Welcome to the Vergecast, the flagship podcast of foundation model foundations, which are a thing that we understand. Uh, I'm your friend David Pierce. Nilay Patel is here. Hey, buddy.
There's so many concepts in this episode of the Vergecast. We literally got an email that was like, can you explain to me what a foundation for a foundation model is? And yes, in fact, that is what we are going to attempt to do here on the Vergecast. Uh, David, have you recovered from CES? How are you feeling?
I've recovered. I think I'm feeling better. You know, this is like the first working week of the year for us for real. So there's a lot of catching up to do because everyone else is one week ahead. Uh, and then everyone got sick because that's what happens when you leave CES. Um, but I think we're ready to go. And then there was a lot of news this week and we just have to start with a big disclosure update. I think
My favorite kind of news story, the news story that is like irritatingly about us and also requires us to figure out how to disclose for a long time.
Yeah. So, uh, Vox Media has sued Google. That's the news.
Uh, and so have The Atlantic, Penske Media, which disclosure within a disclosure. Penske is an investor in Vox Media.
Uh, Condé Nast, their parent company, Advance Publications, but everyone just knows him as Condé Nast. Advance Publications, by the way, famously, uh, owns uh, an interest in Reddit, right? Uh, which is very funny. They incubated Reddit and then did the IPO. Um, but that's the parent company of Condé Nast which owns Wired and Vanity Fair and everything else and then Mlache which owns a bunch of other publications. They've all of these media companies have sued Google uh for antitrust violations. And basically what happened was Google was sued by the government last year for adtech antitrust. So DoubleClick for publishers and how ads get displayed around the web and all this stuff. And they lost like running away. They lost. They're like, "Yep, this is a huge monopoly of how advertising works on the web." That was the case where Google's lawyers had the quote that was like, "This doesn't matter cuz the open web is dying." And then everyone ran that quote. And then Google had to like walk it back and explain that what they really meant was display advertising on the open web.
Yep. Which is the same thing as the open web because if you don't have any money on the open web, that the web itself is under a lot of threat. Anyhow, that case came to an end. The government won. It said Google had an illegal monopoly in adtech on the web. There is not remedies in that case yet. But what you have is proof that Google has an illegal monopoly on web advertising. And so all of these media companies are like, well, we got hurt by that. You owe us money. And so one by one this week, it started with Penske on Monday and The Atlantic on Tuesday and Vox Media on Wednesday and Condé on Thursday. Like just down the line, they've all filed their lawsuits obviously to get attention. We had nothing to do with it. I literally found this out from our company's like press statements uh because they don't tell us. There's a wall that you know the whole ethics policy. There's a big wall between us and them. So it's very funny that they told us like they told everyone else. Uh and we wrote about it of course. But anyway, uh Vox Media, our parent company, is suing Google for antitrust violations in adtech along with basically every other publisher in the world. I'm assuming there's going to be a new one every day. Like who will it be today? It is a pretty fabulously coordinated system here that it's just like somebody is going to file and all the lawsuits are sort of functionally identical, right? That and and you described it to me uh as it was coming out as like they're all just pointing it they're all pointing to the lawsuit from last year being like see they did it they also did it to us can we have some money please just over and over that's essentially what it what it is and everybody wants to negotiate separately. It also I think there is power in doing it a bunch of different times instead of all together all at once. Like publishers are suing Google is somehow like less meaningful to me than this like individual drum beat of publishers suing Google. And so I think it's this is very clearly being calculated on a bunch of different levels.
Yeah. I mean, I'm very hesitant to say anything about it because I don't want anyone to make the mistake of thinking that we have any insight into this strategy or that we are driving the strategy or made any of these.
I don't know anything. So, I can just talk.
Yeah, David like I think we all understand David doesn't know
You at least know the lawyers on our team who didn't talk to you about this.
I've never met any of them.
I spend a lot of time with our lawyers on like defamation claims and you know, pre-publication legal review and all that stuff. So, I do know them. They're all great, but they did not tell me about any of this. and you know they're very careful about drawing the lines appropriately.
Um but it is true they all filed separately. It's all the biggest media companies in the world. I would assume all the rest of them that aren't on the list so far are next. They're all in trade organizations together. They all all these executives have been talking about Google forever. I'm shocked that like the New York Times and News Corp haven't made these same decisions because they beat the same drum all the time. Anyhow, that's it's happening. the the thing that is relevant to us is that we have to disclose that our company is suing Google all the time. Now, y
My instinct and I'm willing to hear from the audience how you would like us to handle this. It really doesn't affect our coverage and Google is vast. So, I think we should disclose it, you know, often and early. We do it all the time. I just don't want every time we half mention Google in a sentence to require us to be distracted by a thing that doesn't affect the cover. Like that's the dance with these disclosures. So I think we should do it when we talk about the web, right? And like web advertising. And I don't know if we need to do it when we talk about Android.
Yeah, I think that's right.
But you tell me. Like I'm open to the feedback. If you want me to do it every single time, we'll do it every single time.
Yeah. Disclosure is our brand after all. Um, email us
[email protected]. This is like a thing I wonder sometimes is frankly if we like disclose performatively to like be proud of ourselves about our disclosures. And so I think finding that balance of like we have to tell you all of this stuff versus like look how good we are at disclosing all of our conflicts uh is a thing I wonder about. So we I do want to hear from people. But I think it was a very funny uh 24 hours because we also I also wrote a story yesterday about how Gemini is just like kicking everybody's ass in the AI race, which we're about to talk about. And I got nice I got nice emails from from people in the know at Google and elsewhere who were like, "Wow, you're really nice to Google right next to like we just our company just sued Google." So this is like this is this is what we live with every day as as journalists. Also, you uh Lauren Feiner covered the trial last year for us, like sat in the courtroom through all of this. And we talked a lot on this show about how consequential that trial is uh for a lot of reasons. The the remedies piece of that trial is still coming. Uh go read Lauren's coverage. Um and I think if you've listened to this show for longer than 5 minutes, you you know how Nilay and I feel about the the state of the open web. The thing we're going to keep talking an awful lot about. Um, but for today, the Google news of the week and I think the biggest news of the week is a couple of different Gemini related things that happened. Uh, and this is where we get into foundation model foundations. Um, the first thing that happened, this was kind of the the like driving first piece of news this week was that Apple and Google co-announced that the next generation Siri will be in in some way powered by Gemini. And I say in some way because I think that the details are really important here and we should get into them because we know some and we don't know others. But basically, Apple has been running this this bake off. Uh Mark Gurman at Bloomberg has covered this really well over the last like 12 months testing all of the different stuff trying to build its own AI internally and eventually seems to have decided that the best path forward to make Siri great and some of Apple's other AI stuff is to just turn it over in a pretty meaningful way to Google's Gemini models. This strikes me as a very big deal but maybe not exactly for the reason people have been talking about. What do you make of all this?
