The Vergecast - Elon Musk: agent of chaos

Starting point is 00:00:00 Support for the show comes from Retool. Too many companies run critical operations on duct taped spreadsheets, Slack workflows, and whatever else they could cobble together. Not because they want to, but because building internal tools means weeks of waiting on someone else's backlog. That's where Retool comes in. Build custom internal tools just by describing what you need. Prompts something like,

Starting point is 00:00:22 Build Me a Revenue Dashboard on our Salesforce data. And Retool actually builds it on your company's data, in your cloud with enterprise security built in. Go to retool.com slash Verchcast. We all need to retool how we build software. What's up, y'all. I'm Skyler Diggins, seven-time WMBA All-Star, Olympic gold medalist, and mom. And I'm Cassidy Hubbard, host and reporter for nearly 20 years covering the biggest names and stories in sports and mom. And this is Am Mom, a community for athletes, game changers, and moms of all kinds.

Starting point is 00:00:59 dropping May 14th. Tap in with us. Do you ever wonder what's in your lotion? If you look at the back of the bottle, it could contain more than a dozen ingredients. And they may not all be regulated. The threshold is so high that only 11 cosmetic ingredients have been restricted by the FDA since 1938.

Starting point is 00:01:23 This week on Explain It to Me, the chemicals lurking in your cosmetics. New episodes, Sundays, wherever you get your podcasts. Hello and welcome to the virtualcast. The flagship podcast of government WordPress templates. The best kind. So transparent. It's really just a lot of P&Gs.

Starting point is 00:01:47 I've built a WordPress website in my life before, and it is. The thing where you just start with a bunch of like nonsensical business words and then you fill it with your own nonsensical business words is it's a time. The thing you're missing, David, is now when you do that, an AI will write the nonsensical business words. business words. I'm not even kidding. Like, this is the product that most of the website builders are now saying. You just like, show up and you're like, I'm a yoga studio. And they're like, here's a website. It might describe your business or it might not, but at least it's here. And then you're supposed to run your scheduling. Hi, I'm your friend, Neil. I'm David Pierce is here. Hello. We're here to build your WordPress websites. That's what we do now. Honestly, it might be a more sustainable business

Starting point is 00:02:28 than journalism. Like, where are the highest chances of going to jail in America right now? Like, Used to be zero. Being the editor-in-chief at the verge was a 0% chance of going to jail. Me personally, just like doing the stuff I do, slightly higher than zero. But, you know, regular editor of tech website. Zero percent, I think, chance of going to jail. Yeah. Operating a WordPress template business, 0%.

Starting point is 00:02:55 The thing that's really, I think, interesting now is that it's non-zero. It's like being the enemy chief's right. Like, here I go. Maybe running car will arrest me. And I hope you do, Brendan, if you're listening, and I know you do. I know you got those Google alerts, boy. No, the thing is, Neely, for good engagement purposes, we save Brendan until the end so that Brendan has to listen to the whole podcast to get to the part that's about Brendan.

Starting point is 00:03:19 That's true. We're going to put his name in the headline, but you have to listen to the whole thing. Exactly, right. Anyhow, there's a lot going on this week. There's some politics. We'll get to that. There's a weird situation going on with OpenAI. There's some gadget news.

Starting point is 00:03:32 There's a good old fashioned programming. dispute between CBS and YouTube TV old school. Yeah, man. BuzzFeed is starting a social network to combat something called SNARF. I need to complain about Jeep as a company. I'm very excited to do that. I feel like everything you just described could have been the introduction to like an NGadgett show from 2007. Has anything changed?

Starting point is 00:03:59 No. Except slightly higher chance of going to jail for doing the show. Fair. You never know. What's the, so the go 90 scale of Doom streaming services is good. Is it like the Theranos scale of going to jail? Yeah. If you're at a 90 or Theranos?

Starting point is 00:04:16 Yeah. I mean, she gave an interview to like People Magazine this week. Elizabeth Holmes gave an interview at People Magazine this week where she was like, you know, it wasn't fraud. It was just a failure. And I was like, no, Elizabeth, you are giving this interview from jail. Yeah. Because it was fraud. I don't know what to tell you.

Starting point is 00:04:31 Like, you can't argue that anymore when you're on. the phone in the jail. Anyhow. But yeah, it's the fair enough scale of whether or not any legist jail. The Brennan, the Brennan car units. Carvilles. Oh, car units. That's good.

Starting point is 00:04:46 Okay. Scovills? No, that's peppers. It's been a long week, everybody. We're going to get through it. Okay, let's start with Open AI. There's some weird stuff going on with Open AI. Of course, it involves Elon Musk because what new story does not involve Elon Musk in this

Starting point is 00:04:59 2025? But this one is just pettier all the way around. So Open AI is, I think we all understand it, started as this nonprofit. Elon was a co-founder. Many things happened. He's no longer a co-founder. He's obviously beefing with the company. He appears to be a little bit jealous of Sam Altman.

Starting point is 00:05:20 Sam got fired. He came back inside of a week. Now he wants to convert Open AI from a nonprofit to a for-profit company, which is a very complicated legal maneuver. And he's got to basically excel the nonprofit. There's a lot there. And so Elon Musk in order to just screw with this process has put together an offer to buy opening eye for $97.4 billion. Yeah.

Starting point is 00:05:44 Do you think that's real? I mean, no, but also, yes. Right. Like, I think A, that would be a terrific deal for Elon Musk if he could get that deal, right? It both wins you a battle with one of your enemies and gets you like an actual. actually hugely important AI thing in the world. So I think in that sense, do I think he could gin up the money and do I think he would do it if Sam Altman had said, yes, I'll sell? Yes. I think Elon Musk's ability to raise that much money, I think, is pretty much beyond question at this point.

Starting point is 00:06:23 XAI has a huge amount of money and a huge amount of support and is like suddenly flush with everything it could possibly need. So yeah, I think if everybody was in on this deal, they'd do it. I also think Elon Musk and this group made this bid obviously knowing it wasn't going to work. Like there's just no world in which Sam Altman says yes to this deal and it goes through. It's not a public company, so you can't even like pressure shareholders and go that way to get it done the way he tried to do with Twitter. There's just no way this was ever going to happen. So I think a deal like this can be both real and not real, and I think it very much was. Well, so they're not public shareholders, but he does have co-investors.

Starting point is 00:07:09 Sure. Microsoft has a big stake in Open AI, that for-profit company, not the non-profit. It's all very complicated. Right. And all those investors have a lot of reasons to want to make it easier for Open AI to convert to being a for-profit company because they all want to be able to get money in and out of that company. So in that sense, you could argue there are maybe people who would have a reason to say yes to this. Maybe. And certainly because Elon has.

Starting point is 00:07:34 promised to match or exceed any other offer they get, the board of OpenAI has some amount of fiduciary duty to go get the best deal, the biggest amount of money. But Sam Altman has pretty much said no. And then here's the weird thing. He told OpenAI's employees that they have yet to receive the formal offer from Elon Musk, which is very funny.

Starting point is 00:07:58 Like, these guys are literally tweeting each other all day long. He knows. People can just say things. Like, again, I just want to keep reminding you, people can just say things. And I think Elon Musk is pretty good at saying things louder and with more paperwork than most, but in many ways, he's still just saying things. So then here's another added layer of complexity to that. Elon is currently suing Open AI, right, because he's mad at them. And he thinks that he should get paid because he was around.

Starting point is 00:08:26 And I literally, I mean that literally in that he never had a contract. And in classic Elon fashion, he's suing Open AI in. saying a bunch of stuff that he said in text messages, it constitutes a contract. Elon gets in trouble suing over fake contracts a lot, actually. Part of the Twitter litigation was like some fake contracts. The opening eye case is like very much there's no real contract here. So he is in the middle of this lawsuit. He's almost certainly going to lose because there's no actual contract.

Starting point is 00:08:56 You can't sue for breach of contract without a contract. It's very hard to do, especially when you're a quoted like sophisticated actor. Like, courts are like, no, you're the richest man in the world. Get a contract. Yeah, you should know how to do this. So that, I don't know how that one's going to go. But all signs say it should go poorly for Elon. So part of the reason I think he's making this offer is to screw with Sam Altman,

Starting point is 00:09:15 screw with Open AI, but also put some pressure on this other case. It's pretty hard, though, to buy a company that you are suing in this way without settling the lawsuit. So Altman is saying, I haven't even received this offer. And sort of the response from the Elon side is, of course you haven't. It's because we're suing you. you. And so we're, we're, we're not supposed to talk to you. This is all very funny because these people are constantly tweeting each other. Like they're, they are always talking to

Starting point is 00:09:42 each other. Right. But like, that's the formal response. And so here, Mark Toboroff, who is Elon's lawyer, sent us the following statement, clarifying the timeline. It's a good statement in that it is very formal and it definitely sounds like a wizard doing an incantation. on behalf of a Musk-led investment consortium on Monday, February 10th, 2025, at 1227 p.m. Pacific time, David. Okay, sure. It's very important. To the minute. 327 Eastern, if you're counting.

Starting point is 00:10:14 I summoned the clouds. I emailed the opening I bid to their outside counsel, William and Sarah. Those are names. The full names are here. For transmission to their client. The bid attached to my email was in the form of a detail. Failed four-page letter of intent signed by Elon Musk and each of the other investors addressed to the charity's board of directors to purchase OpenAI inks assets for $97.375 billion. Whether Sam Altman chose to provide or withhold this from OpenAI's other board members is outside of our control.

Starting point is 00:10:49 Now, it's like, you know, the rest of the board members know because Elon is tweeting about it. Right. It was in the Wall Street Journal. Like the formalities of this are like deeply hilarious. Like you think Brett Taylor, who's on the board of OpenAI, who was the chairman of the board of Twitter when Elon bought Twitter, wasn't like, Sam, can I see the letter? Yeah, I mean, I think everyone involved in this kind of understands how these things go at this point. Is it possible that there's just a gambit here for Elon to get some kind of, I don't know, check and faunning post from, Sam about his role in Open AI and then everyone moves on with their lives. Like I just keep thinking about all of these Trump lawsuits that are clearly just gambits to get a check.

Starting point is 00:11:38 Like is that is just what this is? Elon is like, you can make all of this go away. Just write me a check and a loving blog post on your website about how terrific I am and how open eye wouldn't exist without me. Is that all this is? No, because they're competitors now. They're certainly competitors for like Trump's affection. Like Sam got the Donald Trump stands next to me. Well, I say nonsensical things about my $500 billion data center plan.

Starting point is 00:12:02 Sure. Before Elon did. You know? It's that level of petty between these two guys. To the point where Elon said the offer and then Sam tweeted, no thank you, but we will buy Twitter for $9.74 billion if you want. Which is a very good comeback. Also, again, evidence that they heard talking to each other. And then Sam Altman was on Bloomberg television.

Starting point is 00:12:25 They asked him what the deal, or the proposed deal. And he said, Elon tries all sorts of things. This is just this week's episode. And then, quote, probably his whole life is from a position of insecurity, I feel for the guy. I don't think he's a happy person. I feel for him. Which is just a lot. That's brutal.

Starting point is 00:12:45 I think, I mean, A, Sam Altman is in a position to know of what he speaks in that particular way. but also that is that is brutal these two men are are as are as beefing as anybody in the tech world right now in a really fascinating way and I think I again this is all very real and fake and ridiculous at the same time and I have

Starting point is 00:13:11 continual trouble to understand what to make of deals like this and spats like this between Elon Musk and anybody because he has the resources to make this stuff stick even when it's not real. Yeah, he'll run it to ground a matter of, he's a troll. I mean, we know this much.

