The Vergecast - ChatGPT enters the browser wars

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,

Starting point is 00:00:50 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. 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.

Starting point is 00:01:11 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. This week on Explain It to Me, the chemicals lurking in your cosmetics. New episodes, Sundays, wherever you get your podcasts. Welcome to Broadcast, the flagship podcast of buying Warner Brothers. Every generation of Americans is forced to contend with Warner Brothers being bought, sold, split up, or we combined.

Starting point is 00:01:48 And I believe in my heart that whatever happened to Warner Brothers is generationally defining for you. You know how like, you know, the boomers are defined by the Vietnam War? I believe our generation is defined by Warner Brothers Discovery. Hi, I'm Neelai, your friend. David Pierce is out this week. He's on vacation. So I'm joined here by Jake Hashtrakus. Hey, Jake.

Starting point is 00:02:08 Hey, good to be here. Nilai has some big ideas about what to do with DC Studios, by the way. You don't understand. He's furious about the direction they've taken Superman. Sell it to me. It's too goofy now. Hayden Field is here.

Starting point is 00:02:20 Hey, good to be here. Hayden, you used to work at CNBC. CNBC also sort of like up for sale in a big telecom merger unwinding of whatever's happening with MSNBC is going to turn into a thing called MSN now. Yeah, the logo was a lot to unpack and the tagline. But yeah, it's a crazy time over there right now.

Starting point is 00:02:42 And also, you know, just the fact that, you know, I think they're going to finalize things by the end of this month and then really finalize them by the end of this year. So it's crazy to see all the changes. Yeah, we'll get into it. There's a lot going on in the world of cable and streaming. Just whatever you thought the mainstream media was. It's mostly just a bunch of weird spinoffs and like leverage debt financing deals. for other guys who don't care about anything. It's going to be great.

Starting point is 00:03:06 We're going to talk about a bunch of AI browser news. Browsers are hot. Like, whatever you thought a browser was is now completely up for grabs. There's a bunch of new ones for a bunch of companies that have been making browsers. Google's under a lot of pressure. We've got to talk about Brennan Carr this week on the Verstcast. I'm back at work. And so our buddy is back as well.

Starting point is 00:03:26 And then we got a lightning round. We're going to do it. And the lightning round is going to move quickly through a number of stories from all of us. Isn't that right, guys? That's what I'm told. The Lightning Round is all about. Jake. I mean, you hosted the show while I was gone, so you could do it.

Starting point is 00:03:40 If only, if only there was a mechanism to keep us on track, Eli. Whenever we actually show the Vergecast audience any kind of clock or accountability mechanism, like we get openly harassed. It's a weird relationship we have with you, our listeners, and we love you for it. But before I start, can I just do a PSA with everybody? Do you mind? If you have old search protectors in your house, you should throw them away. Just trust me.

Starting point is 00:04:09 Just go around your house. And if you have a surge protector that you've been caring around since college or before, just throw it away. Just don't even think twice. I'm telling you this because I have one in my house that was somewhere between 15 and 20 years old. And a couple days ago, just went up in smoke. No way. Just exploded. Melted the outlet, singed the wall.

Starting point is 00:04:31 It may or may not have been my fault. I'll come to that in a second, but you know, you see an electrical fire in your house, and you're like, get out of the house. Like, there's no choices. So, like, we grabbed the children. We went out on the street. I flipped the main power. The fire department came. We live, like, right next door to the fire station. It would have been faster if they had just walked. But they came, like, the full parade came. All the neighbors came outside. It was fine. Everything was fine. Everything was fine. Everything was fine. The electricians came the next morning. We tracked away, everything's fine. But that surge protector in particular, I'm guessing, is at least 15 years old, if not 20 years old. And, you know, it's like UL listed. It's a name brand search factor. It was just really old, and it sparked because it might have been loose or there was loose connection inside. And instead of protecting us from a surge, it caused a fire.

Starting point is 00:05:25 And so just throw them away. What did you have plugged into this thing? Nothing. Nothing. Actually, one very funny thing was flagged. into it, but nothing that was drawing any load. So there was an LED lamp, which was off, and this is the funny one,

Starting point is 00:05:38 our brother laser printer, which was off as well, and not printing. So there's no load being drawn to it. A thing that had happened simultaneously is that I was changing light switches in a totally different part of the house and a different circuit.

Starting point is 00:05:53 And like every good dad, I didn't turn the breakers off, which I told a friend of mine that I did this, and I was like, I'm going to turn the breakers off. I got lazy. I've changed hundreds of light switches now. It's over time. And, you know, after a while, I just get lazy and you just stop turning the breakers off because you just do it. I've done hundreds with the breakers on.

Starting point is 00:06:12 And I told him, like, you've got to start turning the breakers off, man. He's like, you're not my dad. I'm going to keep hot watering. It's like, everyone does it. All that really happened was the hot sparked briefly against the wall of one switch. The breaker didn't even pop. You know, it was a different circuit. So I assume this is my fault.

Starting point is 00:06:26 And Max, my seven-year-old was, like, furious at me. And she's like, you caused a fire. And the electricians came the next day. And they're like, these are totally coincidental. Like, we cannot come up with a reason that that caused this search protector to break, except maybe it was loose the wall. It was old. And there was, like, a tiny little, like, whatever.

Starting point is 00:06:43 But, like, that shouldn't happen because the breaker didn't pop. And so I made the electricians go to Max and tell them it was a coincidence. She still hasn't forgiven me. Anyway, the point of the story is, one, you should turn the breakers off. That is the appropriate step. So even if you're going to lie to me and tell me you're going to do it, you should still do it. You should do it. That's fine. I'm going to do it from now, and I promise. I promise. The second is you have old stretch your ass. You got to throw them away.

Starting point is 00:07:10 I don't even know how old any of mine are. Like, I couldn't tell you which ones were new and which ones were like 15 years old. This is the problem. So maybe I have to throw them all out. But like who can do that? I don't know. And that's a lot. My new strategy is every time we like move or like a buy new computer or something, we'll just get rid of the old ones. But like this one, Becky, my wife was like, there's a chance this one is like 30 years old. Right? Like, we have known each other since we were teenagers. So she's like, you might have had this one in college.

Starting point is 00:07:40 Like, this one just has been along for the ride. You know, like that's how long we've known. And it's not like she recognizes the search ride. She's like, there's just a chance that this is the same one we had in Chicago that had moved from apartment apartment. Anyway, the point is it ate itself. Literally, like the back is just melted. and the outlet is melted.

Starting point is 00:07:59 So that's my PSA for everybody. I learned an important lesson, which is, yeah, he's turned the breakers off. I promise the electricians spent two hours here, and they swore up and down. They could not find a logical connection between these two things,

Starting point is 00:08:13 but turn the breakers off, and then, you know, don't have 20-year-old sort of contractors. I'm just putting out there. That's my PSA for everybody. It was too exciting for 4.15 on a Sunday. That's like, it was like two. And we were in and out in 40 minutes.

Starting point is 00:08:28 Like the fire department came. They were like, everything's fine. We were back in the house. The, you know, Matt's was like, why are you such a dummy? Which is, like, great to hear from your seven-year-old. It wasn't like, but it was too exciting. Like, we ramped up too high and we all crashed really hard. That's my PSA for everybody.

Starting point is 00:08:45 A Sunday afternoon. That's rough. It was, like, it's a small town. So, like, all the neighbors can, like, like, contractor calls. Like, I heard there's a firehouse. Like, they're here right now. It's not good. Too exciting.

Starting point is 00:08:56 turn the breakers off, throw away your old search factors. That's what I got for you. All right, let's start with browsers. There's a lot of action in browser world in a way that it's so funny that, I don't know, a year ago, we were talking about Google being sued for antitrust violations and potentially having to sell Chrome because they had such a dominant market share in browsers. And then I'm just looking at the list right now. Hayden, you covered chat, CBT Atlas announcement, opera announced neon, perplexies,

Starting point is 00:09:23 comment is now available for free. DIA from the browser company is owned by Atlassian now. They're putting out more new products. David covered some of that. Reddit is suing perplexity for it's like illegally accessing everything he needs to feed its AI system. There's just a lot of action in like, how do we access the web? And what's the web for? Hayden, you cover the space pretty deeply.

Starting point is 00:09:43 What's the vibe? Like, why is this happening all at once right now? I feel like, you know, companies just have so much FOMO. And every year there's a new thing that they're all trying to put out and they're all like climbing the ladder one rung at a time, doing incremental releases, incremental announcements until they like get there. And they're all trying to basically desperately corner the market share for consumers, for enterprise, for governments. But in this case, it's more consumer focus. And they want to eventually, you know, be in your life from when you wake up to when you go to bed and whether it's at work or

Starting point is 00:10:16 with chat to be depulse, like the first thing you do when you wake up or your browser when you're at work or in your personal life. So I feel like it's just kind of like a mad dash to generate. maybe not profit, but get user attention in any way, shape, or form, and then hopefully it eventually leads to some type of profit so they can pay for some of their compute. But yeah, I mean, it is interesting to see, like, you know, in 2023 and 2024, agents was the big thing. Obviously, it's still huge, and they're all, like, talking about it nonstop. But browsers are, like, the theme of 2025, I think, AI browsers and just, like, you know,

Starting point is 00:10:50 trying to compete with Google. Well, you don't think that browsers and the agents are kind of, two sides of the same coin. Yeah, I mean, they are really connected. It's just funny to see, like, you know, which buzzword you're hearing the most from CEOs. But yeah, I mean, a browser is by nature super agentic. Like, you know, the whole point is that you want the assistant to be doing stuff for you, you know, adding stuff to your cart, booking things on your behalf, you know, chiming in on research, whatever you're doing. But yeah, it's kind of like a next step of making agents actually useful for the consumer, which we didn't really see last year. They were all, like,

Starting point is 00:11:24 doing pretty badly and not really useful for what they were marketed towards this year. They're still not like the best, but they've come a long way in terms of like actually being useful. And browsers are a big part of that. Well, having the agent locally solve some of the problems that you encountered when you were first testing these agents earlier this year, Hayden, right? Because they were being used on these remote PCs that didn't have access to any of your data. And so now they finally do and maybe are more useful as a result. Exactly. That's exactly what's. happening. Like when I was testing these tools earlier this year, they, you know, said that they could

Starting point is 00:11:59 check out on your behalf or, like, you know, access certain services and, you know, buy flowers for your friend or, you know, like look at things on Etsy and add things to your cart. But they couldn't actually do it, at least in my experience, even though that was part of the marketing materials. Now they can. You know, we tested Atlas and it took a long time to add things to your Amazon cart, but it did eventually. And it was faster than expected for like calendar. stuff and emails. And also some research. Like I used it to do some like wedding printing invitation research and compare stuff. And it was faster than I thought, although still not like quick. Well, so the idea is the agent is going to go do stuff for you. And then the fundamental question is

Starting point is 00:12:38 where is the stuff? Like where do the applications that the agent is going to use live and who controls that operating system? And so, you know, I've seen like really impressive demos of Claude where you're like, Claude, go do some agent stuff. Like run some tests in the background or figure out this thing that I need. And that's all in a walled garden, right? That's like happening inside of Claude, basically. And Anthropica has total control of whatever Claude is doing. And so they're able to generate these really cool demos because they can control basically the entire environment that it's working in. Then you end up with like, I'm going to, we're going to run an agent on a browser. The browser will live in the cloud. So I need you to log into your Spotify account on our server.

