What Meta and Anthropic really won in court
01:46:05 Share

Support for this show comes from Salesforce. Today, every team has more work to do than resources available, but digital labor is here to help. AgentForce, the powerful AI from Salesforce, provides a limitless workforce of AI agents for every department. Built into your existing workflows and your trusted customer data, AgentForce can analyze, decide, and execute tasks that

Letting you and your employees save time and money to focus on the bigger picture, like moving your business forward. AgentForce, what AI was meant to be. Learn more at salesforce.com slash agentforce.

Let's get moving.

Summer is coming right to your door. With Target Circle 360, get all the season go-tos at home with same-day delivery. Snacks for the pool party? Delivered. Sun lotion and towels for a beach day? Delivered. Pillows and lights to deck out the deck? That too. Delivered. Just when you want them. Summer your way, quick and easy. Join now and get all the summer fun delivered right to your home with Target Circle 360. Membership required. Subject to terms and conditions. Applies to orders over $35.

Hello and welcome to Grinchcast, the flagship podcast of being the new editor-in-chief of Vogue. I just put my name in the ring. I don't think that would be a good idea, and I love my current job, but it is true that Anna Wintour is stepping down as editor-in-chief of Vogue, the most legendary editor-in-chief of all time. And hey, you know. Hey, Nealite, 30 more years of this?

Hi, I'm your friend, Eli. That's Jake Castronakis. Hello, hello. David Pierce is here. I like imagining Vogue as just like everyone is wearing black on every page of every issue of the magazine. And there's like Packers logos in the background. Yeah, that'd be great. That's my Vogue. And only iPhone 4s are allowed. The most beautiful phone of all time. You would last two issues, but they would be glorious. I've only met Anna once or twice.

And both times she was utterly charming, I have to say. But the second time was at the American Society of Magazine Editors Awards, the ASME Awards. And we had a table right next to like a Condé Nast table that Anna was at and they were terrified of her.

Like she was sitting there and all of the Condé Nast people were sitting there just in terror. And then our table was just like drunk and cheering. We were having the most fun. They kept staring at us like, don't you know? And I was like, I do know this is our vibe. It's our time now. You're the legacy media.

I always like thinking about how she must feel about Devil Wears Prada in the way that Mark Zuckerberg feels about the social network. Because he's, A, been very upfront about how he did not like the way he was portrayed in that movie. I would say history has proven that movie more right over time than I expected. No, Jesse Eisenberg was vastly cooler than Mark Zuckerberg in that movie. I mean, yes. What history has proven is that Mark Zuckerberg is a huge nerd who can't outrun his own essential nature.

Yes. Fair. Which I think was a large part of what that movie is going for. But like, yeah, I think that thing where somebody like very famously portrays you and it changes the way everyone thinks about you has got to be such a weird phenomenon that like most people think when they close their eyes and think of Anna Wintour, they think of Meryl Streep. That's just weird. It's got to be so strange. Yeah.

Uh, and, and, um, you know, I watched the overall product whenever it's on television. I just like, it's a great movie. You should watch it. Uh, and, uh, Becky is always reminding me that I think I'm Anne Hathaway, but in fact, I am now the Meryl Streep character. Uh, it's over, but anyway, look, if you know anyone, um, over there and you want to see Vogue get super weird, uh, I'm open to a conversation.

You know, just a series of feeds. Like, have you thought about going all in on activity? Important fashion magazine. There's a lot there. You can unpack a lot there. The world's most important fashion magazine in the age of Instagram and TikTok is a weird thing to be. They have a monopoly on taste, right? It's the cover of Vogue that's important. The contents of Vogue almost are not.

And so like, that's the dynamic there. And it's super interesting if you're not Anna Wintour and you don't have that monopoly, whether or not Vogue can be the thing that it is in the age of Instagram and TikTok. In the age of all the celebrities being able to publish their own pictures of themselves. It's a thing. I'm just saying it's a thing. All right. We can set that aside. But if you know anyone, and if you know anyone at Conde Nast who has or want to have a really weird conversation about the future of Vogue, uh,

Again, I love this job. There's not another job really for me. I'm trying to decide if it would be more fun to point The Verge's resources at getting you elected president in 2028 or getting you the job as editor-in-chief of Vogue. They're very different futures, but I find the project of both of them very fun. I do think I can shape the direction of the country in either role. It's just in a very different way.

Yeah, which is more effective? All right. There's a lot of news this week, a lot of updates. We got it. There's a big set of AI copyright decisions that Addy is going to join us to talk about and unpack. I would say earlier today, Addy and I had a conversation that can only be described as we got into our feelings about copyright law.

Because of these AI decisions. So we got to talk about that. There's updates on the Trump phone. Brendan's got a new commissioner at the FCC. So he just immediately started doing some of the dumbest stuff in the world. There's a side note, like there's some FTC dumbness involving just straightforward bribery, which is pretty good. A lot, just a lot going on.

Uh, in the world, but we should start with a little update on the robo taxis, the Tesla robo taxis in Austin. I think David on the Tuesday show, Andy was on there like a day in now we're several days in, I would say the obvious has occurred. It was very funny. So we, Andy and I talked about it a bunch and he kept being like, look,

We should be cautious. Tesla should be cautious. But it's only been a day. We'll see what happens. Hope for the best, expect the worst. And boy, did it just kind of go sideways from minute one. So there's a bunch of videos now. There's a media component to this where the only people who seem to have access are Tesla influencers and hardcore Elon Musk ex-posters. And they're all like, we're the media now. And it's like, are you? Yeah.

But then they're all just doing this like incredible live streaming, like everything is being posted. And so even through the filter of a bunch of people who hold themselves up as being Tesla boosters, there's just like endless videos of these cars just like making mistakes. And the net result of it, just from my perspective as a person watching these videos, is that it's just full self-driving, right? Yes, there are safety monitors in the cars.

And there's that one picture of one steering wheel and the background of one photo, you know, like maybe they're all being remote controlled. You know, there's all these rumors about this stuff, but you kind of watch it and you're like, oh, this is just like watching any other full self-driving video. Only,

Like the cars are making mistakes without a driver so that it all feels way more dangerous. Right. They're only they're sitting in the backseat is like a pretty meaningful change in how we're actually applying this technology. When FSD messes up and the person's in the driver's seat, they're like, I'm taking control. And they're like, oh, that was scary. Right. Now it's like, whoops. Yeah. What are these safety monitors actually supposed to do?

just like, like comfort you when like you rear end somebody like it's all right. We, we, we, this, this happens from time to time. There have been a bunch of these. And I think when the safety monitor does and doesn't get involved has been sort of fascinating. Like there, there were, there was the one I think that has been kind of all over the place. And Andy and I talked about this on Tuesday where the,

The car just veers into the wrong lane for a second. And the safety monitor doesn't seem to flinch. This is just like, it's fine. We're just going to go in the wrong lane because we thought the turn lane started before it did. But then there are a couple of other ones where like something actually goes wrong and the car just like stops in the middle of the road. And then the safety monitor does get involved. So it seems very much like the safety monitor is instructed to basically not do anything no matter what.

unless it has like officially broken, which is kind of a wild set of instructions to give to a safety monitor in a car like this in a project as early as this.

But I just I don't know how to read it any other way. Like they should have been getting involved more if they were allowed to be getting involved more. There's definitely one where the car gets stuck kind of like between two cars. It's a tight and the safety monitor drives it out. OK, I didn't see that. That's something. I get it. Right. It's a very limited beta slash pilot of this stuff. There's not a lot of these cars. It's a trial. Sure. They're on public roads with other people and other cars. It does seem to me like the safety monitor should just be in the driver's seat.

especially if it, again, as it appears, it's just full self-driving. Like it's not more, if they were faking it and there was a warehouse full of people, remote driving cars, I'd be like, yeah, safety monitor in the other seat, you know? But like you watch these videos and it feels, anyway, from my perspective, like it's just full self-driving. And then you get to the fact that most people can't access the cars, right? It is just this limited group

at least posting content that is very apt to supporting Tesla. And you have this like war against the media narrative. And I'm just going to say this as many times as I can access is poison. And if you hit capture where the companies are covered, you're going to make bad shit over time and you're going to lose your audience. And that's just the way it goes. And all of these folks were like, the media doesn't just have the wrong narrative about Tesla. If Tesla was confident in the product,

They would just let reporters into it and let them drive around and say, this is amazing. You can just take away Mo because Waymo is confident in its product. It thinks the products are expanding the markets it operates in. This thing where you think Tesla has rewarded these influencers is actually a sign of deep insecurity.

Like the most insecurity. And you can see it. I mean, we, this is, this is our world, right? You see it with every company when they're not proud of their products, they close the distribution, they limit the access. You know, we play these games, but we, our rules are clear. If you're listening to this, you know, our rules is clear. You can't buy us. You can't buy our brand and content. We're unsponsored for all this stuff because that's how we manage the access games. We just draw the lines as clear as we can. And we say, look,

We won't play the games. If we don't get the access, we'll go and do the next thing. And I like there's a video. There's the there's the one video of the woman in the backseat. The car breaks so hard it throws the phone out of her hand. And then she starts apologizing for the car. And it's like, oh, this is bad. Like, this is a really bad media ecosystem. If you think that the reward is the car breaks so hard while you're not wearing a seatbelt, the phone flies. We have it. Do you want to play it, David? I do. I do.

Let's find out. You guys know we're not lying because we have that stopwatch going. Whoa. All right. So we just slammed on the brake.

I'm not exactly sure what just happened, but the car thought it saw something. And this happens in full self-driving. This is something that does happen. That's something that people have talked about being one of the limitations of full self-driving with RoboTaxi is that occasionally it slams on the brakes out of nowhere. You guys kind of saw me react. Yeah.

I mean, you get the idea. What are we doing? It happens sometimes, guys. Sometimes your car just slams the brakes on you and you just sit there. With no, the safety, presumably a safety monitor just saying nothing while this video is happening. It is sort of fascinating that this is what Tesla has conditioned its fans to believe is just how this works. No, I'm saying it's more sinister than that.

I'm saying it is conditioned access on that behavior. Oh, sure. And that feedback loop is actually bad for the company. It's bad for people, right? It's bad for people like great products. It's also bad for the company because they're not actually getting the negative feedback they need. The market is not actually providing the information. They have a bunch of influencers who are afraid of their access being taken away and not getting invited to the next cyber rodeo or whatever. And then they're yelling at reporters who are just trying to do their jobs. Right.

