Taylor Lorenz’s Power User - 'Zombie Facebook' takes over + ScarJo vs OpenAI

Episode Date: May 23, 2024

Have you been on Facebook lately? It’s weird and overrun with bizarre, AI-generated, spammy clickbait. Jason Koebler, co-founder of 404 Media and host of The 404 Media Podcast joins Taylor Lorenz to discuss how this happened and who’s behind it all. Later, Taylor weighs in on Scarlett Johansson’s accusation that Open AI mimicked her voice, a new video game about influencers risking their lives for content, Belle Delphine’s bathwater, and a baby who’s hyped to go to the Four Seasons Orlando. Full video of this episode will be available on Taylor's YouTube channel Learn more about your ad choices. Visit podcastchoices.com/adchoices

This week, Scarlett Johansson goes up against Open AI. The family behind the Four Seasons Orlando Baby speaks out. And our main topic, how did Facebook become completely overrun with weird AI-generated clickbait and who's behind it all? I'm Taylor Lorenz, and that's all coming up right now on Power User. I don't know if you've been on Facebook lately, but it has become completely overrun with really bizarre AI-generated spam. Images with things like Jesus pictured as the wings of an airplane. disabled children who appear to have been washed ashore with signs begging for help. Jason Kebler, co-founder of 404 Media and host of the 404 Media podcast, has been doing a ton of great reporting on all of this.

And he's here to tell us exactly what's going on with Facebook, what it means when one of the defining social networks of our time is overtaken by AI spam and who's behind it all. Jason Kebler is a co-founder of 404 Media and the host of the 404 Media podcast. Jason, welcome to power user. Hey, thanks for having me. So you've been writing so much about Facebook's descent into kind of AI madness. Paint a picture of what is happening right now in terms of what it's like to be on Facebook these days. Yeah. So I actually didn't know what it was like to be on Facebook for a long time because I think like a lot of tech reporters, I forgot about it after the 2016 election and everyone talking about fake news. And it's like, okay, we're leaving Facebook. We're just kind of casually not, we're not going.

there anymore. And over the last year, I've started going there again just to see what is, what's going on. And the only way I can describe it is it's like a dead mall vibes where there's like some people there, but it's a lot of abandoned stores, abandoned accounts, things like this. And over the last few months, there's been a huge push of AI spam into the news feed. Facebook is trying to compete with TikTok. And so rather than just showing you things that you like and that your friends and family like, it's also started to show you stuff that just people on the network are liking. And increasingly, these are AI generated images that some people think are real. And over the last few months has really descended into like the bizarre like shrimp Jesus.

Yeah, we talked about shrimp Jesus maybe a month or two ago. I feel like that was the beginning of a lot of people's realizations that this stuff is getting increasingly bizarre. I've noticed it on my own feed. I go on Facebook not infrequently, and I get a lot of AI generated home photos, which people 100% seem to believe are real. But I feel like the creepiest part is these kind of more bizarre scenes. Can you describe some of them? I mean, aside from shrimp Jesus, give us some examples of the imagery that has become very popular. Yeah. I mean, Jesus in all of his forms are very common. It's like Jesus standing over the bodies of dying old people, Jesus with a birthday cake next to deformed children. There are a lot of like amputee children and mutilated children, which I'm very

sorry, but that's what is happening. Like I'm seeing a lot of that. There's also a lot of Jesus with ripped men, like just really strong like bodybuilder types and as well as with hot flight attendants. Wait, I've seen the flight attendants. And then what is that? And also I've seen like Jesus as an airplane, Jesus as airplane wings. Yeah, there's that. There's also like giant Bibles. Like cops and giant Bibles is a thing. Like cops putting giant Bibles into their cop cars as though they are arresting the Bible. And then the caption will be like, you know, we have to stop this. Show your faith that that sort of vibe. Yeah. Yeah. I mean, I've spent a lot of time just trying to figure out like what is going on and are these people who are posting this real and more importantly are the people who are commenting on it like are those people real i have to say i've been in the comments of some of these posts as i was mentioning the architecture post which are the ones that i've been getting to me those people seem completely convinced when i've looked into the comments of some of the more bizarre ones it seems to be a lot of international people what have you found