Starting point is 00:13:17 And no one wants to do that. So the applications, you can't run them there. Then it's like, well, we'll run them on your computer. But then the model has to, to have access to your local operating system, which appears to be Apple's plan and they can't execute it. Right, this is what they want to do with Syria. And that is effectively nowhere, as far as we can tell. So then you're kind of like, what is the most important next operating system?

Starting point is 00:13:41 And it's the browser, right? I mean, like, the browser has gone from being a place where you view documents and gather information in, like, search the web to where you run the next generation of applications. Desktop apps are basically expressed in browser. now. Even most apps that are native are happening in Electron and they're running in browsers, like custom browsers. So, okay, we'll take over the web browser. That will be the new operating system. We'll have total control of that and our agent will click around the browser and use the

Starting point is 00:14:08 applications for you. And I see how all of these companies just tripped through that process and got to screw we need to make a browser. This is the actual thing we need to do. The question is, like, are these browsers any good? And two, like, are any of these application vendors going to let them do it. Right? Like, you can block browser agents. The New York Times, if you go and try to use Atlas on the New York Times website, it's just blocked because they're in a lawsuit against each other.

Starting point is 00:14:35 That's crazy to me that there's a browser that just can't access some websites because of business terms between two companies that have not agreed on how much the content is valued. And like, that is, we haven't even arrived at like, does this stuff work well? It's more like, what is the operating system that runs all the applications that you might want your AI assistant to use. And it feels, at least to me, like the only possible answer is the browser. Because they can't make a smartphone operating system.

Starting point is 00:15:03 They certainly aren't going to make Windows. But what's left? I mean, not only did they go through the exact same thought process that Google did, which is, okay, we made a search engine. How do we own the surface where everyone does the search? They're even using chromium, right? Like, nine out of ten of these things are just chromium browsers. They look the same.

Starting point is 00:15:25 They work fundamentally the same with all this new AI stuff built in. It's roughly the same product. And the question then becomes, okay, who can actually make this stuff work first? Who can actually make it useful first? I've been tinkering with, you know, each of these as they come out.

Starting point is 00:15:41 And there's nothing here that's making me go, oh, okay, I must use an AI browser right now. Right? Like the AI is still, okay, cool, it can move the mouse a little bit. But like, very, very, very slowly, and I don't trust it. So what is this product?

Starting point is 00:15:55 I will say this. I saw our friend Case Newton yesterday. And I basically had your reaction to all this. And he was just like, you are such a New Yorker. Like, he was made like, you're just like, this is just like some New York jadedness. And he was like, here in San Francisco, everyone's like, the computer is using itself. So I must put that viewpoint into this conversation. It is wild to watch the computer use itself.

Starting point is 00:16:18 Like for you to tell a browser, go do something. That's nice. Yeah. Slowly or confusedly. it does it, but it's still doing it, is like, okay, that's something new. But what you're describing is the gulf between something that is worth trillions of dollars and maybe zero dollars. Like, it's, there's a small difference, but it's really important.

Starting point is 00:16:43 And like right now, it's not at the trillion dollar mark. And we've been seeing like incremental updates on this stuff for so long. I think that's why we're a little jaded too. Like, you know, last year, Anthropic put out computer use. Then we saw OpenAI put out operator. Then we saw ChatGPT agent. Obviously, it makes sense that these would all be like really incremental updates, but it's not coming out of nowhere. So for me, it's, yeah, like, I'll be really impressed when it's really useful for a consumer

Starting point is 00:17:08 and for someone that's not even a power user. Like, it just works well for anyone. So, yeah, right now it's a little slow. They did put out an update yesterday saying they were going to work on that and that, you know, there were a bunch of complaints they heard on the first day. and they're like, we're working on them all. Here's a list, which was funny to see. It was like, you know, 15 or 20 things.

Starting point is 00:17:27 But, yeah, I mean, I don't know. I was impressed with the fact that it was actually able to make purchases because that is not something that Chad GPD agent could do, as we said earlier. So, you know, it's nice that, like, you know, the browser is able to actually complete some of the actions they say it can, even if it takes a while. The thing actually really gets me about the Atlas in particular

Starting point is 00:17:50 is you have to pay either the plus or the pro tier for ChatGBT to have it use itself slowly. And I just think that's so funny. Like, if you would like your web browser to slowly make purchases for you, you can pay OpenAI some money every month. And I get why they're limiting it, right? It doesn't work as well as it should.

Starting point is 00:18:08 And certainly you can have your customer self-select into these experiences and that don't make sense. But to your point, Jake, like, there's no way that you'll ever get to like a trillion dollar evaluation divided by 20 bucks a month for the product that is shipping today. and maybe it'll get better, but it has to get so good

Starting point is 00:18:24 that everyone stops using their web browser, which to me is not the turn that anyone is willing to admit. Right. My experience with Atlas so far is that I did not realize you need to have a paid account to use the agentic stuff, right?

Starting point is 00:18:39 Because this is all they advertise. It's everywhere. So I'm like, oh, cool, go like control my mouse, do stuff on the computer for me. And so I'm like, hey, like, buy this for me, add this to my cart, fill out this form. and ChatschGBT just keeps going,

Starting point is 00:18:51 can't do that. I have no idea. We don't have an Asian mode. You didn't get to the point where I told you how to use DoorDash. It happened to me before I upgraded the plan. It was like, here's how you use DoorDash. Push these buttons.

Starting point is 00:19:01 And I was like, wow, you think very poorly with me. At no point did it go, oh, hey, this is a feature in this browser and you have to pay for it. It was just like, I have no idea what an Asian mode is. I'm like, these features, it's like everything is so close and it's like not quite making the turn to where it is useful. I see the idea, right? Like checking out, filling out forms, it's very tedious.

Starting point is 00:19:26 It would be great to just, you know, say go, go ahead. But at this point, it feels like it is a lot of prompting. Wait, your phone doesn't just fill out the forms for you? Dude, I don't even, every single website. I feel like Autofill forms is like 10 now. No, I think. Autofil forms, I think, has gotten more and more broken as the web has gotten more and more broken. I feel like everything is.

Starting point is 00:19:48 is making me do it by hand now and it's driving me crazy. I could not agree more because it still has my addresses from like five apartments ago. And no matter how much I update these tools, like they will still put my address from like five years ago, no matter what I do. This is how I remember my old address. Yeah, same. It's very useful. This is actually my bigger question about all this.

Starting point is 00:20:10 And it's just like, you know, we've been asking variations of this question in the AI context since we started talking on AI. This is yet another kind of pressure on the web. And I will separate the web into two broad categories. There is the web where you go and find information, and that web we have covered extensively is under enormous pressure to the point where it's almost dead. Right?

Starting point is 00:20:36 The average person today is not like, let me go read a bunch of websites. They're mostly like, let me watch a bunch of TikToks. And Google is sending less and less traffic to fewer and fewer websites. that's a lot of pressure on the, we publish information on the web. That's just one of those things.

Starting point is 00:20:50 And you can just see that part of the web declining. We run the last website on Earth here at the verge, and we understand it keenly, and we spend a lot of time thinking about it, but we also cover it very keenly. And you can just see it. That's one whole thing that AI is doing, right? You're going to stop searching Google

Starting point is 00:21:06 and clicking on links. You're going to start just talking to chatbot and getting information that was scrapped from the web, and there will be a series of lawsuits, and something will happen. Or you will just never do that, and you'll just watch it, of social video clips instead.

Starting point is 00:21:17 Who knows how that's going to work, but that's one set of pressure. The part where the web has been explosively growing is as a place to distribute applications. Right? You don't have your phone on you. You've got your laptop. You want to order a sandwich like DoorDash is there

Starting point is 00:21:30 for you as an application. You want to use any of these chatbots. It's not really native apps that you're using, right? It's you're opening the browser and going to the cutting AGI-I tools, which are all expressed in the browser. Like, the browser is a series of technologies that can run applications in a mix of local and cloud-based scenarios,

Starting point is 00:21:51 that has been just exhilarating for years. And all of them presume that the thing you want to do is you as a person want to use the applications, right? Like, you're going to use the stuff. I don't know what that looks like, that whole ecosystem looks like, when it's a robot doing it. Like, when you don't need to have a user interface, right? Over the summer, we had Sundar, Pichai, and Dakota, and I asked him, like, what happens when you turn the web into a series of databases?

Starting point is 00:22:22 And he's like, that's not going to happen. The web is always with a series of databases, which is the most Google way of thinking about the web possible. But right now, you know, all these agents, they're going to human-oriented interfaces and clicking around them. And at some point, someone's going to just make the turn and say, well, we don't have any people coming here anyway. It's just robots clicking around. we don't need all this design at all, we're just going to have a version of our website designed for the agent in Atlas or neon or whatever comet browser takes over.

Starting point is 00:22:56 And I don't, I just don't, I don't know if the AI companies have thought about that, the same way they didn't think about like, what would happen if we stole all the world's data and didn't pay anyone for it? Right, like there's a term there for the web, which is even the part of the web that is growing and successful, feels like it's now under pressure from agentic browsers. And, hey, I just, I mean, you've been covering stuff you talk to these companies. Have they thought about that at all? I don't think so.

Starting point is 00:23:19 I mean, I haven't heard a thing. And even when people start to get at that in questions with execs and stuff, you know, it's something they kind of just push down the line. They say, like, you know, that's not saying we're worried about. And yeah, I think they're just focused on the short term. And they might not be thinking about that. But it is like a reality that, you know, they'll probably have to reckon with. Yeah, like, what if there are any apps?

Starting point is 00:23:41 Like, it's the same as like, what happens when you run out of train? training data because you script the entire web. Yeah, that's exactly what I was thinking. It's like, you know, that is something that people ask these CEOs and they don't really have an answer. You know, what happens when everything is AI generated and it's all slop and you can't really learn anything else anymore? And then they say, well, you know, they'll learn from actions.

Starting point is 00:24:02 They'll learn from other things like environmental data, like having a advice that's listening all the time. That'll teach them more. So it's like, yeah, they're going to get data from somewhere. But, yeah, it is interesting to think about like what's down the line and, yeah, what happens from an application front. The other piece is, you know, just a few weeks ago, open I had their dev day. They showed bits and pieces of apps inside chat chabit, like fragments of apps showing up there.

Starting point is 00:24:27 I thought it was really interesting. Brian Chesky, who was very close to Sam Altman, was on CNBC, and he was like, it's not good enough for me to put an Airbnb experience there. And he said, I don't want to just be the data layer for an AI bot, which is the thing that we always talk about is the DoorDash problem. Like, if the AI systems disintermediate all the applications, then you are just like a commodity provider of sandwiches and you don't have a business anymore. Like, that's not good.

Starting point is 00:24:52 Nobody wants that. And Chesky basically on CNBC said, this is my fear. It's not good enough. We need to build a robust experience. And I'm waiting for that to happen. And then he was like, I'm friends with Sam. We'll get it. Right?