But you can literally hear the lack of objectivity there. And look, everyone can yell at us. Somehow in my career, I've gone from being reliably called an apple shill to being called an apple hater. Great. Sure. Whatever. You can yell at us about our perceived biases, but we have our rules. You can just read the policy. Those rules exist to at least provide a set of guidelines for what you can expect from our behavior and how we behave.

Think about access and all these other things that we, you know, that are part of our job and our careers and the media that we make. But you can hear literally in these videos, the influencers thinking if my access will get taken away and I need to apologize for this car fucking up. And it's like, no, a real person, a real customer does not stop and say this happens. They just get the fuck out the car and like, that's your problem.

Yeah, the video did make me think about like the terrible Uber rides that I've had in the past. And like they happen occasionally and you have them. But like there's a person in the front seat. You can tell like when you're at CES, I had an Uber driver who drove me at 45 miles an hour through parking lots of casinos in Las Vegas to get me to the airport.

And we were just like launching ourselves over speed bumps. And for the first time ever, I had to be like, you got to stop, dude. Like, I'm not in a hurry. Please stop driving like this. And he did. And it was fine. I went from being like, I'm going to die in the back of this like Dodge Charger. Of course, it was a Dodge Charger. To it was okay because there was a person I could tell. And it was just like, it's...

It's not that these things are uniquely bad drivers. It's that like we have a chain of understanding how this stuff works that we just none of this has made any sense. Again, it goes back to like put the safety monitor in front of the steering wheel. Yeah, but he can't. Right. I know why he should. I know why Elon should put the safety monitor behind the wheel. I know why he can't because the future of Tesla's valuation depends on robo taxis working, not on selling new or better cars. Right.

That's why this whole thing exists. Yeah. Well, the more time will pass, presumably more people get access to this thing. I'm just I've been bemused. It feels like a bunch of people are entering a kind of journalism where the compromises of access will be made clear and some people will learn the lesson and some people won't. And it all kind of feels like speed running to me.

Like, you're about to have a whole bunch of experiences here. But I'm watching these videos and I'm like, oh, the thing doesn't work as well as it should. Right? And if you actually want to compete with Waymo, if you actually want to build that stuff, it needs to work a lot better than this, a lot faster. And like everything with Tesla, maybe it'll just take a long time, but...

I'm going to continue watching this list of Reddit video updates. They're very good. I really it's Reddit has done an admirable job of putting together a timestamped versions of hilariously bad things happening to these people. Yeah, I highly recommend it. By the way, if you want to be an auto journalist and you think Tesla's by all means, I'm just I will just say it again. Access is poison. It will if you chase access, it will kill you.

And I say this to our staff all the time. The less access you need, the more you get. And that's a dynamic that's very hard. Again, I think they're just speed running media in like a very particular way. We'll see how it goes. All right. Speaking of access.

poison and speed runs, there's an update on the Trump phone. Jake, you helped me investigate this. Do you want to take us through what happened with the Trump phone this week? Yeah. Well, if you guys will recall about a week ago, the Trump organization announced Trump Mobile. Trump Mobile announced the T1 phone, 8002 Gold Edition, I believe the name is. I think we came on here and we said,

We don't believe that a lot of this is, is accurate. Um, one of the biggest things was that they said it was going to be made in the United States. Um, Neil, I went on TV and shouted that that was a lie. That's true. I will say I was not on the show last week and I listened to it and I had to, I had what I believe is the very common experience of yelling at the verge cast. So I feel a lot of empathy for the audience, but CBC invited me on. I was like, this is fake. And, uh, lo and behold, um,

David, earlier this week, scrolling through their website, sees that they've changed the website. Used to say, made in the United States. Now it says, what's the exact phrase, David? Oh my God. I'm so excited to read this website to you. First of all, shouts to Simon, who saw this before I did and sent it to me. Can I just read you a bunch of copy from, this is the phone...

of the T1 on the Trump mobile website. It says at the very top, meet the T1, premium performance, proudly American. This is the new tagline of this phone. The T1 trademark, they put trademark everywhere, which I really respect, isn't just another smartphone. It's a bold step towards wireless independence. That doesn't mean anything. Designed with American values in mind, the T1 trademark delivers top tier performance, sleek design, and powerful features, all without the inflated price tag.

designed with American values in mind is what is replacing made in the USA. We also have, um, with American hands behind every device, we bring care, precision, and trusted quality to every detail. Are those just Don jr's hands?

He's just collecting his checks. There's just some American has put hands near a phone. An American? That's all we know for sure. An American hand has clicked the AliExpress order button. That is what happened here. I am telling you, you go to AliExpress, you type in big Android phone, and you will find something that is very close to the T1 phone. And the other thing that David noticed is that they've changed the specs as well.

The phone is not going to ship as the exact phone that they announced a week ago, which means that like they clearly were not confident what they announced.

I think we all said that would be the case, and we were already starting to see it morph in front of our eyes as they realized that it was not correct. David, what is the updated spec? I believe they changed the RAM and they changed the screen size. So the RAM is now gone from the website. It normally, there was, if I remember right, there was a section on the site that said processor and RAM, and it had no processor listed, but it said 12 gigs of RAM. Now that whole section is gone.

So we still don't know the processor and now there is no RAM apparently. The other one, which I think is maybe the bigger signal that something big has shifted here is it has gone from a 6.78 inch screen to a 6.25 inch screen, which I would just point out is a very different phone. If you lop off a half inch of the screen, you've gone from an iPhone plus to an iPhone. Like that's just a different thing. And like,

They fixed a couple of things. Like now it says battery as battery instead of camera, which is very exciting. It's very good. Although I was always excited for a 5,000 milliamp hour camera. It's going to be a sick camera. That camera will power your house somehow. We don't know yet. It also, instead of saying it's going to ship in September, it just says coming later this year. Even the coming soon is gone, which is very funny. The only thing I can think of is that they got

in some kind of legal hot water over the phrase made in the USA, which is a thing that you have to follow certain rules to be able to say about a device or any product. And also that they went from having one supplier to having another supplier. It's the only thing I can think of is that whatever they bought on or thought they were going to buy on AliExpress, now they're buying something else with a half inch smaller screen. There's just no reason that happens in a week other than that.

Especially at $499. Right. Yeah, the render is still the same. The price is still the same. I'm still convinced this phone does not exist and will never exist. But even the thing that doesn't exist...

makes less sense every time. The made in the US thing was just particularly egregious, right? Like from the start, it was wildly obvious that that could not be true. And I guess after enough people asked about it, they realized maybe we shouldn't lie about this one, which I guess I don't want to say kudos on that. But, you know, good, good for stopping lying. That is a positive. It's not not lying.

You know what I mean? The phone doesn't exist. I'll give them credit for not lying when they ship us a phone. I believe, Jake, you ordered one, right? The Verge has ordered two of them. Two? We're $1,000 deep into Trump phones? Well, yes. So Sarah Smithers are...

I take one day off the show, Jake's like, blow a grand on Trump phones. Listen, somebody needs to find out if these phones are real. Sarah Smithers, our wonderful editorial operations director, um,

has spent a week trying to order these things. It turns out for the first week, they did not have an option to order them if you lived in Washington, D.C. OK. And so we had to keep emailing them saying, hey, this is part of the United States.

Could you please accept our order? Because our corporate headquarters is in D.C.? Yes. I understand. So we will find out for all of you if this phone exists, if it is any good, and if they, as the website promises, randomly start billing you recurring charges at some point. Many, many people have asked us to review the phone, and I'm confident we will. I just want to set some expectations that...

It's not like a review of a mid-range Android phone is often a banger. Do you know what I mean? Like, it's a mid-range Android. You already know what it is. The thing that you want from us is to point out the obvious lies, which is like, you almost don't need the phone to do it.

But at best, like the best thing this can be is a competently executed mid-range Android phone. Yeah, that's literally the 100th percentile outcome for this thing is A, it exists, and B, it's fine. That is the best possible outcome for the Trump phone. I suspect there will be a yawning chasm between the device shipped to us, the two devices apparently that will be shipped to us, and that outcome. And that's the point of review, but...

But I'm just trying to... I don't think you're going to open it and say, like, light on fire. I think it's going to be a mid-range Android phone with some shit bundled software. All right, Nilay. Trump Mobile, Project Gen of 5, sis. You have to bet on one as a grand success in the future. Oh, it's Project Gen of 5, sis. Wow. Easily. It's easily Project Gen of 5, sis. And I'm only saying this because it...

To the extent that it exists. You were about to say it exists. And there was about to be a party in that office. Yes. You almost said it. To just set this up for the audience again, because I'm often reminded that we speak in code and inside jokes like a bunch of high school seniors. Project Genofysis is the result of T-Mobile being allowed to merge with Sprint in the first Trump administration. And what they did was they told T-Mobile it had to divest a bunch of assets in Spectrum to dish network

a satellite television operator, which would then create a new from the ground up 5G network so that we would have five national carriers and there would be competition and look around. Does it exist? They launched a network. They called the project on a five system. They're rolling it in a boost mobile, whatever. Most of those phones still roam onto AT&T and T-Mobile because the network doesn't exist. But in some pockets of America, the network apparently exists. That's what I got for you. And so just because they built some cell towers and put some radios on some poles,

Someone will try to turn that into money because they've spent the money. Here, what they've done is they asked ChatGPG to make a mid-range Android phone with the word Trump on it, and no one has to collect on that investment. And so Project Gen 5 says just because there's hardware in the world, like a little bit, has a, you know, on our Go90 scale where zero is alive and 90 is dead, it's quibby.

I would put Trump at 90, just lives at 90, and Jenna Fives is at like 75. Yeah, that seems fair. I don't think I've ever applied the Go90 scale of June streaming services to telecom providers, but it works. Yeah, there's like three of them that are at 10 and everybody else is hovering in the mid 80s at all times. That's just what we do. There's more phones. There's a new Fairphone.