and who are the demographics that are engaging with this content most heavily from what you've seen? Yeah, I think you absolutely nailed it. These are categories I invented and seemingly you also have noticed where there's AI images that look real. Like these are your AI generated log cabins and beautiful scenery and things like that. And the comments on those, they're often engaging with the content of the photo. They're saying like, wow, this is so amazing. Like, I wish I could spend time fishing at this lake. and then you like click into the profile of the people commenting and you can see like patterns of

human behavior where they are interacting with their friends and family they have like comments and wallposts and images and then there are the bizarre images these are the shrimp jesus you know jesus as soldier it's not always jesus but it very often is there's also a lot of like celebrity type stuff like i've seen jason beber as uh like oh my god flight attendant vibes yeah beber back Where is that? It's so weird. It feels like this sort of like manifestation of the collective unconscious of so much of humanity. It's like religion, police, death, disability, all of these things that kind of play into people's worst fears or like strongest emotions almost, it seems like. I think you're right. And I think that the, so the bizarre AI, there is not as much human activity on those posts. Like the comments, are often the same. They often say amen. They often are just strings of emojis. And when you click into those profiles, these are, I'll call them people, but they are accounts that are commenting the same thing on dozens and dozens and dozens of these images, which leads me to think that

they are bots to some extent. But I think what's happening is a lot of them are hijacked accounts, where they were once human and may have been hijacked and now they're like doing inauthentic activity. And that's bad. It's possibly not as bad as if people thought this stuff was real, but it also has the impact of, because of the way that the Facebook recommendation algorithm works, it has the effect of putting this stuff in people's feeds regardless of whether humans are interacting with it or not. I have a family member that I saw comment on one of these AI generated posts. And it wasn't clear if they realized it was AI generated. And I think some of the these images elicit such an emotional reaction that people maybe comment or they see the other

comments and they just kind of post and move on. Is the goal basically just to get those bots and to get that inauthentic activity, to get the post viral, to get it into enough people's feeds that it increases the reach of the page? I think so. Yeah. So Renee DeResta, who is a researcher at Stanford, did a report on what she thinks is going on here. And I talked to her before and after that report. And some of these accounts are pushing people off of Facebook to like scam pages or to AI generated like content farms that have tons and tons of ads on them. So some of the pages are using AI to get followers and to get people to follow the pages and to inject this into people's feeds and then they're pushing people off platform. Some of them are selling what seems to be fake products. Like if you click through, you can go buy like a dog leash that doesn't seem to exist.

and then some of them were just like straight up credit card stealing scams. Others don't seemingly have any scam associated with them right now, which makes me think that they're just amassing followers that they can then use later because it's like if you have a huge Facebook following and you can regularly make things go viral, you can either sell that account or you can pivot it to something different later. This is something, well, I think I texted you about this, but, you know, I think a lot of A.I. Spam is also spread off Facebook onto other meta platforms, obviously Instagram, LinkedIn. I have been getting nonstop these AI generated cat pages now after a friend sent me one of them.

And then I sort of went down a rabbit hole and now they are continually suggested and they're these AI generated narratives about cats. But some of these accounts have hundreds of thousands of followers. And when I was DMing with the owner of one of them, like what is the point? who basically just was explaining that it's to grow his e-commerce store and the page will eventually be flipped into that purpose. How do we make this stop? Because it seems like there's an increasingly higher and higher amount of just AI generated garbage in all of our feeds. Yeah, it's very interesting because like five or six years ago, I was doing a bunch of stories about how Facebook does content moderation.