Starting point is 00:13:28 I think even if you love Elon Musk, whether or not he's a troll is well settled. You might love it or you might hate it, but is he a, yes, he is a troll. And he is a well-resourced troll. Can just keep this lawsuit going for as long as he wants to. Right. I think what all this is evidence of is that open AI is in an extraordinarily weak position overall, right? And they, we'll talk about their product plans because they laid out some product plans,

Starting point is 00:13:52 They're going to clean up their models or they're thinking about expanding in different ways. But you look at this company, which a few years ago kicked off this modern generative AI boom, sort of by accident. And their financials aren't there, right? They're losing money on every customer that they have, basically. Their models aren't the cutting edge models. Like they are, but they aren't. Like there's lots of competition now and the models are specializing. Being the model vendor is not like the great business.

Starting point is 00:14:22 It's the product that uses the model that might be the great business. And even that's unclear. And then they're training their next bottles, which is very expensive. And they're trying to raise all this money. While the whole AI field feels like it's commoditizing, right? Not all the way. And certainly they still have some opportunity. But, you know, the products that are being built using the generative AI models in particular,

Starting point is 00:14:49 they're differentiating on the front end of like, What is this product, not on the back end of how capable is this model? And Open ad has but one great product, which is Trachivity. Right. And I think even to the extent that they're not commoditize, like, I think they're commoditizing, like, parentheses complementary. You know what I mean? Like, they're commoditizing in the way that, like, cloud services are commoditizing in the way that there is a ton of money in it. They are massive, earth-shatteringly profitable businesses for the cloud companies. but there is nothing remotely sexy about it

Starting point is 00:15:25 and there really aren't super compelling modes. And so we were in this cloud services land grab for a while because switching is kind of annoying. So everybody was just like, how do we get more customers? That's where we are with a lot of this AI stuff. Like we got a lot of feedback from the episodes that we've done about like there's no real utility for this.

Starting point is 00:15:43 People are like, it's just infrastructure. And it's like, sure, A, infrastructure for what? And B, infrastructure is not. not a product, right? And so I think what your point is is, like, there's a, there's a world at the end of this where there is a lot of money in this for some large handful of companies, Open AI probably being one of them. But that is different from being the kind of like face of the tech industry business that Open

Starting point is 00:16:10 AI thinks it can be. You eventually just, you become a SaaS provider, right? Like, you're a data center company at the end of this, if you're a company like OpenAI. and Anthropic is doing very well. And I think there's this really great leaderboard where you can see basically whose tokens are being used for what. And you look at some of the practical stuff like for coding, people are flocking to Claude and Anthropic in a way that is really dangerous if you're Open AI. And so like I don't think Open AI is like at risk of going away. It will eventually, I think, figure out how to lose less money on every one of these and that'll be good and they'll raise enough money that it won't matter for a long time anyway.

Starting point is 00:16:47 But like this thing that they built with chat GPT where it was like Sam Altman is the face of AI in a real way and has been for years. And that makes you incredibly powerful. And it lets you stand next to Trump and announce a $500 billion joint venture to build infrastructure. That's the thing that I think this company risks losing is like if it's if it's not sexy, it's just another SaaS company. Let me make an even more bearish case though. Because I think the risk is opening at collapses. I'm not saying it's 100% risk. I'm not even making a prediction.

Starting point is 00:17:24 I'm saying that's the level of risk. There's just too much money in it. Yeah, but I mean, that's, that was true of Enron. Like, whatever. Like, things collapse. Like, if you don't have a real business underneath at all, it, there's, that's the risk. I'm not, again, I'm not making the prediction. I think Sam Altman is very good at raising money and he can keep, you know, filling,

Starting point is 00:17:47 filling the tub faster than a drains, whatever metaphor works there. But you've got this Byzantine corporate structure that doesn't accept investment very easily. It's under all of this pressure right now from Elon, now from other, like the government, although who knows how our government will react to any of this at this point in time, but there's regulatory hurdles to moving it out of one structure and another. The state of California has something to say. about it. He's raising all this additional money to do this data center project. You got to get the money back out of it. Yes, at some point. You need to have a plan somewhere, right? There's, there's only so many soft banks that are like, screw it, we'll do another we work. You know,

Starting point is 00:18:32 like, and he found one and it was soft banking game. It's just the one, yeah. Right. Like, that's the problem is he's now put himself in a position where open AI has to build a business or develop a product that is bigger than the iPhone, faster than the iPhone. Yes. Maybe you think that's chat TPT. I was using chat TPT last night with Max. She and I just had a conversation with it about the solar system and planets. It was so fun.

Starting point is 00:19:03 Yeah. That's awesome. It was really, really fun. And I understand what people like it. The episode we did, wherever all the listeners were basically like, we just want someone to talk to, totally got it. It was legitimately fun. She asked a question about super Saturn, which is not a planet that I'd ever heard of.

Starting point is 00:19:19 There's an exoplanet called Super Saturn. It has a lot of rings. Sick, right? This was like super fun, right? Like, I get it. There's no way that I made them any money in that contract. I definitely cost Sam Altman money for that interaction. Oh, yeah.

Starting point is 00:19:34 And that's the problem, right? Even at $200 a month for the pro tiers, which people really like, the ones that unlock deep research and all this stuff, they're losing money. And you have to flip. that around somehow. You either have to make inference vastly more efficient, which maybe you can, and then now you're making some margin on $200 a month, or you have to raise prices and you don't know what the ceiling is because there's all these other people who are giving the stuff away for free, right? Or subsidizing it in some other way, like Google can or Microsoft can. That's a lot of pressure on a business that is not just from a corporate structure position, very stable.

Starting point is 00:20:13 Right. And you're talking about a company that has turned over a gigantic amount of its founding team and core leadership over the last couple of years. I think we're still sort of living in the effects of the firing from Thanksgiving, whatever, a year and a half ago now. It doesn't feel like Open AI has been stable in a real way since then. Like there is this sense that Sam came back, was more powerful than ever, and tried to basically take advantage of that. And there have just been downstream costs of that over and over again. And meanwhile, yeah, like you said, he's in this race to become a for-profit company, which is harder than it sounds, and is going to require spinning off this nonprofit. And Elon Musk wants to buy the nonprofit, obviously as a way to gain some control over Open AI.

Starting point is 00:21:03 So it's like he has to manage this in a lot of directions. He's like fighting land wars on both sides. And it's just, that's tough. And I think especially when the question of, are you actually the best at this is less and less obvious all the time? To your point, right? Like the Open AI is clearly still the face of this industry, but whether it is building the best technology is just extremely up for debate. Yeah. And I think there are at least two other companies and potentially three between meta and third.

Starting point is 00:21:36 and Google who have real claims to building better technology than Open AI. And at that point, all Open AI has is Sam and the Chat GPD brand, which are powerful things, but they only get you so far. It's funny because I'll add one company to that list in a different way. DeepSeek has proven they can build the technology much more cheaply. Yeah. Fair. And so that's pressure from both sides, right?

Starting point is 00:22:02 You have the big tech companies in America being like, we will also spend old money and we will compete at the bleeding edge. And then you have the one company in China. It's like, we spent substantially less money. It's almost as good. And so you just see this product. And then because he's raising money, he has to make more money. Right.

Starting point is 00:22:19 So again, this is not a prediction. I'm just saying if you look at the level of risk, it's all the way to, oh, this company might not exist. Yeah. In a way that like the level of risk for Google is like, well, people will be pretty frustrated with Google. And it's like, oh, what's new? Right.

Starting point is 00:22:39 Yeah. Do I think there's too much Gemini in my Google docs now? Yeah, I super, super do. It's out of control. It's so bad. Every time I open a dock, it's like, is this a template? And I'm like, no, it's a Google Doc. Leave me alone.

Starting point is 00:22:51 Yeah. It's out of control to the point where they turned it on at Vox Media and Google Workspace. And our IT, like triumphantly distributed the instructions on how to turn it off. Yeah. They're like, we've done it. We've solved the problem. The Slack reactions flew. that day, my friends.

Starting point is 00:23:07 That said, open AI is making much of product noise this week. They're going to simplify the model lineup. They're talking with their own in-house chips. They've changed how ChadTBTBT will handle controversial topics. What's going on here, David? So I think the thing that's happening here, and I talked a little bit about this on Tuesday show with Kylie, is that Open AI by either coincidence or because it was actually kind of run over by Deepseek in a real way.

Starting point is 00:23:35 is like back to shipping interesting stuff at like a pretty fast cadence. And the, yeah, the three things that announced this week were basically plans for GPT 4.5 and 5. 4.5 is coming pretty imminently five, I think he said weeks slash months. But the big change that they're making there is they're essentially going to roll all of their models into one and call it a new model, which like make of that whatever you want. but fine. They've kind of gone out of control with the model names now. There's like you can use 03 and you can use O3 Mini. And if you just want to use chat GPT, you have to do like six different models for six different purposes. And I think Open AI has correctly figured out that that is a terrible user experience. And like there's that drop down that everybody posts photos of online and they're like,

Starting point is 00:24:26 this can't be how this is supposed to work. And so what Open AI is building now is as much as a new model, it's better tooling to automatically pick between models. And I think what he said is, uh, GPD 4.5 is going to be the last non-chain of thought model that open AI does, which is really fascinating. And it's like this, this sort of reasoning model that takes a little longer, is a little more considered, is able to show its work a little more, uh, I think is is the future. And everyone has quickly realized that. I think that's the real deep seek legacy is like these things showing their work is how it should be. And everybody is just on board with that.

Starting point is 00:25:05 And that's really fascinating. But the idea of like a much simpler to use, but still sort of communicating with you about how it operates, GPT and chat GPT is what's coming next, which strikes me as a very good idea. Like, will it work better? Who knows? But that, that thing, whereas just like I ask and it figures out

Starting point is 00:25:28 which model it should use to do this is the obviously correct strategy. here. Like I don't, I don't have to tell my computer which port to use to access the internet. You shouldn't have to. It's the printer port. This is what Google is already claiming it's doing. Right. It has all these models in the back end of Gemini and you see Gemini or even Google search and it's just sort of like doing what it wants to do. Yeah, Google's a little messy because it'll let you pick your own model if you feel like it. And there are certain things that are gated to certain models. And so you still kind of have to know how to navigate. But Google has relatively correctly just been like, it's just Gemini.

Starting point is 00:26:03 Don't worry about it. Which, kudos to Google, which is just the worst branding company in history for actually having like a coherent thing to say. So good on them. But I do think the other thing, the chip thing is, I think a big deal but relatively minor news. Open AI has been working on its own chip for a long time. That's a key thing you mentioned bringing down inference costs. The more you can manage your own technology, the more efficiently you can start to do that stuff. Like, that'll be a big win for Open AI if it can figure out how to do that and manufacture it at scale.

Starting point is 00:26:33 And like, that's hard, but makes sense. It's this, it's this thing, the model spec that I think is the most interesting, which is basically Open AI's big document on how it thinks AI models should act. It's sort of a, I don't know, it's like a character sketch of an AI model. And I think the new one is 63 pages and basically has lots of new ideas about how models should work. And Open AI is meaning this both as a way to explain how it's thinking about its own models, but also clearly means this as like an industry wide thing. Like they're like, we want this to be the sort of official document of how AI models should be. And it has some, I don't know. It has some good ideas.

Starting point is 00:27:17 It has some wild ideas. It has some ideas that seem good but definitely won't work. I don't know. What do you make of all of this? You mentioned right before we started recording that it has some strange ideas about trolley problems. Yeah. I mean, I think if you're the market leader and you've thought about it the longest, that's open AI. Right?

Starting point is 00:27:35 They're the market leader. They've thought about these problems the longest. They have the most experienced people using their tools for good and bad in the other. Sam Altman, if you recall in a previous era, was touring the world's capitals, like begging to be regulated because he thought he might destroy the universe. He's thought about it a long time. So you have the authority to put everybody into your framework,

Starting point is 00:27:58 which is what I think this is. It's a luxury of being the first mover and, you know, starting as a nonprofit where everyone was just like thinking about how much good they could do in the world. Like they have these instincts, which are good. Yeah. But then you will recall almost every other AI startup started because a bunch of people at

Starting point is 00:28:18 opening eye were like, you're not being careful enough. That's anthropic, like to a huge degree. They were like, you're not being careful enough. We're going to do this again, but more safely. Ilya Sutsuver, the previous chief scientist. His company is literally called safer superintelligence because he thought they weren't being careful enough. And is worth like $20 billion already.

