?
Sam:Oh, yeah. I'm looking at it now. Yeah. I guess so. Interesting.
Joseph:We need to improve Does he these videos.
Sam:I don't yeah. It's two left feet. Or maybe he's and now I'm, like, crossing on crossing my feet trying to figure out how feet look. Yeah. I think that's right.
Joseph:Okay. So it's so it's quite good, but it's not perfect. Where how did we learn about this? Because, you know, a few journalists post about this on social media. Right?
Sam:Yeah. So, I mean, like, we say you know, I was like, imagine getting your coffee and seeing this, and it's like, that's literally what happened to us on Monday morning. It's like, I'm getting I'm drinking my coffee, and then every like, all over my blue sky and Twitter and urban relevant workers were leaking this video from inside their offices. So, yeah, it was kind of it was people started leaking it to journalists, from inside the building. Marissa Kavas posted a long version of it.
Sam:Washington Post got a recording of it. A few other people got got recordings. It's, you know, it's a lot of federal workers right now are leaking stuff to the press, because they wanna get the word out about what's happening in there internally. And, obviously, the point of this is, as we've already referenced in this episode, it's a it's a local dipshit moment for sure. It's like protest art.
Sam:But that's the point of it is, like, to get outside of the walls of, of HUD and and show it to everybody. You know, it's not just meant for those workers. So and it's also it's reacting to something that Trump said, on on social media.
Joseph:Yeah. That's that's the King reference. What's that referring to?
Sam:Yeah. So k, the King thing is multilayered, I guess. But Trump on, Truth Social Truth Social, which is still going. It's his the social media platform that he owns. He tweeted or posted, truth, congestion pricing is dead.
Sam:Manhattan, all of New York is saved. Long live the king, referring to himself, like, he saved Manhattan from congestion pricing, which, by the way, is is, is the greatest thing that's ever passed in New York in my time living in New York. So
Joseph:Long live the king about yourself.
Sam:Yeah. He's done a
Joseph:he's done a lot of bad shit. That might be the lamest thing he's ever said.
Sam:It's lame as fuck. And it's also it's and it's been the the topic of conversation for a while on on on Blue Sky, especially, but also on Twitter where people are people are arguing about whether or not to take the king thing at face value or to say, oh, that's a distraction. It's like, well, he's calling himself a fucking king. Like Just
Joseph:take it, man.
Sam:I think I think
Joseph:Take it for all it is. You know?
Sam:I think we're allowed to say that's not cool. Like, you know, imagine Obama being like, I'm a king. Long live the king. It's like people would go ballistic. So that's what that's reacting to.
Sam:And it's also it's implying that the real king, obviously, here is Elon Musk because he's getting, the foot service from Trump. So it's and we, you know, we know this because it's obviously, Musk is coming into these federal offices such as HUD and, proposing or actually cutting federal workers, at HUD, which is something that the AP got a hold of, with documents from internally from there. The administration is proposing cutting half of federal workers at HUD, and that came out over the weekend. So that was right before this happened in the office. And it's they're gonna they're planning on targeting things like disaster recovery, rental subsidies, discrimination investigations, first time homebuyers.
Sam:It's like things that are very popular with people, things that we need. And, obviously, I'm sure people internally are very scared, very pissed off, and feel very powerless. They feel like they can't really push back against what's coming for them, in a meaningful way. So it's a reaction to that, and it's a reaction to, their the treatment that they've had to endure under Doge and Musk. So, yeah, it's, you know, it's making a mockery of the whole thing.
Joseph:Yeah. I mean, that that is, the most likely, reason for this. Right? There's somebody inside the department who went and did this both just based on, you know, having access to those TVs. They may not be exposed to the open Internet.
Joseph:But, obviously, all that context you just gave about firing and and cutting down the workforce, and it's like, well, screw you. I'm gonna put this on the TV. It is one of the better, acts of defiance we've seen in in a while. It reminds me and this happens, like, every two, three years or something, but somebody will hack the, you know, a roadside sign or something that can display a message. And often, you can find these or expose on Shodan, which is sort of an Internet of Things search engine.
Joseph:You just look for particular ports or whatever, and you can usually access or sometimes access a control panel. So it it it pretty sure it lands in that context. Jason, you've been following, as you say, the Musk stuff probably more than all of us and sort of the OPM stuff, and this email went out, and you you did some stuff about the the treasury as well. I'm just wondering whether you've seen any other acts of defiance, and I'm kind of thinking about some of the replies to the OPM email. Have you seen many of those?