And we had all these leaked documents and we kept publishing stories about Facebook's internal rules. and after a period of time Facebook was like, okay, enough with the internal documents, please just come to our office, interview everyone who works on this problem about like disinformation and spam and all of this sort of thing. And so I did this huge story

and I came away pretty impressed with how seriously they were taking criminal activity on Facebook and on Instagram, disinformation, fake news, like all of this stuff. Like that very serious people working on this problem. then Elon Musk buys Twitter, interest rates increase, et cetera, and there are huge layoffs, like, at many tech companies. And the content moderation teams were hit very hard at meta and at other social media platforms.

And since then, it is like anything goes on Facebook, in my opinion. Like, I'm seeing like AI generated hentai constantly, which is just like nude anime women, like injecting the feed, hundreds of thousands of life. Like, this is stuff that would not fly on Facebook two, three years ago. And I used to screenshot all of this stuff and back it up because it was very likely when you saw terrorist content or something really bad on Facebook that they would catch it and they would delete it. And I have stopped screenshoting stuff because I know they're not going to take it down. They are not catching any of this stuff. And that goes for Instagram. It goes for threads. It goes for Facebook proper. Well, some of the most viral AI garbage images on Facebook are now just being reposted for engagement bait on threads and getting, you know, hundreds of reshares or rethreads, whatever you call it. It's taking over. I feel like it's become a plague where it's just infecting more and more of these platforms. And especially I see even meme accounts sharing this stuff because it's just getting traction and it's just cheap and easy likes, I guess. And it's interesting because the AI spam on Facebook has actually made Facebook relevant.

off platform again in a very weird way because it's devolved into such a weird place that people are screenshoting stuff going viral on Facebook and they're putting it on X, they're putting it on threads, they're putting it on blue sky. And it's going viral off platform as like, look at these boomers on Facebook who think AI is real. And then I think that people go to their own Facebook to see if this is like what their feed looks like now. And often it is. And so that that's like, an interesting side piece, but I also think that Facebook has put itself in a very difficult place because it is integrating its own AI features into every platform that it owns. And at the same time, AI that it does not own is generating this spam and destroying its platform. Like

Facebook is, in my opinion, like borderline unusable at this point. And they are not acting on it really. They've taken down some of the pages that I've sent them where I'm like, hey, there are mutilated children that are going viral on your platform. I'm not the internet cop, but this is clearly against your rules and you're not doing anything about it. I'm very curious like how you approach this and they just take it down and don't say anything. And then also these tech companies are between a rock and a hard place in terms of how to take action on it because to say like We don't want AI generated images on our platform. Well, tech companies are generating these images in many cases.

Yeah. I mean, this is really, I think, the TikTok problem because I think what TikTok did is it really made it clear that the future of social needs to be sort of built around discovery and that people actually putting the burden on users to like manually seek out and then follow people and then get updates is like kind of a lot of burden on users. most people just want to sit back and have a steady stream of great content delivered to them without any effort, as is the case with TikTok's For You page. I think the problem is that TikTok has such, you know, invested for years in this recommendation

system that is really hard to replicate. Meta especially has way undervalued, I think, discovery throughout the entire course of its existence as a platform. And that has led it now where like now they're trying to kind of like focus on discovery. but they're really just taking the most sort of garbage. It feels very brute force to me. Two things on that very quickly. So META had an earnings call last month,

and Mark Zuckerberg specifically said that something like 30% of all posts in Facebook are recommended, meaning essentially using the TikTok discovery situation where you have no connection to it whatsoever. And so one of the effects that this has had is like, you can be a farmer in Ohio and you will see stuff that is popular with like a teenager in Thailand or something. And it's had this effect of globalizing Facebook in a way that is that the news feed wasn't before. And you mentioned earlier, it's like a lot of the bizarre Facebook spam is coming from developing countries. And that's just like a fact.