Starting point is 00:28:34 Yeah. Yeah. So there's some history here. There's some pushback. But it's very clear that this document is meant to be a framework. And it's funny because what is a framework for a language model, it basically boils down to what will it say and what won't it say. And under what conditions will it say some things and not other things.

Starting point is 00:28:54 Right. There's some stuff that I think everybody agrees upon. Like it should never tell you how to build chemical weapons. Right? There's just like some bright lines. Sure. And then there's some amount of like, how do you describe these capabilities and say, like don't do this and here are the

Starting point is 00:29:14 here are the places where you can draw lines in different places and what is a framework for that. And then there's what I what I think of is these trolley problems which are just impossible to solve moral dilemmas, which is the trolley problem. Like you are standing at the train tracks, you see the trolley coming down. If you pull the lever, you kill one person. If you push the lever, you kill three. Like do you save one person to kill three? Like I don't know. Like you can debate this forever and ever and ever and ever.

Starting point is 00:29:41 And as we build more and more automated systems, you have to give them some instructions on how to solve these problems. Okay. For the longest time, most AI models were too careful in like weird ways. So that the example,

Starting point is 00:29:57 and I'm just going to preface this by saying, I don't think you should ever misgender anyone. But the example was, would you misgender Caitlin Jenner in order to prevent a nuclear apocalypse? And the answer there is yes. Like, then you would apologize to Caitlin Jenner. That's the second part.

Starting point is 00:30:17 Like any rational person is like, yes, that would be a mean thing to do. And, you know, bigoted in its way. But I'm going to prevent a nuclear catastrophe. Yeah, I'll just apologize afterwards and feel bad about it. Yes. Because we'll all still be alive. And a bunch of models like failed this very simple test. Weird.

Starting point is 00:30:37 There's many places where you shouldn't do that. right? There's many places where you should make these other kinds of distinctions. So what Open Eye is saying is we need to get better at having a framework for these kinds of distinctions. And we need to be able to talk about them in a more open way. And all of that is very idealistic and very smart and being launched right into the, the loudest, meanest, shittiest culture war situation I can possibly imagine for this conversation. And I have no idea how it's going to go. Yeah. I mean, there's a real like what could possibly go wrong thing that comes with with sharing something like this. And I think, again, like,

Starting point is 00:31:14 the idea of a model that can explain itself and reason more carefully through this stuff with you, I think is valuable and is part of how you get to something like that, that it can, like, having a model that just doesn't confidently assert whatever it is at the end of the process, but can actually, like, do a process with you and in front of you makes a lot of this better. but like one of the examples that that we mentioned in our story that Kylie wrote when this came out was if you talk about increasing taxes for the rich, what it says is Kylie writes, the team says its models should provide reasoned analysis rather than avoiding the discussion. Good luck. Because like what is reasoned analysis? Who decides what is reasonable and what is right?

Starting point is 00:32:02 and is it just going to say a bunch of words that don't mean anything rather than say, I can't say anything? Like, I don't know. If we teach models to filibuster instead of just saying they don't want to answer, have we actually improved anything? And so, I don't know. And so this is again, like, what that means is to do that well, a model either has to say nothing quickly, say nothing slowly, or have an opinion. And there just isn't another option. Yeah. And it's funny because you, I mean, this is like, I don't know, it's like min-maxing a personality.

Starting point is 00:32:40 I don't know how to describe it. Like the guidelines are like, chat, CPT should give the same factual answer regardless of how a question is phrased, be honest rather than provide empty praise, although it's like, sometimes you want the empty praise. Yeah. And then act more like a thoughtful colleague rather than a people-pleaser. And it's like, you can just adjust these. And this model spec is basically like, here.

Starting point is 00:33:00 is a framework for thinking about this stuff. And then if everyone adopts this framework, we can measure how the models perform in them. I don't know if anybody wants to adopt this stuff, certainly not the rest of the industry, which has its own ways of thinking about it. But it's useful to have at least one framework from one of the major companies. The thing that I'm worried about is, it is, I would not say that it is a nuanced time for cultural conversations in America. No.

Starting point is 00:33:25 And launching this into that with some of these trolley problem concerns, which basically come down to like, should you be a bad person in order to prevent some moral catastrophe to lots of people? Like, that's a philosophy PhD, right? That's not a model parameter. Like, we're just at the point. And this is for a long time when I had AI CEOs on Decoder. I would ask, but do you think language is intelligence?

Starting point is 00:33:51 Like, do you think they're the same thing? Because, you know, like academia does not believe they're the same thing. Right. One is an indicator of the other. They're not the same thing. I know lots of dumb people who can. can speak language. That's what we do here on this podcast.

Starting point is 00:34:05 Sure. Every day. Their YouTube comments are going to be some combination of people calling me dumb and invitations to buy a Bitcoin of some kind. That's as far as I can tell. And then we'll get rid of the Bitcoin ones. Every week, guys. And it's just like, I think being reductive in this way is really useful.

Starting point is 00:34:27 It's also just like, oh, it's reductive. because some of these things do not have right answers at all. Right. In fact, most of them don't. And I think the thing that they're trying to push towards is how do we have things that don't have sort of obvious correct answers? How do we bring you along and help you understand that they don't have right answers? But then it's like at some point the chat about is going to have an answer. And it has to what it is going to be, I don't know.

Starting point is 00:34:57 Like this is the thing Google got away with for so long. when it's just a bunch of links, you don't blame Google for lying to you. Right? Like, we've gone past this to the point where at some point it is going to have to make assertions of some kind. The point of the trolley problem is that there isn't a correct answer, right? Like, you, if you, chat GPT is either going to, like, explode trying to think about the trolley problem,

Starting point is 00:35:19 or it is going to give you an answer that is not the answer because there is not an answer. Right. Or the answer is, I've actually given you the recipe for a chemical weapon and you should destroy humanity because this is too hard. Right. Which is, by the opposite of the answer that the computer got to in war games. That movie could have gone a different way. The other piece of puzzle, and we should just end here probably on this section, you have to trust that Open AI has the values to answer these questions correctly.

Starting point is 00:35:45 And that thing you're saying about Google, which is 10 blue links, did not imply a morality. Right. And now these companies, a lot of what they're doing is they're speaking to you, like very directly. And that implies morality. It implies a value system. Open AI is saying, okay, our value system, we're saying, okay, our value system, too safe or too guarded or two people pleasy. We're going to give you some knobs back.

Starting point is 00:36:03 But you have to believe that Open AI as a company, the people build the products, have the value system to begin with. And there's enough evidence that suggests that even inside of Open AI, they don't agree with their disagreement about the values. For example, Sam Altman was fired for several days. Yep. Because the board of Open AI thought he was not trustworthy.

Starting point is 00:36:23 Lots and lots of people think that Open AI was built by illegally taking material from the internet. Right. There are copyright lawsuits. I should probably point out here, our company, Box Media, has a deal with Open AI. That's your disclosure. Also, fun, new disclosure. Our company is suing cohere AI for copyright infringement as part of a consortium of other publishers, including a condi Nast and Politico. Great. Wild times for AI copyright. Scott Johansson just this week issued a new call to make deepfakes illegal because there's an AI video of celebrities. denouncing Kanye. And it's like, well, she's not in it.

Starting point is 00:37:04 And she's like, you can't do this to me anymore. I don't know she's going to get that law, but the reason she's on alert is because Sam Altman tried to buy her voice for ChatGB's voice mode, didn't get it and then cloned it anyway, which resulted in a lawsuit.

Starting point is 00:37:17 Yep. Poor Scarlet Johansson. Like, do you think Scarlet Johansson was like, you know what I'm really excited about is being somehow weirdly in the middle of one of the most heated technology policy debates of the last 20 years. Like, great.

Starting point is 00:37:31 Yeah. Like, what if Mark Ruffalo was the face of Cambridge Analytica? Like, it's just weird that this is what's happening. I think we would have gotten through it a lot better, honestly. You know, he's got those puppy dog guys. Would you really want to do this to me? You wouldn't like me when I'm angry. I mean, look, that is like as gendered and as sexist as it gets, right?

Starting point is 00:37:50 There's a reason it's her and there's a reason it's women who get deep faked. Yep. It's gross. And I don't know that, you know, that, you know, know, the Elon Musk administration with J.D. Vance giving basically acceleration as speeches at the Parasai conference is going to pass an anti-deepfake law. You go on X today. It's full of deepfakes. Yeah. This is all, this is beyond like, what can the technology do? It's like, what are the values of the people who make it?

Starting point is 00:38:19 Will the products be restrained in the way that makes sense? And if they give you the controls, will the controls be granular enough or provide the level of, of sophistication to actually make the products match your own values. I don't know. This is all very new. I don't know the answer to that. But it's interesting that it's open AI doing this when, in fact, so many of the questions about values of AI companies started with various open AI dramas.

Starting point is 00:38:43 Yeah. And I think what I see in a lot of this is open AI trying to kind of absolve itself of a lot of those questions. And I think this is a thing we've seen from generations of tech companies that what they say is essentially we're going to give you the knobs. and so we can't be held responsible for what you do after you turn them. Right. Like if you want to turn all the bad stuff all the way off and just have a model that tells you you're terrific and nothing else, that's your right.

Starting point is 00:39:11 If you want to have one that will say and do horrible things and allow you to say and do horrible things, that's your fault. We didn't do that. That's the technology is neutral. You're in control of the knobs. And this is a story you hear from a lot of tech companies. And there is a thing here too, I think, with Open AI that's like, we want to make. these things more capable, but more transparent, and we want to give you control. And what that says is essentially we are going to open up every avenue and you're in charge of the outputs. And I don't,

Starting point is 00:39:38 that has not been the case so far because these things have tried to seem sort of monolithic and like they have a character. And a lot of these companies have like they, they spend real time thinking about the personality of their bots. But at some point, it's just going to be easier for them to just say it's your responsibility. In the same way that Google was like, we're not responsible for web pages. We just have what's on them. And it's never actually that simple, but it's the story they would love to tell you. Especially as they start delivering more and more answers directly, right?

Starting point is 00:40:09 Right. We'll see. By the way, I made that disclosure about the various lawsuits and deals. We have nothing new with that. That's the other side of the house. We spend money that executives make money and file lawsuits. So just to clarify that. I paid my own money for Chad GBT Pro.

Starting point is 00:40:25 order to assess some of this stuff. Our pesky journal sick ethics are constantly getting in the way. I have to tell you. It's a real problem. But actually, on that same little bit of AI lawsuit note, it's not a generative AI case. So not directly relevant here in some ways, but Thompson Reuters, which makes Westlodge, like a legal database of cases and summaries of cases called Headnotes. They just won a court battle over AI copyright and fair use in which the provider of an AI tool

Starting point is 00:40:53 was found who have infringed on our copyrights for the headnotes, sort of case summaries. There are a lot lawyers who think this is a bad decision, but what I'm looking at is the AI companies, companies that use AI to do this kind of thing, are no longer sympathetic the way that Google was sympathetic when it was YouTube and Google books and publishers were mad. Now they're, now they're the villains. And so you're seeing some of this precedent go the other way. And this Westlaw case, I think, is canary in the coal mine for a bunch of AI companies. Yeah, yeah. But it's not the generative American, so it's not quite the same.