I think training a foundation model is very time-intensive and very costly. It is also not that much of an advantage right now to have your own. That's the best unless you're going to sell it to lots of other people because that's how you would make the money back from it. And that's the business that all of the model providers are in. That's Anthropic and Google and OpenAI just down the line, right? They all spend a bunch of money to train models or they find ways to train models that are really good on the cheap like DeepSeek. And then the goal is to sell that capability to lots and lots and lots of application providers. There's some question about whether the models will crush the application providers. That's like part of the bubble. Like if the model is just good enough and the chatbot just good enough, will you ever need to build a specialized application?
But that's a lot of complexity that Apple doesn't need to manage because Apple is itself an application provider, right? Like Apple makes hardware because it wants to make great software. This is the founding DNA of the company. And so you don't need to run the cloud infrastructure to do it unless you have some reason to do it. And it has always just made sense to me that Apple would turn to some external provider to do this.
I think a lot of the conversation about is Apple behind in AI was they didn't see it early enough to build the capability for themselves as they want. Like Apple wasted a bunch of money trying to do a self-driving car.
Mhm. They should have spent every one of those dollars back then building data center capability and training their own model that could be as private as they want. And they just didn't see it. That was Apple's miss like straightforwardly. They put a lot of local compute neural engine stuff on their phones and they didn't see the other revolution that was happening along the way. Fine. It all that occurred. We never got to see Jony Ive's car.
Sure. which again I would remind everyone the bestselling car in the country is like a midsize crossover and Jony designing a mid-size crossover was fundamentally doomed from the start.
Um anyhow I just don't think it matters if the back end is Gemini or OpenAI or Apple's own model that they built because the the thing that no one has really made is the product that brings it all together and does all of the stuff that needs to pay off the bubble. I could caveat this. I think there's important Apple lost and Apple is violating its own principles arguments to be made but the point it is Apple is in a position to build the product that pays off the bubble and I don't think the model underneath it is the important part of that product.
Okay, I have two thoughts about that. One is I sort of foundationally agree and I think I actually don't buy the theory that what Apple should have done 10 years ago instead of trying to build a car was build a bunch of data centers. I think sure I think what Apple should have done 10 years ago is like probably sign this deal with Google. Do you know what I mean? Like at this there were there were opportunities to do this a long time ago and to understand that this was going to be a meaningful thing. I mean the the attention is all you need paper was in 2017 and a lot of people then were like this is going to change things and it was like people Google was on top of that in a lot of ways and had weird shipping cadences that got all screwed up. I think for Apple, what it should have done a long time ago is start building AI products, not fail to build a model and then panic and say, "Oh, we need to build a model." And now arrive here. I think it should have arrived here a long time ago because that's like a heated agreement to me. That's this is what I'm saying. Like Apple didn't care about Siri. They stopped caring about Siri and nothing Google did made anyone switch to an Android phone even though assistant was better. And then you know all that stuff all the all the helpful Android stuff that they were announcing. None of that made a dent in iPhone sales and they just did monopoly stuff with the iPhone. They got lazy and they fell behind a disruptive change to the user interface which is what everyone says. That is what Sundar Pichai says it's a platform shift. That's what Satya Nadella says it's a platform shift. Sam Altman doesn't say it's a platform shift. He says no one in the future will have a job and I will have all the money. and he literally says that out loud, but he's basically saying, "It's a platform shift where all the benefit will down to him. I just hired Jony." Apple is like, "I don't care about a platform shift. We're still selling a bunch of phones." And they didn't get ahead of it. And so what I guess I'm saying they should have built a bunch of data centers. What I'm really saying is they should have made Siri good and looked at the core technology that was available to make Siri good and then done all the stuff to make Siri good on their own terms.
I think that's right. And I think your point about Apple being uh the company to make this stuff into useful products, I think is like uh is historically true that Apple is very good at being the company that takes kind of a lump of clay technology and turns it into something that people use. And like uh I would not say there's a ton of recent evidence that that's the case, especially with AI, but I think about like the what uh what they did with Spotlight on Mac OS, right? this idea of like, okay, this is actually going to be a sort of function engine for your computer that you can use to just start actually doing stuff without having to open apps and do stuff in the apps. This becomes a new interface to new parts of your computer. What Apple needs to do is build a hundred different versions of that kind of thing and it should have started doing it a long time ago. And so I think this is where we agree is that I think Apple there are going to be a lot of people who are like well Apple is the company that famously talks about how it it makes its own hardware and its own software and that AI models deserve to be kind of the third pillar of that stack. And an argument I already heard is like okay well Apple invested heavily in Apple silicon for years and years and years and years because it believed that it mattered and that it couldn't do all the other stuff it wanted to do without it. Aren't AI models the same? And I think the answer to that is still somewhat undetermined. But I think it's at least in my mind I think the answer is no. It's not the same. And that actually these things are racing towards commoditization really really really fast. And that actually the best thing you can do is say we're just we're not going to build the data centers. We don't need this. We're we're going to buy it from somebody else. We have quite a bit of money it turns out. And our job is to build the stuff on top of it rather than try to like reinvent every single wheel for ourselves.
All right. Let me make this much more complex because the thing you're describing is the called the Cook Doctrine.
And Tim Cook said this in 2009 in a moment where he was the acting CEO of Apple and it was unclear if Steve Jobs would come back from medical leave, right? Yeah. Uh and so he needed to like put his stamp on the company and like you know play play the Steve Jobs role a little bit like boldly proclaim his values. And it's it's a long statement and it's everywhere, but I'll just say I'll just read the first part. We believe that we're on the and this is the Cook doctrine. We believe that we're on the face of the earth to make great products and that's not changing. We're constantly focused on innovating. We believe in the simple, not the complex. And then here's the important part. We believe we need to own and control the primary technologies behind the products we make and participate only in markets where we can make a significant contribution. So AI just that's two parts of a sentence, right? We need to own and control the primary technologies found in products we make. Okay. Apple should own the AI, right? Like obviously that this is a primary technology. Y but also we participate only in markets where we can make a significant contribution. Can Apple make a significant contribution to the state-of-the-art in the models?