Like if you go and look at who's running it, it's like often an administrator in Vietnam, administrator in Bali, like something like this. And I think that that is a significant change to how Facebook worked before. I'm not saying if it's good or bad. It's just it's different. And then the other thing, you're absolutely right. It's like Facebook has not tuned the for you algorithm like TikTok has. It's like I feel like if I like, if you go on Facebook proper and you like one AI generated image, it's over. It's over. And on TikTok, I feel like it has more of a memory where it's like, okay, I'm getting a mix of things in a way that doesn't feel true on threads or Instagram or Facebook. It's just a better recommendation system too. And I think what

you're saying about it becoming a more global platform is also 100% correct because so much of this stuff is really just like, wow, this has a certain amount of traction and momentum. Maybe it came out of Indonesia or whatever, but it's getting enough momentum that now we're just going to blast it into everyone's feed because that is going to get people to engage. and that is what we want above all else. It's disturbing and kind of scary. I feel like it's become a meme, especially with like the shrimp Jesus of it all

and like just some of the absurdity of this stuff. But it's so exploitative and depressing. And I guess I keep thinking about some of those images, especially of the disabled children and the darker stuff and kind of the way that vulnerable people are always exploited for content. And now we have like AI generated sort of vulnerable people being exploited for content. It's gross. Yeah, I think you're absolutely right.

And another thing that we're seeing a lot of is, as I mentioned, there's like AI-generated hardcore hentai, like on Facebook. But Facebook and Instagram have been like historically extremely aggressive about policing against sex workers, for example. But then you have like these spam pages that are, you know, breaking all of Facebook's rules that are somehow evading detection. Have you seen the breastfeeding ones? Yeah, the breastfeeding ones.

ones are crazy. I love the fake. They just have like increasingly fake looking dolls. It's like these only fans women for people that haven't seen and they have their, you know, they have their breast out. But because Facebook allows breastfeeding, they can, I guess, show more of their breast without getting taken down. So they hold these like fake looking dolls up to their breasts and just kind of like dance around or jiggle around or like pretend to breastfeed, I guess in an effort to not get taken down and it seems to work. Yeah. Yeah. And I don't like more power to the. them, I totally agree with you. But there are so much AI-generated breastfeeding content that is like I say deformed and I don't say that lately. I'm saying like these children have like 16 eyes.

And like sometimes the babies are also pregnant and it's just like their hands are like melding into the mom's face and no human looks like this or has ever because they are just so bizarre and AI generated. And I just saw like a half pig, half human baby breastfeeding content on Facebook. And it's clearly AI. And it's just like this is a monster. It's like it's a monstrosity that has been dreamed up by an AI and has been posted to Facebook. And it goes viral because it's so weird, I guess, or because it just has tons of bots clicking. It's disturbing. But I do think that there's a time when this stuff will become more realistic looking. I mean, we're already seeing as you reported as well, like these AI deep fakes on Instagram, right,

where women's faces are being replaced. Yeah. And this one is, I actually find this to be more disturbing because it's very, very hard to detect. And so it's more of a traditional like deep fake where there is a video that has been altered to be AI generated. And traditionally, this has been used to put the faces of celebrity women onto the bodies of porn stars or adult performers rather. You know, that's, that was the mechanism of abuse for a very long time. But what we're seeing now is these AI influencers where the face is AI generated and the

face of this person. They're not a person. They're an AI generated, you know, women. Like, they have an Instagram account and they are an Instagram influencer and someone is taking their face and is putting it onto the body of a real person. And we've seen these go viral on Instagram over and over again. I am positive that the people engaging with these are real. Like, there's no doubt because they're very obviously humans.