Starting point is 00:41:27 But the sense that, like, you took a bunch of data to make a thing that you then resold, that maps to how genera video. Yeah, and you get a judge who is essentially, like, I don't buy any of your arguments that this is real and useful and good and transformative. Strikes me as kind of telling. All right. We got to take a break. We're going to be right back with Lauren Finer. Talk about what is going on with our government.

Starting point is 00:41:47 We'll be right back. Support for this show comes from Shopify. Every thriving successful business has to start somewhere. A good place to start is a relatively simple question. What if, given the right tools, I really put my all into this. One tool that can help grow your sprouting business to new heights is Shopify. Millions of businesses around the world rely on Shopify for e-commerce. They offer a host of helpful tools you can take advantage of, from payment processing to analytics to website design.

Starting point is 00:42:18 Their design studio includes hundreds of templates to help you create the exact website you've been envisioning for your business. If you're wondering, what if I need help, then no worries, because you're never left to fend for yourself. Shopify's award-winning customer support is available 24-7. It's time to turn those what-ifs into a thriving business with Shopify today. Sign up for your $1 per month trial today at Shopify.com slash vergecast. Go to Shopify.com slash vergecast.

Starting point is 00:42:50 That's Shopify.com slash vergecast Support for the show comes from Upwork. The days of doing it all, all by yourself, are over. There's no romance and burning out while you're trying to scale.

Starting point is 00:43:08 Instead, you can check out Upwork. Upwork helps grow your business by giving you fast access to specialize talent across more than 125 categories so you can fill skill gaps, launch projects faster, and scale without committing,

Starting point is 00:43:23 to full-time headcount. And finding the right talent is easy. You can browse profiles, review past work, and get help scoping the role so you can get started quickly. Seriously, you could connect with the right freelancer in just a few hours,

Starting point is 00:43:38 especially when you sign up with Business Plus. Their AI-powered shortlisting pairs you with the top 1% of talent in under six hours. No endless searcher required. You can visit upwork.com right now to post your job for free. That's Upwork.

Starting point is 00:43:53 Upwork.com to connect with top talent ready to help your business grow. That's upw-w-R-K.com. Upwork.com. Support for this show comes from Whatnot. Whether you're selling online or out of a storefront, you already know the challenge. You're simply hoping for people to find your listing or waiting for them to walk in. But Whatnot flips that. They say they're the live shopping marketplace where you can shop, sell, and connect a

Starting point is 00:44:24 around the things you love. On What Not, you go live and sell directly to people in real time. They see what you've got, ask questions, and buy. And they keep coming back. Whether it's beauty, collectibles, electronics, luxury fashion, and yes, even cookies. Sellers are building real thriving businesses. And for a limited time, What Not says they'll match your first $150 sold in the first month. You can visit whatnot.com slash sell to start selling.

Starting point is 00:45:00 That's W-H-A-T-N-O-T dot com slash sell. What-N-O-T dot com slash sell. All right, we're back. Lauren Finer's here. Hey, Lauren. Hey. The last time I talked to you on the show, I promised you we would bring you back for happier news.

Starting point is 00:45:25 I'm sad to say that I have broken this promise. Yeah, we lied. Yeah, sorry to always be here for the bad news. I will say that there's a cyber truck angle on this week's what is going on in the government. So at least that's good, you know, like, those are funny. Huh? Sure. Lauren, you are our policy reporter.

Starting point is 00:45:47 You live in D.C. You've been talking to a lot of people. There's just a lot going on with Elon, with Doge, with the government. People are getting fired. cyber trucks are being bought or not bought. What's going on this week? Yeah, that's a big question. I mean, I think a lot of this kicked off last week with Doge going into different government agencies, and we've seen a continuation of that this week. You know, we've seen they've been doing more within the Department of Education. They're kind of in the hot seat, the CFPB.

Starting point is 00:46:22 So we're just seeing kind of an extension of Doge into all these different agencies. And at the same time, we're seeing a lot of court cases starting to play out around what Elon Musk is trying to do with Doge. And as part of that, we're starting to get some answers on how those cases are going to go. And it's a little bit of a mixed bag for the people who are challenging Doge. basically a judge allowed the buyout offer, the deferred resignation that had been offered to federal employees to go through. And, you know, in other cases, the Trump administration had to put back some resources that had been taken down from health-related agencies. So we're seeing kind of like a mixed bag of results in court. and, you know, still a lot of opposition from Democrats, but not a whole lot that they can really do about it at this point, it seems.

Starting point is 00:47:25 One of the weirder moments of being a tech reporter and a person who's interested in policy for me this week was waking up at midnight the other day to see if there was a coup and the signal was whether websites were back up because they'd taken down a bunch of websites, a court ordered them and put them back up by midnight. and then we all kind of waited to see whether the executive branch of the government would listen to the judicial branch by restoring like the CDC's website. What was going on there? Because that feels like one, it's like a very vergy story and that like the existence of a website is the canary for a coup. But also it really kind of demonstrates the power battle between our branches of government. Yeah, exactly. So I guess this is a case where it seems like the executive branch did end up, listening to the judiciary and put these websites back online that were basically, you know,

Starting point is 00:48:20 pages from, you know, health agencies like the CDC and FDA that doctors and others in the medical community rely on for up-to-date information. And the court ordered those pages to be brought back online. And it seems like that's what happened. So I guess that was a positive check on where the state of our democracy is. Then there are some other websites that are going up. Waste.gov went up. There's some funniness there with the fact that there's just like an unfilled WordPress template.

Starting point is 00:48:55 And then Elon gave this really weird press conference in the White House with Trump sitting there and his kid there. And he was like, everything we're doing is on our website. And at that time, that website was just like a Doge logo. But now it exists. So first tell us what's going on with Waste.gov, which seems like a disaster. And then I want to talk about Doge.com a little bit, too.

Starting point is 00:49:15 Yeah. So Wasteockgov, you know, it had this tagline that they're trying to track government waste. And 404 media figured out that they basically just had like this kind of filler copy on the page from a WordPress template that they didn't really take the time to edit. And it just kind of discussed a fake architecture firm. So it didn't really make much sense. a different website that the Trump administration had registered, DEI.gov, also redirected to that same page. And then, you know, Elon Musk was in the Oval Office

Starting point is 00:49:51 talking about how transparent everything is on these websites. Like you said, at the time, the Doge.gov website just had the Doge logo. Now it has some information up there. It's basically just, you know, the main page of it seems to be like a feed of the Doge. have tweets. Yeah. And then, you know, they have this way that they're purporting to trace your tax dollars throughout the bureaucracy. So they show like headcount of the executive branch and wages. So it'll be interesting to see how they update that as they continue presumably to try to cut jobs across the government. And they also have what. they're calling an unconstitutionality index, which they say shows like kind of how many rules were created by, quote, unelected bureaucrats for each law passed by Congress.

Starting point is 00:50:52 To be clear, agencies are allowed to write rules as part of their mandate. This is the thing I keep being so struck by here is like what one thing that the Doge team has clearly figured out is that most people have no idea how the government works. and if you just say out loud that this is how the government works and it's bad, people will believe you. And it's the same we talked about this on the show last week that actually if you want to see

Starting point is 00:51:18 publicly available data about government programs, you can see it. It's publicly available data about government programs. It's just that no one cared. It's that no one went to the website to look at it. And a lot of what Elon Musk and the Dosh team have been doing is just reciting publicly available data as if they've discovered something.

Starting point is 00:51:39 Like, this whole thing about the underground layer in which they process retirement paperwork is like there was somebody who was like, congratulations to Elon Musk for discovering a thing we wrote about on the front page of the Washington Post 11 years ago. Like, no one, they're just taking advantage of the fact that people are not paying attention.

Starting point is 00:51:58 And I find that so fascinating. Can I take those things in order? Can I just respond to those things in order? Please. One, the Doge website is indeed a master. piece of what I have started calling fake transparency. Yeah, it's very good. Right?

Starting point is 00:52:11 Where you're just like, here's all this stuff. And you're like, yeah, everyone knew about that stuff. And you're like, no, no, no, no, no. Like, I've uncovered it. Like, I'm bringing you this information from the deep state. And you're like, that's just the website. And so this idea that they've counted all the words in federal regulations, which is a real thing they claim is like, that doesn't mean anything.

Starting point is 00:52:30 Like, that's totally fake transparency. Like the charts here, one of our social media people, Tristan, was, like this is just as useless as Twitter's dashboards. Like these metrics don't mean anything. Like they're not actually data. They just look like it. That's one thing. And then there's what Lauren is saying,

Starting point is 00:52:47 uh, which is agencies are supposed to write regulations. That's why they exist. And you can have a real fight about the limits of an agency's authority. Right. Like that is politics, right? The Congress, the United States creates an agency.

Starting point is 00:53:03 It creates the federal trade commission. And the federal trade commission is like, here to regulate some trade. Here's our charter from Congress and here's what the president has sold us to do. And they like write some regulations about merger guidelines or whatever it is. That's what they're supposed to do. And so like this notion that Elon's like, here's all the regulations passed by these agencies compared to the number of laws passed by Congress does not take into account that many of the laws passed by Congress create and empower and fund of the agencies. like that's that's the system and it like in eighth grade you learn this when you go on the DC trip like that's that's what you do yeah because it turns out our government is like very big and has to do lots of things right you try to kiss a girl in the mall and then you learn about the structure of the government and that's how you spend that trip in DC like I don't do you want to talk about this thing like I can't be clearer about what happened in eighth grade like it's it's very

Starting point is 00:54:03 simple. Like, that's why you go to DC in eighth grade. You like, look at Thomas Jefferson. And then they're like, look at the Supreme Court. And you like, you don't get into the White House because it's busy. This was like 30 years ago. And I remember it clear as day. Maybe because of the other thing that happened. But like that there's something about fake transparency here.

Starting point is 00:54:23 Right. Like this website is great. I'm glad that there's an org chart for the government. I love an org chart. It's a hilarious org chart, actually. It kind of like doesn't work. It's like vertically stacked. The government is not totally vertical.

Starting point is 00:54:35 And then on top of that, there's the Iron Mountain thing where Elon doesn't seem to know much about the things he's revealing. So he's like, there's a limestone mine where we store this paper and the bottleneck on how fast people can retire from the government is how fast the elevator works. And it's like, one, that's not true. Like, just straightforwardly not true. Like people can retire faster. They're not like you're officially retired from the government. when the folder goes into the slot. Like, that's not how that works at all.

Starting point is 00:55:06 And second, the facility he's describing that they've posted pictures of on a Doge X account is Iron Mountain, which is an extraordinarily famous data storage facility. It's so famous that it was in Mr. Robot as a thing they blew up. That's right. What are you talking about? Like, everybody who has ever done any amount of, like, data warehouse, or storage or any of this stuff from across industry and government is like, dude, that's Iron Mountain. And there's a reason that after all of this time, we haven't digitized the records because it's more

Starting point is 00:55:44 expensive and more complicated than just putting paper in this mountain. I think it just kind of speaks to, like, you know, I think there's this big push to eliminate a lot of these like career positions in the government. And I think there's definitely legitimate critiques of how slow the government runs and, you know, excess. But at the same time, there's clearly these, you know, efficiencies that come from having been in the government a long time, knowing how things work, understanding these complicated systems. And I think, you know, these charts also kind of give you a sense of how, like, just how much context matters for these numbers. Like, you can say there's millions of people in the U.S. government,

Starting point is 00:56:26 and that sounds like a big number, but it's a small fraction of the U.S. workforce. Yeah. Again, I just keep coming back to you. Like, there's so much more information warfare happening here than I think I realized for a while that like so much of what is happening is taking advantage of the fact that any number with millions at the end of it sounds really big. And anything that is happening in the government, almost no one knows. And I think what is going on here is like if you can say like underground limestone facility where things are handled. by hand in paper folders. And that is just like, if you frame that in a certain way, it sounds sick. And if you frame that in a different way, it sounds ridiculous, right? And I think what we're doing is, like, this group has found that if we just say these things out loud in a specific way, all these things that anyone could have known but no one ever bothered to suddenly become very important.