So like which part of the cook doctrine do you think is operative here is very important.
Right. And I would say most people are focused on owning control of the primary technologies behind the products we make because if Apple loses control of what the models can do or where the models are going or what Google thinks is important, that changes the destiny of the iPhone. And I that to me is tremendously important in a way that Apple being beholden to the Intel chip roadmap drove that company crazy until they made Apple silicon.
Yeah. No, I but I think inherent in that divide is the question that a lot of people have been asking, which is is this Apple making a sort of prudent responsible decision based on available technology or is this Apple recognizing that it's failed spectacularly to compete and giving up?
Yeah, I think I think it's the second one.
You think it's you think it's fully the second one? I think Apple was really convinced that it could execute all of the things it wanted to do with Apple Intelligence either locally on the phone or with its own models running on private cloud compute and it can't. We know it can't because it delayed all the products and reset the whole project, right? Like there there's just evidence that their first idea whiffed and failed.
Yep. And actually, you know, the story of AI up until about now is you need bigger and bigger training sets in order to make the models capable. And Google, as we will discuss, has the biggest training data sets of them all. And it has finding ways to make those training data sets even bigger and more relevant. And I think this is very important and this goes into how the deal actually works. Google has also apparently done a bunch of work to run Gemini on Apple's servers. And that seems very important in the scheme of how Apple will work. All this announcement is super weird. By the way, Apple barely announced it. They gave one quote to Jim Cramer anonymously, not even sourced to a name spokesperson, which is what we would demand. Cramer just said, "Sources tell me this." And then Google tweeted it. Google tweeted on X of all places in the middle of the whole X Grok scandal. Google tweets, here's a joint statement from Apple and Google. And the statement's super weird into it, but that's how they announced this news. They It wasn't like a They knew it would make waves, but they didn't want to make it a big deal.
Yeah. I mean, I think that is partly weird, but also partly just kind of a recognition of what this deal actually is, right? And every bit of reporting that we've seen suggests that this is not some big partnership, even on the scale of what Apple launched with ChatGPT, right? where if you ask Siri a question, it it might punt it to ChatGPT and you can sort of interact with ChatGPT inside of the Siri interface. That by the way is just a trash product. It doesn't work at all. It never routes things correctly. It doesn't come back with the right answers. Like that thing sucks. And Apple has said all along that it wants to have other providers there. And I would be shocked if we don't see Gemini as one of those providers. Um no, all that's getting thrown away. I would be shocked if any of that survives this deal.
So maybe that I would say if if you're Google that is like the 10 out of 10 best possible outcome, right? That Gemini just becomes Siri. You you you address Siri but you talk to Gemini. That is what Google wants because I think I think this gets to what is the foundation of a foundation model.
So but I think what what this actually is is it's essentially a technology deal, right? Google in this case becomes a technology provider in the same way that you know if Apple signed a gigantic deal with AWS, we wouldn't suddenly be like Amazon and Apple are crucial forward-facing. It would just be like no, they're just renting a bunch of cloud storage from Amazon. Um it's it's a meaningful deal in terms of like the the mechanics of the industry, but it doesn't like change how we should perceive these two companies. What this actually is is it appears to be Apple paying Google somewhere in the range of a billion dollars a year. Again, this is not official, but Mark Gurman has been reporting this for a long time. Um, for some version of the Gemini model to run some places, and this is where it gets messy, right? Because Apple has its own uh private cloud compute. It has it, it does have its own data centers and systems that it runs that will, I think, have a version of the Gemini model that runs on those things, but may not talk back to Google. But then there's also a cloud technology portion of this agreement that suggests that in fact a lot of this stuff will just be routed through Google Cloud and
Wait, can I can I just read the statement to you? I feel like we're already trying to explain a thing that from the plain language of the statement makes no sense. I'm just going to read the statement.
And again, Apple barely issued this statement. Google tweeted this and it says a joint statement from Apple and Google, which is wild. Apple usually speaks for itself. But anyway, here's a statement. Apple and Google have entered into a multi-year collaboration under which the next generation of Apple Foundation models will be based on Google's Gemini models and cloud technology. These models will help power future Apple intelligence features, including a more personalized Siri coming this year. Okay. The next generation of Apple Foundation models will be based on Google's Gemini models and cloud technology. And that is all going to happen fast enough for a new Siri coming this year. So is Apple forking Gemini and then making its own version that it will call Apple Foundation models? Is it going to just call Gemini in the cloud? These are pretty important questions.
Yes, the only hint that these questions have an answer comes at the end of this long statement. Apple intelligence will continue to run on Apple devices in private cloud compute while maintaining Apple's industry-leading privacy standards. So whatever is happening here is going to happen on Apple's private cloud compute and somewhere locally in the devices. What is unclear and left unsaid and I think open to interpretation is whether when you want like the more powerful Gemini or the next version of Gemini to do something that only Gemini can do, whether Apple has access to that as well or whether this is happening today and Apple's going to take Gemini and run it on its servers. I don't know, right? And I don't think anybody knows. What what seems to be true is that the idea of Siri basically just having its backend labotomized and replaced with Gemini is not exactly correct but is maybe functionally kind of correct that you won't you won't interact with sort of the character of Gemini but the basic underlying technology
For all intents and purposes I think is going to be Gemini and that's a big deal right Gemini Gemini 3 in particular is by pretty much every measure the best overall large language model on the market. Um it it is being used by lots of other companies. Apple is hardly the first company to go do this. You you can just go have Gemini if you have a credit card. You can just go use it. But Apple putting the the Siri name and character on top of that seems like what is likely again if if they're trying to ship this thing in the next couple of months. I don't know what else you can do. It's not like they're not going to use Gemini to build some complete new Apple thing. They just tried to do that and it didn't work. I think it's just going to look like Gemini.
Yeah. I mean, this is You can't do this this fast. Right. Right. Like you can't take all of Gemini, rework it to be your own model, which is what this half implies. You basically just have to call Gemini. Like you're going to run Gemini. Google's already done a bunch of work. They have their own riff on private cloud compute which they obviously built for Apple. Like this whole time as they've been building it and talking about it. It's like, oh, this is obviously for one customer, the only customer that cares about this.