Starting point is 00:25:01 Like, he's Brian Chesky. And that's how he was. He's a character. How do you see the difference between open a icing? All the applications will live in chat, GPT, and Open AISing. We have a browser that will go out in the web and use applications for you. Like, those seem like different bets, and I can't tell which one they're more confident or excited by. Yeah, I mean, I think that I don't know which one they're backing, but I do think that they're backing both and seeing what happens,

Starting point is 00:25:29 and then they're going to go one way or the other. Right now, it feels like most AI companies are kind of taking a pancake at the wall approach, like throw it all out there and see what does well and then what might generate a profit and, you know, pursue that. because they don't have enough compute for all this stuff anyway, and they don't have enough resources either. So it's like, you know, they're putting out a bunch of stuff, seeing what works, what does well, what people complain about, and what people actually use,

Starting point is 00:25:52 and then just kind of trying to go from there is my sense. So, yeah, I think, I mean, they were really excited about all the chat GPT apps. I think that those are the ones that I see consumers talking about right now and using, like the Zillow one where, you know, they demoed someone, you know, looking for a house in a certain area and then narrowing that search and asking follow-up questions, like which ones are this many square feet and which ones have a backyard for my dog, things like that. So the follow-up questions within the app seem to excite people. And then, yeah, they have a suite of apps coming out.

Starting point is 00:26:25 They're also allowing individual developers to work on apps for chat cheptie, and then they're going to, you know, decide whether they distribute them or not. So that's what I've seen, like, actual people more interested in so far rather than the AI browser. People in tech are really excited about the AI browser. But yeah, that's kind of anecdotal. But we'll see which one comes out on top. I think this tension shows the tough spot they're in right now with the development of these systems, right?

Starting point is 00:26:52 Some of the premise of this agentic AI stuff is, well, no one would give us access to the platforms. No one would build all this stuff for us to integrate with. So it's actually easier to make a computer that can use itself and operate everything like a human would than to go and get all those partnerships. And they got a little bit of the way, but it's not all the way,

Starting point is 00:27:16 and it's not reliable enough, and it's not fast enough. And, you know, Open AI is, by virtue of being Open AI, is pretty powerful, right? ChatGBTGBT has a lot of users. And so now they're taking the other route and they're trying to go direct to the companies, being like, hey, why don't you open up?

Starting point is 00:27:33 Hey, why don't you give us access? And this is where, to your point in a lot, you get the DoorDash problem where it's like, okay, maybe these companies want to because, you know, they probably get a cool little stock bump by announcing an Open AI partnership. Yeah. I looked in the Sam Altman's eyes and he said, we are all going to be rich in your stuff. Yeah, exactly, exactly. But what happens if they say no? What happens if the companies just don't want to give them access to this stuff, right? Then their browser will come to your front door and turn you into a data layer, whether you want to be one or not. That's the possibility. It kind of reminds me of all the

Starting point is 00:28:04 media company deals with AI. So same vibe. Yeah, it's just, you know, when you're saying more people are exciting about the app fragments inside of chat media, it's like, yeah, that's the interface. That's the thing they control, right? Like, it's not we're going to reinvent the browser, which is a really difficult thing to do. People have a lot of ideas what web browsers are and what they should be and how they should work. And when Apple moves the address bar from the top to the bottom, there's like a five-day news cycle. Like, that is a pretty difficult thing to go and reinvent. Like, the browser company tried and they basically gave up. and started over with the premise of, okay,

Starting point is 00:28:36 what if it's AI first instead of just like, I mean, the browser company's pitch was, we know this is an operating system. What if we build a browser that's app-centric instead of document-centric, and they got nowhere with that idea? I mean, people liked it, but they didn't, like, take market share. Now they're like, what if it's AI first,

Starting point is 00:28:51 and they have it lasting and backing them so they can take a longer time to figure it out? But all, I think Open AI is basically saying, well, either you can integrate with us on our terms or make a business deal, or our browser will just come and do it for you anyway. And unless you're in New York Times and you're suing us and you care to block us, like you're going to want that traffic because your Google traffic is going down.

Starting point is 00:29:12 That's at least what it feels like for Open AI. I think for opera and perplexity and the others, I think they would just like more market share. I think they were just like more users. But Open AI has like the powerful base of users right now. Yep, exactly. So my last question here is, what's Google doing all of this? Because they're the ones who are under the most pressure, right?

Starting point is 00:29:31 All the users that we're talking about are, Google users almost definitionally, right? Whether that's in Chrome, whether that's Google search, whether that's just being accosted by Gemini in Gmail, like a mugger in the night, just jumping out of alleys, being like, like, do you want me to summarize this? Like, Google has a lot of power here, and they are having Gemini start to do some agentic stuff in Chrome. Are they moving fast enough, do you think?

Starting point is 00:29:56 I don't think so, because, you know, I remember they made a really big deal about a lot of agentic announcements recently. But a lot of those announcements, when I pressed them on exactly when they were going to roll out, were months away. And they didn't even have a real timeline for exactly how long it would be. So they basically were doing a marketing employee of saying, hey, you know, don't worry. We're having agent stuff. It's going to do tedious tasks on your behalf. We have all this in the pipeline. But it was way down the line. So, you know, they introduced some stuff day of. But all the real agent stuff that was actually going to compete with what's already out there. is months away at best, it seems. So, yeah, I mean, I think they're not moving quick enough and, you know, they're doing their best to compete by saying what they're going to do eventually. But, you know, it's interesting to see the timelines

Starting point is 00:30:46 and, you know, if they're behind or not. Do you think Shem and I can compete with Chatsybt or any of the other models on agentic tasks like this? It's hard to say when we haven't actually seen it do that much yet, you know, comparatively. Like I've seen the other models do. a lot more because they've been releasing incremental stuff. Google is still, you know, a couple months away from even releasing some of the comparable

Starting point is 00:31:10 tools for Gemini. You know, I would think that it can because, you know, it's a model that is pretty powerful. But, you know, it's going to take a lot of trial and error too, just like the other companies went through. So I think, you know, they might just stay perpetually behind, at least for a while. Yeah. The other two companies I want to bring up here really quickly are Apple and Meta. Apple is nowhere on Agenda K.I.

Starting point is 00:31:34 Right? There's no other way to describe it. They announced the Siri product that was supposed to use the app on your phone for you. They control the operating system. They control the user interface. Famously, they control application development and app distribution.

Starting point is 00:31:48 Like, they have all the things that everyone else wants, and they don't have a foundation model. They don't have the ability to do it for whatever reason. So they got to figure it out. But I look at Safari. in iOS 26. And it actually feels the most ominous to me. Because if you look at how they've redesigned Safari

Starting point is 00:32:09 in iOS 26, it's primary purpose, like the primary intent of the visual design, is for it to look good when it slides up in an app to let you log in. It looks beautiful. When you're in an app, you're like, I want to log in, and the little Safari sheet slides up and you like hit the button and it goes away

Starting point is 00:32:28 and you've logged in on the web, it looks beautiful. They have optimized the designer Safari for that thing so clearly, and it looks great. And then everything else is like, oh, we made all the tradeoffs against making this a great web browser because we actually optimized it to look like a login sheet inside an app. But we haven't given it any ability to go and browse the web for you. Like, why would we do that?

Starting point is 00:32:49 We want Siri on the phone to do all that for you. So you're like, oh, Apple is deprecating its own web browser. right? They're paying less and less attention to this, and the intention they are paying into it is to make it appear to be more and more part of the app, less and less an app on its own. I'm sure there's an apply people listening to this. I'm sure they're all going to yell at me that Safari is a conjugal, and it's just as good as any other browser. But like, look around, man. The opening eyes browser is using itself. It just uses itself for you. Like, Safari can't do that. And I wonder, like, again, these two pressures on the web,

Starting point is 00:33:25 the document web, the information web, under such pressure that it might already be dead, and we're just pretending it's not. And the application web has always been under pressure from Apple. They don't want you to use web apps on the iPhone. They want you to use native applications on the iPhone. So always pressure there. And now it's under pressure from automated browsing assistants.

Starting point is 00:33:46 Like, that's when I say it, like, it feels even more ominous. Like, there's no out. There's no Safari. based out for the real web. Like, even that is pressure. I wonder, have you heard anything about Apple getting back in the game on AI? Because I haven't heard a word since WWC. Yep, I haven't heard a word since then, honestly.

Starting point is 00:34:05 And I've been looking. You know, WWDC was a little underwhelming, a lot underwhelming, actually, when it comes to AI announcements. I was sitting there ready to cover a bunch of stuff, do follow-ups. It was like, where was the AI? You know, there were a couple announcements, but they were, pretty underwhelming. I mean, I don't know, you know, translation, some health stuff. Yeah, but yeah, I mean, it seems like they're kind of perpetually behind and I don't know when

Starting point is 00:34:32 they're going to catch up. But yeah, I have not heard a word since then. The only thing that I've heard is a whisper that the Siri app intents framework will somehow play nice with MCP model context protocol, which is a give up, which is Apple saying they no longer have so much control over their app developers, the app developers have to do it their way. They're going to use what is the fact that is Anthropics. MCB belongs to Anthropic, but it's an industry standard now, right? Apple giving in to that feels like a watershed moment for like, oh, we don't have power over this ecosystem anymore. That's the only thing I've heard.

Starting point is 00:35:05 And that even that is like not quite confirmed yet. Yeah, I could see that. Then we should talk about meta really quick. Meta has all the resources and money in the world. And they, from what I gather, keep spending billions of dollars to hire single AI researchers. and then firing 500 AI employees. And it's sad. Like, the whiplash of META's spending all of the money it has to hire, like, one guy.

Starting point is 00:35:28 And then, like, three weeks later, fire 600 people, which it just did this week. I can't explain it. Can you explain it? Well, I can tell you what they told me, which is that they're really, I mean, they are really, you know, trying to control the narrative. But I will say this was planned. Like, it was something they told me, you know, months ago or maybe a month ago, time all runs together. But when I last talked to them, they're like, yeah, we're planning a restructuring.

Starting point is 00:35:53 Those internal memos that came out a month ago or more about how they were going to like kind of put all their resources into TBD Lab, the one with all the glam high profile researchers. That was, you know, reported. And they said they were going to do a restructuring.

Starting point is 00:36:08 They said they were going to pause hiring for a lot of the other teams and that there would likely be layoffs. And so this is that, basically, from my understanding. I mean, it's way bigger than we thought in terms of the number of people. But it makes sense because they spent so much money on all of these researchers, you know, like compensation packages. So, you know, their point of view is, hey, we're going to stop the red tape. You know, there's going to be fewer people making decisions.

Starting point is 00:36:35 You know, we'll move faster. We want to be operate more like a startup, whatever. You know, we'll see. But it is interesting that, you know, a lot of this for meta is about PR. It's a about, like, you know, hey, we're doing the cutting-edge stuff, come work here. And so when you cut 600 rolls, you know, that really impacts that. And whether people want to work there, we've also seen some people go to meta and then leave immediately. So, you know, it's all, basically it's just a big arms race for trying to get all these researchers and engineers to come and, you know, sometimes you lose them.

Starting point is 00:37:10 It's wild out there. But meta's bet on what is the interface for computers is their glasses. Right? I mean, they're not building a browser from what we can tell. I don't know what their agentic ideas are. But if you think agentic AI systems that can go and do stuff for people are what will eventually pay off all this investment, are they also waiting on their glasses to be a big hit? Like, what are all these high-paid researchers doing?