And some of these influencers have hundreds of thousands of followers. And it's really hard to detect because with regular, like, still images, you can do a reverse image search. You can kind of like figure out where it is coming from. And with fees, you don't really know who's body. they are stealing, which sounds very dystopian, but it's the truth. It's like it's very hard to detect where the original video comes from. Yeah. And the goal is to usually just generate OnlyFans subscriptions or... Yeah, they just link to OnlyFans or they link to Fansley or FanView, which are just like OnlyFans competitors. Some of them link to Patreon. The goal is to get people to

pay $10 a month for AI generated porn. Talk about why this matters. Because I feel like there's also a lot of people that, you know, okay, maybe they see those posts that are like, ha ha, look at these dumb boomers on Facebook that think this stuff is real. But maybe they don't see a lot of it in their own feeds. And who cares if Facebook goes into the graveyard? You know, most of us don't use it anymore anyway. Well, I think that we're seeing AI generated content. Like many news outlets are integrating AI into their newsrooms and are using it to replace journalists. I think that more of the internet is becoming AI. generated or AI assisted in a way that is very concerning. It's like AI spam often outranks human generated content and often that stuff is pretty low quality. I think that people are very annoyed with Google recently in terms of like, oh, you search for something and you have AI content and you have ads. Then you have like a bunch of lists that have been kind of SEO reverse engineered.

And I think that there is this field. like the internet is getting a lot harder to navigate and to find things that you want to find. So there's just the general like AI is adding a layer of crap to everything. And then the other thing I will say is that I posted on Facebook, I said, hey, tell me if you've seen AI generated content in your feed. I was curious when my friends and family had seen. And a bunch of people sent me stuff that they had seen. And I was like, okay, that's AI spam, AI spam, AI spam.

But then a bunch of people also send me images and content that I knew was real. And the way that I knew that it was real was because I went to this artist's Instagram and they have videos of themselves painting a picture or drawing like painstakingly showing their process. And I think that it is creating this alternate problem where people really don't want to be fooled by AI. And so their guards are up to the point where they are saying that real, stuff is AI generated. And I think that there is this issue of like people are just having a hard time deciphering what is real and what is not. It reminds me of the problem with bots as well. I don't even think the main damage done by bots is necessarily always done by bots themselves,

but just the perception that basically any sort of activity can now be viewed as inorganic. I was covering the student protest stuff on college campuses a couple weeks ago. And I was interviewing older people about kind of what they thought. And so many of them said that, oh, you know, these alleged students posting online, they're Russian bots, they're Russian trolls. It's all AI. It's Russian disinformation and bots. And of course, you know, does Russian disinformation exist? Sure, to a point, there's, it's absolutely, you know, these students are 100% real. But I feel like just by introducing doubt into the equation, it almost makes it harder for people. to discern what's real or fake because they reflexively view everything as fake or everybody that they

don't agree with as a bot or I think you're absolutely right like it's different but the industry plant episode you did a few weeks ago where it's like any you can kind of like denigrate anyone you don't want to by saying oh that's a bot that's AI they're a plant they're fake you know we've been in this world for a decade at this point but people are creating their own realities and their own narratives and I think that we're long past the point where we have like a shared reality but I think it's getting to the point where even if you are very media literate it can be hard to tell what is real and what is not or what has been altered and what is not and it's like I have spent hundreds of hours looking at AI generated images and I feel like I have a

very good detector for it but there are some things I see on Facebook and I'm like I don't know like I think that this is AI but I don't know and of course the technology is getting better every day and it's getting in some cases harder to tell every day on facebook it's actually getting easier because this stuff happened like it started out very realistic and now it's gotten so surreal and so weird that it's like okay this is nuts um but i think that the pendulum will swing back to a point where it's like okay is this image of protests real is this image of war in gaza real like we've already seen some of that stuff where people are saying these are bots these are fake this is ai and

that's not always the case. It's not even usually the case. Is there any effort from the platforms to kind of label certain content as organic or not organic? Do they even have those capabilities? Yeah. So meta has said that it is going to start labeling AI generated images as AI generated images. And it said it was going to start doing that in May. You know, there's still time. But they rolled out a new tool where you can label your stuff as AI generated. but it's voluntary.