Yep. Okay, I'm going to read you the cook doctrine again. We believe we need to own and control the primary technologies behind the products we make and participate only in markets where we can make a significant contribution. This is the primary technology behind the product he makes. It's the first part they got rocked and they are having to turn to an outside provider for the primary technology for the product they make and that's fine and every company does this stuff. It's not like Apple makes OLED displays like whatever, right? They have a lot of say in how those OLED displays are made, right? They're the biggest buyer of OLED displays. They basically tell Samsung what to do, right? They basically tell TSMC what they're going to do and what their manufacturing road map is going to look like. Are they going to be such a big client of Google that they get to tell Google what to do? Because here's the really interesting flip. Antitrust rears its head. Part of the antitrust case with Google search, not adtech, the one our company is suing Google over, but Google search was Google was paying Apple to be the default provider of search results on the iPhone because that in many ways kept Apple from competing with Google in search, right? and it provided an enormous stream of data to Google that no one could compete with because it was making search more more relevant. In this case, the reporting is Apple is paying Google a billion dollars a year to be the model provider for Apple intelligence. That suggests to me that Apple will one day walk. Apple hates paying other people money.
Apple hates paying its suppliers money.
And there's already been some reporting. There was a good uh I think it was a Financial Times story saying that Apple continues to hope this is a temporary agreement that it it wants to build its own things. But I also think there is a there is an interesting case to be made and I think uh this is the case you make if you want Apple to look good in this deal that maybe the model isn't primary technology that again maybe maybe we get to the point where choosing AI models is like choosing cloud storage where it there are pros and cons to different things but it is it is fundamentally the same technology run the same ways by the same people and the only stuff that matters is around the edges.
Yeah, maybe that is where it lands. Maybe these models stop doing the thing where they leap frog each other every 6 weeks and we hit this sort of plateau where what you get from Anthropic and what you get from OpenAI and what you get from Google and what you get from DeepSeek and what you get from Llama are close enough and that actually if you're Apple being the sixth option is a waste of your time and resources. Um I I think all recent evidence suggests if that if you believe the tech industry and where all the money is flowing, the belief continues to be that the the model itself is primary technology and that if you own the model, there's other stuff you can do on top of it that's very powerful. But all the way back to what you said at the beginning, the only reason to do that is to sell it to other people. And Apple doesn't do that.
No, that's not a thing Apple does.
Yeah. I mean like iCloud runs on all the big cloud providers. Apple has some of its own iCloud infrastructure.
Like Apple doesn't rent space on iCloud to enterprises.
I'm sure it does. I'm there's some enterprise IT person who's going to tell us what their enterprise iCloud, but that's not what we mean, right? They're not competing with AWS. In fact, they are a client of AWS cloud.
And so I think if if you're Apple, it's it's not impossible that that is where it will land. That's saying, you know, it's like being a wireless carrier. Like let let others do it and we will we will live on the technology that exists. I will point out that Verizon went down for 10 hours yesterday.
It's it's a swing, right? If you're if you're Apple and again, all of the last several years of Apple trying desperately to do this itself suggests that it does not share the opinion that it doesn't really matter to do it yourself. It just didn't and couldn't. And so here we are. We'll see.
Hey, I mean your point is that between the Apple deal and then Google announcing it's going to do personal intelligence where it'll use your data from Gmail and search and YouTube and photos, which we should talk about a little bit, but your point is between those two announcements on the same day, you're you're looking at Google taking a commanding lead in whatever the AI race is. I think at this point, if you believe in the biggest possible version of AI, which is very cook doctrine, right? It's it's we want to own all of the pieces of this and we think that these things add up to a whole that is greater than the sum of its parts. No one can do what Google is doing. It it maybe maybe it's impossible to do what Google is doing, but certainly right now no one can. Google has essentially infinite computing resources. Like Sam Altman is out here begging, borrowing, and stealing his way to try and get what Google already has in terms of compute resources. Google built its own chips, so it's not even reliant on Nvidia's computing resources. It has TPUs, which it used to train Gemini 3, which again by most measures is the best model that exists. Uh, Google has all of your data that exists in the world. This is the other announcement you're talking about. It was called personal intelligence. And the idea is that when you ask Gemini, rather than having to like give it information about you in order to answer the question, it can go comb through your email or see your what's in your photos and actually just get those answers for you based on your own data. That's very powerful. It also has crucially places to put all of this now. And this is where the Siri thing comes in, right? Like if if you just think of Siri as the new Apple character on top of Gemini, that becomes super duper powerful for Gemini because it gets all of the query data. It gets this incredible access to people. One thing I'm very curious about to see is as the idea of personal intelligence becomes more and more baked into Gemini, is that a thing that exists in the Siri version of Gemini? like can I I just all of this gets weird because it just gives Google more and more power. But so Google is in this position now, especially as it continues to put Gemini and AI into search, which it will. And if if I were really going to go out on a limb, I would say this is the year that AI mode becomes the default Google search interface.
Well, yeah. All the publishers are suing Google.
Like they got nothing to lose.
Disclosure, Vox Media is one of them. Uh but goo Google is desperate to do this. Like make no mistake, Google is desperate for AI mode to be the main interface for search. They they believe it is the right answer and they will put it there the minute they feel like they can get away with it.
All right, let me issue my my scathing criticism of Google. That is also me agreeing with you. I I think that Google is desperate to put AI mode in search. Absolutely agree. And I I think they're going to do it pretty fast. The reason I think they're going to do it is not because they think it's the best product. It's absolutely not the best product. I know it's not the best product because their own examples of all the crazy things you can do are always always always buying Yeah, it's always buying stuff. And the reason it's always buying stuff is because yes, they have a bunch of investors and they need to convince shareholders that all this AI spend is going to result in some revenue and here's a transaction that we can put ads in front of is revenue. Like you just read between the lines like look at this thing we're going to monetize. But it's also that Google has two main revenue lines. Three if you count Google Cloud, but it's it's way smaller in the the scheme of things. They have Google Search, which is the monster. They've got YouTube, which is growing, and then yeah, they've got Google Cloud and like other bets, right? Yeah, Google Search is the monster. It's it is the money at Google. And if you think the web is in decline and search is in decline and whatever, you got to move to the next thing pretty fast. And you the way Google search is monetizes obviously ads everywhere. There's a reason that traditional search is just cluttered with ads and spam up and down the stack, right? You look at that list of results, you're scrolling halfway down the page on mobile before you see the first organic result.
Google has to get in front of that, right? They got to replace all of that money, all of the AdWords money and put that into AI. So, of course, they want to shift to AI mode because then they are in total control of that results page. Every word on that results page is generated by Google directly and they can monetize it in all kinds of ways that are way more exciting than just you bought an adword and here's the first result that's spam.