Starting point is 00:37:37 That is the question. I mean, a lot of memes came out after vibes came out. Their AI, you know, generated video, social media thing. So, yeah, I mean, we'll see. They say they have a lot in the pipeline. It seems to me, like, so far it's been vibes, you know, themed chatbots on Facebook. It's funny when you say vibes and you mean the app, but I keep interpreting that as just general vibes from meta that they're going to be doing a good job. It can work both ways.

Starting point is 00:38:08 Because that's mostly what they've shipped. Yeah, I would say it works both ways. Yeah, they're operating on vibes right now in terms of the actual app and the, you know, word vibes. But we'll see. I mean, they said they had a lot in the pipeline. We have like Alex Wang leading stuff. So, I mean,

Starting point is 00:38:26 I do think they're leaning far into glasses, but yeah, we'll just have to see. I mean, the glasses were more impressive than any of the other iterations. You know, the verge reported that. But yeah, I mean, it's definitely going to be a wait and see situation. It's only been a couple months. But, you know, I'll be watching what they do

Starting point is 00:38:42 for sure, especially the ones that were paid the most. Everything better does. feels less like a thing they truly believe is the future and more like a thing they deeply, deeply want to be the future, right? Like, they're doing the glasses because nobody else has done it and because they lost out on smartphones, right? They did the metaverse because of, I don't know, the exact same reasons.

Starting point is 00:39:05 They needed some other new thing to hook people that other people weren't using. The glasses were a good pair of headphones, right? Are they a good AI interface? Like, maybe they still need to prove that out. But if they're going to mean AI interface, they need applications. This is the thing that I, the reason I want to talk about meta last year was, I understand that Apple has all the puzzle pieces and like literally no ability to put the puzzle together.

Starting point is 00:39:32 Like, they just can't do it. Like, they don't have a model. They don't have a team that can do it. But they have the applications. They have the user interface. They have the ability to build the next generation hardware. Like, they can do everything. And they're, like, paralyzed by their inability to attack.

Starting point is 00:39:46 actually execute the product. Google could do it. Google is paralyzed by the fact that I'm pretty sure the Gemini team has never even met the people who run the Nest Hub. It's just like, Google is Google, and they're paralyzed by their own stasis and like desire to kick off 50 projects and never complete any of them, like Google. But they have all the pieces, and they own the next most important application layer, Chrome. Right? You can just, okay, that's one. you can see Open AI being like, okay, we're going to do an application layer inside chat GPT because we have enough market share to demand

Starting point is 00:40:20 that Uber show up and do a demo at our Deb day. That's cool. Zola's going to do a thing. That's going to be cool. We're also going to make a browser to hedge. Perplexity is basically just like, we'll do anything. Anything that you want. Perplexity is there to say they're going to be the future

Starting point is 00:40:36 and make you big promises. And on and on and on. And then you get to Meta, which is spending all of this money, which has taken these big shots at post-phone devices. And I just, they don't seem to have a thesis, right? There's nothing that holds that together.

Starting point is 00:40:51 Do you remember when Zuckerberg wrote his like manifesto on the future of AI in July? He said personal devices like glasses that understand our context because they can see what we see, hear what we hear, and interact with us throughout the day will become our primary computing devices. And he wrote this whole manifesto on it. And I remember thinking at the time and people were kind of. kind of like memeifying it. It was like Mark Zuckerberg says you can trust him with super intelligent AI and it'll come in the form of glasses.

Starting point is 00:41:20 But I remember thinking at the time like, okay, this is what they want to happen, but do they really believe it? So back to what you were saying, I mean, it seems like, you know, they're working really hard to make their vision of the future that they hope happens happen. But maybe operating with blinders on about what people really want and will use broadly. Well, but also, like, where does meta have any credibility to, like, the universe of application developers to say, put your apps here, and we will do a good job to help you make the most money. And they haven't done any hedge. Like, Open AI is saying that inside chatyptee, but their hedge is the browser, right? They've made both bets. Whereas, I think, meta has some weird fork of Android running on the headsets.

Starting point is 00:42:06 and there are some games in Meta Horizon, and that's it. Like, there's nothing else going on there. Like, there's not a rich developer ecosystem that's pushing the platform forward to the point where you might as well say it's dead. Like, I look at that and like, and I really thought, like, supernatural

Starting point is 00:42:23 with the fitness app they bought, I was like, oh, this is actually a killer app for this device, and they bought it, and they laid off a bunch of people and it's mostly dwindled. And it's like, oh, they even let that go. Even the thing that was working, they just sort of like let it with,

Starting point is 00:42:36 into the ground because Mark Zucker got distracted by the next big thing. And, like, that's the turn. If I, like, had to put the sharpest headline, and so he knows what he wants and he has no idea how to get it. Right. He wants to be in charge of the primary computing platform for millions upon millions of people. He has no idea to how to get that there because he actually doesn't, he doesn't know what it is.

Starting point is 00:42:56 He just knows he wants it. Right? It's a slight twist on what you're saying. He knows what he wants. It's power. But it doesn't matter what that power looks like. As long as he has it, right? Like, and that he has no plan to, like, to make it happen in a way that it feels iterative.

Starting point is 00:43:13 Like, you start with, I have to invent the wheel, and then I have to put the wheel on a skateboard, and then I put an engine on the skateboard to make it a clock. Like, he doesn't have that. He's like, we're going to make glasses and put them everyone's face. It's going to listen to you all the time, and it's to be magic. I think that's why he made all these hires so that they can, like, you know, do the iterative part for him and come up with a plan to achieve his vision. I know, Jake, you've been just, like, grinning this whole time.

Starting point is 00:43:35 What's going on in that brain? I can see your face as I've been saying this. No, no, I mean, this is the problem for meta, right? Nobody trusts them. And they are just constantly chasing the next big thing, right? The glasses are very impressive, but I've said this again and again. What they made is a very impressive pair of headphones, and they keep calling them these AI devices and a gateway to AI,

Starting point is 00:43:57 and they are just simply not that, right? Right now they have one pair that is a pair of headphones, they have another pair that is a weird Android phone on your face with no apps. And it's, don't get me wrong. It's a remarkably impressive Android phone. It's really cool to put an Android phone on your face. It's not even an Android phone. It's like a WhatsApp single purpose.

Starting point is 00:44:20 Yeah, right, but this is it. It's got WhatsApp, and it, like, maybe kind of has Spotify, and that's about it. And it's like they need to be able to do more. And so, yeah, I mean, it's funny. like what did they hire Alexander Wang to do to figure out a product? Because they don't have one. And they seemed like they were early with mama, but that fell apart pretty catastrophic.

Starting point is 00:44:47 Fickly. So, yeah, it is not clear to me what product they have in AI right now. They have a bunch of, you know, fires going. They have a bunch of ideas that they are working toward, but they have the steepest uphill battle to climb of any of the big tech companies because there is just a remarkable distrust of meta and their products.

Starting point is 00:45:11 And they don't control any of the existing big platforms already. And so they're fighting against the network effects that Apple and Google can use to build this stuff up when or if they pull it together. Yeah, I'm legitimately shocked that meta hasn't realized it needs to make a browser. That's kind of where I'm headed, right? It's like the idea that you're going to make a thing

Starting point is 00:45:31 that can use other apps and services for people is such an important part of this whole puzzle and meta has taken no shots at that, like zero shots on goal. They're going to need to rebrand the company again if they want to get a web browser that people use.

Starting point is 00:45:45 I don't think that's going to happen. I think they've done that once. It's going to be called brow. Yeah. That's great. Like I said, these browsers are all out. You can go download them. Let us know what you think of them.

Starting point is 00:45:56 They're kind of interesting. They're fun to play with. The computers are using themselves. It's wild. It's a legitimately interesting thing that is occurring, but the twin pressures on the web are super real. Like, whatever the web is today, I would say 18 months from now will be totally different, like completely different and potentially unrecognizable, because more and more web apps will decide to either allow these agents to use them or block them. Like, that's a weird choice that's coming for a lot of web apps.

Starting point is 00:46:23 And then on the information side, well covered, we don't have to get into that. It makes everybody sad. But it's already happened. All right, we've got to take a break. We're going to come back and talk about some happy news, which is what's happening to a bunch of legacy cable companies. We'll be right back. Support for the show comes from Framer.

Starting point is 00:46:42 Framer is an enterprise-grade, no-code website builder, used by teams at companies like Perplexity and Muro to move faster. With real-time collaboration and a robust CMS, with everything you need for great SEO, not to mention advanced analytics that include integrated A-B testing, your designers and marketers are empowered to build and maximize your dot com from day one. So whether you want to launch a new site, test a few landing pages, or migrate your full.com, Framer has programs for startups, scaleups, and large enterprises to make going from idea to live site as easy and fast as possible.

Starting point is 00:47:22 Learn how you can get more out of your.com from a Framer specialist or get started building for free today at Framer.com slash verge for 30% off a Framer Pro annual plan. That's framer.com slash verge for 30% off. Framer.com slash verge. Rules and restrictions may apply. 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 00:47:57 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 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, especially when

Starting point is 00:48:27 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.com to connect with top talent ready to help your business grow. That's upw-w-rk.com. Support for this show comes from Whatnot. Whether you're selling online or out of a storefront, you already know the challenge.

Starting point is 00:49:00 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 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.

Starting point is 00:49:33 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. That's W-H-A-T-N-O-T dot com slash sell. What-N-com slash sell. All right, we're back. I feel like we should talk with the Galaxy XR for five minutes, based on the conversation we were just having.

Starting point is 00:50:13 I did just say nobody wants to wear Android on their face. And with that, the Galaxy XR is here. So the XR, it's basically a Samsung Vision Pro. Like, I don't know, even, it even looks like one. It is as shameless as anything Samsung has ever done. It's cheaper. It's $1799 compared to the Vision Pro is $3,500. It's lighter.

Starting point is 00:50:38 There's no removable strap. There's no eyes on the face, which is truly the most mystifying decision Apple ever made. It has the native Netflix app. That's only a bullet point because Netflix was like, why would we ever be in the Vision Pro? And they were completely right about that resource allocation.

Starting point is 00:50:53 I don't know. What do you think, Jake? The smartest thing they did here is that they made it, as far as I can tell, out of plastic. And so it is able, right? It is still a weird, ugly headset on your face. that puts a mobile operating system in front of your eyes. But at least now it's not really, really, really heavy on your face, right?

Starting point is 00:51:12 Apple is just like constitutionally opposed to not making everything out of metal. And so the Vision Pro is this heavy thing that people can only wear for like an hour. Right, I have these like super light and thin pair of glasses, regular normal human glasses. And these get tiring on my face, right? I take them off pretty frequently throughout the day. And I feel like that is a major problem. for headsets because you do not want to keep them on all day and you have to work for a lot of hours during the day. So this made a smart decision, right? They stole the exact same design as

Starting point is 00:51:48 the Vision Pro and they just made it lighter. And that seems to probably solve a lot of the issues. Now, whether anybody actually wants to live in this thing for eight hours a day, that still seems to be an open question. Or whether there's a reason to. Right. Like beyond comfort. V did a quick The thing that really grabbed me was, oh, it's so much less powerful as well. Like, if they cut down the feet, it is Android phone, whereas I think the Vision Pro specs up against a Mac, right, especially now with an M5 chip, like literally a MacBook Pro chip in there, because it's meant to be a whole ass computer. And I think this is meant to be kind of like a display, like a consumption environment.