Okay, no one's going to do that. No one's going to do that, especially these spam pages that are trying to fool people. So they're not doing a great job. Open AI has said that it's also going to try to do something like this. Some people have suggested that we just assume everything is AI generated. And then like our phone cameras and, you know, our Photoshop programs and whatever people use to make art or whatever they're using will label real things as like, this is real. and this has not been altered. I don't have a lot of hope for this because I think what we've seen is that

a lot of this stuff is coming from off of the platforms and it's being posted to the platforms and it has the specific intention of trying to deceive. And I just think that it's going to be a cat and mouse game for quite some time. None of it bodes well. To me, a lot of this AI stuff is a little bit of the nail in the coffin for Facebook. I don't see Facebook becoming more usable. It is so overrun with garbage. I do think that that's coming to Instagram and LinkedIn and these other platforms, too,

just as we move into this recommendation-based internet. But we'll see how it goes. Yeah. Yeah. Well, Jason, thank you so much for joining us today and for chatting. Yeah, thanks as always. Thanks for having me. All right.

I'll be right back to talk through some of the biggest news stories this week. All right, I'm here with my showrunner, Zach Mack, to talk about some of the biggest stories this week. How's it going? So last week, OpenAI unveiled this new AI voice assistant as part of its chat GPT update. And the voice on this voice assistant sounded eerily similar to Scarlett Johansson, who played an AI assistant in the movie Her, which is, of course, about a human man who falls in love with his AI assistant. Like, are these feelings even real? Or are they just programming? And that idea really hurts. People called out the similarities between Open AI's voice and Scarlett Johansson's voice almost immediately.

But this week, we heard from Scarlett Johansson herself. She basically posted this statement that said, Open AI came to her in September, asking to use her voice in their AI system. And she declined. Then, two days before ChatGPT 4.0 launches, Scarlett says that Sam Altman called her agent asking her to reconsider. But before they could even respond, ChatGPT 4.0 was released with a voice sound. a lot like hers. Hey, chat GPT. How are you? I'm doing fantastic. Thanks for asking. How about you? How's your day going?

I'm doing great. So I... Now they did give this voice assistant a different name. It's called Sky, not Scarlet. But right after it launched, Sam Altman just tweeted one thing, and it was the word her, seemingly referencing the movie. Open Eye still says that Sky's voice is not an imitation of Scarlett Johansson, but it belongs to a different professional voice actress using her own natural speaking voice. Zach, do you buy any of that? Yes, I buy that it is a different actress. Like, yes, I buy that it's not Scarlett Johansson's voice, but it's so clearly meant to resemble Scarlett Johansson's voice and the whole vibe that she created in the movie her.

And so much so that they're reaching out to her asking for her permission, she's saying no, so then they're just like, well, we're going to do it anyway. It kind of reminds me a lot of those old Navy commercials that had that Kim Kardashian look-alike, but it wasn't actually Kim Kardashian, and then Kim Kardashian sued them, and there was a whole big out-of-court settlement. I mean, brands have long used celebrity likenesses to market products, sometimes legally, sometimes illegally. To me, I think the reason that this is such a compelling story to so many people is because we know that OpenAI, especially, but all of these large language model companies are using training data from all over, often that they have not paid for in license,

right? This is the whole controversy and lawsuit between the New York Times and Open AI. And so I think what this shows is that a lot of these companies, specifically Open AI and Altman, are willing to blaze ahead whether or not they get approval from someone to essentially replicate their likeness. And I think that's what's concerning. Totally. It seems like, yeah, all these companies at every every opportunity are cutting corners and just, yeah, blowing past restrictions. They really don't seem to give a fuck about copyright. Well, it's the whole move fast and break things mentality in Silicon Valley. And, you know, Sam Altman, who previously ran Y Combinator, they had that saying as well, where they basically just encouraged people to worry about regulation later, right?

Worry about crossing your T's and dotting your eyes later. The most important thing is to push tech innovation forward at all costs. And I really am eager to see how this conflict gets resolved. But I do think that it's commendable that Scarlett Johansson is standing up for her creative work and voice. So a new video game that I'm kind of excited about just launched. I haven't had a chance to play it yet, but it's called content warning. And you essentially play as an influencer. It's a horror-themed game. And the objective is to go on these scary missions and get footage to upload to Spook Tube, which is a version of YouTube for the game without being killed. You earn views based on how frightening your content is, and the game's premise plays on basically how people will do anything for content and fame.