It's coming and I know it's coming because they they got to move the revenue before the old thing dies completely. And then you just look at the personal intelligence blog post and our team like freaked out about this. The commenters noticed this. Everyone noticed this. The example in the what can we do with personal intelligence was whoever the product manager was said I needed new tires for like my 2019 Honda Odyssey and I like asked Gemini and it looked at my old road trips and it suggested allseason tires because I I traveled to like Oklahoma or something and it's like dude what other tires would you buy for a 2019 Honda Odyssey? Like you don't you don't need help buying tires. Most people in the world with Honda Odysseys just get whatever tires came on the Honda Odyssey. Like no one's upgrading. Like I know there are like hot rod dad van guys and I encourage you to reach out to me because I have a lot of questions for you. Um but like it's just a nonsensical query, but you have to find a transaction. And so every example Google gives is always a transaction which leads them astray in how they even structure the examples. like what's the thing I need to buy that you can inform by my history of Google photos of you're going to end up with I bought tires for my car that I took on a road trip.
And I there's just something about all the AI mode stuff that is so nakedly commercial that I I think will actually trip up Google in the end.
Well, it's the only idea that anybody has about how to monetize that is in theory also kind of part of the product, right? Because everything else you're describing is just we're going to stick an ad in the middle and everybody's going to hate that or the result itself will be an ad.
Right. Right. You're like how do I get this stain out of my shirt and it is
Just like buy Tide.
Right. Right. Shopping at least feels like a sort of payment system people understand, which is people are going to pay to put their things in front of me and I will either buy it or I won't. Like that's a trade people understand on the internet and it is, I think, the least gross feeling option that these companies have.
Yeah. I'm just going to put, but pay attention. Every time Google talks about search or AI mode or any of this, the examples always end in a transaction. And maybe it's a couple steps down the road, but there's always a transaction at the end of the rainbow which just feels gross.
Like it's like, can you just be cool for once in your life?
Yeah. Anyway, we should move on. But there's a lot more about all of this to talk about. I think we're five months away from WWDC where Apple will presumably talk about all of this in a very different light. We're going to spend, I think, the next several months continuing to do all of this AI ramp up stuff. I have reached a point where I'm like a full nihilist about AI and I'm like, in the name of journalism, I will give all of my data to every AI model that exists. So at some point this will probably arrive.
But for now it's going to be fun. And one thing to have to mention, OpenAI didn't get this deal and they leaked to the Financial Times that it was a conscious decision to not try to get the Apple deal, that they're going to build their own hardware that we'll discuss.
That's like how The Vergecast made a conscious decision to not win best podcast at the Golden Votes this weekend.
We didn't apply. So, yes, it is exactly the same. Actually, we decided not to.
Listen, we're the good people here. What can we say? We would have won if we wanted to.
Yeah, that's what I'm saying. Watch out. We're coming for you next year.
All right, we should take a break and then we're going to come back and we have just a lot more stuff to talk about. We'll be right back.
Support for this show comes from 1Password. When you run a small business, you have a lot to worry about, but one thing you shouldn't have to stress over is keeping your business safe from cyber criminals. With 1Password, you can wave that stress goodbye. 1Password provides centralized management to make sure your company's logins are secure. It's a simple, effective solution that can be rolled out in hours, whether you have dedicated IT staff or not. 1Password is designed to meet small teams where they are, but it's also built to grow along with your company. That way, however complex your security needs may get, 1Password will still stay with you every step of the way. And with 1Password's enterprise password manager, your team can securely store and share developer secrets and other sensitive data while simplifying the move to passwordless authentication with passkeys. You can take the first step to better security for your team by securing credentials. You can learn more at 1password.com/vergecast. That's the number one, password.com/vergecast.
Support for the show comes from Bilt Rewards. These days, you can earn loyalty points on just about anything you buy. But until now, you couldn't earn points on what, for many people, is the biggest monthly expense: rent. Bilt is the loyalty program for renters that rewards you monthly with points and exclusive benefits in your neighborhood. With Bilt, every rent payment earns you points that can be used towards flights, hotels, Lyft rides, Amazon.com purchases, and so much more. And starting in February, Bilt members can earn points on mortgage payments for the first time. Soon, you'll be able to get rewarded wherever you live. Plus, using Bilt unlocks exclusive benefits with more than 45,000 restaurants, fitness studios, pharmacies, and other neighborhood partners. You can redeem Bilt points for Bilt home delivery services, gift cards at over 120 brands, even student loan balances. It's simple. Paying rent is better with Bilt. And soon, owning a home will be better with Bilt, too. Earn rewards and get something back wherever you live. Join the loyalty program for renters at joinbilt.com/verge. That's j-o-i-n-b-i-l-t.com/verge. Make sure to use our URL so they know we sent you.
All right, we're back. So before we get into the next bit of news, we should probably talk about one of the stories that's been kind of bubbling the last couple of weeks. We haven't really had a chance to talk about it with all the stuff going on. And that's what's going on with Grok and X in general. This story sucks. We should just say that up front. It's a real bummer. Everyone looks bad in this story, but there's been some interesting stuff this week. The basic setup here is that Grok, X's AI bot, is just running around happily making deepfaked inappropriate pictures of anyone who asks.
Young, old, male, female, doesn't matter.
Anybody. And it has become like a meme on X to reply to any picture and be like, put them in a bikini or, you know, worse in some cases.
Yeah. I mean, some of the details are truly horrifying. And this has just become both one of the biggest and also just kind of the ugliest story in our space right now.
It's a story of cowardice and corruption at every level of content moderation, the economy, leadership. It's bad. I mean, you recall like ages ago, Google got in all kinds of trouble because the Gemini image generator would make like racially diverse pictures of Nazis and that was too woke and they had to pull it down because of a woke backlash.
This is like dead at the heart of a thing that everyone says they agree on, right? You should not sexualize children. You should not harass women just left and right.