Starting point is 00:52:26 Yeah. I mean, this is like, I apparently can extend a Samsung laptops display. So if you're like really living that Samsung life, I, it feels like a lot of companies began building these during the pandemic and just had some huge dreams that this is the future of computing. And I actually think the most interesting thing about this is, right? Like, Samsung seems to have done a lovely job, right? They cloned the Vision Pro mostly one-to-one, has most of the same features, it seems like it.

Starting point is 00:52:59 And I think, like, fundamentally, the question still remains whether people want to spend a ton of time in these things. And right, when you've made a general purpose computing platform that is also not powerful enough to be a general purpose computing platform, it is just not clear to me what the market is for this. It's a great hedge, but it is fascinating to me that right now Google has one partner on this platform. And it is Samsung against Apple, and it is not even clear how committed Apple is to this. Yeah. And Lenovo isn't even there, to your point. Like, Lenovo will do any old weird shit. Right.

Starting point is 00:53:37 Where are you at, Lenovo? I'll put the Galaxy XR on Neely's Unified Scale of wearable bullshit where the X axis is utility and the Y axis is fiddliness.

Starting point is 00:53:46 So anything that you wear in your body, if it's too fiddly and not useful enough, you end up in the bottom quadrant and you fail. Or you're not very fiddly at all and very useful,

Starting point is 00:53:56 like the Apple Watch, and you end up in the top quadrant and you succeed. You get a huge fiddliness penalty for face. right? The only device in history

Starting point is 00:54:07 that has overcome the penalty of face is plain old glasses. Massive utility for people who need them. Very, very fiddly face penalty, but the utility overcomes the penalty of face. I'm just putting the Galaxy XR in the chart. And the Vision Pro so fiddly, right? External battery pack,

Starting point is 00:54:27 the XR has an external battery pack, but the Visioner had external battery pack had weird app support, was heavy, was bulky, had fake eyes, very, very limited utility, just instantly in the bottom corner, didn't work, right? That's my not enough utility. Oh, and face, huge face penalty.

Starting point is 00:54:44 Put the XR on the scale. I mean, the XR, it's the exact same product. Like, it is the same thing. They, I don't think most people could tell these things apart if you put them side by side. Okay, okay. So now assume that this, there's an AI agent inside the Galaxy XR.

Starting point is 00:55:03 Gemini inside the XR is tremendously useful. Does that change your... I... They keep advertising it. They keep being like, the thing you don't understand about the Galaxy SR, it's built Gemini first.

Starting point is 00:55:14 And I'm like, what does that mean? It's on my face. I'm... I'm putting it on my face for an AI to go and do something. If, like, if AI's going to do it,

Starting point is 00:55:23 then I don't want to be using technology at all. All right, Aiden. It's always funny to me when a company says something is like Gemini first or AI first. Even the Starbucks

Starting point is 00:55:33 CEO the other day said Starbucks was going to become an AI-first company. So I laughed at that. But yeah, I mean, I don't know. I guess it's less fiddly, obviously, because it's lighter. But yeah, otherwise it's the same. Brutal. All right, Gemini has not provided enough utility to overcome the face penalty. That is one of the more Vergecast sentences I've ever said. I'm very proud of it. I stand by it. We're going to review the XR. We'll have V on to talk about it. And we'll go through the chart again. As always, Nealized scale of wearable bullshit remains undefeated. In the sense that you can't defeat it. All right, let's talk about cable news, our favorite.

Starting point is 00:56:13 Warner Brothers Discovery is up for sale. The company just straight up announced that it's going to sell itself, whether all together or in parts, it seems very much that Paramount Skydance wants to buy it. CNBC reported that there have already been three offers. One of Paramount's arguments here is that, Warner Brothers can't sell itself to anyone else because every other company is so big

Starting point is 00:56:37 that the regulators will say no, but Paramount's just a wee little, just in any bitty, and so maybe it'll go to it's like that. That doesn't seem right. This comes after Warner Brothers had announced plans to split itself into a linear business, like a traditional cable business,

Starting point is 00:56:54 and a streaming business, and kind of remix all the assets between Warner Brothers and Discovery so that you would have a profitable streaming and studios business, and then a decaying cable business. Earlier in the show, I mentioned, that is essentially what Comcast is doing with NBC Universal.

Starting point is 00:57:08 They're spinning off a bunch of the old cable networks, including CNBC and MSNBC, into a cable division and keeping a bunch of streaming stuff, including Peacock. There's a lot here. And then just today, there's news that Netflix, Amazon,

Starting point is 00:57:20 and Apple are interested in buying Warner Brothers. What do you think is going to go down here, Jake? First off, nothing instills more confidence in me than the idea that we might get a media conglomerate, that is three companies run by the son of a, you know, wealthy tech mogul who just wanted to build a media empire, right? Like, it is not clear why Paramount Skydance, which is actually just a started life as a skydance, which was not itself a huge and hugely successful company, should now have Paramount and Warner Brothers discovery and what that is going to do for Warner Brothers and allow them to. know what it will do. It will do the same thing that happens to every company it buys Warner Brothers. It will destroy it from the inside with Rod. Yes. Can we just go through the list

Starting point is 00:58:10 really quickly? I just want to emphasize that point. Every generation gets its Warner Brothers merger, and then something important dies. And so, you know, my generation, I would, you know, the Gen X Dial-Up Generation, we got AOL Time Warner. Everything is dead. All of those things are dead. Like they, that, that was, there's a great book. Caras Swisher wrote, I think, The best work of her career, there's the book she wrote about the AOL Time Warner merger. It's called There Must Be a Pony in Here Somewhere. And that is a quote from, I believe, an AOL executive who said,

Starting point is 00:58:43 we put together a huge pile of shit, and we decided there must be a pony in here somewhere. Like, that's that one. It killed that company. Then they spun it back out. Then AT&T bought it, a deal which gave us only the 4-3 gray scale Justice League movie. That's not a good outcome.

Starting point is 00:59:03 that basically killed 18 in a variety of ways, then Warner Brothers had it, and it killed Discovery. Every generation gets its Warner Brothers deal. Buying Warner Brothers is full on a midlife crisis move for corporate executives. They want to be kind of cool. They're like, oh, there's some good assets in there, right? When I see Apple on this list, I'm like, stay away, you guys. Like, you know what's coming here.

Starting point is 00:59:29 Like, it seems good at first. I understand, right, all of these companies have a bunch of money and they see a company with a bunch of, right, everybody loves some good IP in there. But, right, this company is a disaster right now. And it clearly doesn't know what it's doing. They're in the middle of splitting the company apart, right? Because it is not functioning in its current state. The only way they're going to not do that is if they can make a sale first. And so these suitors are lining up to buy a company that is so unsure of itself that it is already splitting itself up and trying to get rid of a bunch of assets that aren't performing well. So, yeah, like every generation is about to get its, you know, it's Warner Brothers. Like, the cycle is about to continue. I will say the somewhat terrifying thing is that the cycle appears to be exhilarating, right? Like, the Warner Brothers Trail of Destruction.

Starting point is 01:00:31 is only getting faster and wider as time goes on. Like more and more people, I believe that by 2045, every American will have been Warner Brothers Discovery. Now every generation is going to get two Warner Brothers Discovery deals. These are defining, when the distribution tries to buy the content, these are defining moments and media. And what's really interesting here is David Ellison, Larry Ellison, that family, as Paramount Skydance,

Starting point is 01:00:59 also part of the bid for TikTok. So you can imagine a world in which you buy TikTok and we're just running the clips of top gun on TikTok, but actually making money instead of just pirating them with like the little guy in the corner. The Justice League cut, it's going to be good. Right. That's just like a theory you could have of media in 26 that all the kids are on TikTok and you're going to program TikTok with all the great content from your studios. And we're all going to make those weird vertical microdramas.

Starting point is 01:01:30 Yonko Rutgers wrote about them for. us this week in Lopass. Like, there's a lot of investment in weird vertical microdramas because they're distributed on TikTok. I don't know. I don't understand any of that, but that's a thing that's happening. You can see the thesis, right? You can make the argument that we'll

Starting point is 01:01:45 buy TikTok and program it with a bunch of Warner Brothers content, and that will make us powerful. I will just warn you that was the exact same thesis, literally the same thesis that AT&T had, only inside of TikTok, it was preloaded garbage on mid-range Android phones.

Starting point is 01:02:01 That was the entire plan. They were going to buy Warner, and they were going to take Game of Thrones and split it into, like, episodic chunks and put a weird app on, like, bundled into mid-range AT&T Android phones, and that was not going to hit your data cap, and then they were going to win. They said this to me with a straight face.

Starting point is 01:02:22 They said this to the United States Court of Appeals with a straight face when they faced their antitrust litigation, and the Court of Appeals was like, I see the vision and let the deal go through. And I, like me, I'm just like, have you ever used a mid-range Anderad film? Like, this is never going to work. But this is the plan only inside of mid-range Android phones. It's going to be TikTok.

Starting point is 01:02:40 Do you think it'll work? No, the other thing that's like devastating here is like Netflix is on this list. The thesis, allegedly, right, Zazlo's thesis is, okay, if we combine these companies, we will have the content to compete with Netflix, right? We will be able to. Oh, Netflix is just trying to drive up the price. Yeah, yeah. They're just being dicks.

Starting point is 01:03:01 They don't want that. They're smarter than to get to. They've got plenty of reality on their own. They don't need this. They're fine. Yeah, yeah. Like, they do not need, they don't need the DCU. Amazon I get, right?

Starting point is 01:03:14 They need the content. They have distribution. Apple, we should talk about Apple in one second, but they're obviously loading up on content to make their service more attractive. The weirdness next to this is all, the other part of the plan for all these companies is they will take the money,

Starting point is 01:03:29 and put it into streaming, which is profitable or growing, at least, and they will allow the cable assets to decline. Very much just sort of like nakedly the thing they're all saying. The problem is if you'll get a streaming business, boy, is that starting to strain under its own weight. So just this week, HVo Max is going to raise prices

Starting point is 01:03:49 for the third year in a row. Hulu with Live TV now costs $90 a month. CNN, All Access, are going to try it yet again, and that's going to be $7 a month. So you're looking at it. consumers are under an awful lot of pressure all over the place, and all of the streamers are raising their prices. Like, does that feel sustainable to you? Hey, how many streaming service are you paying for? Would you pay all these bills? I split a lot of them, I will say. And some of them I secretly

Starting point is 01:04:14 still have for free. But, you know, they're cracking down on that. I would, okay, so I split HBO Max with, like, two other people. Let me think. Netflix, I split with like two other people. Oh, my God. You have no obsex. Not telling the truth. And then I'm trying to think, like, honestly, yeah, I get some of them for free secretly through old jobs and stuff. And I canceled Disney plus. So, you know, I don't know what the amount is. I should look that up and know for my budget.

Starting point is 01:04:46 But, you know, I'm not willing to pay a lot for any of these. But if I'm splitting it with, like, a bunch of people, I'm fine to pay like $5 or $10 a month. It's probably like 30 total a month, like all things consider, maybe like 20. But yeah, I mean, I don't know. I'm a big fan of the free trial, too. You know, some of them, I only need them for like one season a year, one week a year when something comes out. And then I just watch everything in like three days and then I cancel it. Here's my agentic AI product idea.