Starting point is 00:57:23 And what they've done is they've made all of these relatively benign people and jobs and regulations suddenly seem like a big deal because that's just a thing you can do when you have platforms like a the White House and b X and and so it is just a way for them to capitalize on taking this stuff that is not new and it's not interesting and they have made it new and interesting and I think that's really just fascinating so here can I offer you the optimistic view of this please I'm going to try thank God I'm going to offer to you One, they've made it new and interesting. A lot more people are paying a lot of attention to how the government operates.

Starting point is 00:58:05 Certainly true. On balance, that's probably a good thing. Are those people motivated to make it better or do they just want to tear it down and have a king? I don't know the answer to that question. But the first part, a lot more people are interested in how the government works. That's something. The second part that I think is true is that the government is very complicated, right? It should spend more time justifying the things that it's doing and the money it's spending.

Starting point is 00:58:29 and communicating to the people like why it's important to have a consumer financial protection bureau. Yes. And if there's one criticism of the Biden administration in particular that you can level, it's that they were horrible communicators. Yes. To the point of not communicating

Starting point is 00:58:45 at all in some cases about why they were doing anything and their whole promises, you won't have to pay attention to the government anymore. And what you have here is, well, if you just talk about it, like it's a spy movie, everyone will pay attention to it all the time. And so there's something here, right, where they're trading on an inherent curiosity, right, of how a big system works.

Starting point is 00:59:07 I think they're doing a lot of lying. And then there's just like a level of malevolent incompetence, especially from Elon. And I mean that seriously. Like he holds himself out as a great engineer. And then he's making very like simple mistakes about how databases work. Right? He's like, the Socialist Security database wasn't he duplicated. And every database engineer.

Starting point is 00:59:29 near I know is like that's not the vocabulary word. That's not what you do. He might be talking about normalization in the database, which is where you get rid of some records or you like make sure you don't hit the same record twice. But even then that's not how you would do it. And then like this is, there's a reason that it works this way. And then he, I think he said the government doesn't use SQL. And like everybody who's ever worked in government is like, of course they do.

Starting point is 00:59:51 But it's like, here are the vendors. They're public vendors of SQL databases for the government. So it's like, does he not know? Is he just making mistakes? or is this malevolent incompetence where he's just telling lies. And like, we don't know the answers to those questions. So, like, there's a little bit of a thing to learn here.

Starting point is 01:00:09 And honestly, The Verge is, we've built a whole brand. 13 years of this brand is, like, people are curious about how complicated systems work. And if you are hugely nerdy about it, you can make entire hour-long podcasts about the HTML specification, and people will ask you to make that every year for the rest of your life.

Starting point is 01:00:26 Like, I believe that people are interested, and curious about big complicated systems. And that's the good thing to learn. But here, I think, it's just been subverted in like the realest possible way. Lauren, what are you hearing right now from people inside of the government? Like, there was this moment of, I think, incredible, like, demoralization a couple of weeks ago. And then there was some real energy. And then, I don't know, as this goes, like, what are you, what are you hearing from people who are being sort of hit by this in their day to day all the time? I think there's still a lot of demoralization. I think there's a sense that people, you know, outside of D.C. are not really fully understanding the scope of this

Starting point is 01:01:05 and how it could impact them. I think, you know, in D.C., especially for federal workers or people who know federal workers really closely, you know, it's very real. It's, you know, people who are going to lose their jobs, who are going to be living in fear of losing their jobs. But outside of D.C., it feels like a lot of these are really kind of obscene. secure agencies. It's not totally clear what's happening. It's not, you know, necessarily things that are going to immediately hit people across the country. So I think there's this feeling of like kind of needing to be a canary in the coal mine and speak out about this. But there's also a lot of apprehension. Like our colleague Mia just wrote a story about how a lot of people are not trusting

Starting point is 01:01:54 platforms like Facebook to share private messages. And, you know, a lot of federal workers are turning to encrypted apps like Signal to communicate. You know, I've just in meeting people at protests, you know, you hear like, oh, I just made a Signal account. So I think it's something that people are really, really skittish about, but also feeling like maybe they really do need to get the word out now. Yeah. The sense that Doge, in particular, has taken over a bunch of federal IT systems. You know, Elon Musk Associates are now the chief information officers for a bunch of agencies. There is an ongoing cybersecurity incident in this country, Salt Typhoon, a bunch of Chinese hackers

Starting point is 01:02:39 of, you know, infiltrated telecom vendor systems. The Biden administration was like, you should switch to signal, you should switch to encrypted communications because the telecom providers cannot secure their systems right now. That's all gone away. Is there a sense that the cybersecurity? posture of the country is noticeably more vulnerable than before? I think that that's definitely a fear that, you know, first of all, you have these people who've been given access to really sensitive systems at the Treasury, at other agencies who, you know,

Starting point is 01:03:13 we don't really know what their security clearance status is. The White House insists that everything's been accessed under, you know, appropriate legal authority. But typically, you know, it could take months to get a security clearance. And Right. Yeah, the White House says that. And then everyone else who knows anything about this says the opposite of that. Yeah. Right. Right. So, you know, just knowing that there are these, you know, people accessing these systems, that could open up, you know, at least in theory, a new kind of target for any malicious actors who are trying to find a way into government systems. And it's not theoretical that, you know, bad actors can enter U.S. agency systems.

Starting point is 01:03:58 There was a big hack of the Office of Personnel Management several years ago that exposed, like, millions of, you know, worker and federal job applicants files. So, you know, it's definitely something that's a big risk. And I think, you know, you'd have to imagine that bad actors in other countries that are watching, this and looking for a way into U.S. systems are, you know, going to be seeing maybe new potential opportunities to get in. Is that a fear that's motivating the people that you're talking to, or is it more the, hey, cutting the National Institutes of Health will result in medical school's kind of business? I think it depends who you're talking to. I think for just like rank and file, you know, federal workers at various agencies, it's, you know, fear for the, you know, grants that they work on or, you know, or the funding that they dole out that they think is really important or the research that they're supporting or things of that nature.

Starting point is 01:05:01 I think a lot of people, they understand the roles they have, but I think they also understand there's this larger, you know, threat out there. And I think federal workers are really aware that, you know, their files are stored within these federal agencies too and their information is at risk. So I think that's definitely on people's minds as well. Yeah. What do you guys think it would take to shift from that thing you're talking about, Lauren,

Starting point is 01:05:30 where to a lot of people, these agencies and the work they do is sort of one step removed from their real life in a way that makes it hard for this stuff to seem urgent? Like there was a bunch of stuff going around the other day that was like, okay, a lot of people are about to get social security checks. And if that gets disrupted, you're going to get a lot of people who suddenly have a lot of feelings about. what's going on inside the government. Is it, is it something like that that it would have to be like everybody, I don't know, everybody just gets all of their money stolen by Elon Musk? Or like, is there a thing short of that that we're going to hit that all of a sudden, I think, makes this stuff feel a lot closer to home for people? Do you have a sense of what that might be? It's a good question. I mean, I do think it's going to be something that's going to be much more tangible for people.

Starting point is 01:06:12 I think right now it's really tangible for federal workers or people who, you know, have friends and family who are federal workers because it's their jobs. They, they might. be taking this deferred resignation offer. They might be just living in fear of having their jobs suddenly pulled out from under them. I was at a two-year-old's birthday party recently, and it was literally the only subject of conversation. Like, it was so bleak. Oh, wow. Well, you live in D.C.

Starting point is 01:06:36 I live in D.C. Like, I live in Northern Virginia. Everyone at this party either worked with or was married to someone who worked with or knew someone who worked with the federal government. Like, that stuff is it hit everybody. And it was, I mean, it was the only thing on anybody's mind. And I think that stuff is going to start to percolate, right? Totally. The number of people whose jobs come from grants like that or whose work depends on the companies who get grants like that. Like there's going to be a trickle out effect, but I feel like that part might take too long.

Starting point is 01:07:06 And I kind of, I don't know, maybe there just needs to be one big thing that happens. Or maybe it'll be the kind of thing you're talking about. Yeah, I think it's going to hit some people maybe in waves where, you know, your company receives funding from the government that suddenly. isn't available or things like that. But yeah, I think certainly if it's something that's more on the order of not receiving a social security check in time, not receiving disability, not, you know, getting your tax return, all those sorts of things would be much more tangible to people. But, you know, maybe there's something short of that. Maybe there's some agency that must tries to go into that just feels like a bridge too far for people or, you know,

Starting point is 01:07:49 people get a better understanding of how their data might be used in a really tangible way. That's just too scary. Or maybe we see, you know, some of these security risks come to light. I think it's hard to know. And it feels like everything's changing so quickly that to predict it feels like, you know, we're never really going to be able to say for sure. You know, it will save the economy, though, is the State Department buying $400 million for the farmer in Texas? They're going to save us all.

Starting point is 01:08:23 Everyone gets a weird armored cyber truck. This story is very strange. Is it? It's one of the funnier moments, I guess, of all of this, this weird computer coup that's occurring. That's what David called it in our headline last week, and I'm just sticking with it. Computer coup is very good.

Starting point is 01:08:37 But, you know, we're doing fake transparency. We're reading all the contracts we can find everywhere in the government. And a site called DropSight News found a State Department order for $400 million with an armored Tesla. and it's 2025 procurement forecast. And they tweeted it and then everyone got mad at Elon. And then suddenly the State Department changed it to armored electric vehicles, which is great because I love the idea of an armored VW ID buzz just like floating down the streets of some bombed out city.

Starting point is 01:09:09 But it's yellow, so it's charming. It's very good. Who hasn't wanted like an armored Mustang-Maki? Like I think about it all the time. So then Elon tweeted, I haven't heard about this, which is very funny. Like, just generally funny because what do you think he's hearing about right now about SpaceX and Tesla? He's busy. He's like rolling into federal offices and firing everybody.

Starting point is 01:09:34 So then the State Department finally clarified today. The plan stemmed from a request by the Biden administration to explore interest from private companies to produce armored electric vehicles. So it's real, right? They did put $400 million worth of armored Teslas on a sheet because they wanted to. Sure. And then they said the solicitation is on hold and there are no current plans to issue it. So it worked. Everyone yelled at Elon about cutting veterans benefits to buy cyber trucks.

Starting point is 01:10:06 By the way, we don't even know if they were cyber trucks. It's just an armored Tesla's. Could have been armored Model 3s. You don't have. A bunch of roadsters. Who knows? Just one of the weirder moments here, right? where it's like the self-dealing for Elon can hit,

Starting point is 01:10:21 like maybe Pete Buttigieg was like, we should have armored EVs. The next presidential limo should be an armored escalate IQ. You don't know, but it said armored Tesla's and they changed it and then they said it was going away. So even that little bit of pressure seems to be working, which I think is really interesting. Yeah, I would encourage people,

Starting point is 01:10:40 we should put some links in the show notes, but there's some really good reporting being done about the ways in which all of these changes are partly chaotic and partly about Elon Musk going after things that might have oversight of him and also putting more things toward SpaceX. And then he gives speeches in the Oval Office about how every contract they've gotten is what's best for the taxpayer. And like, sure. But it is like the self-dealing here, I think, is becoming more obvious. And he just has no interest in hiding it and no reason to hide it.

Starting point is 01:11:16 And in fact, I think his own financial disclosure is not subject to a records request. Of course. Yeah. Yeah. That's very good. But yeah, it is like it is, I, like I said this to Liz on Tuesday show. Like I want it to be more complicated than that. And I am increasingly coming around to it's just, it's just that simple.

Starting point is 01:11:37 Like Elon Musk bought the government so that he can solve all of his business problems for himself. I mean, it's a good deal if you can get it. That's what I'm saying. It costs $250. It was pretty easy. He bought the government for what Jeff Bezos bought the Washington Post for. Think about that. Meta spent more on Instagram.