Deepfake pornography is bad. These are things people agree on, right? And there are bipartisan bills in our country to stop deepfake pornography. There are regulations in the world and Elon is just letting Grok do it. He is throwing tantrums about it basically every day, saying this is free speech. They just want to censor me. He will comply with the law. He's letting Grok speak for X, which is stupid because Grok can't actually speak for X because it's not a person. It's not a person and it's hallucinating the responses. It's not actually the policy of X. But media organizations, like the mainstream media especially, is like, here's what Grok said about what it's doing. That doesn't make any sense. Grok spent a while lying and saying that the image generation features are behind a paywall. But that's just a thing it was programmed to say on X that you could just go to Grok and just do the thing. There's also workarounds on X where you could click the button and still get it to do it. Today they say they've stopped the image generation from being inappropriate in the countries where it's illegal, including the UK. We have reporters in the UK. That is not true. You can still just get around it trivially. Elon does not want to stop this.
No, it is very clear he thinks this is fine. And in fact, he's like putting like Keir Starmer, the prime minister of England, in a bikini on Grok to like protest the English government being mad about it. We're going to do an entire episode of Decoder on this. We're actually recording it right after The Vergecast today. It is a problem at every single layer of content moderation all the way up to Apple and Google who have built empires on needing total control of their app stores ostensibly to protect people from bad actors.
Yeah. Why does Apple say it needs total control of the app store? It needs to take 30% of every dollar that almost everybody spends on that platform. It's because if they don't, bad things will happen to people. This is the worst thing.
Yeah. I mean, it doesn't allow porn apps on the app store. And what is this in practice if not a porn app?
Yeah. It's just a straightforward harassment vector by which real people are getting deepfake porn made of them. It is fully against both rules about the app stores. I will say again, we have a whole episode of Decoder coming up on every layer of the stack being corrupt and cowardly about this. We ran the headline: Tim Cook and Sundar Pichai are cowards. Liz wrote that piece. Of course, Liz wrote that piece. Just a little inside baseball. We do no surprises just like every major news organization does. We tell people what's coming if we're going to put their names in a headline. We told Apple and Google. I will tell you it is business as normal in our newsroom.
I think everybody knows what's going on here and I think they know they just have to take the hits. So people are like, they're going to take their access away from you. Like, I don't think they are. I really don't. And if they do, I'll let you know because that would be even more cowardly. But I suspect that they've made the calculation that that criticism is less costly than if they stand on their principles and Elon Musk gets mad at them. And I think that is just ridiculous for two of the most powerful companies on the face of the earth.
Yeah. I don't know what else the dynamic could be if it's not that. If it is, it is fear of what happens if you pick the fight is worse than being correct and picking the fight. Especially on this, yeah, on this specifically, there is one correct moral outcome and it is to say this rule that we've had, that we've built our empire over enforcing, we're going to enforce it again and you will have to force us to back off as opposed to backing off right away. And that's us who's criticizing them. Everyone. Charlie Warzel at The Atlantic wrote a great piece that we'll link. A bunch of advocacy groups have sent open letters.
Again, to your point, people agree on this. This is not a complicated question about, you know, free speech and people being mean to each other on the internet. This is not a complicated moral line to draw. No, this is a bad thing that should not be allowed. Well, the one person who sort of disagrees is Tim Sweeney, who thinks app stores are bad and thinks calling for app stores to use the power he thinks they shouldn't have is just as bad. But I don't think that's actually the case. I think in this case, you're like, all of society's gatekeepers should make this stop. And whether that's the government or the people who have the power, that's fine. And I don't, especially when the app stores in particular have fought their way for a decade through antitrust, through angry developers by saying if we don't have this control, we won't be able to stop these bad things. And now here in the face of the bad thing, they are rolling right over. Anyway, like I said, a full episode of Decoder is coming. That's, I think, it's next Thursday. We're going to, like, every piece of the content moderation stack that we've talked about for over a decade is implicated here and I'm sort of going to go through it, but that's coming next week. We'll keep talking about it here too because it's a funny thing in that there's been a lot of talk and a lot of sort of movement, but also the platform is in the same place that it was. Like, let's be so clear: for whatever Elon Musk is saying, for whatever their policy team is saying, you can still do the stuff. Our team has heroically been testing these things in some like truly heinous, awful ways to prove how easy it is to do it. We'll keep covering it. We'll have more to talk about here too.
Yeah. If I can issue like one strident instruction to my peers in the media, to every influencer in greater, you cannot trust what Grok says about X's policies. It's not real. That is not a meaningful way to do any reporting is to ask Grok about itself.
No, it is in fact nothing.
And right next to that, you can't trust what Elon is saying. You have to verify that what he's saying is true.
And that it's just very, it's hard to break out of these patterns. I get it. But in this case, it's obvious that he does not want to stop it.
And it's deeply bizarre that the way to verify it right now is to ask an AI to make pictures of you in a bikini. But shouts to the folks on our team who have been doing so. And I'm so sorry. This is what our job is. All right. Let's talk about the other piece of the news I really want to talk about here is Meta, which I think you can sort of call the metaverse. Now, the news of this week is that Meta laid off a bunch of people in the Reality Labs division. It closed three of its VR studios. It is essentially like taking all of the resources it has been putting into the metaverse and moving it towards AI. The company continues to think that wearables are the thing. It is just moving every dollar it can find toward smart glasses and AI. And most of those dollars are from the Reality Labs team. And this to me feels like the end of a story that we've been covering for most of a decade now, that Meta believed in the metaverse as a concept so much that it paid a bunch of money to Palmer Luckey to buy Oculus. It basically, not basically, it renamed the company to Meta. A story that we had. Great job, Alex Heath. And it essentially tried to corner the market to the point where there was an FTC trial about whether Meta should be allowed to buy this company called Within, which made Supernatural, which is one of the most interesting VR apps that exists. So Meta basically makes this big, gigantic, industry-shaking bet that the metaverse is going to be the thing, and then kind of slowly and sort of all at once here has just decided to walk away. Can I connect the dots from the App Store and Grok to Meta and VR and Mark Zuckerberg?
Sure. Jesus. They're not hard to do.
Okay. Mark Zuckerberg hates Apple. He hates the fact that Tim Cook through app store policies can affect Meta's business. And Tim Cook hates Meta. And I think he hates Mark Zuckerberg. And in fact, Apple has many times just directly screwed with Meta's businesses. Yes, they did app tracking transparency. They ruined Meta's ability to do app install ads. Like they took billions off Meta's bottom line and they had to rearchitect the business to make that money back. Straightforwardly, Mark Zuckerberg hates the fact that Tim Cook was able to do that to his business through iOS level permissions and app store level policies. At one point, Meta was running a VPN called Onavo to get data on what users were doing, which is super sketchy. Super sketchy to give people a VPN in the name of privacy and then track every single thing they do in every single app to figure out what apps to buy or compete with.