It's modular? Is this what this is? It's a modular phone? I love a modular phone. I find Fairphone totally fascinating. This company has just been like sort of chugging along in the background doing great work on making like upgradable, repairable, more workable

well-made and thoughtfully designed smartphones than almost anybody. And it's not like a huge hit and their phones have been not amazing because they're trying to do the right thing more than they're trying to build great phones and those two things are hard to do together. But this new one's really interesting. It's the Fairphone 6. It is smaller and lighter but has more battery. It's more powerful. They've just made a bunch of the trade-offs you want them to make for the first time. But they're also doing...

the like modular accessories thing. Like, do you guys remember the nothing phone that came out and had the little loop on the side that you could like stick a lanyard or like change the case and screw that stuff off? That's coming to the Fairphone. They're starting to do stuff like that. But the thing I think is really cool is they're shipping this phone in...

um basically two different ways one is with uh android like normal android and fairphone has a pretty like refined take on android but they're also shipping it with a button that essentially lets you switch to like a minimalist mode of android which is kind of like as we haven't learned a ton about it but it seems like basically a power do not disturb feature which i think is super clever um

And then the other one is it's shipping with a de-Googled version of Android, which is called EOS, made by this company. I believe it's called Mirena. And essentially, they've spent several years now building out a version of Android that is totally bereft of Google services. And that's one you really have to make a moral commitment.

It's like it's like being a vegan. You know what I mean? It's like you have to do it on purpose and live with the consequences. But you're doing it because you believe it's the right thing to do for the world. It's interesting. I suspect there probably is a lot of overlap between those two markets. Like they're going to be a bunch of vegans who hate that. I just made that. Oh, I was not. I didn't mean vegans and people who are against Google. I thought I meant people who want to repair their phone endlessly and who want to get away from Google. But yeah.

It is. I think the bummer about this, you know, I think there's a very exciting device, right? Like this is, I think the best looking for a phone yet. It looks like a normal phone. It's like an upper mid range Android device. It looks great. The colors are fun. Um,

I like that it's opinionated with the button that does the minimal launcher. But if you want to get it in the US, you have to get that de-Google-ified version. Oh, that's the only one in the US? Yes. Oh, that sucks. And that is a problem, right? Yeah. Because choosing to go without Google services means missing out on a lot of the phone experience, right? Just like everything's in the Play Store. Just good luck, right? Like...

Most people that I know use Gmail. Yeah, where do you get apps for this thing? I'm like clicking around this website. Where do you get apps for it if you don't have the Play Store? Sideload them. They have a proprietary app store and they have some stuff, but it is, I mean, it is like vanishingly small number of apps compared to the Play Store. Yes, they're sideloading and there's, you know, Android is Android. There's other app stores, but the Play Store and then Play Services provide a whole lot of capabilities to Android apps? Yes. Well, and I think increasingly,

So does Gemini. And so I think we're now hitting this place. Like there was also some news this week about Gemini getting more and more integrated with the system and starting to do some like cross app stuff like we've been talking about with like the future of Siri. And so...

The non-Google version of Android and the Google version of Android are getting much further apart over time and not closer together. So I think the case here for something like Mirena is going to get harder, I think, because you're just going to have to make a broader set of sacrifices. Because I think if I don't want to have a Google account, it's not that hard to get a Proton email account and use...

DuckDuckGo and like you can relatively successfully replace most Google services with something else without trying all that hard.

But if you believe in all this AI stuff, and especially like the way it might change the way that you use your devices, taking that out is going to make all kinds of other stuff just break. Yeah. It's like, oh, I don't use HTTP on my web browser. It's like, well, you can't use the internet. And it does feel like a lot of the AI stuff is heading that way. Oh, David, you explained this in your piece about Marinette. They're just downloading the apps through Google. They're just piping you to the Play Store.

And then they have a clone of Play Services that runs. Yeah, I mean, I knew they had their own clone of Play Services that doesn't feed any stuff back to Google. I had forgotten that they just take apps from the Play Store. Fair enough. Yeah. I mean, you wrote this three years ago, so you're forgiven for not remembering about the weird open source Android clone. I have thought this company is interesting for a long time.

I mean, I'm fascinated by this and I'm fascinated by the button that turns it into minimal mode on the phone. Like it's a switch. Love that. Yeah. But yeah, I mean, this whole, the fact that you have to commit to an entirely secondary operating system, that's a, it's a rough, it's a rough ask. Totally. But I think the, the, there's also like something in this combination of like more repairable, you can swap out some parts and this like,

but growing accessory ecosystem that like there's something there if you're a person who wants to sort of muck around with your phone from time to time that I just think is very cool. And to Jake's point, this is the first one that is like on its own a pretty compelling phone, which

which is awesome. This is the thing. I feel like in the past, the Fairphones have always looked a little bit lesser than, and I think this one, obviously we'll still have to review it, see if it all holds up. This is the first one that you look at and it looks like a modern phone. It does not look like you're going to dramatically miss out on things with the exception of those living in the United States. So that's a bummer, but I'm really excited to see how this pans out. All right. There's one more phone that I'm desperate to talk about this week. It's the Titan 2.

which the only way you can describe this is that it's a BlackBerry that runs Android. That's what it is. And it looks exactly like an old BlackBerry. It has a screen and a physical keyboard. And I don't know who this is for. It has already raised a million dollars on Kickstarter. So this is the thing I keep, and it's done big traffic on the site. People are into this thing. And I think

My overwhelming theory about devices like this for years has been that people have real nostalgia for this stuff, but not actual desire for it. It's the same thing as small iPhones. Everyone on earth is like, oh, God, remember when the iPhone 5 just fit perfectly in your hand? Wasn't that so great? And then you present them with a phone that small and they're like, why is this so small?

I don't want this. The battery sucks. What are we doing here? I've always assumed that that is the same thing with physical keyboards. People remember their BlackBerry because it was their first smartphone and it was really exciting. And it was like a simpler time in the tech world. And that was all very cool. But then things like the Clix keyboard come out and they're like sort of a compelling novelty and they're fun and exciting to have and play with. But it's not

A thing that is like a mainstream viable product. But then this thing comes out and raises a million dollars on Kickstarter. And so I'm like, there's clearly some people out there who want a giant ass slab of an Android phone with a big keyboard on it. It looks delightful. There's two things about this that I think are special that we need to talk about.

One is that on the backside of it, there's a second screen that shows notifications and a clock. It sort of just looks like an ambient Apple Watch. It's pretty cool. It's interesting. Normally, you're like, I want to have my phone out but not have it out, so I flip it over. And this way, you flip it over, but you still have your phone out. So it's actually, you're just not being rude. You're still being rude, which I think is delightful. And then they also advertise multiple times that they have an infrared port

And they do not explain what that means or what you do with it. You obviously, you turn off motion smoothing at the bar, Jake. But is it an outward infrared port or is it an accepting infrared port? Can I control my phone with a remote? Oh, I see. Is it an IR glass or an IR window? They do not clarify. They just say phone calls with my TV remote. That's what you're talking about.

That would be pretty innovative. I don't know if it would be useful, but... I'm going to go ahead and say it's an IR blaster. I mean, that's why this thing has raised a million dollars. Yeah. People just want to go to the bar and click around the menus. This is officially the most 2013 phone we've seen since 2013. The standard was called IRDA, I-R-D-A, and you could very, very slowly beam contact information between phones, like dumb phones back in the day.

It's a real thing. Yeah. It wasn't great and no one used it, but it was called Erda. And there was a time in gadget blogging where like if a phone didn't have Erda, like the commenters were like, no Erda.

I will say, to this thing's credit, I think one very compelling thing about it is that you can buy it right now on Kickstarter for $269. And I think that, for like a half novelty of a phone, is a perfect price. Two Trump phones or four Titans? Oh, four Titans. It's not close. Give me a Titan and a Fairphone over the two Trump phones.

I'm a happy guy. There is also actual motion footage of this phone existing, which cannot be said. There's not a single frame of the Trump phone existing. All right. Some more gadgets to round out our first segment, which is very gadget heavy. I point out that we're all over the place with gadgets. Microsoft and Meta announced an Xbox VR headset that is just a Quest 3S that's green and comes with a controller.

One of my favorite things to do this week has been to troll Reddit.com.

watching people be confused about what this thing is. It's so funny. So two things are true, right? One, this is the best looking Quest there has ever been. The black and green color works for the Quest in a way the white, beige-y thing kind of doesn't. The combo of Xbox stuff and Quest makes sense. Shipping it with a controller that's already paired makes sense. This thing is a gaming machine. They're treating it like a gaming machine. It's going to stream Xbox games. Makes perfect sense.

There are so many people who just think this is an Xbox that will play Xbox games. And it turns out it doesn't. And we, this is what we've talked about before. Like the list of Xboxes that don't play all Xbox games is vast. And this is one of them. It will, it will stream some games. If you have Xbox game pass ultimate, which requires jumping through a series of hoops. I cannot even begin to describe to you. And it's,

like this thing is it both is and is not an xbox in a way that is driving a lot of people on the internet absolutely up the wall wait everything at this point is and is not an xbox the state of being an xbox is probabilistic this is what i mean yes but this one's called xbox like every morning i wake up and think to myself am i an xbox and i go through the list of what games i support and yeah and many days the answer is yes yeah it's beautiful for you is

Is your Samsung frame TV an Xbox? It could be because you can run Game Pass on it, right? So wait, I have not paid a lot of attention to Game Pass. So this is just a VR headset and it's got the one app in it that runs in 2D, right? So you're in the VR environment and it puts up a big fake TV, a big virtual TV, and then it just runs Game Pass on it? That's right. It's not even running VR games? No, the Quest does that anyway. This is just...

purely like a branding exercise. And again, I think the branding exercise makes a lot of sense, but it is just, it is just a branding exercise. This is, this is a, like, I don't know. Imagine some celebrity picked a color scheme for a quest and it did that. This is that.

It's just called an Xbox. So people assume it's going to be an Xbox. Well, it's an Xbox in the sense that Xbox Game Pass Ultimate is Xbox. If that's an Xbox, we're all Xboxes. That's just what I'm saying. Every morning I wake up. Yeah. All right. It's very confusing. I'm trying to figure out who is making money here. Microsoft seems to have some very long-term plan here to redefine what the Xbox is. And they are really overshooting for what they can actually deliver on today. Yeah.

That seems to be it. I do not see how Xbox is benefiting from this right now when they're just confusing everybody. So my understanding is that Xbox came in second place the last two generations, right? Yes. The PS4 and the PS5 blew the Xbox away, and then the Switch, and presumably the Switch 2, is throwing away gray, and that's going to be its whole other thing over there. And it feels like Microsoft is reacting to coming in second place as though they have failed so completely they need to burn it all down.