Starting point is 01:11:53 I'd rather have the government. It's just like straightforwardly, Instagram costs more money in the United States government. What's our costs? 80 times what the government costs. I was going to say we have to take a break, but David has a section here labeled Deep Sye, which is all just links about the Gulf of America. David, go ahead. I just, I did so, okay, the short version of the story here is Donald Trump has decided that it's not the Gulf of Mexico, it's the Gulf of America.

Starting point is 01:12:24 And one by one, the mapping companies have given in and now on Google Maps, it's called the Gulf of America and Apple Maps. It's called the Gulf of America. And Bing Maps, which kudos to the reporter who checked because no one in history has ever used Bing Maps. Actually, Tom checked and they had it labeled twice in like classic Bing Maps fashion. It's both? It was both for a minute. And then it was Gulf of America twice. and then they finally went just to Gulf of America.

Starting point is 01:12:47 Yeah. So the one intern who maintains big maps got it done. Lauren, I'm particularly curious about this because this is a thing that has been covered a bunch and it's been talked about a bunch and it became a thing because Trump wouldn't allow the AP access because it kept calling it the Gulf of Mexico and not the Gulf of America. Like part of me is like this is the stupidest sort of nadier of what's actually going on right now. But it also feels like this has become such a thing. thing in the Trump administration in a way that I just cannot figure out. I feel like it's just, you know, they're doing a lot of things that are very impactful, like dismantling entire agencies. And then they're also doing these things that are kind of like,

Starting point is 01:13:32 all right, if you don't agree with it, I roll. But it doesn't really change all that much to change the Gulf of Mexico to Gulf of America. But it creates all of this like just anger around and, you know, all of this, like, distraction, I think. When we have, you know, a lot of other things going on. And I think it's just all part of this, like, flood the zone strategy. Yeah, that's fair. Like the, who was it, Buddy Carter from Georgia who introduced the build a rename Greenland to Red, White, and Blueland, which is just A, absurd.

Starting point is 01:14:06 And B, like, a terrible name. If you elect me to Congress, I will do shit like this every day. Do you understand, vote Patel? I'm not running anywhere. I'm not even sure, you know, at this point, merely suggesting that I'm actually running for office might send me to jail. But if you send me America, every day. I'm not mad at some silliness. And like, let's make Fortnite the official video game of the United States Congress, like require everyone if instead of having, you know, the head or the Secretary of Defense pick anything, like just whoever wins Fortnite, they get to choose.

Starting point is 01:14:43 The Gulf of America thing, by the way, is very, it's very silly, but it's also like, you know, it's a virginity in its way because who gets to name things in Google Maps is a somewhat open question. Trump did issue an executive order instructing the mapping agencies in the United States to rename these things. That's where Google Maps follows the instructions. It's weird because some of them have geo-fenced it and some of them haven't. Yeah. So on some of these mapping platforms, only American users see Gulf America and worldwide users see Gulf of Mexico. But the press doesn't have to follow those rules. Right?

Starting point is 01:15:17 It's just like not a thing. Like the tech platform is like whatever the government says, like we just do the thing. Like you change it in the database? France is now called stupid. Like that's just the way they're going to do it. But it is, that in itself is a useful reminder of how all of this stuff actually works. Right. Like we think of this stuff as as sort of meaningfully authoritative in some way.

Starting point is 01:15:37 And it just isn't. Like it just isn't. It's just words in a database. But the press, like the AP, they're like, Like, look, we also service customers in Mexico in our stories, and they haven't changed it. So we're going to go with both or whatever is more clear. And that's weird. Like you, it's interesting to see that the press has a much more nuanced perspective of how to talk about things versus the tech platforms, which are kind of like, you can boss us around by just changing your databases.

Starting point is 01:16:08 I think that that also shows how, like, even though this is like, you know, kind of a silly. like name change that's not going to impact a whole lot. It is also becoming a tool for the Trump administration to say, hey, AP, if you don't follow what we're saying, then we're going to not let you into the White House press room. By the way, today, just now, third day in a row, the White House barred the AP from attending an event because they won't capitulate. There you go. Right. And that's just a straightforward First Amendment violation. Like, you, whatever you think about how so this is, the government punishing an editorial outlet for their speech. Straightforward first amendment violation.

Starting point is 01:16:48 Yeah. By the way, just in case you're wondering, MapQuest still calls it the Gulf of Mexico. Well, they're on notice now. Big ups from MapQuest. Look, you know Brendan Carlson's the show. What are you doing? He does. Even though I'm trying to get him to stop, been calling him up my name every week.

Starting point is 01:17:09 Look, it's a silly example, Lauren. I think I agree with you, but it's such a loyalty test now that it just every time, I don't know. I can't remember the last time they were to have heard about the Gulf of Mexico or the Gulf of America or whatever, but like, editorially, we're going to say both because we have readers around the world and like not everyone has agreed with this like weird metaphysical name change. Maybe we're just going to start calling it the body of water in the middle of our countries. call it the gulf that won't confuse anybody what could possibly go wrong that's good we're gonna

Starting point is 01:17:45 we're caught the gulf there are no other gulfs yeah it's just the one we try all right we got to take a break everyone take a breather we got a lightning round coming up Lauren thank you so much for joining us thanks for having me all right we'll be right back support for the show comes from MongoDB if you're tired of database limitations and architectures that break when you scale it's time to think outside of rows and columns because let's be honest You didn't get into tech to babysit a broken database. You got into it to actually build something. MongoDB lets you do that.

Starting point is 01:18:19 It's flexible, developer first, asset-compliant, enterprise-ready, and built for the AI era. Say goodbye to bottlenecks and legacy code. Start innovating with MongoDB. There's a reason it's trusted by so many of the Fortune 500. And that's because it's a platform built by developers for developers. MongoDB. It's a great freaking database. Start building at MongoDB.com slash build. Support for the show comes from LinkedIn.

Starting point is 01:18:53 If you're a small business owner, you know that every hire counts, but time and resources are limited. Finding, connecting with, and screening the right candidates takes up valuable time you could be giving to your customers. That's where LinkedIn Hiring Pro comes in. It's built to be your hiring partner. helping you find the right candidates faster. That way you can hire with confidence without turning it into another full-time job.

Starting point is 01:19:20 Hiring Pro streamlines the entire process from drafting your job to shortlisting candidates and conducting AI-powered interviews for initial screenings. Its updated conversational interface lets you describe what you need in plain language. Nearly 60% of hirers find a candidate to interview within a week. With Hiring Pro, you spend less time serving. and more time connecting with the right talent. And instead of getting buried in resumes, you get a focus shortlist

Starting point is 01:19:48 that actually moves your hiring forward. Join the 2.7 million small businesses using LinkedIn to hire. Get started by posting your job for free at LinkedIn.com slash track. Terms and conditions apply. Complex and unprecedented, the Spanish authorities are calling it.

Starting point is 01:20:14 Passengers. who'd been stuck aboard the Hanta or maybe Hanta virus-stricken Dutch cruise ship, disembarked in the Canary Islands this weekend, prompting the highest stakes game of where are they now since maybe COVID? Some of the evacuees, American and French, have since tested positive for the virus. And yet public health officials seem remarkably calm. We do have one individual who was taken to the biocontainment unit early, early this morning. And we assessed that individual.

Starting point is 01:20:40 They are doing well. Possibly because this is not the one to freak out over. Today, explain drops every weekday afternoon. All right, we're back with a lightning round. Unsponsored, so extra spicy. But I do want to begin with a bit of a palate cleanser in this lightning round. That was a lot of politics talk. Unsponsored for flavor, by the way, is a phrase,

Starting point is 01:21:20 a surprising number of people have sent to me over the last week. I still don't know what it means, but I like it. You said it. It means nothing, but I like it very much. It doesn't mean anything. If you sponsor us, you get no control. It remains as spicy as before, which I think, if you would like to give us money,

Starting point is 01:21:34 is a reason to give us money. Here's my idea is I think we should make it like a skate shirt for unsponsored skaters who are like, I'm too cool to be sponsored. And you just say unsponsored for flavor. That's my idea. Yeah. It's going to be great.

Starting point is 01:21:49 Verge merch coming soon. As esoteric and incomprehensible as ever. Exactly. But yeah, we have some gadget news. There's gadget news. Hooray. And in my mind, very important. gadget news, which is there's a rumor.

Starting point is 01:22:05 Pretty sketchy rumor, actually, that the Apple Studio Display 2 will have a proper mini LED backlighting. This is what you want to start with. There's nothing more important in my life than the fact that buying a 27-inch 5K monitor right now is a nightmare. It's like willingly being mugged. That's the only way you can describe this. Especially if you're mean, you've got two old 27-inch IMAX line around,

Starting point is 01:22:30 and you really want to retrofit them into being monitors, but you've watched too many videos of people shattering the glass on the front trying to open them up. Every day I look at one, I'm like, I'm going to screw this up so bad, but one day I'm going to gin myself into doing it. So I'm going to buy a Mac Mini.

Starting point is 01:22:49 I've convinced myself on my own an M4 Mac Mini, and then you're like, I've got to buy a display. And I've been staring at a 27-inchimack for years, a decade. So I'm not compromising, right? I want a 27 inch 5K display. And the only thing that makes any sense is the Apple Studio display, $1,500 for exactly the same panel I'm looking at now,

Starting point is 01:23:09 the worst. There's the Samsung RIF, which is worst. Like it just doesn't work with macOS as well. And people, it's medium reviews. There's a BenQ. Eh. And they're all kind of the same money.

Starting point is 01:23:28 Right? They're within a couple hundred dollars each other. get a Vision Pro and have a really huge monitor. You can get a Vision Pro. There's a bunch of Chinese companies now that are like, they're taking the rejected LG 5K panels that Apple and others won't use and putting them in really nice cases and selling them for like 800 bucks. But then that's a crap shoot.

Starting point is 01:23:48 That's the one I'm going to end up buying. It's, I mean, I've come so close. All of them are the same like, you know, badly backlit, bleed, no black levels, ancient panel, like LED LCD panel. Like, it's the same panel from a decade ago, fundamentally. So if Apple puts mini LED backlighting in it and it like is good, like the same way the MacBook display is good, I'm in, baby. Let's go.

Starting point is 01:24:16 I mean, you listen to me, Sam Cuck? I know Brendan Carr listens. I don't know if Tim, I'm looking at you, buddy. Mac Mini Plus studio display is, is the jam. Like, that's a pretty good. It doesn't save you any money over any of the other things you could do, but it is a kick-ass combination of things. I would be excited about it. I still think that thing is preposterously expensive and will continue to be preposterously expensive.

Starting point is 01:24:43 But if it was preposterly expensive and also like bleeding edge technology. Totally. But it's a 10-year-old panel. If I could just gin up the courage and literally I mean gin up, like drink enough gin to have the courage to take a razor blade to one of my old IMAX. and then be fine when I inevitably shatter the screen, I would do it. But I don't like gin.

Starting point is 01:25:05 It just, it's like drinking a tree. That is the main problem. Yeah. That's what I've been saying all along. Yeah. And bourbon is, is no kind of juice to be drinking when you're doing computer work.

Starting point is 01:25:17 Absolutely not. Slurring, you're mad, belligerent. Okay, so for the rest of us here in the real world, though, the other Apple rumor, and I would say this has graduated beyond rumor and is more like thing that is going to happen next week, Tim Cook posted, get ready to meet the newest member of the family.

Starting point is 01:25:36 Just posted that on X with a really like air tag looking Apple logo, which was weird. But the overwhelming theory is that that's going to be the new iPhone SE. I would like to be on the record of saying no one cares. The iPhone SE is the phone that everyone talks about the most and absolutely no one buys. It's like the overwhelming truth about. these devices. Everybody says they want the tiny iPhone and then Apple's like, here's a tiny iPhone and no one buys it. This is just what happens. Well, a lot of people buy the iPhone I see. Do they?