Starting point is 01:05:14 Then I'm looking for investors if you're out there. An AI agent that just cancels your subscription on time. So you sign up to watch Yellowstone and you tell the robot, like, I'm just going to finish watching Yellowstone. When I'm done, just handle this for me. That would be the most useful AI agent I could ever use in life. I'm saying, and I will sell it to you as soon as someone who knows how to code shows up and gives me money and does all the work. Can I tell you how you can code this thing? Let's vibe code it.

Starting point is 01:05:42 So vibe code an AI robot that just turns you off of subscriptions. Truly the nightmare scenario for these companies. I just say like a lot of, there's a lot of devastation coming to this sector. You can feel it right now. Like I said, the Versant spin-off of Comcast, it already feels like the devastation is coming. Like, there's already, like, NBC News is staying, I think, with NBC Universal. And that means the newsrooms that are leaving don't get access to NBC News, and they have to have their staff up or lower their ambition.

Starting point is 01:06:13 There's just like a lot of questions about how all of that will work. And that's the declining cable business. Like, that's the business that has to figure out how to get to consumers directly. And so my sort of like hottest take is this is the stuff that we have been describing as the mainstream media for 30 years, right? These big, fancy networks that are distributed over cable pipes.

Starting point is 01:06:35 And I think the turn is coming where they're not, where they're not the thing that most people watch, where they're not the thing that set the narrative. And it's, I don't know, YouTubers and TikTokers and streamers that are the mainstream. And I don't think that anybody's ready for that flip.

Starting point is 01:06:51 But you just look at these. business deals, and like none of them make any sense if you believe that there's life left in the cable networks, because they're all betting on the cable networks dying and the investment going to streaming in some way. And that's just, is anybody ready for that to be true? Like, I don't, I don't think so. But the economics are suggesting that that will be true very, very quickly. Do you guys have cable? Because I don't. I have YouTube TV for which I paid an ungodly sum of money, and I don't know why. And then, like, you know, three times a year we watch CNN.

Starting point is 01:07:26 And that's it. You know, like, why do I have? We have football. We have sports. That's what it's for. But it's entirely too much money. And I will say, at the beginning of this football season, I said, I'm not going to do it. I'm not going to pay for YouTube TV and a Sunday ticket.

Starting point is 01:07:39 This is too much money. These children have to go to college. And then I tried to sail the stormy seas for like two weeks. And then I was like, screw it. I'm paying the money. Like, I don't have enough free time to figure this out. Like, it's not good. There's a great Android, if you have an Android TV,

Starting point is 01:07:54 there's a number of browser apps that are optimized for video playback. They're custom browsers that you can download on the Play Store on your Android TV, and they're optimized for full-screen video playback. I'm not suggesting that those browsers are used for going to shady websites

Starting point is 01:08:12 and blocking the pop-ups in the shady websites so that you can watch full-screen video. I'm just saying there's a number of browsers on the Play Store for Android. Android TVs that are optimized for video playback. And you might think to yourself, boy, I could save a bunch of money. And then you might find yourself wasting two hours

Starting point is 01:08:28 and just paying the money. Like, that's very much the calculation that I did. But that's why we have YouTube TV. Jake, do you have cable? I do not. And I try to subscribe to like only one or two services at a time. And just like, like Hayden, I'm binging, I'm in and out as long as I remember to cancel.

Starting point is 01:08:46 And I will say, a thing that reminds you to cancel is the company raising the price yet again. And so, yes, literally yesterday I canceled HBO Max. So thank you, HBO, for the reminder that I have been accidentally paying you for a few months without watching anything. You want to pay off your trillion dollars of AI investment,

Starting point is 01:09:05 $40 a month AI agent that just cancel streaming services on time. It'll be worth it to everybody. I don't know what's going to happen next. I just know that every generation gets its Warner Brothers deal. And they are literally defined it. Like the media ecosystem

Starting point is 01:09:18 reshapes itself around a Warner Brothers deal every time it happens, and it feels like we're going to have another one. So get ready, kids. Like I said, mine was AOL, and it shaped me for years. I still, to this day, and I don't, I can, I'm going to say it out loud because I don't think they even know how to remove it. To this day, I get a Time Warner discount on my AT&T mobile service because I was an AOL Time Warner employee when I worked at Engageant.

Starting point is 01:09:44 That's crazy. That's just a weird fact. I used to work for AOL Time Warner. That's nuts. And AT&T gave us a discount when it bought Warner. And I don't think they know how to turn it off. I don't think they'll. I don't think, I don't, I don't, I don't, there's a, there's a mainframe in a, in Virginia basement running cobal.

Starting point is 01:10:03 It gives me a tiny little discount. Full disclosure. I get a weird discount because I used to work for AOL Time Warner and the companies have merged and unmerged enough times. Every generation gets its Warner Brothers deal. We're about to face another one. Generation. All right, let's take it on the break.

Starting point is 01:10:19 We're going to come back. We're going to get a wetting around. We're getting it out of here. It's going to be amazing. We'll be right back. Support for the show comes from Anthropic. Not every question has an easy answer. And the ones that are really worth asking usually come with a healthy mix of inspiration and backpedaling,

Starting point is 01:10:36 aha moments, and quiet meditation. When you're working through one of those problems, you want a partner to bounce ideas off of and figure out where the deeper issue lies. That's where Claude can help. Claude is the AI for minds that don't stop at good enough. It's the collaborator that actually understands your entire workflow and thinks with you, whether you're debugging code at midnight or strategizing your next business move. Claude extends your thinking to tackle the problems that matter.

Starting point is 01:11:05 Plus, Claude's research capabilities go deeper than basic search. It can have comprehensive, reliable analysis with proper citations, turning hours of research into minutes. Ready to tackle bigger problems? Get started with Claude today at cloud. combe. That's clod. com slash vergecast

Starting point is 01:11:29 and check out Claude Pro, which includes access to all the features mentioned in today's episode. Claude.aI. slash vergecast. Support for the show comes from LinkedIn. If you're a small business owner, you know that every hire counts,

Starting point is 01:11:46 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. Hiring Pro streamlines the entire process from drafting your job to shortlisting candidates and conducting AI-powered interviews for initial screenings. It's updated conversational interface lets you describe what you need in plain language.

Starting point is 01:12:25 Nearly 60% of hirers find a candidate to interview within a week. With hiring pro, you spend less time searching and more time connecting with the right talent. And instead of getting buried in resumes, you get a focus shortlist 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.

Starting point is 01:12:52 Terms and conditions apply. Complex and unprecedented, the Spanish authorities are calling it. Before the disembarko, asymptomatikas. 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.

Starting point is 01:13:21 We do have one individual. who was taken to the biocontainment unit early, early this morning. And we assessed that individual. 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. Look, I know I came back to the show last week, triumphantly,

Starting point is 01:14:02 and then I was completely unprepared and went into Brennan Carr's dummy. I got a lot of notes about it. So it's back this week, America's favorite podcast within a podcast, Brendan Carr is a dummy. Now, I know whatever one expects me to talk about here on Brennan Carr's telling you. Brennan's been very busy. He's been very public with wanting to regulate speech to America. That's what the man wants to do.

Starting point is 01:14:21 You all know it. You've seen it. His name is famous now because of the Jimmy Kimmel stuff. When John Oliver is doing a Brennan Carr segment, I know that we've penetrated the public consciousness here at the first cast. My job is to not tell you what you already know about Brennan Carr. My job is to tell you that this man is so stupid, he's so bad, fundamentally bad at his job, that even when he wants to regulate speech, his own ghost will get in his way,

Starting point is 01:14:44 in surprising and honestly hilarious ways. So this week on Brennan Carr's a dummy, I submit to you that he has decided that the FCC will somehow preempt all state regulation related to artificial intelligence. If you remember, Hayden, I think you and Lauren Finan covered this bunch, in the big, beautiful bill, there was a provision that would preempt all state regulation

Starting point is 01:15:08 of AI for 10 years. and it looks like it was going to go through and there was some outcry from both sides of the aisle and I got blocked and it's not in it anymore. So then Brendan, White Knight that he is to the tech industry, said, I'll do it. I'll find a way for the FCC to block all state regulation of AI.

Starting point is 01:15:27 Okay, here's how he's going to do it. As far as I can tell, he's got some, you know, and like what I would call sort of the classic Stephen Miller, Russ Vaught, like, project 2025 way.

Starting point is 01:15:42 They found an obscure section of the law that says one word they like, and they're like, the whole law is about this one word that you've never thought of before. So he found 47 U.S. Code 253, which says, no state or local statute of regulation

Starting point is 01:15:53 may prohibit or have the effect to prohibiting the ability of any entity to provide any interstate or interstate telecommunication service. But here's the thing. He's got to say that AI is a telecommunication service. He's got to classify AI as a telecommunication service.

Starting point is 01:16:09 so the FCC can regulate it and then preempt all the state regulation about it. If you have been a Vergecast listener, you know that the second I read this, I turned in, I melted. I turned into a literal puddle because I spent, I don't know, the first 10 years of my career, Drake here at the Verge covering net neutrality, which boiled down to a fight over whether broadband was a telecommunication service could be classified as a telecommunication service. and Brendan Carr and his old boy, Ajit Pie, fought tooth and nail to prevent broadband

Starting point is 01:16:44 from being classified as a telecommunication service so the federal government might regulate it and prevent ISPs from throttle in traffic and all the stuff they wanted to. So if broadband, the literal pipes of the Internet are not a telecommunication service, there's almost no way for Brendan Carr to claim that AI is. Because AI, somewhat notably, runs over broadband.

Starting point is 01:17:06 So the ghost, the deregulatory past Brendan Carr is preventing new and improved MAGA Brendan Carr from preempting state regulation in AI because he already deregulated broadband so hard that he has no ability to preempt state regulation. This is like crazy. I feel crazy talking about this. This circle is nuts,

Starting point is 01:17:29 but this is how dumb he is, that his past efforts to appease the ISPs are now in the way of his. efforts to appease the AI companies. And he will never admit it. He's just going to keep saying that 253 gives him whatever power he wants. But fundamentally, he has to say that AI is the exact kind of thing he said broadband isn't. And if you can make the argument that AI is somehow more of a telecommunication service,

Starting point is 01:17:57 more pipes than broadband itself, well, please send me a note. Brandon, as always, if you're listening, I'd love to have this debate with you, either here on the birdcast, the back of the bar, an alley, wherever you want it, buddy, we can do it. But I cannot, literally cannot see how Brendan Carr can claim that AI is a telecommunication service after he spent the majority of his career arguing that broadband was not. And so we're going to cover this more because he's obviously going to keep trying. But I'm telling him he's just a dummy.

Starting point is 01:18:29 I'm just telling you's dying. Brendan, you're always welcome on the show. That's been Brendan Carr's a dummy, America's favorite podcast with my podcast. Nilai, that was lovely. Can I just ask, do you need to get a Title II in there? I said I couldn't do it. I spent so long covering net neutrality. Like, so long.

Starting point is 01:18:45 And whether broadband was a Title I information service or a Title II telecommunication service is the heart of that entire debate. Yeah. Title I information services were like old dial-up AOL where you would buy access to a service and it would contain content, right? AOL had email and forums

Starting point is 01:19:02 in title, two telecommunication service are just the pipes. It's like your phone lines are just pipes. And the ISPs argued for years that they were information services. That because they provide things like free email addresses and web storage, this is what Verizon argue. We're not pipes because you get five gigs of free cloud storage on Verizon.net or whatever that was. That was that whole fight for a decade up and down to every court in the land.