Right. And like we will reconceptualize the very concept of an Xbox and turn into a PC console gaming situation on handheld cloud streaming service on VR headsets. And then the actual console is like of little value to them. Yeah, I think that's right. That's kind of what this all feels like. But I'm just saying the Quest 3S is not a huge hit for Meta.

There's been reporting that they've kind of lost interest in Quest because their glasses are the success, which we'll talk about in a second. The Ray-Bans and now the Oakley smart glasses feel like their future. And so it's like, okay, an unsuccessful VR headset is being paired with an unsuccessful gaming brand to accomplish success.

What? The thing about it is I actually think in a certain way, calling this thing an Xbox makes it make more sense because it is a gaming machine. Like everybody who works on VR stuff will tell you this is not true, but VR is a gaming platform. It just is. But this doesn't play VR games.

This is the same as one of those X-Real glasses that just puts up a fake TV. Yeah, that's also an Xbox. Apple Vision Pro, also an Xbox. Do you know what I mean? What isn't an Xbox? Help! But I think to the extent that this just screams, this is a large screen on which to play games, I'm not bothered by it. But the thing you're describing is why people are getting tripped up because it's like, oh, this should now play my Xbox games.

And it isn't that simple. And so we're in this position of like nobody, many more people know what to make of Xbox than know what to make of Quest. So I can actually see why bringing Quest and Xbox together makes sense for shipping this headset. Except it's not fully an Xbox. No.

I just, I get turned in circles, but ultimately this thing is like a limited edition and looks very cool and not that many people get it. So it's like whatever. But I do find it sort of fascinating that they're like leaning into this thing is for games and Microsoft is like games are for everything. And they're like, oh, I see why you're friends. Like, sure. Yeah. Just on the one hand, you know, Microsoft spent last week or I guess two weeks ago talking about the, you know, the ROG ally.

And being like the future of Xbox is Windows games on the go. We also support Steam. And one turn later, they're like, it's a limited cloud gaming library on the quest. It does. It just feels like they've they've they've lost sight of what they want Xbox to be. So it's everything. And yeah.

Who knows? Agreed. The Meta smart glasses, they are a success, at least according to the metrics Meta keeps saying. And they've expanded them this week. And now they've got Oakley smart glasses. Same basic concept as the Ray-Bans, only they look like Oakley's.

And they shoot 3K video, slightly higher resolution video because they're slightly bigger is basically my understanding of the product. Yeah, pretty much. The thing I thought was really interesting, Alex Heath talked to one of the folks on the team building this stuff and they confirmed what we have always suspected, which is that people use these glasses for two things. One is audio and one is camera. And they call them AI glasses, which is very funny. It's like, well.

We're going to call this thing a thing no one uses them for. But like this, this positioning makes sense. Doing it with Oakley makes sense. I don't think these glasses are particularly good looking, but I also can't pull off Oakley's in general. So I'm not,

like a great person to judge. When you went on the eighth grade trip to DC, did everybody buy fake Oakleys? Oh, a hundred percent. Yeah. I've owned so many fake Oakleys in my life. We called them. Would you call them? Jokeleys. Oh, we were, we were Folkleys. Oh, very good. Yeah. They were good. But yeah, I think like this, this thing that is doing continues to make perfect sense to me. Uh, it's, it's still not clear. Uh,

how you bridge the gap between this and the Orion glasses that we talked about before, because they're like the technical breakthroughs required to get from one to the other are vast and unsolved. Uh, but just this thing where like meta has a hit at relative scale doing this basic kind of stuff. And they've just decided to lean into, this is the thing that everybody wants. We will figure out how to stack capabilities on top of it over time. Um,

Feels like the right move to me. It also feels really clear that their partnership with Luxottica has just continued to pay dividends here, right? Like you need glasses that are naturally chunky enough to fit this hardware. And they have managed to find multiple styles here that adapt well to this additional hardware that are already available.

that people know and love, right? These are very popular glasses. They're very popular brand names. Expanding to Oakley was a really smart next move after Ray-Ban. And so I, right, like this is a challenge that every other company getting into the glass space is going to have. If you put out knockoff pair of Ray-Bans or Oakleys, people will be able to tell that they are a knockoff pair. Mm-hmm.

Right. Meta's biggest advantage here is it is just selling Oakleys with some headphones in them. Right. That's not a bad product at a bare minimum. Right. Nobody has to use the AI. Well, I think Meta is desperate for a platform where their AI assistant is the first party assistant.

not an app on the iPhone. And they've settled on his glasses. I will say they, um, they transition lenses are available, but all the cool photos are the reflective orange lenses and Meta's continued attempts to make transitions. Very cool. Uh, continues to amuse. That's all I'm just saying, especially because if you're wearing transition Oakley's many decisions have been made along the way. Vindication is coming from my dad. Who's been wearing transitions for like a decade. Um,

The world is coming around to Joey. I don't know if you saw Joanna Stern's video with Craig Federighi and Greg Joswiak, but Joswiak was wearing transitions in the video.

And about halfway through they tinted because they were in the sun. And I was like, well, this is, I mean, it's fine. Like do it, do you, whatever you want to do. But the fact that you can't control them actually sometimes causes problems. Maybe it's a defense mechanism where he's like, these questions are too hard. I just black them out. Like, I can't see you. Just like you can't see Siri. I don't know. I like the fact that it's camera and, and, and headphones exactly the same problem as Alexa had for years. It's timers and music.

And it's great. I think people like having cameras on their face. I got a TikTok ad today for what can only be described as a necklace with a MagSafe mount. Huh. Because people mount their phones here when they drive their cars and do stuff and they want the first person view. I get it. I get why people want cameras in their glasses. I don't think Meta has actually. I would love to know the rate at which people who buy these glasses also use the AI stuff.

Is meta convincing 2% of those people to be like, I love the meta AI assistant? How many people are looking at stuff and being like, what is that? Because that's the use case they keep coming up with. Again, I think the real thing people want to build is Orion, the augmented reality glasses. And even Orion doesn't actually do augmented reality yet.

We're long, long away. But if you love an Oakley and you like glasses, Joanna Stern, speaking of Joanna, loves her Meta Ray Vans. Loves them. Can't get enough of them. I wear mine all the time. Just as sunglasses that happen to do other things. Like, that's great. Do you have the transition lenses? No, mine are just sunglasses. So I just get to be the cool guy who, like, wears sunglasses inside the coffee shop for a while. Does the Meta app support multiple glasses? The last time I tried, I could only be paired to one at a time. I mean,

I mean, these are just like, right. These are, these are problems with the form factor that you're going to have to overcome. Nobody wears the same pair of glasses all the time. Right. Like even people who wear glasses, like prescription glasses, I have two pairs of glasses because when you go to the store, my glasses are always like, do you want two pairs of glasses? And then people have sunglasses, have lots of different kinds of sunglasses. Like being locked into one pair, one style sunglasses forever is not great. And now they have multiple styles. I don't know. I'm just saying.

On the chart of wearable bullshit, you still have to put the thing on your face, man. Utility has to be much higher than it is right now. But you love them. They're great. And they've solved the face problem because they just look like sunglasses. That's the thing. That's to Jake's point. The hardest thing about this is just provided to Meta by another company that is fabulously good at making glasses people like. And...

rather than do what Meta tried to do and what Apple is trying to do and what traditionally tech companies have always done, which is like build the hardest, most expensive thing and then try to like trickle the tech down to cheaper prices. Meta was just like, we're going to do the least.

and then try to build up from there. And I actually think in this particular case, in part because of, you know, the problems that come with Nilay's chart of wearable bullshit, that's the right move. Like make a thing people like and then figure out how to get it to do more stuff strikes me as way more compelling than I'm going to build a helmet for your face and then figure out how to make you like wearing it. The Apple Vision Pro, everybody. Yeah.

All right. We got to take a break. We'll be back. Speaking of meta, we got to talk about their big AI win and tropics, big AI win, but they're also losses. We'll, we'll be right back. We'll explain. Support for this show comes from Indeed.

you need to make a new hire and fast. But you know that the hiring process is slow and tedious, unless of course you use Indeed. With Indeed's sponsored jobs, you can stand out from the crowd and connect with the right person fast. With sponsored jobs, your posts jump to the top of the page for your relevant candidates, so you can reach the people you want faster. And when you look at the numbers, they say it makes a big difference.

According to Indeed data, sponsored jobs posts directly on Indeed have 45% more applications than non-sponsored jobs. Their data also says that, in the past minute alone, 23 hires were made on Indeed worldwide.

There's no need to wait any longer. You can speed up your hiring right now with Indeed. And listeners of this show will get a $75 sponsored job credit to get your jobs more visibility at Indeed.com slash Verge. Just go to Indeed.com slash Verge right now and support our show by saying you heard about Indeed on this podcast. Indeed.com slash Verge. Terms and conditions apply. Hiring Indeed is all you need.

Support for this show comes from Salesforce. You might remember a time not long ago when AI wasn't all that helpful. But today, AgentForce, the powerful AI from Salesforce, can analyze, decide, and execute tasks autonomously, operating at speeds and scales no human workforce could match.

These AI agents represent a new world of digital labor that not only handles monotonous, low-value work, but orchestrates and carries out high-value, multi-step tasks. This isn't just another step forward. It's an enormous leap, redefining how work gets done and what's possible for businesses and their employees. AgentForce is adaptable,

autonomous and proactive, and of course, totally integrated into Salesforce. So they're truly part of the team. That way, you and your employees can focus on the tasks that actually move your work forward. AgentForce, what AI was meant to be. Learn more at salesforce.com slash agentforce.

As a contractor, I don't pay for materials I don't use. So why would I pay for stuff I don't need in my mobile plan? That's why the new MyBizPlan from Verizon Business is so perfect. Now I can choose exactly what I want, and I only pay for what I need. Right now, with MyBizPlan, get our best price, as low as $25 a line. Visit verizon.com slash business to get started today. New lines only. Price per month with 5 plus lines. Includes auto pay and paper free billing and promotional discount. Taxes, fees, economic adjustment charge, applicable add-ons, prices and terms apply. Guarantee applies to base monthly rate and stated discounts only. Add-on prices additional.

