WIRED Roundup: The New Fake World of OpenAI’s Social Video App

0:01Welcome to Wired's Uncanny Valley. I'm Wired's Director of Business and Industry, Zoe Schiffer. Today on the show, we're bringing you five stories that you need to know about this week, including our scoop of how OpenAI just launched a social app dedicated completely to AI-generated videos. I'm joined today by our Senior Culture Editor, Manisha Krishnan. Manisha, welcome to Uncanny Valley. Hi, Zoe.

0:28Our first story is about the thing that I feel like our whole newsroom is talking about, possibly the whole country is talking about. It's the government shutdown. As of Wednesday this week, the U.S. government has officially shut down, at least for now, and many federal workers are being furloughed until the government reopens. Our colleague Victoria Elliott learned this week that employees at the Small Business Administration, or SBA, received a template from with language for their out-of-office email, and they were advised to blame Democrats. Of course they were. How explicit was this template?

1:05Well, let me just read a little bit of it to you. So it says, quote, I am out of office for the foreseeable future because Senate Democrats voted to block a clean federal spending bill leading to our government shutdown that is preventing the U.S. Small Business administration from serving America's 36 million small businesses. So pretty explicit, I guess is what I would say. Yeah. If I got that as an out of office, I'd be like, I'm not reading all that. Like, that's very wordy. It's petty. It's kind of funny, but it's also dark. It just feels like yet another example of how explicit the government is.

1:43There's no semblance of even trying to like unite the country or the two parties anymore. I know. Like, federal workers are supposed to be boring. We're not supposed to be hearing these takes from them. And one thing that Tori pointed out in the story was that this actually could violate the Hatch Act. And that is a law that sets limits on the kinds of political activities that government employees can engage in. And so, like, by forcing them to do this out of office, it could actually be forcing them to break the law. Yeah. And I just think, you know, for a few years now, we've heard a lot about the quote unquote deep state, like this idea that these federal employees are really political and they're trying to like push their political agenda.

2:26Obviously, when you actually report on this stuff, you talk to people, they're like, we've worked here throughout many different administrations. Our job is not to weigh in on political issues. Our job is to like do this stack of boring paperwork or whatever. And that seems like it's starting to change. They're being asked to almost take a side. So employees at other agencies have received more standard out-of-office directives, but also a lot of these agencies are still using this kind of spicy political language themselves. There's pop-up banners on their websites blaming the Democrats. The Department of Housing and Urban Development has one that says, quote, the radical left in Congress shut down the government.

3:07The agency will be using available resources to help Americans in need. And the DOJ website has like a pretty similar banner. Yeah. To see that from the Department of Justice, I think it just kind of shows you where we are right now because these messages, they're insane, but they're becoming so normalized. And I feel bad for the federal workers. I mean, you would know better than me because you guys did that brilliant feature speaking to federal workers who were impacted by Doge. They're losing their jobs. They're being forced to put out these super politicized messages. It's just so, I mean, uncomfortable.

3:46It doesn't even begin to kind of describe it. I think one thing that they talk about a lot is just how quickly the norms of government have changed. There was a standard way of operating previously. It didn't always happen. Like there were people here and there who flouted the norms. But for the most part, it was like, if you were a federal worker, you weren't supposed to be super outspoken. You weren't supposed to be wildly political in public necessarily as part of your job. And those things have changed really fast. And I think even the people who've kept their jobs feel like something real has been lost in the process.

4:19I'm not normally one to cape for decorum and etiquette. But I mean, in this situation, I really wish that some sense of normalcy would return. Yeah. Our next story is kind of a doozy, and I am specifically very curious to get your take on this. Wired contributor Laura Bullard wrote a deep dive on Peter Thiel's obsession with the Antichrist. So for a few years now, Thiel has been going around giving lectures about his doomsday theory of how we're all way too concerned with technology. And as a result, we've failed to pay attention to the far greater threat, which is the coming of the Antichrist.

4:59This figure will allegedly unify humanity under one rule before delivering us to the apocalypse. So I am very curious to hear what you think. I mean, I certainly think it's convenient that he's trying to be like, oh yeah, don't worry about technology. The apocalypse is coming. That's what you should be worried about. It reminds me of that I think you should leave meme where he's like in the hot dog and trying to pretend he didn't cause the car crash. I mean, I guess each of us are free to have our own apocalypse theories. I kind of feel like we might be in it right now. But the difference is that Teal is one of the most powerful men in the world as the co-founder of PayPal and Palantir and an active Republican Party supporter.