I think this is one of the most fun ideas for a game this year. I love horror and I love influencer stuff, so I kind of want to check it out. I love how some of the best horror content that we get has a lot of creative social commentary wrapped up into this. So it's sort of you're playing this exciting game, and it's causing you to also think about our society and how we are willing to risk our lives to get content. Yeah, totally. I agree. It kind of reminds me of when Logan Paul went into the Japanese suicide work. He didn't risk his life, though. He didn't risk his life. That's true. I mean, I saw Jeff Wittick posting today. He's back coming out against David Dobrick saying, oh, you know, Snapchat, I can't believe you're partnering

with David Dobrick. You know, this is the influencer who bashed my head in for a stunt. For people that don't know. Jeff Wittick is somebody that used to work with David Dobrick and was in his kind of vlogs. And this stunt famously went wrong where he was swung around on this piece of construction equipment and it bashed his head in and he lost, I think, partial eyesight. He has brain damage. And I mean, David has retreated to Snapchat where he's making millions, this ongoing suit, I think, between Jeff and David. But it's just a reminder of the lengths that some of these content creators will go to create content. and the danger that that puts not always themselves,

because they're not always the ones taking the risks, but these secondary characters, right, the Jeffs of the world. So it'll be interesting to see. I was sort of upset at the response to the tweets that Jeff posted because people were being so cruel about it. They're like, shut up. You've been talking about him for years. Like, it's your whole personality.

And it's like, well, yeah, if an influencer swung me around on a rope tied to construction equipment and it gave me permanent brain damage, like, Yes, I would still be talking about that four years later, you know. I think people just don't want to confront that reality. They just want to keep the good times rolling. They want the stunt content, but they don't want to deal with what happens when the stunt goes wrong. They don't want to see that. They just want to see the guy jump from this building to the other building successfully.

They don't want to see the take in which someone fell. Okay, here's a cute and happy internet story as a pallet cleanser. E.J. Dixon, a reporter at Rolling Stone, interviewed the family behind the four Seasons Orlando baby? TikTok. Have you even seen the TikTok, Zach? I just watched it today. I have a question for everybody. Who wants to go to the Four Seasons Orlando? Me! The mother is asking, who wants to go to the Four Seasons of Orlando? And this baby really cutely raises her hand and says, me. I think it went viral because people aren't used to seeing

sentient babies. And the baby seemed weirdly sentient and aware of the Four Seasons Orlando. the parents themselves didn't put it on social media. It was actually like a family friend or relative or friend of a relative who posted on TikTok just thinking it was really cute. It obviously blew up. And so far it's been really positive. The Four Seasons Orlando has actually reached out to the family and, you know, apparently they like to go there. They live not too far away. So I think it's a cute, happy story. That's great. Love a cute, happy internet story where kids aren't being like totally exploited. Great. Well, let's not forget that the majority of content featuring kids on the internet is kids putting themselves on the internet. So the idea that kids are being exploited in social media content is a, that's like my pet peeve. What I like about this story is that there's no hate. The 13-month-old has not been milkshake ducked yet. And it's a cute story. And when E.J was asking the mom, you know, how does this baby know about the four seasons Orlando? Apparently they were doing renovations on their house and they took advantage of these. in-state rates that the hotel has. So they would go there sometimes and use the pool. And so Kate, the baby,

became aware of it as a place. So while it seems kind of absurd to be like, who wants to go to four seasons, Orlando, like that is actually a place that she's really familiar with because she's been taken there

a bunch of times. So I thought that was... All right. Now she gets to go again. I want to grow up as a baby being taken to the four seasons. Adopt me. I didn't know you were in age play.