Offers end June 30th, 2025. All right, we're back. Addie Robertson's here. Hey, Addie. Hey. I'm excited. Usually when you and I talk, something very dark has happened.

Like a lot of times you're on Decoder and we're like, well, Addy's here. And we both have the same reaction. I'm excited to talk to you about AI and copyright because this is just chaos. Like it is delightful chaos in our nation's courts as the AI copyright cases begin to wind their way to some decisions. Oh, yeah. These are fun. Can I back us up here a little bit and just like

build from the ground up a little bit because i confess uh i have had trouble keeping these two cases this week separate in my head uh and then also trying to figure out what they have to do with each other because it seems to be a lot so uh addy can you just like walk us through these trials a little bit first let's do anthropic first because that that decision hit first this week uh just remind us quickly like what that case was about uh and what was decided this week this

This is David saying, you know, I can't immediately get in this feeling. So that's just what happened. Yeah. Neal, I was so fast to like, let's talk about the whatever the fourth factor of copyright law. And I'm like, can I remember what books are first? Yeah. So there are two groups of authors here, and both of them found out their work is almost certainly ingested into the training data set of a chatbot company. In one case, Meta, in one case, Anthropic.

And they both sued. And there are kind of two arguments. One argument is these things were pirated in the first place. And then in some ways, the bigger argument is, well, even if you get these things legally, you are putting these into a training set that is then using them to power a very lucrative product. And that's unauthorized. That's copyright infringement. You need to get permission from us and pay us. OK, so this became legal.

Nilay, a fair use thing, kind of all the way down, right? It's very important to separate what specifically is fair use. And the courts here don't agree. There's a lot going on in these, but it's important to just think about AI from the technical perspective. There's training the model, and then there's inference, where you get outputs from the model. And in these cases, it's all about training the model. Is AI

In the Anthropocase, literally buying a bunch of used books, ripping off the covers, scanning them into a database, and then using that database to train a model is all of that fair use. Not if you ask the model to produce some copyrighted book, is that copyright infringement? And that's a big distinction. The reason people hate lawyers is because lawyers are constantly making extremely pedantic distinctions between things.

But in these cases, those distinctions are wildly important because the judges react to those distinctions between training and inference, between inputs and outputs, very, very different. So both of these cases were fundamentally about training, right? Both of these cases are sort of fundamentally about training with a side, which is actually also very important, of you pirated these books in the first place. It's kind of the difference between, again, in both cases at first,

They were like, all right, we're going to download all these books and we're going to train on them. And then Meta kind of kept it there. And Anthropic went, wait, we need to fix this. We're not comfortable with this. So then, as Neil, I mentioned, they got a bunch of books and they destroyed the books after scanning them. So then there was a copy in the cloud. And this turned out to be a really hugely important distinction for reasons besides the training. If I told you to just go buy a million books, would you know how to do it?

I've been thinking about this all week. Anthropic hired one of the people who ran Google Books. Like they found the one person who knew how to do it. So the Anthropic case is really interesting for that reason. They started by saying we're just going to steal stuff and they pirated books and they referred to the problem they had is like legal business slog.

That's in the opinion in the legal business law is like, should we pay money and acquire things legally? And I'm like, no, we'll just steal it. And then they came back around and were like, we're going to scan a bunch of books. And the judge in that case, adding my read was that he cared a lot about acquisition. He was so mad. He's very focused on did you acquire these books legally or not? He makes a distinction between the legally acquired books and the pirated books.

And then he spends a lot of time being like, okay, once you have the stuff, what can you do with it that's fair enough? And the first thing, and this is really just, again, pedantic steps down the line. The first thing that Anthropic did was they made a big database of books. And that is, you don't think about that in the context of you're training a model. But the judge spends a lot of time being like, okay, the first thing you did was make a database of books.

And then you use that database. You can use that database to do all kinds of things. You can just search the database to see what books you have. And he, you know, there's a lot of like consumer rights history here, a lot of Google history here, ripping the covers off of books and scanning them into a big database. If you wanted to do that to all the books you owned, you would be like, I can do that.

Right. Like if the judge here had found that buying a million books, ripping the covers off and scanning them into a database was illegal, like a lot of things would break. I think like there's some part of this where copyright is such a maximalist control over what you can do with the things you can buy.

The only answer you can get to is you're allowed to buy a million books, rip the covers off of them and scan them into a database. As long as you don't let other people access the database. Yeah, it's pretty notable here that they went from, OK, yeah, obviously it's illegal to acquire these digital books, but there are many legal ways to acquire digital books. And it is notable they didn't do that because you have the most maximalist freedom as a consumer with a physical copy. Yeah, it gets like you if you want to buy a bunch of Kindle books.

Like they come with DRM, they come with license restrictions. You can, as I was saying, you can buy a bunch of eBooks without that, but you still end up with like terms of service agreements. If you just go buy a book, you can kind of do a physical book. You can kind of do whatever you want. So the judge says, okay, you made this database. That was fair use. Ripping the covers off the physical books and scanning the database. That's fair use. You're allowed to do that. Stealing books and making a database. That's not fair use.

Stop there. The end. You're in trouble for that. Is that even a fair use thing or is it the stealing the books? Well, he's made the distinction between is it fair use to buy a bunch of physical books, rip the covers off and scan them? Okay, that's fair use. Is it fair use to steal a bunch of books?

No. There are sort of a bunch of separate charges here. There are the charges of, all right, is it copyright infringement to do this? Or is it also separately copyright infringement to pirate all these books? Right. So then Anthropic tried to fix it by saying the database is not important. Training the model is important.

And the judge said, okay, training the model from your legal database. That's fair use. That's very transformative, especially because the authors one conceded that trading the model was just like reading a book.

Which is a big thing to concede. Right. Which has been an argument of the AI industry for years, right? That it's like, it's the equivalent of reading a webpage and talking about it. And this is going to come up over and over again. And I think this is why Addy and I both think these are fun. The lawyers blew it for the authors, like left and right. And the judges are kind of like, you know, you guys are blowing it. Like, try again, but don't blow it. So the lawyers conceded that it was just like reading a book. And then they did not argue the outputs of the model would affect the market for their work.

Right. So they stopped. And this is why it made the big distinction between inputs and outputs. They stopped at training. They did not get to. If you ask Anthropic, if you ask Claude to make our book, it'll just make our book. Or if you ask Claude to summarize our book, it'll just like spit out the answer to our book. And that will ruin the market for a book. They just didn't go there. And the judge is like,

This is where you need to explain the different tenants of copyright and fair use stuff. Okay. So, David, the four factors of fair use. Here we are. We got here so inevitably. You've told me this so many times and it just goes fully. My eyes glaze over every time you say this to me. So, I'm going to try and listen this time. It's 17 United States Code, Section 107, the four factors of fair use. So, this is when someone accuses you of stealing their work. You say, no, no, it's fair use. And so, it's what's called an affirmative defense.

You say, yep, I did. I did sample your song. This is fair use. Yep, you're right. I made a copy, but it's fair use. And then the judges are supposed to evaluate this on a case-by-case basis. They have four factors, the purpose and character of the use, the nature of the copyrighted work, the amount and substantiality of the portion used, and the effect on the use of the market. And-

How important each of these factors are just goes up and down over the years. The Supreme Court issues fair use copyright decisions. You get law review articles being like the third factor is no longer in vogue. Like it's everyone's always arguing this. But here, what everyone should be arguing about is the fourth factor, the effect of the use on the market.

If you train a chatbot on all the books in the world, and then you can just ask the chatbot about those books, and it will tell you the answer, the market for those books will inevitably dry up. And this, if I'm remembering correctly, is like the New York Times whole argument in its case against OpenAI, right? That if you can get OpenAI to regurgitate New York Times articles-

You can. Why would anyone go to The New York Times? I mean, I would be careful with that because in that case it is specifically output. They are saying that you could get them to regurgitate exact text. And here it's kind of more a you can generate a bunch of stuff that imitates us and that infringes on our ability to sell things or also that it can summarize and it can give answers. And that can even if it's not really direct, clear, you're copying our stuff that's still diminishing the market. And then in the anthropic case.

The authors didn't allege the outputs, right? They just didn't argue about it at all. They were totally focused on training and they didn't get to outputs. And the judge is even sort of saying throughout the opinion, you didn't talk about the outputs. So your argument is merely training the model will have some effect on the market for your work. But like, what is it?

You didn't argue. It's not here. Like, we don't have a record for it. So I'm just not going to talk about it. So we're totally focused on training and not on inference, right? We're totally focused on the inputs to train the model and not on the outputs of the model itself. And so the authors lose. He says training is fair use. This is totally transformative. You can legally acquire books. You can put them into a database. You can use that database to train a model. Making the database is fair use. Making the model from what's in the database is fair use. And then stop.

Right. And like, there's a big yawning gap there of, well, now you have a model. What do you think is going to happen? And the answer is like, oh, you're going to make a bunch of stuff that competes with the author's original works. And so that's a real puzzle. Like who knows what happens next? And he even leaves it open. Like if you want to lodge the next thing, you can lodge the next thing. And then on top of it, he says, you, what you are not allowed to do is pirate a bunch of books and make a database.

He doesn't even get to, and then you can train the model. He just says doing that first step where you stole a bunch of books and put them in a database, that's not fair use. You can't do that. And that is a nightmare. Like you might read this as training the model is fair use for your legally acquired physical books that you've ripped the covers off and scanning your database. What it does not say is it's fair game for anything. And so if you, for example, say,

have opened YouTube and downloaded a bunch of YouTube videos. I'm looking at you open AI. Well, you violated YouTube's terms of service. The copyright and all those videos is owned by the YouTubers. You have a, you have a database of content that you're training Sora on that you don't have the rights to. And so this anthropic decision has opened the door to being like, we're not even going to get to training as fair use. We're going to get to your illegal database of YouTube videos.

And you can't you're not allowed to clean up that problem by saying training is fair use. So this first decision, very complicated, like very technical, very readable. We've profiled Judge Alsup before. Sarah has profiled him. He writes code. He's he's in the weeds of the technical details here.