5:47Yeah. Just to give you a little bit more backstory. So Laura traced Teal's belief to his relationship with an Austrian theologist, who in turn was influenced by the writings of legal theorist Carl Schmitt, who himself was tapped by the Nazis back in the 40s to justify Germany's slip from democracy to dictatorship. Schmitt believed, as Thiel seems to, that the Antichrist's evil is an attempt to like unify the world. And this, although it sounds positive, is actually quite negative in their view. This unity would lead us to be brainwashed and eventually it would be our downfall. Now, the irony of the co-founder of a data and surveillance technology company having these societal control fears has not gone unnoticed.

6:29Thiel recently had a conversation with The New York Times journalist Ross Douthat, and there was a very awkward moment where this was brought up. My very specific question for you, right, is that you're an investor in AI. you're you know you're deeply invested in palantir in military technology and technologies of surveillance and technologies of warfare and so on right and it just seems to me that when you tell me a story about the antichrist coming to power and using the fear of technological change to sort of impose order on the world i feel like that antichrist would be maybe be using the tools that you were building, right?

7:13Like, wouldn't the Antichrist be like, great, you know, we're not going to have any more technological progress, but I really like what Palantir has done so far, right? I mean, isn't that a concern? Wouldn't that be the, you know, the irony of history would be that the man publicly worrying about the Antichrist accidentally hastens his or her arrival? There are all, look, there are all these different scenarios. I obviously don't think that that's what I'm doing. I mean, to be clear, I don't think that's what you're doing either. Ooh. I wish that Ross didn't chime in at the end there where he was like, I don't think that's what you're doing.

7:52Like, just let him be awkward because he was stumbling. He obviously felt very uncomfortable with that. But you have to wonder, like, who is he worried that is going to, like, be uniting? You know, this extremely rich, powerful figure. Is he just worried that normal people might try to, like, take back some power? And I think that this does raise a really interesting point about our level of media literacy right now and all the slop that we're being fed and how, like, I mean, I've seen it in my own family, like my own, my dad, just to be honest with you, like, the things that he believes that are real now.

8:27And so, yeah, I think there is absolutely like a danger to all of the slop that we're putting out there combined with a lack of education. Right. Yeah. No, I completely agree. I mean, I had two thoughts. I listened to this podcast when it came out because I have a brain disease that makes me really like this type of content. And one of my takeaways was just having reported on Elon Musk for years and years and years, I was really struck by how when you listen to Elon, you can tell that even if you disagree with every single thing that comes out of that man's mouth, he does seem to believe what he is saying, no matter how nuts it sounds.

9:06I did not have that feeling with Peter Thiel. I was like, I can't tell if this is a genuine belief. I also just like on a literal note was like, is this whole thing a metaphor or are you like a literal person is going to come? Like, I just was so confused. And Ross was taking it very literally, very seriously. I know. I haven't heard that much Antichrist talk since Catholic school. And even then, I was mostly disassociating. So something else that Laura's reporting brought up was this idea of the counter figure of the Antichrist, the catacomb. and it's supposedly a figure that can withhold the Antichrist and the end times.

9:46And when Teal was asked if Trump was the catacon, he refused to answer. Who would be your personal catacon, Zoe? I really struggled with this because, I don't know, that's a lot to put on one person right now. It's hard to think about someone who could unite the country, much less the world at this point. I did think of like two different things. One was when the astronomer CEO was caught on a kiss cam at a Coldplay concert, snuggling with his head of HR. I felt like that moment really did bring the country together. Everyone was kind of united and laughing and interested. And there wasn't a lot of different takes.

10:28It was just kind of like everyone was on the internet watching this thing play out. The other thing, I don't actually watch South Park, but I was just listening to a podcast with Wesley Morris last night about it. And it does sound like the South Park audience seems to kind of cross political lines, although the show itself is extremely political. Yeah, this season has, I mean, this is the first time I've watched in like 15 years, but this season has been pretty spot on. It's almost like they're like reading Wired and like satirizing everything. I feel like if Trump and Eric Adams started a talk show, maybe that could be the catacomb.

11:04Because like everyone would find it funny across political lines. So maybe South Park will save us after all. Switching gears literally for our next story, our colleague Ariane Marshall reported that Tesla has been encouraging drowsy drivers to use the full self-driving or FSD mode on their cars. Contrary to its name, this feature does not actually allow cars to drive themselves. It just assists drivers in doing a variety of basic tasks. The manual for the car says that the driver needs to be ready to take over at all times, but drivers are reporting that in-car messages are appearing to tell them to do just the opposite.

11:41The messages say things like, quote, drowsiness detected, stay focused with FSD, or quote, lane drift detected, let FSD assist you so that you can stay focused. Yeah, that sounds dangerous. It sounds like they're kind of like, hey, hey, you want to take a nap right now? Let FSD kick in. No, they should be like blasting music, blasting the AC, make it like a spin class in there to wake you up. Tesla has made changes to its technology to make it more difficult for inattentive drivers to use FSD. Back in 2021, the company started using in-car driving monitor cameras to determine where their drivers were sufficiently paying attention while using FSD.