Okay. So Belle Delphine is a podcast. popular content creator, YouTuber. In 2019, she earned around $90,000 famously for selling these little jars of her bathwater. People were paying her through PayPal, which ended up taking action against her account and locking up all of that money. PayPal's current terms of service said that the platform only allows U.S. transactions for certain goods and goods that are physically delivered to the customer. The bathwater was apparently a gray area for PayPal. Nobody had really tried to sell Pathwater through the platform before at scale.

And Delphine claimed that PayPal told her that each sale was considered a violation and would cost her $2,500. So that's $2,500 per order. Obviously, she combated this. It's been a couple years battle. She posted a whole Twitter thread about losing this $90,000. And PayPal took action and actually reinstated the money. So she was very happy about that. she said, I knew it would be a better news story to say that I made so much money from selling my

bathwater. So I just kept it a secret. Ultimately, I'm still glad I did it, since it was really funny at that time on the internet when it happened. And I did still mail out all the bathwater that people bought from me, which I did all by myself. But F PayPal, I guess. And then, of course, PayPal came in and made the situation right, which I'm happy about. Well, first of all, I just can't believe she sold them in jars. I remember when she sold these jars of bathwater. I can't remember if I ended up writing about it. I just got mine. I just got mine in the mail. Yeah. I mean, look, I think selling unconventional goods is part of being a content creator these days. Everyone is looking how to monetize and coming up with unique and kind of outside of the box products is actually

how you succeed online. And I think this says a lot about payment services and their struggles to handle influencer businesses. Nobody knows more about the restrictions that payment systems implement and how that shapes the internet and commerce more than sex workers. And I think this is a good example of just, you know, PayPal sort of arbitrarily deciding that this was a violative offense to sell this bathwater. And I'm glad that now, I think also through the normalization of selling random things and content creation, generally that has happened over the past five years, they were able to make it right. It's weird that bathwater sort of violated PayPal's policies. Yeah, I think it's just because it's, you know, this like me of like gamer, you know, bathwater or whatever.

Yeah. But yeah. I mean, I think anybody should be able to sell, you know, whatever good that they produce as long as it's within, you know, slightly ethical guidelines. Do you remember when someone was selling on eBay air from a in a Ziploc bag from a Kanye West concert? Yes. Like it was the air from McCann. Yeah, I do.

I think people have tried to sell vibes before, too. there's another TikToker that has different energies that she keeps in jars. I think people were trying to sell energies on Etsy at one point, too. I mean, there's a lot of people selling sort of dubious stuff. With Bell, I will say, I don't think people cared even as much about the exact bathwater. It was more about cementing this pariscial relationship, right? Or cementing this bond. When you buy a product like bathwater, it's like, what are you going to do with that bathwater?

Probably not much. but it's like, yeah, it's this intimate purchase and it sort of makes you feel like you're closer to the content creator and I think that's what people wanted. And man, 90K for bathwater? She's going to have to pay a lot of taxes. I got to start taking baths again, man. Well, a lot of that's going to go to taxes.

She paid for the shipping and packaging already. So I'm glad that she was able to recoup, you know, her costs. Let's not forget. All she had to do is take a bath. No, she had to create these jars, though. She packaged them. She had labels made. she had to ship them in these waterproof containers.

Like, it took effort. I have something that I have one box, Zach, that I have not mailed in like months. I was supposed to return this one thing in November and I still haven't mailed it. So like, anybody that can mail out $90,000 worth of bathwater, that is a skill in itself. She probably does not have severe ADHD. I bet if I gave you 90K, you would go mail that shit. All right. That's it for the week.

Thanks so much for joining me, Zach. Of course. All right, that's the show. Power User is produced by Travis Larcuck and Jolani Carter. Our video producer is Brandon Kiefer. You can watch full episodes every week on my YouTube channel at Taylor Lorenz. Our executive producers are Zach Mack and Nashat Kerwa. Power User is part of the Vox Media Podcast Network. If you like the show, give us a rating and review on Apple Podcasts, Spotify, or wherever you're listening. We'll be back next week with another episode of Power User. See you then.