What you have is training is fair use if you have a legally acquired set of content to train from. There's also kind of a backwards looking and a forward looking issue here, which is, say, in the future, if you're kind of just starting from scratch, then the upshot of this is, well, OK, you've got to buy a million books. That costs a lot. It's a finite number. In backwards looking, if you've already done all this stuff, the potential damages for each work are just

like ruinous if you add them up to a large enough number and so anybody who's starting from scratch with a model that is starting acquiring all this stuff legally like that's one economic model any of the big players that's a whole different one yeah because they've already done all this and again open ai nobody has a lot of money but every youtuber in the world can read this decision and say well you you stole our videos

And it wasn't, you can't buy physical YouTube videos and rip the covers off of them and send them into a database. You access them on YouTube, which has terms of service that says you cannot do this. So this is already like puzzle one. This decision, you read it and you're like, okay, there's a way forward if you pay a bunch of authors. And maybe it doesn't even work, right? Used books don't pay the authors, but there's parts of copyright law that say you're allowed to buy used books and do whatever you want with them. And this ruling reaffirms that.

It does not say you can just pirate a bunch of stuff and train it and call that fair use. Okay. So here's a bunch of liability for the industry. Then literally the next day, a decision in the meta case comes down, which Addy, I'm reading is-

Basically being like, no, actually, Judge Alsop got this wrong. Like, this is all theft left and right. It's basically Judge Alsop got all the fundamentals of this right, but he weighted the transformative factor, overarching factor, and the fourth economic market value factor totally wrong. And that makes him completely wrong on all of it. Yeah. Go through this case. So this case is slightly different, but it's a bunch of authors suing Meta. And-

if I'm correct, Meta got caught using torrented books. Like, straightforwardly, they just torrented the books. From Library Genesis, I believe, yeah. The collection of books that was just this very large corpus of pirated books, they torrented

torrented them there's debate about whether that included them uploading things which one typically does when you torrent things the classic cedars versus leechers debate it's our nation's copyright law yeah they have to define seating and leeching in there there's a whole section about that but they're specifically not dealing with the copyright uh like was this direct copyright infringement right now except to say yeah look it basically was

So there's that. And then after that, these authors sued in a pretty similar way, except that they also specifically said that

The issue here is partly that it is recreating our work. So they did focus on output. And then also that there is this market potentially for us licensing our work to you and that that's a market that you're taking from us. The judge called both of these clear losers. He basically thinks everyone here is wrong.

So the meta won the case. Meta won the case just handily. And it's because the judge says, look, those are the only things you focused on. And both of those are bad arguments, partly because they couldn't get Lama to actually spit out more than 50 words of the direct text. So they were trying to say, look, you're verbatim producing stuff and it just didn't work out. And then the second one, because you just can't necessarily say you're

yeah, if I were allowed to license this, I would make a lot of money. That doesn't necessarily mean you should be able to exploit that. And so then after saying,

look, concluding just this argument is terrible. These lawyers are bad. Please don't do this. They said there's this other argument that would have been way stronger. And it's that the fundamental use of these things is to flood the market in a way that destroys the ability of authors to make a living and completely diminishes the thing that copyright is meant to do, which is encourage human creativity. Basically, AIs are creativity killing machines that

And that should mean that it outweighs any potential transformative value. So I want to just quote from the opinion because it is brutal. This ruling does not stand for the proposition that Meta's use of copyrighted materials to train its language models is lawful. It stands only for the proposition that these plaintiffs made the wrong arguments and failed to develop a record in support of the right one.

He's just saying you screwed this up. Yeah, you did it wrong. You did it wrong. And then mentions, by the way, these guys only they only represent 13 artists, 13 authors. So, you know, there are a lot of other authors and maybe Meta scan them. It probably did. Just saying that. Yeah. Yeah. Wait. OK, so let's talk about this here, because the reaction to this this week has been fascinating because I think it's been sort of a Rorschach test for.

depending on who you are promoting here. All of the AI folks are like, this is a landmark victory for AI. This puts us on the path to do all the stuff that we needed to do. But then it also sounds like there's an easy read of this, that this is like a small victory for these companies who won this case on a technicality, but have just had the door thrown wide open for everyone else on earth to sue them better and win. Right?

Which of those things is it? Or is it both of those things? It's the second one. It's so by far the second one. I think, you know, Addy and when Sarah's on and I, whenever we're talking about the legal system on the show, we are constantly reminding people that the courts are not computers. They are not deterministic systems.

You cannot predict what will happen. And so I think the AI companies are like, this is fair use. And then they got an outcome that kind of looks like this is fair use. And they're like, see, we predicted this correctly. We were right all along. And what you actually have is a court saying these lawyers sucked. And if you had better lawyers and you argued these things differently, we would have reached different outcomes. But we can't reach those outcomes in the absence of your better lawyers.

And that is a maddening to people. I know it's maddening to people because I have the Blue Sky app on my phone and I open it from time to time. And it's just people being mad that these lawyers blew it. And they blew it. And the judge is saying flatly they blew it. I don't know why in the Anthropic case, the lawyers happily conceded that training an AM model is like letting school children read books to learn things. I wouldn't have conceded that. I do know that it costs a lot of money to build an expert record that

about how AI systems work versus the brains of six-year-olds or whatever, and they may not have had that money. But you can't know, and that's why it's not deterministic. Like, a lot of decisions got made along the way, and then we have these rulings. But there are more cases to come. As Addy's saying, the court in the meta case is basically inviting other authors to try to make the better argument that they have laid out.

You have the New York Times, which is over a million dollars deep into its open AI litigation that's watching these outcomes and will change its arguments. Then you have Disney, which is a very late entrant into all these copyright lawsuits. They're partnered up with Universal, and together they're suing Midjourney about outputs.

They like they're they're watching these cases develop. They're watching these arguments and they're like, oh, we need to focus on outputs. By the way, I need to disclose the Universal's on my Comcast. NBC Universal is a part of Comcast, an investor in our parent company. Our parent company has a deal with OpenAI. One of the reasons these companies have deals with OpenAI is because they want to set markets for licensing. They want to set rates so these lawsuits can proceed on whatever way they're going to proceed. You can see it's not settled.

These are just the first little skirmishes. And a lot of times what's happening is that the first plaintiffs are authors, are creatives who don't have the money for the most lawyering who need to get to outcomes quickly. And they're getting outgunned by the lawyers for the huge technology firms. That's not always going to be the case. Record labels are also in on this. The record labels have sued Tsuno and Yudio. Although they're starting to settle some of those on the side. Yeah.

I'm just saying, like, the legal system is not just feministic. It is also reactive to what happens before, like, necessarily. It's precedential. It's reactive to these cases. And the arguments in the future cases will shift around these decisions to attack their weak points or, in the case of the meta decision, to take the open invitation to get it right. Right.

I would also say we're at the point where basically this is kind of every judge for themselves. At the point where you appeal and then you start getting appeals court cases and then circuit courts start saying, all right, here's roughly how we think about it. Then judges have kind of precedents that they are trying to draw from. And of course, if the Supreme Court says something, then that's something that you have to wait pretty heavily. But at this point, they can kind of just say stuff to each other. So it's also not like this one ruling that, say, also pans down and

is a law that everyone else has to follow. Somebody else can come up with their own theory of fair use, which is pretty much exactly what we've seen. Right. And fair use in the law is supposed to be case by case. Right. So one fair use decision is not supposed to affect all the other ones. In reality, because of the money and time and risk involved, they get treated as precedent. But here, everybody has incentives to reopen those doors again and again and again.

So, I don't know. Addie, what do you think? I think this is a bunch of small, narrow rulings that open the door to much more chaos to come, not definitive reckoning. But maybe you feel differently. I don't know. In part because I feel like sometimes court cases are a little bit just they're reflective of a larger vibe and culture that it's like I think we've seen with Section 230 cases that it's

It's not necessarily that judges all impact each other, but that if you start seeing a bunch of rulings around one kind of particular interpretation of the law or one kind of like weighting of something, that sometimes means you are seeing a pattern that's going to continue. It's correlation, not causation.

And also, I don't know, I'm actually wondering what would happen if Ulz had a case that was more specifically output based. Because on one hand, yeah, it opens up a bunch of whole new cans of worms. On the other, some of his comments did seem like they kind of indicated it doesn't matter if somebody is, say, writing...

a book with an AI system that competes with yours, that doesn't inherently mean that they are, that is unfair. Yeah. He's got a line that's like copyright law does not protect you from competition. It's supposed to inspire more people to create. And there's a, you know, there's a very deep, meaningful tension in there because the mechanism by which copyright law is supposed to inspire you to create is by letting you sell your stuff.

Right. You create stuff, you get rights to it. Other people can't take it for free. They have to pay you money. But at the same time, if they take your stuff for free and then make stuff that competes with you, like that's competition. There's something really weird in there. But there are also a lot of pre-AI cases where I think people would be a lot more sympathetic to the idea that you should be able to say, okay.

copy things without having to get permission. Like the fact that you can report on facts that somebody says in a news story without getting permission from the original writer of that story. Like the fact that you can do a pastiche of an artist and that artist's style is not necessarily a thing that's automatically copyrightable. I think these are all things that when humans are involved...

we actually are pretty okay with the idea that creativity is cumulative and people build on each other. I think the thing that's really interesting about the meta ruling is that it specifically makes the argument that scale is different, which is something I see a bunch of people make pragmatically, but I don't think I've seen in the court system as much that the argument isn't actually, well, these things are fundamentally different and not competitive because the writing is somehow different or because something's not necessarily learning. Um,

It's because you can create all of them so quickly that they will flood the market and they will destroy the market for books. Specifically, he thinks romance novels and spy thrillers. Romance novels come up a bunch in this.

The whole, the entire like romance novel, like fanfic community right now is like deeply insulted right there at war with AI every day. Yeah. Romance authors are starting to use AI and romance readers are just furious about it. Well, it's funny because you, we've talked about this many times. So if you go on those forums, like now people just say tropes all in a row and request stories and like, oh, that's what AI is for. Like, I just want this to happen. Yeah.

wild interactions in that community. The scale problem is really interesting because if we just play this game right now on the show, the moral dilemmas of copyright law become plain. So if I say to you, David, can you rip a CD onto your computer without paying extra money to a record label? What would you say? Yes. Can you rip 10 CDs?

Yes. Can you, David Pierce, rip 1,000 CDs to your computer without paying record labels additional money? I guess in theory, yeah. If you had the hard drive space, could you rip a million CDs to your computer without paying the record labels? You, David Pierce? Yeah, I think so. Can Meta do that? No.