12:25It seems at odds with their past efforts to build more safety around their self-driving features. This is like a pretty delicate time for Tesla. For years, the company has been accused of making products that can be allegedly defective in certain ways. This past August, a Florida jury found that the company was partly liable for a 2019 crash that killed a 22-year-old woman. The crash occurred when a Tesla Model S driver was using an older version of the company's driver assistance software called Autopilot. At the same time, Elon Musk and the company's board of directors have put FSD at the center of the automaker's strategy.

13:02So Musk has promised that the feature will transform into a truly autonomous driving system by the end of the year, although that's looking unlikely. And Elon Musk is generally known for promising pretty aggressive timelines that then he blows past again and again. One more before we go to break. Wired science reporter Emily Mullen reported this week that scientists made human eggs from skin cells and used them to form embryos. This is a huge deal because it could mean a new way to treat infertility for people who want kids. But to be clear, none of the embryos were actually used to try and establish pregnancy.

13:41And it's unlikely that they would have developed much further than the womb. but it's still a really big deal because it could one day be used as an alternative to IVF. I mean, that's pretty incredible, but how would it work exactly? Well, one of the reasons IVF can fail is because of poor egg quality. I feel like you and I probably both know people who've been in this scenario who've tried to go through egg retrieval and just couldn't get viable eggs because the quality of eggs declines with age and it's obviously a major factor in infertility. But if IVF patients could have a ready supply of eggs generated in a lab from a sample of their skin, it could vastly improve the success of IVF and allow more people to have babies.

14:23But this potential new method also raises significant ethical questions about how the technology should be used. In a 2017 editorial on the journal Science Transnational Medicine, bioethicists warn that it might raise embryo farming to a scale that is currently unimaginable. I mean, I think on one hand, like you're absolutely right. I do have friends who've struggled with infertility. And so it is really fascinating to see what new options could be on the table. But you also do have to wonder, like, we're in this time right now where, especially among the ultra wealthy, like there seems to be a lot going on in terms of like fertility and trying to have the best genes in your children.

15:06And I just kind of wonder how this could play into that, how this could be potentially used unethically, and who might be hurt by that or even exploited by that? Yeah, I think it's a really good question. I mean, this is a big thing, like you mentioned, the Silicon Valley elite, in addition to being kind of obsessed with longevity, are like pretty into some of these new methods for selecting embryos or testing them in different ways. I think Sam Altman has actually invested in a startup that is doing something similar. In certain ways, I think all of the embryo selection stuff, there's a certain degree that I totally get about it.

15:43And then when you take it too far, I personally feel like it's just a misunderstanding of what creates a good life. There's no guarantee that if you select the embryo that is the chart topper on all of these specific tests that it'll necessarily be happier or better off than one that ranks a little lower. Wow, that was very deep and philosophical. And I agree with you, though.

16:11Coming up after the break, OpenAI just launched a social app for AI-generated videos. We'll talk more about that scoop in a minute. Stay with us.

16:27This week on the political scene from The New Yorker, Trump's rupture in the world order. Europe caught between two adversarial great powers. That's basically dialing back the clock to not only pre-World War II, but really it's a pre-20th century view of the world. And I would say it's a world of permanent insecurity that we're looking at. Join me, Evan Osnos, and my colleagues Jane Mayer and Susan Glasser every Friday on The Political Scene, available wherever you get your podcasts.

17:05Welcome back to Uncanny Valley. I'm Zoe Schiffer. I'm joined today by Senior Culture Editor Manisha Krishnan. Let's dive into our main story. So earlier this week, my colleague Luis Matsakis and I learned that OpenAI was preparing to launch a social app for AI-generated videos. This app has now come out a day after we broke the story it actually launched. It's a standalone app that uses the video generation model Sora 2. I've been using the app a little bit since it came out, and it has some pretty familiar features, like a vertical video feed with swipe-to-scroll navigation and a tab that has a For You recommendation page.

17:44So it's basically TikTok, except everything you're looking at is generated by AI. Okay, first of all, congrats on this scoop. This story is so crazy. I just kind of want to know how you felt using it? Because I saw some of the videos that you were posting in Slack. And honestly, I was like, is that Zoe? So a couple of things. Basically, you open the app and it really encourages you to create content because it doesn't want people just mindlessly scrolling. So if you press this plus button, it asks you to verify your identity. And you do that by like filming a video of yourself, essentially reading out these numbers on the screen, which allows the app to generate your likeness, like a digital avatar of your face and body, and also your voice.