Starting point is 01:19:31 And Ajit Pi and Brendan effectively won the fight. right, right? They are in charge, and they said, as hard as they could, we are not going to allow broadband to be Title II services. They wrote themselves out of the equation so hard that they lost their own ability to tell states what to do. They lost those cases. So California, New York passed net neutrality laws, and the ISPs sued, as they always do, and said federal law preempts this. And then courts are like, no, no, no. Ajeet Pai erased himself. He took himself out of the picture and now the states can do whatever they want. federalism, baby. And so now we're doing it with broadband, but only this time they want the power. It's very good. If you're like into this kinds of weedsy stuff, like you just have Brendan Carr thinking he's so smart,

Starting point is 01:20:16 but he's tied himself into a knot. Sorry, you said Title II and I went off another round. This was supposed to be fast. I was provoking. I was, no, this is one of the most insane turns imaginable for the FCC. Just truly, oh, what if, we just pretended none of that ever

Starting point is 01:20:36 happened? The main thing people were mad at us for for a decade. What if we just go back on that for a very narrow use case? It is almost as though it would be useful if these were dumb pipes. It would be a lot.

Starting point is 01:20:53 Anyway, here we are. Brendan, if you want to talk about your plan to regulate AI by saying it is more of a utility than broadband itself, please, I welcome you. But I'm back, buddy. We're going to do this every week. Your Google Alert's about to go off.

Starting point is 01:21:08 I know you read them. All right, lightning round, unsponsored for flavor. Hey, let's start with you. What's your lightning round of him? Okay, so one of mine is the fact that there was like a friend AI protest in New York recently. You know, it's a little like necklace that you wear on your neck and it's supposed to offer commentary on your day. The Verge reviewed it and said it sucked and couldn't hear anything and also was like weirdly combative. of and, like, sassy in a way that was not fun.

Starting point is 01:21:36 I will say V reviewed it, and I can see how V got an AI assistant to be, like, weirdly combative and sassy. It was very sassy. It also just couldn't hear anything that was going on. Like, it was rough. And so, as we all know, if you take the subway, there are nonstop friend ads all over the subway, and the graffiti has been hilarious. It's, like, truly works of art.

Starting point is 01:22:00 People are getting so creative. It was so intense that the New York Times even did a story on the graffiti itself on the ads. It was like they spent over a million dollars on this subway campaign. And yeah, you can't go anywhere without seeing a picture of the friend device in the subway, basically. And so, you know, even though the device came, you know, was announced a long time ago and then, you know, but the subway ads actually only started in September. So that's why this protest happened. It's like everyone's been talking about it because of the ads.

Starting point is 01:22:30 And so this coalition of people that are, you know, really upset with the fact that it's being called a friend and saying like, oh, I'll never, you know, leave the dirty dishes in the sink. I'll always be there for you. Stage this protest in Manhattan and just said, hey, like, everyone come and share your thoughts. Let's talk about why you should have friends in real life instead. And the funniest part is they invited the friend founder, Avi Schiffman, and he took a red eye and came to the protest. And then he also inexplicably, like, you know, took a ton of pictures and videos of the protests and shared them on X. And then he actually sat in a circle with the protesters, apparently, after the protest and, like, signed a handwritten document that he would never sell a friend to, like, a large tech company for surveillance purposes. There was a lot going on there. It's a crazy story. I talked to some of the protesters after the fact and added some of those details in there.

Starting point is 01:23:24 But, yeah, it's interesting. Did he sell one friend to anyone at the protest? No, he was definitely just there to like, he apparently brought a film crew with him. Oh, God. To, like, film it. So, yeah, I mean, it was definitely a trip hearing about this. I wish I'd been there in person. But, yeah, they also apparently, like, had a homemade, like, cut out or, like, facsimile of the friend, and they had someone wear it.

Starting point is 01:23:51 And then they all, like, kind of, like, ripped it apart ceremonially. they had a big banner that you could write on with Sharpies about why you didn't like the friend. But then they also had like a conversation with, you know, the founder and talked about what they liked and didn't like. And it was a very wide-ranging, apparently like long protests. I understand like what this will sound like in the context of protest coverage, writ large in America. Are we sure these weren't paid protesters? I thought the same thing. Believe me, my first question, I DMed Avi Schiffman and I was like, hey, did you, you know, plan this?

Starting point is 01:24:24 because the verbiage on some of the posters was also like, hey, New Yorkers, I heard you got beef with me, like come say it. So it seemed like it was planned by the company to me at first. But then he said he didn't have anything to do with it. Then I heard from the actual protesters who contacted me and were like, no, he didn't plan this. We planned it. We just didn't want our names out there, really.

Starting point is 01:24:45 So, yeah, it seems legit now that I've spoken to the actual people that planned it. But it is interesting because at first that was my thought. Yeah, I will say shine bright, friend, that is as much attention as you will ever get, because I don't think the core LOM technology can do the thing your product says it can do. This is the problem, right? Like, you're making a lot of promises that the steam engine inside your product cannot actually deliver upon. It's a long way to go before you can just have a friend that listens to all the time. I'll also never get over how much they spent on the URL, you know?

Starting point is 01:25:18 How much they spent on the URL? I was on leave when this happened. $1.8 million for the domain. For a URL, that's nothing. Yeah, but I think it was like most of their funding, right? Yeah, but how much revenue are they making on these things? Proportually. All right, it makes sense.

Starting point is 01:25:33 All right, Jake, what's yours? Okay. I don't know if any of you tried to use the Internet earlier this week, but it was not available. Most of it was not working. I feel this happens like every, I don't know, other year, there's just a massive AWS outage. That's Amazon's web server. arm. The entire internet runs on this thing, right?

Starting point is 01:25:55 Especially East One. It's AWS East One is the one that always goes down and it's the one that takes the whole internet down. This is the biggest. I think it's maybe their oldest one. And just people, right, tech companies rely on this thing. It is one of the most critical pieces of internet infrastructure. And when it has a small issue, things big and large start to break.

Starting point is 01:26:16 Right. So, uh, Fortnite wasn't working. Snapchat wasn't working. This also leads to some weirder issues, right? Alexa wasn't working, which meant that Alexa alarms weren't working, which meant that people didn't get woken up, right? There were all kinds of... And those people were asleep in their eight sleep beds,

Starting point is 01:26:32 which were no longer working because the sleep requires a cloud connection to work. That was the craziest to me. Yep. I saw some smart cribs were not functioning. That's not good. And it's like, right, like every single thing points back to AWS. And, you know, according to them, this was a small error. that just cascaded out of control.

Starting point is 01:26:52 But this is the thing that happens every couple of years. And it easily wipes out millions, if not billions of dollars worth of revenue for all of these companies involved because people just cannot use their services for most of a day, right? Amazon was having trouble, I think even well beyond the actual day where this happened. I think it started around like 3 a.m. was mostly wrapped up about like 15 hours. hours later, it's just this is such a critical piece of infrastructure that when there is even a small hiccup, the entire web fuels it. And I think, you know, we tend to think of the web as this, you know, impossibly distributed network. And it's like, nope, it's actually most of it's on Amazon.

Starting point is 01:27:40 Most of it's in one data center. Yeah. Yeah, yeah, yeah. Even Signal went down, which was crazy. But yeah, I remember exactly when it happened because I was woken up out of a deep sleep by like, a blaring alarm because my fiance was on call. And, yeah, apparently the whole night, all on-call software engineers were on Reddit talking about it and being like, hey, did you see this? Did you see this? Is this normal? Is this normal?

Starting point is 01:28:04 And one guy was like, yeah, it's my first day ever being on call in my life, and this happened. So, yeah, it's definitely a crazy, crazy sedge. That's very good. If you know any information about why it's always East 1 that seems to have the most trouble, we're kind of like looking into it. Let us know. Reach out. There's something about East 1 in particular.

Starting point is 01:28:23 I think, as Jake said, it's the oldest. It has the most dependencies. I'm very curious if we can get some insight into why it's always East 1. Because it is one of those things that everybody can see, but nobody talks about enough. I would like to do a story on it. All right, mine, I'm just, mine is just like straightforwardly a shameless plug. Mary Barra, the CEO of GM, was on Dakota this week. Talking about all of GM's announcements, along with Sterling Anderson, GM's new chief product officer.

Starting point is 01:28:47 They're putting Gemini in the cars, the 20 and 20, escalade will have eyes off self-driving, so level three. So right now, if you have Super Cruise on, you have to look at the road or it freaks out. If you look at the side mirror in my car, my GM car for too long, it freaks out. And I'm like, I'm literally looking in the mirrors. But if you're not looking dead ahead in the road,

Starting point is 01:29:06 it freaks out. But level three coming, eyes off. You know, they're going to put basically a data center in the escalate because the escalate is big enough. It's like a bunch of announcements. And then, of course, what do we talk about? Talk to our carplay. and they are fully doubling down on not having car play.

Starting point is 01:29:22 They're going to take it out of the gas cars, which drove its own little news cycle right on top of, you know, level three self-driving and whatever else they're announcing. I was like I really liked talking to Sterling Anderson. He's new. I think he has a thesis for why he's doing it. Here's what really got me, though. I have a new GMEV, and we have a Cadillac Vistic.

Starting point is 01:29:41 You can talk to Google Assistant in it. My daughter loves asking a car to play Taylor Swift. It is now her favorite thing to do is to be in a car and to just screen a Google assistant to play whatever music she wants. And I just always see her, K-pop demon hunters or Taylor Swift. My Spotify is a disaster zone. K-pop demon hunters. That's all of my Spotify has been playing at all.

Starting point is 01:29:59 It's going to be my whole Spotify wrapped. We got to get you out of Max because it's going to be, that's, she is going to be a K-pop demon hunter for Halloween. So am I. If I can figure out the costume. There, yeah. So we'll get everyone together. Anyway, all I'm saying is they talked about moving on from CarPlay and building a

Starting point is 01:30:17 platform in the car and Gemini-powered voice assistant, and you can chat with Gemini about where you're going and what the best restaurants and the town you're going to end up with it. And I didn't ask one question. So this is my lightning around item. It's the question I did ask that, like, the second we hung up, I was like, oh, no, I forgot to ask the question that ties altogether. If the car is driving itself and you don't need your eyes on the road, why wouldn't you just look at your phone?

Starting point is 01:30:42 And I'm like kicking myself for having not asked this question. because if your thesis, we've got to get rid of this stuff so that you're using the systems in the car, which is the thesis. It's Rivian's thesis. It's Tesla's, it's everybody's thesis.

Starting point is 01:30:55 But then also the car can drive itself. Now your thing in the car is competing with your phone. More directly than ever. You know what they need? I did, it just, like, I literally, you know, like you're in the shower,

Starting point is 01:31:07 like a day later and you come up. I was like, oh, no. So that's my liner. I missed it. They need original content. They should buy Warner Brothers. This is how it goes. Anyway, a lot of discussion about CarPlay.