Yes, super sketchy. Apple revoked Meta's entire enterprise development certificate and took down every single one of Meta's internal apps on iOS. People were not able to like order food in the Meta cafeteria because Apple thought Onavo was so bad. This is scandalous levels of power being used against a nonsensical competitor.
This is real. Apple has done all this. You can look it up. They just have attacked Meta left and right using the power of the app store. They have threatened to pull Instagram over some of the teenage girl issues that Instagram has had. Yeah. I mean, so much of Apple's in particular sort of privacy is the thing messaging is directly aimed at Meta.
Well, it used to be aimed at Google, but that's a little too cozy now. So now it's all been aimed at Meta.
And everyone still thinks their iPhones are listening to them. They've accomplished nothing with this messaging because the ads on Instagram are still good enough. It doesn't matter.
But that dynamic between Meta and Apple is a dynamic born of the app store. Mark Zuckerberg knows that Apple's authority over the App Store, rightfully or wrongfully, is a check on his business and we've just been in that dynamic for 15 years. And he hates it.
He openly talks about, he's said it on The Verge, in interviews on Decoder. He has said, I hate this. Like once people realize how much power Apple has, they'll be mad. Meanwhile, Apple is so afraid of Elon and Trump that they won't use the power and in one case they should. So there's just naked hypocrisy there. I'm only bringing this up because that's why Zuck bought Oculus, right? He said, I need to buy the next, I need to get away from the iPhone. I need the next thing and I will control that platform and then I will turn off Apple's developer certificates. Like he, this is the motivation. The man is motivated by power.
So he bought the headset and he said, this is going to be the thing that takes over from the phone. And he did all of the metaverse and he would talk about the embodied internet because he thought he could get you away from the phone. I think now he thinks he can do AR glasses before Apple gets the AR glasses and that will be the next platform and he will be in control of that and he's motivated by that. He is not motivated by making these products great. He is not motivated by finishing the thought, by investing in the things that he bought in this case. He just wants to get out of Apple's shadow. He wants to own the thing after smartphones. Yes. And is so unworried about what that thing is. And I do, I mean, I think he's like, if it had been, I was just about to make a cryptocurrency joke. Meta got really into cryptocurrency. Like David Marcus showed up and did a whole thing at Meta about Web3. Like they don't care what the thing is as long as they don't have to play in somebody else's yard for the next one because they perceive that it cost Meta a lot of money and it objectively did. But here I pulled this one quote from the original FTC complaint against Meta for this fight about whether it was allowed to buy Within. And just to put everybody in the moment, this was when the Quest was growing really fast. There was a sense that this was like an exciting platform, at least for video games and stuff. And every time anybody would release anything cool on that platform, Meta would immediately buy it. And it just, every good app that came out was swallowed up by Meta. Including Supernatural, this like Beat Saber workout thing that I think you and I both had pretty long stints of using if I remember correctly.
Oh, yeah. I want to talk about, I mean, reminder this is all in the pandemic. Like everyone's at home.
It was really hard to buy a Peloton.
Yeah. Supernatural came out. It was a fitness app in VR. A lot of people had bought Quest because what else? At least you can get out of your house somehow. You know, at least you and your partner don't have to look at each other for like, well, I guess I'm in Jurassic Park now. And it was, we should talk about Supernatural more directly. But it was in that moment that just the context was the pandemic and Zuck went to buy and the FTC, the Biden FTC tried to stop him.
Yeah. And so here's the line from the complaint. It says, Mr. Zuckerberg has made clear that his aspiration for the VR space is control of the entire ecosystem. As early as 2015, Mr. Zuckerberg instructed key Facebook executives that his vision for, quote, the next wave of computing was control of apps in the platform on which those apps are distributed. Making clear in an internal email to key Facebook executives that a key part of his strategy was for his company to be, quote, completely ubiquitous in killer apps. Completely ubiquitous in killer apps is a kind of bar and b, the whole thing, right? It's not. I am so with you in the sense that I don't think Mark Zuckerberg was ever a true believer in the metaverse so much as a true believer in I will spend whatever it takes to not lose the next platform war.
Yeah, absolutely. And he was spending that money. And so the tragedy here to me is Supernatural. And I really feel like it's tragic that as part of all these cuts to Reality Labs, the Supernatural division is no longer going to invest in new content or features. They're just going to make it on life support.
It's on life support, which means it's dead.
And the reason I say it's a tragedy is I interviewed the founder of Supernatural, Chris Milk, in 2021 on Decoder.
Like a legend of the VR industry, that guy.
A legend of the VR industry, one of the most creative people around, a legend of the music industry. And he had come up with Supernatural after playing Beat Saber. Supernatural and Beat Saber obviously very close to each other. By the way, justice for Beat Saber. Like Beat Saber was a monumental shift in how people think about VR. It's like one of the first and most important VR applications to ever exist.
People are like, oh, this is so cool.
So he was playing Beat Saber.
But anyway, Chris Milk had been playing Beat Saber. I asked him about the influence and he was like, oh, I can make an actual fitness app out of this. No, it won't just be an accident. So he basically set out to build a Peloton, like a Beat Saber Peloton. And he hired a bunch of fitness teachers. He called them athletes. He had a bunch of choreographers. Like he built out an entire studio. He has a background in the music industry. He figured out all of the licensing for all of the music in Supernatural. And not only was he growing, he was growing in like a normie way. Actually, play the clip. I just want you all to hear the stats for Supernatural in 2021.
Fitness is the killer use case for VR. It will be the first driving force of mass adoption through a normal consumer audience. Just as an example, what we're seeing from our membership base where 50/50 split women and men where I think over 60% are over 40. This is not what a typical VR demographic looks like.
It's the understatement of the century.