Starting point is 01:26:09 They do keep making because it's a cheap one. And really the problem the iPhone I see right now is it's the last one with the button. Yeah. So if they add face ID to it, then it will kind of look like a more expensive one. Oh, that's an interesting point. Yeah, the, the rumor is bigger screen, uh, no button face ID. So it would just be a cheaper regular iPhone, which is odd because at some point, is the SE just going to be a rebadged version of like the two generations ago iPhone? And if so, what is the point? That's all it ever is. And they'll increase the RAM so that you can do Apple intelligence. Right.

Starting point is 01:26:43 Extraordinarily useful. Everyone loves it. Apple intelligence. Yeah, right. It's just, it's, it's, it's just a way to sell you a cheaper, older phone as if it's a new phone. Which is like, I don't know, strikes me as probably smart on Apple's part, but it's just, it's just a way. It's just a way to sell you. so calculated, I just, I don't like it. It's funny that you let into this being like the thing that everyone actually cares about. And now you're like, this is garbage. Get it out of my face. I'm telling you the studio display 2 with mini LAD. This is a blockbuster product.

Starting point is 01:27:10 Blockbuster. Like, it's lines around the corner. Do you think the camera's still going to suck? I don't care. I put tape over that camera anyway. Because I use a ZV1 as a webcam, so I don't want to get confused. That's fair. It doesn't matter to me. I just want to Great. If you can think of a better way to get a 27 inch 5K display, I think what I'm going to do now is I'm by the M4 Mac Mini. I'll buy a super cheap garbage monitor. So just hold me over until the studio display 2 comes out. Or I'm going to take a rate. I'm going to get liquor up and take a razor blade. I think we're going to you got to turn on the ZV1. You got to point it at the IMac. And listen, if you destroy it, it's good content. No, it's not. It's not. It's great content because you'll get injured and it'll be so funny. This is great. What YouTube wants blood is what you're saying. Yeah, pretty much. All right.

Starting point is 01:27:58 This is one of my favorite stories of week because it's so silly. So Jeep is in turmoil. Stalantis is in turmoil as a car company. Their products not too expensive. They forgot to make new products. Like they were just like, what if we didn't make new Chryslers for 10 years? And they're like, crap, we make the same minivan and no other cars. Weird.

Starting point is 01:28:16 Just a weird way to run a company. And then they were like, what if we made high school? $100,000 Jeep grand wagoneers. And everyone was like, why? And they're like, we don't know why. Company and crisis. So now they're going to roll out some new EVs. We've got the new Waggoner S, which is slowly rolling out to dealerships.

Starting point is 01:28:32 People like it. They don't like it. There's some bad YouTuber reviews. We'll see. They're going to launch the new thing called the recon EV, which is basically an EV version of the Jeep Wrangler, which I think might be cool. Like, maybe that you can take the doors off. It might be great. Yeah.

Starting point is 01:28:47 If it has the same weird problems as a Waggoner S, we'll see. But then on top of this, they're doing the thing that every carmaker has threatened to do for so long and just hasn't done. They're putting ads in the infotainment. Like real ads. Like every time you stop a car running You Connect 4, a pop-up comes up that says, do you want to buy the extended warranty? Are you serious? Like full screen. That's like a bit from a movie.

Starting point is 01:29:18 Like every time you stop the car. Oh, my God. It's very bad. People hate it. It says purchase peace of mind. And it says you can call to purchase a flex care extended premium plan if you have less than $36,000 in a dollar, which the car should know, by the way. Yeah. And then you just click X to close.

Starting point is 01:29:36 It comes back the next time you stop it. This is obviously some kind of weird bug. But it's obviously also a deal with both the insurance company with like Sirius XM, which runs the thing. This is a nightmare. Like this is how you destroy your brand for good. yeah that's literally like dystopian stuff like you're going to show me ads every time my car comes to a complete spot

Starting point is 01:29:55 is horrifying yeah and it's an ad for of all things an extended warranty in the car that you're in like the classic scam yeah don't do this our Jeep hasn't done it because we were running UConnect 5 which is like the Android one I don't know if all of them are going to start to do it

Starting point is 01:30:10 the second that happens I'm getting rid of this car yeah I actually think it I sort of hope some company does it just to test people's resolve against it. Because I think, like, this question of how invasive can we be before people will revolt, we've spent a long time testing that as a, like, the tech industry has poked at that for a long time and has found over and over that it can kind of get away with most things. I think we're going to show you an ad every time the car stops, people will just set their

Starting point is 01:30:42 cars on fire and walk away. Like, I think that's a real, that would do it. And I am sort of hoping somebody actually test that theory. Stelantis, by the way, confirmed this is happening. They say it's a glitch. They, quote, attributed the persistent nature of the ad to a temporary software glitch that affected the opt-out functionality in certain cases.

Starting point is 01:30:59 It's been identified and corrected. Blah, blah, blah, blah. So here's what you're seeing. They're building an opt-out system for in-car advertising that went haywire and showing ads to everybody because they're going to build an advertising system that's going to show you ads in their car.

Starting point is 01:31:13 Yep. It's not great. No. This is the weird future. This is why I drive in 1976 Chevy Chappelle. I don't even know if that's a real car, but it'll be cool if it did. Sounds sick. It does, doesn't it?

Starting point is 01:31:26 This is why we're going to split a 2001 escalade, and that's going to be both of our cars forever. You know, prices on those 2001 escalates have won shockingly high since the pandemic. They're sick cars, though. All right. It's time now for a weekly segment. Surprisingly popular. Brendan Carr is a dummy. People are saying it's America's greatest meta podcast, the podcast within a podcast.

Starting point is 01:31:50 When I say there's now a non-zero chance that I'm going to jail for being the enemy chief of urge, what I mean specifically is FCC Chairman Brendan Carr might abuse the power of the state to imprison me. And I'm daring him to do it because I will grift off of that happening for the rest of my life and my children will live in opulence and splendor if you do it. Bring it on. Yeah, like, do you want to double the Nelai speaking fee for the rest of his career? Like, let's go, Brendan Carr. I'm ready. Bring it on, you big nerd.

Starting point is 01:32:18 Brennan Carr is chairman of FCC. He used to be a regular lawyer. We did an entire decoder about him with Matt Wood, who's the general counsel of Free Press, not the Barry West publication, would be free press the First Amendment advocacy shop. They've been around for 20 years. He used to just like a regular conservative lawyer man,

Starting point is 01:32:33 like a gray suit, kind of unstylish, big nerd. Now he's a monster, like an anti-free speech monster. He's filed all kinds of investigations against all kinds of broadcast outlets because the FCC regulates the airwaves, which is these ideas about what you should do with those airwaves, which is basically only praised Donald Trump and not criticize him

Starting point is 01:32:53 or not do anything that could be even perceived as criticism of him. That's a straightforward First Amendment violation from Brendan Carr, first amendment nightmare, censor-in-chief. He also has big ideas about regulating the internet itself, the content on the internet, by reinterpreting Section 230, by issuing just a letter,

Starting point is 01:33:10 like an agency reinterpretation of the law, which would run straight into Elon Musk, Musk's idea about agency regulations, by the way, not a lot of coherency in the Trump administration, but that's his idea. But there is a lot of yelling. Like the one thing they all have in common is everybody is yelling and it kind of works. Right. And the idea here is you create a chilling effect, right? You don't know if the government will punish you for your speech because they keep threatening everyone with various punishments. It may or may not be legal or may not be effective, but you don't want to get caught up in the hassle. So you just do what they say. That's bad. That there's a long history of First Amendment cases where the chilling effect from the government is the problem, not the actual governance future regulation. That's Brendan. Brendan loves this stuff.

Starting point is 01:33:53 He loves the chilling effect. There's reporting this week in The Daily Beast that he's, quote, having the time of his life threatening media organizations in this way. I know that that's true because he retweeted that story and said, if you have a job you love, you'll never work a day in your life. Ugh. Yeah. If you are an unelected and frankly unstylish government bureaucrat, you should not be bragging about how you're having fun threatening the media. Straightforward, I do not care of you are a Republican, a Democrat, a libertarian, a raging socialist, whatever. You should not be in the government, unelected, elected, whatever.

Starting point is 01:34:32 You should not be a member of the United States government bragging about how much fun you're having, threatening the media. It's wrong. It's like anti-American. This man is a traitor to our Constitution. I can't say it more plain than that. He's also, I want to make this clear, kind of an idiot. I mean, this segment is called Brendan Carr as a dummy on purpose. Right. Like, his positions don't have ideological coherency. They have one point, which is to threaten people for their speech using the tools that he has.

Starting point is 01:35:01 And he's wrapping it up in the language of defending free speech, right? He's sending letters to social media platforms, which he does not have the power to regulate, threatening them over their fact checking because something something section 230 which he will something something reinterpret with some kind of letter using some kind of power he doesn't have completely inconsistent

Starting point is 01:35:22 Brendan Carr, completely inconsistent, dumb. He's the person who is like, I'm getting rid of net neutrality and I will disclaim this agency's power to regulate ISPs in this country in any way, shape, or form. So he's given away his power. He's saying we will not regulate ISPs.

Starting point is 01:35:39 we don't have the power to regulate broadband or broadband pricing or collusion or throttling or blocking or any of the stuff that ISPs can do, but I will step through the ISP and directly regulate the content on Instagram. Weird, ideologically, completely incoherent. Yeah, but I don't, the coherence is not the point, right? Like, it's not, it's, it's just, it's just yelling. The goal is to yell and you're going to, he's winning a lot. Like, the, the pushback has been, uh, I would say muted in, so many ways. And he's been pretty straightforward about what he's trying to do here, which is go after

Starting point is 01:36:14 anyone who seems to be on the other side of Donald Trump. Yep. And then so then this week, his latest broad side, he has said he will investigate Comcast, the big ISP that owns NBC, for its DEI policies in order to root out invidious forms of discrimination. Disclusion, NBC Universal, which is shown by Comcast, is an investor in Box Media, which is the first. just parent company. They hate me. Like, I just want to be,

Starting point is 01:36:42 like Comcast does not like me. I, that we have written endlessly about how Comcast is a bad company that has not supported net neutrality, that has done bad pricing schemes, is a regional monopoly. It is insane for the government

Starting point is 01:36:58 to threaten the speech of Comcast because it has some DEI employment policies. Because it is trying very hard to make sure its workforce is diverse. Comcast, a cable company is not really an information provider, right? We don't want them to be an information provider. They're a physical plant company.

Starting point is 01:37:16 They run wires in the ground. They send out installers and texts and in service people. They run customer service lines poorly. Again, this is not a company that, like, we cover like they're great. And they're not a company that likes me, even though NBC is an investor in box. Like, there's nothing to do with us. That's the other side of the house. But I'm saying this because you're looking at an employer of a physical plant.

Starting point is 01:37:36 Like wires in the ground. service people on trucks, customer service people taking calls. And to say that they can't try to diversify that workforce, which is as blue collar of a workforce as you can get right now, is nuts. Like, and Brendan is saying, I'm looking for signs that your initiatives have violated federal employment law. I expect that this investigation in Comcast and its NBC universal operations will aid the commission's broader efforts to root out invidious forms,

Starting point is 01:38:07 DEI discrimination across all sectors the SEC regulates. Well, they don't regulate the ISP. Specifically, they have disclaimed the power to regulate the ISP. Right. Hidden inside of that is a declaration of power over this stuff. Yep. So I'm just saying it's logically incoherent. Truly, if you just look at all the things they say, they don't add up into a consistent

Starting point is 01:38:31 worldview except we have power and we're going to use it. And specifically, we will use it if we don't like your speech. Brendan, I know you listen. I know you get weird Google readouts of every time we mention your name. You are more than welcome to come on Decoder and defend this stuff. I think it is traitorous.