Starting point is 01:31:23 The one thing that really gets me, you know, I read all the comments on the thing, people are very mad, right? People feel a lot of ways about CarPlay. I personally, I think it's fine. It's like not the best, it's not the worst. Most of the carmakers have not adopted CarPlay Ultra. They don't want to give up control of their customer to Apple,

Starting point is 01:31:38 not specifically. We've heard about that. Ford, Joanna interviewed the CEO of Ford, and Dakota while I was out and leave. He basically was like CarPlayer. Ultra isn't good enough, I've given Tim Cook the feedback to which as a listener, I'm like sitting in my house being like, good luck, Jim, you know, like, go see if that, see if that goes well for you. But the reaction of GM stuff is like, good luck, we're never buying a GM car or you're just

Starting point is 01:32:00 trying to use your market power. And I'm just curious, I'm wondering if people will give us feedback. One, there are lots of carmakers who will happily sell you a car with car play, right? Like, Ford will do it. Hyundai will do it. Kia will do it. Volvo will do it. Like, all the other cars have carplay in them. Honda and Toyota famously didn't, and they caved and they added it. The Honda Prolog is a Chevy Blazer with CarPlay. Like, you can literally one-to-one buy the same car, just with Honda logo and CarPlay, you're fine. So, like, why that reaction, like, GM made a market decision, and maybe they'll succeed or they'll fail.

Starting point is 01:32:33 And their EV sales were up this last quarter. So it seems to be working out fine. So there's a split there, and it feels like the split that we have with AI, where whenever we write about AI, everyone's like, we hate it. and then the usage numbers are off the charts. And there's something there that's really interesting to me. So I'm curious about that. Why it feels like one company making this decision is such an affront when you can just go buy another car?

Starting point is 01:32:55 I would love that feedback. And the second one is, is there ever a car system that would be so good that if the car was driving itself and you didn't need to look at the road, that you would pick it over your phone? Because that's the one that's like, all of them are going to crash into this.

Starting point is 01:33:09 Tesla, if FSD gets so good that you don't have to look out the, the windshield, because right now FSD makes you look at the windshield, why would you ever use their screen instead of your phone? And I have not yet heard of answer. This is a question I ask, so I'm dying for the feedback. All right, one more set of lightning around items. Hayden, what do you got? So Open AI was really under fire recently for sending sub-peonuts to a ton of nonprofits in their like battle against Elon Musk, the lawsuit stuff. A lot of the nonprofits only had like one to three employees. You know, it was a smattering. I think it was seven

Starting point is 01:33:42 in total that received them over the last couple months. And the craziest part was not that they sent them, because, I mean, they're just trying to figure out basically who is supported by Musk in terms of, like, you know, bad-mouthing or, you know, pouring cold water on opening eyes restructuring. But the thing that was interesting was how wide-ranging the subpoenas were. Like, they were trying to get every document, it seemed like, that related at all to opening eyes restructuring, any communication. So, like, with one nonprofit, for example, like, they put out a bunch of, like, web pages about Open AI's, you know, history and journey and investors and stuff like that as a public good. And Open AI wanted to know every single communication they've ever had regarding its own restructuring or anything like that. So, I mean, that would involve them turning over, you know, tons of communications with journalists, ex-employees, investors, and just research they had done.

Starting point is 01:34:41 And, you know, that nonprofit in particular had, like, I think one employee at the time. So it would have sideline a bunch of them, too, all these requests. So they're pretty wide-ranging. Open AI's response was, oh, like, you know, this is pretty normal. It's supposed to be a back-and-forth. We asked for more, and then they come back at us. But that's how it's supposed to be when it's two large corporations doing, like, lawfare against each other, not, like, a big guy and then, like, small nonprofit.

Starting point is 01:35:04 So, yeah, it caused a lot of drama online. And even some Open-AI employees came out and said, that they were not really happy with that. So, yeah, it's been interesting to read them myself and talk to all the people that received them and where they were when they got them and what's happened since. A lot of them have just said,

Starting point is 01:35:22 we're not doing this, or they've hired, like, pro bono, legal aid, so that's been helpful. But, yeah, it's been crazy. Did you see that there was another story in the same vein, Open AI, subpoenaed the family of the boy who committed suicide after talking to chat Chhabit,

Starting point is 01:35:38 and they subpoenaed the eulogies at his funeral? Yeah, the first. full attendee list of the funeral, everyone that spoke, videos, everything else. Yeah, it was a lot. Is that just their law firm as a strategy, which is to harass people under the ground, or do they have an answer for that one? I haven't asked them their explanation yet, but it seems like from what they said about the other story, that they would say, well, this is relevant because they're suing us, the family. And we want to see if, you know, maybe some people were saying, well, Maybe it's because they want to see if there were other incidents that led to it.

Starting point is 01:36:14 Maybe people would mention their speech, whatever. But I really, I would love to hear their explanation. It's a weird way. Open Eye is just in all kinds of weird zones lately. There's something about that company to be done. Someone should do something at Opening Eye. That's the word chest. Jake, what's your second one?

Starting point is 01:36:33 Yes, big, I think, earth-shattering news. Pitchfork is beta testing a comment section? and the ability for users to submit review scores. They're going to average them all up and allegedly display them alongside their own review score. I think this is going to have a monumental cultural shift. I think this is going to lead to a lot of chaos on pitchfork.com. This headline should have been pitchfork admits defeat to Kalk fans.

Starting point is 01:37:07 Fandoms have defeated pitchfork, is the answer to this question. This is, the best pitchfork scores are the ones where you look at them and go, I don't understand what on earth happened here. And I think this allows for too much correction, right? It allows for too much market correction. You don't want that. It's an editorial product.

Starting point is 01:37:28 It's trying to tell a story. That story is a little chaotic and unexpected. And, you know, honestly, at a regular basis, I've been reading Pitchfork for probably 20 years now. on regular basis, I scroll to the bottom of their reviews to look and see what the heck the commenters think and they are not there because they have never had a comment section. And so this feels particularly notable in a music magazine

Starting point is 01:37:50 because, to your point, Eli, there has been a lot of, I don't know, perhaps bullying from pop fans who want to, you know, be more respected among the critics and, you know, pitchfork among, I think, all of the music websites have had to reckon with that. And this is certainly one way of going about it while also potentially increasing engagement on a dying web. I'm just going to say this with no further explanation and no further discussion. Pop-timism was a mistake. That's it.

Starting point is 01:38:27 That's my verge cast, I'll take. And I think Jake knows what I'm talking about, but I'm not going to let him say another work. What's your life of a showgirl review? I have a seven-year-old dog. and I declined to comment my true feelings on Taylor-Sifting album. There's a big story here about fandoms, and what fandoms have done to cultural criticism across the board. Most major publications are losing their credits.

Starting point is 01:38:50 Like, New York Times, restructured all of its cultural criticism, and they don't really have critics in that way anymore. The Life of the Showgirl, I will note, launched with some very, very positive early reviews embargoed. There's just something happening to that ecosystem, which is really weird. And I think, as far as I don't know how pitchfork works anymore. But like back in the day, pitchfork reviewers didn't pitch reviews.

Starting point is 01:39:14 They pitched scores. They're like, we're going to review this album and I'm pitching you the score. And then they backed into the review from the score, which is crazy. And like that was their brand was the score. Notably the score is it's a hundred point scale. Right? It's 10.0. So you have a, the idea that you can score music on a hundred point scale is bananas to begin with.

Starting point is 01:39:36 Anyhow, I'm excited for them to try. We'll see. But I hope they have a sophisticated bot detection network. That's what I have for you. All right, here's my last one. It's on the same theme as every generation gets its Warner Brothers deal. Tom Warren pays a lot of attention to Microsoft, as usually does. Notice that Xbox president, Sarah Bond, did an interview with Mashwell this week.

Starting point is 01:40:02 Mashwell asked her if the next-gen Xbox will be more like a gaming PC. and Bond said, I can tell you, you're right, the next-gen console is going to be very premium and very high-end period in experience. You're starting to see

Starting point is 01:40:11 something thinking we have in this handheld, but I don't want to give it all away. And so Tom has had a lot of reporting over the past year that the next generation Xbox will basically be a PC. And here we go.

Starting point is 01:40:23 They're going to do it. They're going to make a Windows-based Xbox and putting your living room because that's where they're in the handheld. And I'm just going to point out every generation gets Microsoft trying to put Windows in the living room. It doesn't matter who you are.

Starting point is 01:40:34 It doesn't matter. What your lived experience is, if you are alive, sometime in your lifetime, Microsoft will try to put windows in your living room. It's just the eternal return. When I was a child, Bill Gates, pre-Iphone, Bill Gates would show up at CS every year and give a huge speech about smart homes and convergence. And he would say, I'm going to put windows in your living room. And then we would have CS coverage in the local news stations would be like, look at this knowledge navigator video, like, whatever nonsense they were doing. And then he would show at CS and there would be a demo and the local news stations would breathlessly cover

Starting point is 01:41:07 and they'd be like, look at this concept video Microsoft put out of a guy with a pocket PC ordering raspberries and his nonsense, pure nonsense. It's coming again. They're going to try to put Windows in Living Room. You recall the Xbox One was Windows in Living Room. Do you remember it ran three operating systems? It had a hypervisor.

Starting point is 01:41:23 It ran Windows and it ran the Xbox operating system and had a cable tuner because Microsoft can't stop trying to put Windows in living room. And if you're a young person listening to this, I promise you. One day your kids will have the experience of Microsoft trying to put Windows in living room. And you will tell them, this happened to me.

Starting point is 01:41:40 I know what this is like. That's it. That's all I got for you. I don't know what's going to happen this time. Maybe it'll work. History suggests that no one wants Windows in their living room. You don't know what's going to happen this time, really? I mean, I said this on Blue Sky.

Starting point is 01:41:53 They never tried this on Xbox before. And a number of people sent me the logo for Microsoft Windows XP Media Center Edition. Like, that's how long we've been doing this. I'm excited for them. I don't know why anybody wants Windows in living room. I don't know why anybody wants Windows in a handheld. I just know that deep in Microsoft's DNA, like, at the base layer of the organization,

Starting point is 01:42:16 there is the desire to put Windows in a living room. Like, that's why Microsoft exists. It's to get Windows into living room. By hook or by fucking crook, they're going to put Windows in living room, and maybe they'll back off and they'll forget about it, and they'll think that I forgot about it, but I didn't forget about it

Starting point is 01:42:30 because it keeps happening to me, and it's going to happen to your kids, and it's going to happen to your kids, kids. That's just life. We'll see what this new Xbox is. There's a lot going on with Xbox. There's much more to talk about it than that. There's a lot of reporting this week about their profit targets

Starting point is 01:42:43 and why they cut so much, that-da-da-da-up. But at the end of the day, what does Microsoft want to do? They want to put Windows in a living room. And I'm happy. Sometimes you have to be yourself. All right. If you know more about Microsoft XP Windows Media Center Edition, please let me know.

Starting point is 01:42:58 I'm here to talk to you about it. That's it. We got to wrap up. We're so over. David isn't here. I'm just like out of control. I'm excited to be back. We're back.

Starting point is 01:43:08 It's the end of the year. There's lots going on. Hayden, Jake. Thank you so much. That's it. That's Virgcast. Talk to roll. The Vergecast is a production of the Verge and the Vox Media podcast network.

Starting point is 01:43:22 It's produced by Eric Gomez, Brandon Kiefer, and me, Travis Larchek. If you have questions or comments, we love to hear from you. call the hotline 866 Verge11 or email us Vergecast at theverge.com. All right, thanks for listening. See you next time.

The Vergecast - ChatGPT enters the browser wars

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