That's remarkable, right? To get a 50-50 split, men and women, 60% over 40, to be early adopters of new hardware. And so, Supernatural was so important that Apple and Meta were in a bidding war for it. And everyone, this is before the Vision Pro came out, everyone thought Apple wanted it to drive Vision Pro sales. And you can imagine how that might have gone very differently if Apple had succeeded here. But it is just a tragedy that Meta bought it, didn't really invest a ton into it. The Supernatural audience was all immediately upset about this, right? Like, yeah, no one wants this acquisition to happen. They were cheering on the FTC case. The FTC lost, whatever. And then Meta has done nothing but drain Reality Labs of funds and with it kill the thing that everyone saw as the killer app for this new hardware because Mark doesn't actually care about the metaverse. He cares about power and he cares about being first to whatever he thinks the next platform is going to be. And if Supernatural had just been on its own, growing in this way, maybe it would have gotten bought by Peloton. Maybe it would have built its own ecosystem of hardware that was more specialized for fitness. You can think of another hundred things that would have happened that would have been better for an audience of consumers who were significant enough for Apple and Meta to want to buy the platform to get healthier, right? And like they're getting robbed of that now. The thing that helped them get healthier, the thing they were excited about is going away because Mark would rather pay for glasses. And I just think that that's a tragedy. Like, yeah, I rant and rave about antitrust and buying stuff, but this one in particular is like, oh, this was a great product. This thing was demoed on Late Night with Jimmy Fallon. Like, it had crossed into mainstream appeal and Zuck lost interest. And that just sucks. And the thing is this whole run that they went on put it in a position of just being able to do it until they lost interest, right? Like Meta has so cornered the market that it didn't become a competitive space. Nobody else is trying to do that anymore. And there's an argument to be made that it's because it was a bad idea all along and nobody was ever going to do it. But that market never even had a chance to materialize because Meta just completely cornered it from the first minute it could. And then it was then existent based on Mark Zuckerberg remaining interested in doing it. And that was never going to work for any of these companies. I mean, the other thing, I think the best VR game ever was Batman: Arkham Shadow and that studio was bought by Meta and was also hit by these layoffs. Like this whole thing, this whole ecosystem of people who were the only ones who had a chance to make this into something viable. And again, I think there are lots of reasons to think the metaverse is a bad idea. I think immersive gaming, just boil it all the way down, that basic thing is very cool. And like, wouldn't it be awesome if I had a Supernatural app for my Nintendo Switch that I could play on my TV? Like, this should have had a better run than it did and Meta just swallowed it whole and then just kind of gave up. And that sucks. That feels bad.
Yeah, I think I slightly disagree with you. Only in the sense that I think general purpose VR headsets might have always been a dead end, you know, like the idea that I have a face computer, like whatever, you know, like that's the axis of wearable. Like it's so fiddly and you have to care for it and it's on your face and some of the stuff is going to make you dizzy. There were these bright spots and maybe you think gaming was a brighter spot than fitness. I just saw what people were doing in fitness and I heard that quote about the demographic and we went and looked at those communities on Facebook and Reddit where people were gathering to talk about Supernatural and it was just, it was the thing you want from technology.
Right. It was a bunch of people who were not there for technology. They were there to like lead better lives and support each other and like do that thing that in fitness communities is often very like rewarding. And Mark didn't care about that at all. Like you, I, and I just think that's tragic.
I also think like they should have gotten away from the Quest and made their own hardware and like had a cheap VR headset to do this in or like whatever kind of motion tracking. Like there's a million things they could have done.
This is what I mean by Switch game. Like there, Supernatural could have been a lot of places that weren't a VR headset.
Yeah, sure. I mean, it's dancing. Like it was, it's dance fighting. There should be more dance fighting in the world is basically what I'm saying. But um, and like that's cool. Like Crayons and I were both Supernatural fans like in the pandemic. We were both like dance fighting every day. Like it was like great. I was like, I'm not at my house. I'm not like trudging along in this treadmill. I'm like having a good time in a different environment.
Supernatural was selling a lot of Quest headsets. Like I.
And that's why they bought them.
It was the killer app in so many cases for people.
Yeah. Especially was fun. But Becky was wearing a.
She, there's, she, I'm like, do you want a new laptop? She's like, get, I get this technology out of my face. That's her reaction to most ideas about tech.
Did she have a Kindle for like 12 years until you had to bully her into upgrading?
Yeah, upgrade her and that's going to, she's have it for another 20 years. She like wore the headset all the time. We bought a dock. It lived in our, in the place where we work out and she would take it off and like use it. And that's like, that's what I was like, oh, we should get this dude on Decoder.
Again, the tragedy of Supernatural aside, the thing you're looking at is I think the essential character of Meta is not about taste or what the products are or the experiences people are having. It is just about power. And I think in this turn, if you're like, why won't Meta win in AI? Why won't they win in the next platform shift? Well, why are they letting their platforms be overrun by slop? Well, it's because I don't think they actually care about the products. I really don't. And I think that the companies that do care about the products tend to be the companies that always, even to this day, they still tend to be the companies that win. And I see them pulling back from Reality Labs. And it's because it didn't win and they don't actually care. They don't actually have the persistence and the drive to execute. This is why Horizon sucked. Like it was so bad. It was so, so, so bad because no one knew what good looked like or cared to sit down and rigorously think through what good looked like. This is why Mark Zuckerberg just hired Alan Dye from Apple.
Because he doesn't know what good looks like.
Yeah. And I don't know that he cares.
I think that conviction is overrated in a lot of these cases. That having a thesis clouds you from chasing whatever feels like it's going to be the next thing. Can I just read you maybe the single bleakest part of this entire thing? Which is that there was a Bloomberg reported that there was an internal memo from Andrew Bosworth, who is Meta's CTO, about a shift in the metaverse strategy. Meta is careful to say it's not ditching the metaverse entirely. It still believes.
Renamed the company Meta, right? What are they going to do? But this is just the last line of our story about this and it just hits. Rather than being a VR-first social platform, the metaverse will be focused more on mobile devices moving forward.
It's just tough. And I think what's funny is to some extent, like if you sort of broadly define the concept of metaverse, it's happening already. Like Fortnite and Roblox and Minecraft, like this idea of people being in virtual worlds is already happening and it's huge.
Oh, get out of here. That was the argument they were making five years ago.
No, the metaverse is you got to protect, you need to, you got to be in it. You got to be in a virtual environment where you're buying and trading NFTs or whatever the hell.
I think that was the thing that Mark Zuckerberg was undeniably wrong about.
Yeah. But the idea that Fortnite is a metaverse is nonsensical.
Oh, you're totally wrong. Fortnite is the metaverse in all of the ways that it is useful to be the metaverse. All the other stuff is dumb.
But the definition, it's a video game. Like I can draw a distinction between I live my economic life in a virtual environment and I'm playing this video game where there are occasionally concerts and brand integrations.
I mean, how many people live their economic lives in Roblox? We could do this all day.
Too many people is the answer to that question.
That is true. That's the PSA at the end here. Stop playing Roblox, everybody.
By the way, it's very funny that Apple still makes the Vision Pro, has not announced any pullback, and is shipping more and more VR content for it.