18:33And then it's pretty easy. You just put in a prompt. It kind of has the chat GPT issue where you're like, creating a good prompt is just like actually harder than you would think. Okay. So obviously there's a novelty to being able to sort of create your own likeness and make it do weird things. But do you think it could sustain your interest in the same way that TikTok does. Because like, I don't know, one of the things with social media, which our culture writer Jason Parham was talking about, you know, we kind of built our followings by leaning into being authentic and like having a personality.

19:06And this just seems like such a radical departure from that because you're not even using your own content. Yeah, it's weird. I mean, it unlocked something because Meta tried to roll out an AI-generated video app recently, and it's a complete ghost town. Like it was a total flop. I actually think that there's reporting that the company might have rushed up the release because they heard that OpenAI was working on the Sora 2 app. But regardless, people weren't really using it. I think one thing that OpenAI did really well is that it put you and your friends at kind of the center. And there is a real novelty to like watching you and your friends in like AI generated, invented scenarios.

19:46And you know, you can make it look realistic. You can make, you know, you and all your friends like mermaids swimming in the sea, like, and there's something like funny and interesting about it. That said, when you look at the actual feed right now, at least my feed is mostly OpenAI employees. And the main character of the app completely is OpenAI CEO Sam Allman. He's in like every other video. Yeah, I don't think anyone really wants a CEO to be a main character. Are there like guardrails in place for, I mean, obviously I'm thinking of like deep fake porn and like just things like that. Like what's the deal with that?

20:23Yeah. So it does have guardrails in place for that. Like you can't generate like a nude image or video of someone. I think if you even try and prompt explicitly romantic scenarios, it could block that. You also can't add someone else to a video. Like I couldn't create a video with Sam Altman unless he had set his settings to say like anyone can generate my likeness. My personal settings are set so that only I can generate my likeness because I just feel like that's a really scary genie to let out of a bottle. But you do get a bit of control. Interestingly, copyright seems like it's a bit of an open question for the app.

21:01Reese tried to generate a video of Taylor Swift or Darth Vader and the app blocks you saying that it violates the platform's rules, but then people were able to generate videos with Pikachu in them or other kind of famous copyrighted characters. I wonder if they base it on who they think is going to be most litigious. I don't know. I'm so curious how they like came up with these rules, but I think we'll keep reporting on it because it does seem like the copyright concerns are obviously going to be front and center. And also people are just going to be trying to like, you know, for example, I saw some people trying to create videos with different people promoting crypto coins and stuff.

21:42Like that seems like it's currently you're able to at least get around certain filters. So I feel like OpenAI is going to be playing whack-a-mole a little bit for the next few weeks as it sees different ways that people are trying to get around the current rules and generate videos that could be problematic. I'm curious how it's going to end up competing with TikTok. and also will it change behavior? Like TikTok has literally changed how people act in public. And so I'm kind of curious your take on what do you think they're trying to do with this essentially? That's a really good question. I mean, I feel like a bunch of AI firms are looking at AI entertainment as kind of a future of social media.

22:23Interestingly, like TikTok has really leaned away from AI generated content. Like it doesn't seem like it's something the platform is trying to like invest in significantly at all. And so I think, again, OpenAI took a step forward by saying, oh yeah, allowing people to put themselves and their friends in these scenarios, in these videos could be somewhat of an unlock, but it's very much TBD whether this will actually take off. Well, I will be following your recording and I want to try it myself. I'm not going to lie. I do. No, you should. You should. I know. Okay, Manisha, thank you so much for coming on today.

23:00Thanks for having me.

23:25On Tuesdays, join me to hear interviews with influential leaders in the industry like Calvin Klein, Daniel Roseberry, and Jonathan Anderson. On Thursdays, join head of editorial content at Vogue, Chloe Mal, and head of editorial content at British Vogue, Chomenadi, as they explore style and culture through the lens of fashion with guests like Martha Stewart, Kamala Harris, and Tracee Ellis Ross. The Run Through with Vogue, new episodes every Tuesday and Thursday, wherever you get your podcasts.

24:00That's our show for today. We'll link to all the stories we spoke about in the show notes. Make sure to check out Thursday's episode of Uncanny Ballet, which is about how delivery apps like DoorDash are betting on robots being the future of delivery. Adriana Tapia and Mark Leda produced this episode. Amar Lal at Macrosound mixed this episode. Fran Vandy is our New York studio engineer. Kate Osborne is our executive producer. Kondi Nass, Head of Global Audio, is Chris Bannon. And Katie Drummond is Wired's Global Editorial Director.

24:35150 years ago, they were hunting us down to kill us, and now they're hunting down immigrants to deport them. This is First America, the true story of how the United States came to be and how we got to this present moment. Listen to First America wherever you get your podcasts.