Like, why? This is the problem, right? You write the rule that says it's okay to like rip CDs or we got a lot of Plex users on the show. Wait, wait, wait. Fair use can account for this, right? Part of the consideration of fair use is whether something is commercial or not. Right.

Part of the consideration is whether it's commercial, but in the Anthropic case, Judge Alsup says you're allowed to just make a database. If what you want to do is make it easier to search, make it easier to run different applications on, it's totally fine for you to buy a million books, rip the covers off, and scan them into a database. It's the next use that gets you into trouble.

Right. It's OK. Is it training the AI model? Is that fair use if you have the database? He says, yes. If you were to say, I've now made this database publicly available for anyone to read all the books or listen to all the music in my database, you would go to jail. Like that's the Internet Archive is in trouble for their database of books that they acquired. They scanned it and they let everybody have access to. So the questions get really difficult.

The second you start not just adding scale, but adding different uses of what they're going to do. And some of those uses are commercial change the equation. And this is the problem. This is why everybody hates copyright law. This is why everybody hates lawyers. You start to get incredibly pedantic about what happened when, to whom, who made the copy, when, where does the copy live? And everyone's like, it's just a computer, right? I can just like open it up and get whatever I want. And then underneath that, all of those questions determine who gets paid. And I think here, again,

The big problem with the anthropic fair use ruling is, one, if you buy a million used books and train a model that generates billions of dollars in value, the authors who wrote those books get nothing. That's just the way it works in our society. If they pay a one-time fee to buy a million brand new books and then train a model that generates billions of dollars in revenue, the authors essentially on that curve get nothing. They get paid one time. Is that fair? Yeah.

I truly don't know the answer to those questions, and these rulings kind of don't get you there. I don't know. I think that the meta-ruling is interesting because it does actually kind of cut through all this. Because everything in copyright is very angels dancing on a head of a pin, like is putting something in RAM a copy? But the overall argument is the whole reason copyright exists is to encourage human creativity. And so the ultimate bar should not be, is this fair or is this hitting this particular target?

place where it starts being uncomfortable or is the company that's doing this making it commercial. It is when you look at this thing holistically, does it serve the purpose that the law is meant to serve? And that's not mostly how courts work. But I think if you are considering this from a larger kind of philosophical sense, that is interesting and worth considering. Sarah and I have often crashed out because we both started as like

copyright minimalist warriors in our legal careers. And we've now like horseshoe theory around to like copyright law is good, actually, which is crazy. So the argument that you would make from the other side, and it's a good argument, is that human beings do not need copyright law to be creative. This legal system that we've invented is there to create scarcity and economic value for creativity, which certainly incentivizes some people to create

But if actually, if you look around, our instinct is literal human beings every day is to create. We will do it for free. People wake up every morning and they have access to creative tools on their computers and they cannot help themselves. Like it is just the, our most essential human nature is to make stuff. And that's in real tension here, right? Like it, it,

It's hard to square the court saying we have to preserve the essential nature of copyright law so that there will be creativity. And then literally those judges go home and they open the Instagram app on their phone. It's like the most free creativity that has ever existed in the history of humanity. I mean, somebody at Andreessen Horowitz is going to play that clip that you just said for everyone who works there. That's the crash. That's the I mean, that is the argument that these companies make in favor of all of this is that, no, we're not diminishing anything because everybody would do it for free.

Right. And then they make that argument from their yachts. I will make the counter argument to this, which is that while people will create a lot of the purpose of creativity is to share that thing with other people. And if AI creates a situation in which it is virtually impossible to distribute your work or to get anyone to see your human made work, I think that you can say that is really meaningfully chilling creativity. Right.

The problem is that I don't think copyright laws as it stands can solve this. There's this really bizarre tension in the meta opinion where, okay, if you take its premise as true, that this will completely destroy the market for books, then the only thing copyright law can do is say,

OK, well, some people have to get a few dollars a month as residuals for this to create this entire creativity killing database. Really, the only way that you can take this argument as true is to assume that also copyright law is going to be powerless to change it. It might be.

I mean, one outcome of all these cases is they hit the Supreme Court and the Supreme Court says, fine, training is fair use. Maybe that's an outcome. I think it is much more likely that as these cases careen towards that risk, because no one knows how that will go, that actually the incredibly well-moneyed interests here will go get a law, right? That Disney and OpenAI and Google, all these companies have to deal with each other. They will go, Nashville will.

is a powerful economic force in copyright law. Where does that leave the next generation of artists, though? Because the way that I see that gaming out is, yeah, all the record labels that currently have artists sell a bunch of data to companies. And then companies create this database and they train on it.

and they train a bunch of, I don't know, music apps and apps for writing romance novels and all of that. And then they create training data based on that, that then they sort of cycle into their system over and over, which is what they're trying to do right now.

And then at that point, where does the next generation get paid? I don't know that there's actually a way that happens, especially if there is this situation where, say, AI generated content floods everything. Yeah, I don't know. I that I think that will be the heart of the negotiation fundamentally, especially because private equity is in the mix like.

The creative industry is not clean. It's not pure. It's a lot of essentially banks now. And so, you know, private equity companies have bought Bruce Springsteen's entire catalog for five hundred million dollars.

They're not going to take a one-time training payment for that. They bought that because they think they can exploit that music for years to come in various ways. I don't know. I just think that's not actually satisfying because that means Bruce Springsteen has a steady stream of income, but that there is no market for the next person. No, Bruce Springsteen got his one-time payment. I mean, this is why it's particularly gross. Bruce Springsteen has half a billion. Okay, yes, fair. It's the private equity company that made the investment is going to want to return over time.

And that return over time is the thing that ideally makes any new work also valuable. And that's tough. I don't know how any of that plays out, especially because I don't think this gets litigated. I don't think it's solved at the level of Supreme Court. I think Congress has to write a law in that you can already see how messy that will be. And the battle lines of that law as these court cases wind ever forward are literally the huge entertainment companies that

basically own this kind of lobbying. This is what they do. And they're very, very good at it. Going up against Disney to walk back copyright law is not historically a successful effort. Going up against Nashville and the music industry to walk back copyright law

Not historically a successful effort. In fact, some of the only AI-related laws we have are like the Elvis Act in Tennessee that prevents people from making AI Elvis impersonators. Well, and things like the Disney one seem more sort of viscerally straightforward in that you can just go to Midjourney and be like, show me Simpsons, and it'll just do the Simpsons for you in a way that I think...

Things like these cases are a little just feel more abstract because it's like something goes in black box, something else comes out in that one. It's just something goes in and then that same something comes out for free and you can do whatever you want with it. And I feel like that.

We're just going to look at that differently. Whether we should or not, it feels different. The pictures always win. Legally different media are also different. Like you can quote a bunch of a book, but if you want to sample three seconds of a song, you're on much shakier territory. I would actually chalk that up to the fact that the record labels in Hollywood are better at this than the book publishing industry.

The news media is not great at this historically, and those industries are really, really good at it. And part of it is, to David's point, you have a more visceral emotional reaction to that picture is the same and that song is the same. They are good at fighting these fights. So my suspicion is...

here at it might not be satisfying is that you get to some solution where there's still value for the individual song. There's still value for the individual video. It might be less than before and there might be more supply, which will crash that value because there's only so many people and only so many minutes to consume songs and videos in the day.

But I don't think you get to a place where this set of artists gets paid once and no future artists get paid because that will kill these businesses and they're not existentially stupid. They can be stupid. They're not existentially stupid.

My really dark take is that the reason Disney is suing Midjourney is because Disney is about to use all of its IP to generate infinite Star Wars forever. So I don't know. I think we're just shifting the people who are going to be making these things and flooding the market. Disney learns wrong lesson from Andor is a real headline. It's right there for you. We need to take a break. We'll be right back.

This episode is brought to you by State Farm. Knowing you could be saving money for the things you really want is a great feeling. Talk to a State Farm agent today to learn how you can choose to bundle and save with a personal price plan. Like a good neighbor, State Farm is there. Prices are based on rating plans that vary by state. Coverage options are selected by the customer. Availability, amount of discounts and savings, and eligibility vary by state.

Shop 4th of July savings at the Home Depot right now and get up to 40% off, plus up to an extra $600 off select appliances with free delivery, like Samsung. From all-in-one washer dryers to smart refrigerators, upgrade to tech you can trust. With Samsung appliances, the Home Depot has what you need to simplify your routine. Don't miss 4th of July appliance savings at the Home Depot. Free delivery on appliance purchases of $396 or more. Offer valid June 18th through July 9th. U.S. only. See store or online for details.

Put us in a box. Go ahead. That just gives us something to break out of. Because the next generation 2025 GMC Terrain Elevation is raising the standard of what comes standard.

As far as expectations go, why meet them when you can shatter them? What we choose to challenge, we challenge completely. We are professional grade. Visit GMC.com to learn more. We're back. The lightning round, which David says has to be two minutes long. Two minute lightning round. Because we just spent an hour crashing out about copyright law. So let's talk about Brendan Carr for 45 minutes. America's favorite podcast within a podcast. By the way, lightning round unsponsored.

If we're in a Tesla and it crashes, we will point out that it crashes. I'm just saying. You can't buy us. I've noticed a lot of people have started posting on various social platforms at me that you can't tell us what to do. That's all I've ever wanted from the lightning round is unsponsored for flavor. Is people reminding everyone that you can't actually tell us what to do and you can't buy us. I think when you die, I'm going to write, you can't tell me what to do on your tombstone. And it's going to feel right. It is the animating thesis of the Verge.com.

That and figuring out what's going on with the frame TV. Okay. Brandon Carr is a dummy. America's favorite podcast and podcast. It's time. This is a short one because he did something dumb, but it's like predictably done.

Like he did the dumb thing that Republican commissioners always do at the FCC. So a couple of things happened. One, everyone left the FCC, presumably because Brandon is so unpleasant to work for. Does Anna Gomez still work there? I haven't checked on this in a minute. She still works there. She is not yet fired by Donald Trump. Okay. The only Democratic commissioner on the FCC. And she literally will tell you she does not know why she has not been fired. And she's on tour saying Brandon is a bad guy who's violating the First Amendment.

Wild stuff over there. So there were just the two of them, which is not a quorum for the FCC to do its business. Nathan Symington, who left the FCC, wanted this guy Gavin Wax to be appointed. He didn't get it.