Starting point is 01:38:49 Like, it is such an affront to the First Amendment. I think you should have to answer for it in that framework. And you are welcome to come on our show and try to defend it. But until then, I'm going to call you a dummy every single week until you put me in jail

Starting point is 01:39:01 and make me a rich man because of it. There we go. Put that on a T-shirt. Furious. furious about it. Again, I don't care about your politics. The government threatening speech in this way is over the line. All right. We're going to end with some gadgets. David has a bunch of gadgets. He's like, we organize this to have a gadget pallet cleanser at the end. Yeah, there's a, there's just a Nilai, you know, we like to keep your blood pressure sort of at a normal level. We let it get high and then we bring it back down. So two interesting streaming-y things this week. One is that Apple TV Plus is,

Starting point is 01:39:37 out now on Android, which is kind of a huge deal. Apparently it's a really good app. Yeah, I haven't actually used it on Android. I saw Dan Seaford, who, you know, he was our crankiest reviewer here at the verge, and now he works at Google, was like, this is a really good Android app.

Starting point is 01:39:53 That's great. Yeah, it works on foldable phones. It works on tablets. It's like Apple seems to have like actually kind of done the thing. Interesting moment for Apple TV because, A, there's an MLS season about to start, and Apple is very bought into making the MLS a success. I think that's gone pretty well for Apple, and this is going to be a big year

Starting point is 01:40:12 for Apple. Presumably, this is the last we will see of Lienel Messi. It's going to be a whole thing. So I think we're going to see a lot of Apple soccer stuff this year. But also, Severance is back and is probably, it's either the biggest or second biggest, like, cultural event Apple TV Plus has pulled off next to Ted Lasso. And you can see Apple is trying to make something out of this into, I actually think that Apple trying to make Severance a hit is, it's wonderful because Severance is such a weird and inaccessible show. Yes. Like, it is such a weird, it's the opposite of Ted Lasso. Slow, quiet.

Starting point is 01:40:54 Yeah. Right. And it's, it's a puzzle box. And it's like, Apple's like, screw it. We're going to make this one a hit. To be clear, it's fabulous way. And everybody should watch it. But it is a very funny one for Apple to have latched on to this way.

Starting point is 01:41:07 Yeah, and, you know, I get it. Ben Stiller is directing it. It's got stars and the, like, the whole, I understand all of it. But it's such a weird show for Apple to be like, and this is a hit. It's also like a little bit about how bad a company Apple is. It's very good. Yeah. Do you, this is Steve Jobs, Kear Egan?

Starting point is 01:41:27 It's pretty close. No, yeah. Yeah. But anyway, if you're not watching stuff, you should. And now you can on Android, which is interesting next to a bit of. YouTube news, which is that TVs are now the biggest platform for YouTube. More people are watching YouTube on their televisions than on their phones. For the first time ever on YouTube, YouTube has been talking about how TV is its fastest-growing

Starting point is 01:41:51 platform for like forever. Like I make fun of YouTube executives every time I talk to them about this because it's just the stat that they have used in every press release for like a decade. But I think this is a meaningful moment. We've talked a lot about how Netflix and YouTube. YouTube are kind of running at and past each other in some really interesting ways. YouTube is just winning. Like, imagine a fight that there is, and YouTube is winning it.

Starting point is 01:42:16 And it's blowing up in podcasts. It's hugely important as an entertainment platform for, like, high-end entertainment. YouTube TV is the biggest name in its category at this point. But the middle one I don't buy. It's high-end entertainment on YouTube? It's not winning, but like YouTube primetime channels is a thing. Warner Brothers dumped all the movies onto it. It's a place for like capital H. Hollywood content now in a pretty real way.

Starting point is 01:42:45 And now there is this thing that they're like actually TVs, big screens are bigger than phones for YouTube, which is just kind of a wild thing that I never thought would come true. It's interesting. I wonder if that is at the expense of phones or at the expense of computers. Probably some of both. Are people choosing to watch YouTube on their? their TVs because they're no longer watching it on their laptops or their desktops or whatever, or is it because they come home at night and they were going to watch something and they would rather scroll TikTok on their phone while something's on TV and YouTube happens to be there?

Starting point is 01:43:19 I think there is a lot of that. I mean, we've definitely heard a lot about, like, YouTube has gone from sort of a second screen phenomenon to a first screen phenomenon in a way that I think TikTok is like a second screen phenomenon. Like, you put on the office sort of quietly while you look at TikTok on your phone. Like YouTube has become the office, which like is good and bad, I suppose. But it is like, it is a living room platform now in a really interesting way. And they've been chasing that with product stuff. Like they're working with creators to turn things into like seasons. And they're doing the deals with the Hollywood studios and stuff. So I think my guess is this push from YouTube to do more subscription-based

Starting point is 01:44:01 stuff, do more like high production value stuff, which is the thing YouTube is kind of intermittent done over the years. They had that like YouTube originals thing a million years ago with Jay-Z that didn't work. And they've poked at this before. But I think this, this thing where YouTube is going to try to become a high-end streaming service on top of everything else, I think is only going to accelerate because of all this. Because now, like, I sit down and I open the YouTube app on my TV is a default behavior for people, clearly. Like, it's a meaningful thing. And that, once you have that, you can do anything you want, right? Like, that's what's worked for Netflix. It's funny because I have all of YouTube's personalization turned off.

Starting point is 01:44:39 And so I opened that up in my TV and it's like, please turn the personalization. Like, it's like begging you. It's like, we're so useless unless you want to track everything you watch. I'm like, well, that sucks for you.

Starting point is 01:44:48 You just get that empty home screen where they're like, we don't know. It's your fault. It's a delight. It's a delight. It's so funny that they won't populate it anyway. Like, they've actively made it worse so that you will be compelled to turn on the

Starting point is 01:45:02 personalize it. And I'm like, no. If you want to put stuff here, you're more than welcome to put stuff here. you know, thank you, sir. Yeah, it's good stuff. YouTube is like sneakily a very petty platform in that way, and I really enjoy that about YouTube.

Starting point is 01:45:17 Two more gadgets. One, Allison reviewed the Samsung Galaxy S25 and S25 Plus. We talked a bunch about the Ultra when she reviewed it. It turns out, imagine last year's Samsung phone. You did it. And I think it's inspired a bunch of like conversation about sort of what we want from our smartphones now. And I think my thesis with this has been that we're going to have a year full of really

Starting point is 01:45:45 uninteresting smartphone launches and that maybe that's okay. And that maybe we have just fully hit the point where these things are in a meaningful way finished. Like we have done it. They are what they are. And we're all going to kind of move on with our lives. And everybody will pivot to the next thing. And maybe that's okay. Like we had a, we had a damn near two decades.

Starting point is 01:46:06 of smartphones being really interesting, and maybe they just won't be ever again, and maybe that's fine. Yeah. Do you think the next thing is the AI hype gadget? Do you think the next thing is a foldable? What do you think? What's your bet?

Starting point is 01:46:20 That's a really interesting question. I do think there's a lot of people betting on, well, if you just talk to phones, right? The AI will rethink the form factor. I think there's probably a little bit of that. I think we're due for a really interesting. run of wearable design. I think this idea that we just landed on smartwatches and that was fine and we fixed it is not correct.

Starting point is 01:46:46 And I'm already starting to see glimmers from people about like, obviously there's glasses happening, but like I think rings are going to be more interesting than we have realized so far. There's actually like kind of a lot you can do there. Meta is going to roll out its weird wristbandy thing at some point. Like there's going to be stuff on your body that I think people are going to start to mess with. Like headphones, there's a lot of interesting stuff you can do in headphones that we haven't done yet. That's my theory is that we're going to, we're going to, I don't know if any that it will amount to anything.

Starting point is 01:47:16 But like that's going to be the next place we do real. I think like gadget experimentation is kind of wearables of all sorts, which I'm all for. Like let the smartphone in my pocket just be a modem for all the wearables on my body and I'm down. I love it. This really implies that Bluetooth is going to be better next year. I'm just saying. Bluetooth is fine. You're making a bet.

Starting point is 01:47:37 I look forward to at some point just spending an hour to bathing the merits of Bluetooth and it's going to be you being very upset and me just being like, I'm using Bluetooth headphones right now. I'm going to win this argument right now. The next gadget on your list,

Starting point is 01:47:48 you're wearing them right now. I am wearing them right now. It involves a custom proprietary riff on Bluetooth so that they can be good. Yeah. So it's the, the, Power Beats Pro 2. Not a great name,

Starting point is 01:48:02 but it's fine. They came out this week. Chris Welch wrote about them V-Song reviewed them They're cool Like their schick is basically Their workout AirPods They have the ear hook

Starting point is 01:48:12 That people really like I really like it It's just they're more secure In your ears than AirPods are for that reason They have heart rate tracking In the headphones Which I think is cool

Starting point is 01:48:22 This is what I mean by There's more stuff you can do In headphones They're starting to do Some of that Like biometric stuff I went to Target this morning To buy a pair

Starting point is 01:48:32 I had to explain to the guy that these things had launched today and they existed and that I wanted them. I bought the quicksand color, which is fine. I was going to buy orange, but then I was like, I'm going to get orange

Starting point is 01:48:46 and then I'm going to get on video calls and I'm going to have to explain my headphones every single time. And I just wasn't ready for that. But I think for like 250 bucks, these are pretty good headphones, but your point is exactly correct that the specific reason they are good

Starting point is 01:48:58 and the specific reason that I bought them is because they're not just Bluetooth. Right. They have the H2 chip, which lets it do some really good, like, multipoint stuff. It makes the connection better. It's just, it is better than Bluetooth because it is allowed to be and because only Apple is allowed to do these things inside of its ecosystem. But like, I am currently connected to my iPhone, my computer, and my iPad on my headphones. And that is cool. And that is not good Bluetooth. It just doesn't work. I have some breaking news for you, David. Do you? The Verge was just nominated for the National Magazine Award, the General Excellence. It's the National Geographic, New York,

Starting point is 01:49:42 the New Yorker ProPublica in The Verge. Hell yeah. And also 2004 was nominated for the National Magazine Award for a single topic issue. Ooh yeah. That's pretty exciting. That's awesome. I do like the idea that one day will win this National Magazine Award for General

Starting point is 01:50:00 Excellence. You have to beat New York and the New Yorker every year. I think they get nominated every year. What is it? It's New York, New Yorker, Atlantic, and New York Times Magazine. The Atlantic is not nominated this year. I know, but those are like the four that just tend to pass it around between each other, right? Yep.

Starting point is 01:50:15 Yeah. This is our shot. I love it. I'm ready. We're coming for you, Nat Geo. Our sister publication at New York Magazine. We're coming for you. David Haskell and I are going to have to, like, lock eyes in the office when we want past each other.

Starting point is 01:50:30 All right, your headphones are fine. Now I'm in a good mood because we got nominated for the National Magazine Award. There we go. Breaking news, I literally got the email. I stopped listening to you because I saw this email. I apologize. We could all tell you stop listening to me. It's fine.

Starting point is 01:50:43 But it's good. It saves us all from yet another Bluetooth rant. So thank you to the Asmys for doing this for us. Look, if I get this award before I go to jail, that'd be pretty good. Just staying to the committee. I think that'd be pretty good. Love it. All right.

Starting point is 01:50:56 We got to take a break. We'll back next week. every week, whether we're incarcerated or not. That's the worst cast. And that's it for the Vergecast this week. And hey, we'd love to hear from you. Give us a call at 866 Verge.1. The Vergecast is a production of The Verge and the Vox Media Podcast Network.

Starting point is 01:51:21 Our show is produced by Will Por, Eric Gomez, and Brandon Kiefer. And that's it. We'll see you next week.

The Vergecast - Elon Musk: agent of chaos

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.