Jason:Or Yeah. So before we move on to that, there there's it's kind of interesting that like, we don't know exactly how this was made, but there's a nonzero chance that it was made with Grok, which is x x's, like, you know, AI thing all about, which I don't think has a very good video yet. But, like, Grock has been in the news the last week or so because I believe there's, like, a new model or a new new launch that came out that is allowing you to make celebrities, like, with no guardrails whatsoever. And so there's, like, a nonzero chance that Grock was used here in some way to to make this, and then, you know, the video was made in some other way. So just to to go down the technical aspects of how how it was made, we've seen pretty big, like, leap forward in AI video generation from prompts in general.
Jason:Like, Emmanuel's been doing a lot of reporting on that. There'll be some articles about that soon. I've been sort of following this as well. And just anecdotally, I'm seeing a lot more AI generated video in the real world. Like, I'm just kinda, like, walking around outside, and I'm starting to see AI slop videos.
Jason:I was in an airport in Mexico the other day because I would just went to Mexico, and, the, like, national guard there is using AI slop to, like, scare travelers into not doing drugs in Mexico, which is pretty wild. So I think that we are going to see more of this stuff deployed in the real world, whether it is, like, a act of protest or whether it's just, like, advertisements, general slop. Like, I I think that we've been somewhat, lucky so far that a lot of this has just happened on social media and the Internet, and I think that a bunch of, like, stupid companies are going to start showing it on billboards and, like, walking around. You're just gonna, like, see it. So be on the lookout.
Jason:If you see any, like, AI stop in the real world, please send it to us. In terms of, like, other types of acts of disobedience, yeah, there there was, like, a lot of spam and just, like, mean things being said, in the response to the what are the five things you did this week email address because
Joseph:Can you explain what that email is just for those who don't know?
Jason:Yeah. Like, Elon Musk via OPM, which is the Office of Personnel Management, sent an email to all federal workers on Saturday saying, what are five things you did last week? Respond in bullet point form by the end of day Monday. And then Elon Musk tweeted, I'm gonna fire you if you don't respond. And yeah.
Jason:Cool. Just normal stuff. And the email address that they were being asked to respond to was not, like, rejecting emails from people outside the government, so anyone could email it. And, therefore, just people started spamming it. You know, we I did a story sort of about not the spamming, but just that entire saga and how, like, messy it was, earlier this week, which you can go check out.
Jason:But there's been some responses there. We've had a couple very public resignations, you know, like the attorney the district attorney in New York resigned, and then someone else resigned, as, like, related to the whole Eric Adams saga, which I don't I don't know enough to get into at the moment. But, basically, you know, Eric Adams is trying to get out of, being prosecuted by throwing immigrants under the bus, more or less, in New York City. There there was a public Slack like, not a public Slack, but a resignation in Slack that went to an entire agency that we wrote about last week. And so there's been some, like, like, public resignation type things.
Jason:I would say as far as just, like, things that will piss off Elon Musk and Donald Trump, I I think the AI generated toe sucking video takes the cake at the moment. So, you know, the bar is raised. If, you know, now it needs to to be raised again. We'll see what happens. But
Joseph:There was also, you know, the classic obey thing where you put the glasses on and you see the messages. There was a couple of posters of Trump and Musk where it's like, obey, you know, put the billionaires first or whatever. Don't know who did that. I didn't I didn't look into it. But, yeah, people are mad.
Jason:Yeah. I mean, there's been a lot of protests, like, from nongovernment workers, like, town hall at town halls all over the country. There's been, like, tons of protests and really angry town halls and and things that, you know, showing people are are not super happy about this. There's been more in person protests from regular constituents, and there's also been a bunch of protests at Tesla dealerships over the last few weeks. So, I mean, in general, I would say, like, protest is back in terms of, like, from within the government itself.
Jason:I think that I think the ones that we covered are are the major ones, although there may be a few others I'm not aware of.
Joseph:Yeah. And then just the last thing I'll add is that this isn't really from in government, but some federal employees who have been laid off or fired are now, you know, launching various websites to be like, hey, here's our story, that sort of thing, which isn't exactly the same, but the you know, in the same sort of vein.
Jason:Yeah. Some of the you the workers' unions from within the government are starting to do things. There there was a federal government government worker protest as well, like an anti Doge protest in DC last week. You know? So the people doing that are are pretty brave because, you know, they're trying to purge the government at the moment.
Jason:So, if you hear more things, just please let us know. We're gonna continue covering it.
Joseph:Yeah. For sure. Reach out to us on a signal. Alright. We will leave that there, and I'm gonna play us out.
Joseph:As a reminder, four zero four Media is journalist founded and supported by subscribers. If you do wish to subscribe to four zero four Media and directly support our work, please go to 404Media.co. You'll get unlimited access to our articles and an ad free version of this podcast. You also get to listen to the subscribers only section where we talk about a bonus story each week. This podcast is made in partnership with Kaleidoscope.
Joseph:Another way to support us is by leaving a five star rating and review for the podcast. That stuff really helps us out. Or just tell your friends about it too. This has been four zero four Media. We will see you again next week.