This woman named Olivia Trustee is the new Republican commissioner of the SEC. She was confirmed by the Senate. She has a long history with everyone. There were statements about her like we would confirm her in a normal time because she's a Republican commissioner, Republican presidents get Republican commissioners. But Brandon is such a weird threat to free speech that we're opposing her nomination. That's where Brandon has taken this organization.

This woman who would have sailed through a normal times was opposed by various Democrats and various free speech groups. And literally they're saying she's fine, but Brendan sucks. So we can't let Brendan do anything. You can't let him have another vote. She got through party on vote. So it goes immediately, immediately. Brendan moves to open comment on letting companies own more broadcast stations.

So we have this concept in the United States called media ownership rules that say big companies can't own all the TV stations for a variety of reasons. First, you just want your local media to have a relationship to you, right? That the TV stations and newspapers and radio stations that serve your area should have some skin in the game where you live, right? They should be responsive to their local communities. And this is like deep in the law and the policy where you want the local media to be local.

You don't want national media to ignore your concerns and do all this other stuff. This has waxed and waned over the years, but you can kind of feel it, right? If you go on social media, there are not local media ownership rules. There are not rules that your algorithm has to show you what's happening in your community. Everything is just national culture war all the time because that's what plays in social media.

Weird. It's a weird outcome. So we have these rules from ages ago that say big companies can't just buy all the TV stations in the national program. They have to be locally owned. And then we have this other rule that says there's a percentage. The same company can't own all the stations in your town. The same company can't own all the stations and all the radio stations, all these favorite. So you have all these media ownership rules. Brendan thinks they're dumb. He thinks we should open it up and let everybody own everything. And his argument is...

It's not smart that these companies have to compete with the Internet. So we should let them own everything so they can get big enough to compete with Google and Facebook. And the problem there, of course, is when you open the social media apps on your phone, you are not subjected to local news. You're subject to the national culture war, which is the exact reason the policy exists. Brendan, everybody, he's just like just walked into the literal opposite of the reason for the rule.

We should make everything National Culture War so that local news can compete with Google is not the right answer at all. Make local news do National Culture War is what we're after now. That's our boy, Brendan. So that's Brendan. He's a dummy. As always, Brendan, if you can defend any of this stuff, you're welcome on the show. I'd love to have you. It's been a minute since I've issued this threatening invitation to you, and I'm issuing it again. You're welcome. On Decoder, on The Verge Cast. We've only got weeks to go.

David, I believe this is your last show before baby. This is my last Friday show. I, uh, I will be on Tuesday's show, which due to the magic of technology, we have already recorded. This cake is already baked. Exactly. Uh, but yeah, this is, this is the, the, the last one I'll be on for a couple of months. Yeah. And then I, I, I'm weeks away here as well. So our children are coming and Brendan, we will let our children interview you.

I think they can take you, buddy. That's a good idea. I mean, Max can take me. So yeah, I think she can. She can definitely take Brendan. But you're invited anytime you want. You got a couple weeks to go. I'll come back special just for you, Brendan. I do have, by the way. It's going to be so great when Jake interviews Brendan while we're out. I think Jake can take Brendan too. I'm not worried about it in the slightest. Just a little bonus round. Not the FCC. The Federal Trade Commission did some dumb shit this week.

They are allowing Omnicom, which is a big ad agency, to buy another big ad agency. But they're issuing a consent decree that says they're not allowed to keep ads off of X, basically. So they're literally stepping into the free market and saying you have to spend money on platforms where you disagree with political views. And then we'll let your merger go. What kind of free market is that? Where the government is like, we will let you merge.

But you are not allowed to choose how you spend your money. You are required to advertise on objectionable content is such a funny thing for the government to require. Yeah. The advertisers, by the way, the clients are allowed to still make their decisions. But the ad agencies, which fundamentally what the ad agencies do is help people spend their ad dollars. They are not allowed to make those decisions.

Perfect. I can't imagine how that will ever get. How will anyone get around this? This incredible rule. So they're like, here's here's the people. Here's the Nazis. And then just another another one just to throw it out there. Paramount is being sued by Trump for the CBS interview. You know about this. He says the Kamala Harris interview was deceptively edited. There's some mediator that's proposed some settlement, 20 billion dollars, and it's held up.

Because Paramount executives and their lawyers cannot figure out how to pay this money without being on the hook for bribery charges because it's just bribery. This is a real problem. If you don't believe me, you can read the notoriously leftist Wall Street Journal editorial board saying, don't do this, Paramount, just fight and win this lawsuit.

Yeah. When Rupert Murdoch is like, don't be an idiot, like go win your First Amendment argument. And he agrees. I'm just letting you know. All right. That's politics for today. Don't do bribery. Brendan's an idiot. Got it all on my system. David, do you have a palate cleanser? Jake, let's do let's do one each before we get out of here. All right. What do you have? I am super excited about this one.

Have you guys ever looked at your TV and thought, not enough pixels here? I can tell you for Neelak, the answer is constantly. Every day. The answer is always every single time you think, this isn't enough.

Guys, we're going to 16K. It's happening. The HDMI 2.2 spec is out. Now, there are some drawbacks, and I know this is going to be a hard one for the gamers out there. It's only 60 hertz. So I know nobody's going to want this just yet, right? How is this going to look on Bravia Core? That's all I want to know. Until we can get 16, 120...

You know, it's not for me. But the HDMI forum is on top of it. They're weirdly ahead of the curve here. I do not know what the current demand is for support for 16K 60Hz video. I'm not sure how much of that exists. I don't know where I would get it. The YouTubers aren't yelling at us to film in it yet. I love, by the way, that they've only revved it to 2.2.

Right. Because we're at HDMI 2.1 right now. Right. Yes. 16120 is only worth 0.1. They've only revved it to HDMI 2.2 to get to 16120 16K 120. But what is it going to take you guys? 32K 360.

HDMI 3. This is the goal. We got to go to the HDMI forum headquarters and just yell numbers at them until they concede that it's time to rev the main version. I love this. I'm all in. I'm really excited to buy a bunch of expensive HDMI cables and connect them to my $300 42-inch TCL Roku TV. This is what is most important. As soon as these cables are available, I don't care if you're still running 1080p.

Any HDMI cable you buy from here on out has better be a 16K cable because one day you're going to want to change out your Xbox and it's going to be too hard to get to. You're going to need to make sure you have that 16K support just in case. My $19 Roku Express is going to look sick now. It's going to rule. Yeah. Okay. So Andrew Osofsky wrote this story for us. Every paragraph of the story is great because Andrew is a great gadget writer and he wrote it very dryly. And if you just read it on its face, this is the work of people who...

Who have no connection to reality. So those cables that you're talking about are now officially called HDMI 2.2 Ultra 96 cables. Sick. Sure. There's a new standard they support called the Latency Indication Protocol or LIP. That's a sick band name, by the way. The Latency Indication Protocol is like a swing band that needs to exist. Why is it called LIP? Because it matches up the audio and the video. So the audio matches the lips. Oh, that's good. It's good. It's pretty good. It's good stuff.

And some might even say somewhat more important than 16K. Having your audio synced to your video. HDMI 2.2. Mostly here on the Verge cast, we love a good backer name. So I will take it. I'm all in. I'm just saying. It's a new top end. Ultra 96 is the new top end. You get high speed, premium high speed, and ultra high speed, and now Ultra 96. I'm happy. All right. What do you got, David? Microsoft is killing the blue screen of death.

finally after 40 years when your computer explodes it will no longer show the blue screen of death it will show the black screen of death

Microsoft is, is redesigning it. A thing that I learned this week, uh, thanks to a story Tom Warren wrote for the verge in 2014, uh, was that Steve Ballmer wrote the text of the blue screen of death message because he didn't like, he, he got the blue screen of death and he thought the, the wording didn't seem right for when it was going wrong. So he, they, the windows team challenged him to go do a better job. And he did, uh, he wrote the blue screen of death message, which is wonderful. Uh,

But now they're changing it to something much less fun and interesting. It just now says your device ran into a problem and needs to restart. And then at the bottom, it actually shows you the code of what went wrong, which is vastly more useful than what the blue screen of death did, but is much less fun. And the black screen of death just doesn't

Doesn't roll off the tongue in the same way. And I will be very sad to not have my computer just randomly run up this thing. You still got to use the acronym, right? It's still the BSOD. It's still the BSOD, which I assume was intentional. But like, why? Just a tragedy. I agree. There's no other way of looking at this. This is like a canonical thing in technology. Everybody knew what the blue screen of death was.

And now they're getting rid of the sad face. Yeah, dude. I thought they were just changing it from blue to black. No, this is what I'm saying. It's devastating. It's very bad. I mean, it used to feel bad for you. And now it's like, do either of you ever watch the office?

Yeah. Do you know the scene where it's Kelly's birthday and Dwight just puts up a sign that says it is your birthday? That's the black screen of death now. It just says your device ran into a problem and needs to restart. Your computer used to feel bad for you. It was like, I'm sorry, I blew it and I'm going to fix it now. And now your computer is just like, well...

I'm dead. And other great moments of sort of dead ahead tech writing that read is very funny. Tom wrote this story and he wrote it. It's very, again, it's very dead ahead. It's got some good quotes. The whole thing did great reporting. The last line dead ahead.

The changes to the blue screen of death are part of the broader effort by Microsoft to improve the resiliency of Windows in the wake of last year's CrowdStrike incident, which left millions of Windows machines booting to the blue screen of death. And it's like, oh, part of that effort was making the screens black. Do you think it's so that if you see a lot of them all at once, it's not as obvious that they've broken? That's what I'm saying. Yeah.

Right. You can like look over a room full of them and they don't all, it's not, it doesn't immediately scream. We're all destroyed. It's like how Tiffany has the trademark on its Pantone blue. You see the blue screen of Jathric, that Windows PC has crashed. And now it's just a black screen. You're like, maybe that's Linux. Like it's, it's the wrong thing to have a trademark on. Own your shame, Microsoft. That's the French cast, everybody own your shame. We'll see you next week. Rock and roll.

And that's it for The Verge Cast this week. And hey, we'd love to hear from you. Give us a call at 866-VERGE-11. The Verge Cast is a production of The Verge and the Vox Media Podcast Network. Our show is produced by Will Poore, Eric Gomez, and Brandon Kiefer. And that's it. We'll see you next week. ♪