And build your system. Sonos, love the sound, feel the system. Computers are getting smarter, Silicon Valley's getting weirder, and AI is taking over. At least that's what everyone keeps telling us. Welcome to Machine Gods, a new NPR podcast that will be your field guide to the AI-powered future. We're veteran tech journalists with front row seats to the industry. And twice a week, we'll make sure you get smarter about AI, so you know what's real, what's hype, and what might come next. Listen to Machine Gods from NPR on the NPR app, or wherever you get your podcasts.
Support for the show comes from ServiceNow. AI is moving fast across the enterprise, but without visibility, it's just chaos. Different tools, different models, different teams using AI in completely different ways. ServiceNow turns that chaos into control. With the AI control tower, you see all your AI across the business in one place. What it's doing, what it's done, and what it's about to do. So you stay in control. To put AI to work for people, visit servicenow.com.
Hello and welcome to The Verge Cast, the flagship podcast of suing your way to the top. I'm Jake Castronakis, executive editor of The Verge. David and Neelai will be back later this year. We've got a lot to talk about today and a really good crew. Senior reviewer, the song's with me. Hello. And senior reviewer, Alison Johnson. Hello.
Lots to talk about. Dish continues to be a complete disaster. The U.S. now owns part of Intel, which is a questionable investment. Elon actually sued Apple and OpenAI. I thought that was just a random tweet that was going to go nowhere. Senior news editor Richard Lawler is going to join us later to talk about all that. But first, let's talk phones. A lot of phone news this week. And I think most importantly, the two of you have been testing the Google Pixel 10. Lots of new stuff there. It's got magnets in it. Tons of AI. There's MagicQ, which runs on device, makes suggestions about what you're doing.
ProRes Zoom, which uses AI to reach 100x. Both of you have been testing these. Allison, your review just published yesterday. How was it? How was the Pixel 10 Pro? Basically, good. I would say you look at all that together, it's pretty good. Google's been on a good trajectory with the Pixel phones. I think they're feeling more and more flagship.
You know, they got more expensive, but it feels like the price is justified. So the 10 Pro, just to review, you know, we get Tensor G5. Qi 2 is built in with Magnus the EQV. She's doing a low QVC segment over there. Excuse me, you're taking Jimmy's money. I am. I'm showing Jimmy how it's done. See, I'm holding the phone right side up. It's not upside down. Yeah, this is for the YouTube viewers. Yeah, my TLDR is the AI is starting to actually make sense.
starting to be a little more helpful than just a random smattering of like, it's over in this app. It's in the screenshots. Gemini is going to bother you every time you open a Google Doc. And don't run out and buy it for any of that, but it's a very good phone. Yeah, you said some of that in your review. You're like, it's actually helpful this year. And then I was like, well, didn't they have some helpful stuff like in the past? And I was like, oh, right. Like it's it really was like a lot of like weird like gimmicks, grab baggy stuff. And it's like some of it makes sense. And a lot of it like there's the add me feature for photos where you can like hand your phone to somebody else and then it'll put you in. And I'm like, this is a really lovely demo. And I would love to know if they have data on how many people actually use that. It's like not a situation that happens that often. It's like it happens when you're on vacation. Are you going to really like take time to be like, hey.
good friend of mine i know we could just hold our arm out take a selfie but there is a specific feature on my phone that you don't know how to use it i need to learn how to use but in this moment we can and like There were demos. It felt like there were demos. And I think you're kind of right. It seems like this year, maybe Magic Q in particular, there's actually the start of something that you actually would use. Yeah. Kind of. I think Magic Q worked a little bit better for Allison than it did for me because we would be talking on Slack and they're like, can you send me this message? So it was like faking it a little bit because you just want to see if it's going to. Because I think you have to prompt it in such a way that it will be like, aha, information is being asked for. But that's not necessarily how people convey it.
To you? So like Allison in your review, and this was real, she slacked me. She's like, can you ask me when my Japanese breakfast concert is? And I did. And it worked for her. And then I was like, oh, can you ask me when I'm seeing Heather's? And she did. And it didn't work for me. So I was like, but it's in my calendar. The thing that I think feels limiting right now, and I'm curious if they'll ever be able to expand this, is I think Allison, you said it only works in Google's own apps. Like it only has access to Google data.
Yeah, and kind of small list of them, like Calendar, Gmail, Messages. Like Messages is kind of the most frequent place I saw it. It can suggest things in Gboard. So you can see, one I had just now was I opened Spotify and it came up with a suggestion from the Pixel Screenshots app of like some band names I might have been searching for. I was like, that's actually helpful because I'll screenshot something that's playing on KXP. I'm like, I should listen to that later and I never do anything with it. So that kind of thing. There is a way for them to kind of like...
creep into other apps, but it's mainly in the main Google apps right now. Which, like, in fairness, you know, they do contain a lot of your life, but it's still, like, pretty limiting if it's, like, okay, it's calendar events. Yeah. It's, like, I think the one that they mentioned was, like, oh, and you can get reminders for your Google Keep app. And I was, like...
Oh, I know. Yeah. Google keep for reminders. But okay, I'll try it. So on Tuesday, I was at Shakespeare in the Park, which if you're not a native New Yorker, you wait in a really freaking long line at Central Park. And this year, it's a stacked cast. So I had to get to Central Park at 445 in the morning. This is ludicrous. It's ludicrous. You messaged me at like some ungodly hour of the morning. And I'm like, sorry, why are you awake this early? Oh, yeah. And I was doing an edit because I was sitting in line.
at like 445 in the morning. I was like, I'll just do a little work now because I'm going to be here for the next seven hours trying to get a ticket to this, this free ticket to the show. But the whole point was I was there with a couple of friends and one of my friends turns to me and she's like, remind me. that I need a sweatshirt tonight because it's going to be the same temperature. And when we're in this outdoor theater, I need a sweatshirt. And I was like, how about I remind you by you telling me in a text message to remind you to bring the sweatshirt because I need to test this Magic Q thing. And she gave me this look like, are you fucking for real? And I said, yes, please, please. I just want to see if this works. And it did. And I got a little Google Keep reminder that said at 8 o'clock, remind me.
friend to bring a sweatshirt. And I was like, oh, and then at eight o'clock, it did not pop up. So close. It was so close. But or at least I didn't notice it. So, yeah, the way it feels to me is a little bit like, you know, when auto like one time passwords started auto filling from text messages. It was huge. Yeah, it was great. I was like, this is awesome. I don't have to do anything. And then you just kind of forget about it. And that's just how phones work now. It sort of feels like it's going to be that like I did have more success with it having stuff popping up.
But then I don't I don't think that's something I was like, well, thank God for this. You know, I was just kind of like, oh, this is super nice. And I will just get used to this and take it for granted. Probably. I mean, I whenever like a password doesn't autofill for me, I'm like, this is the end of the world. Yeah. Like I can't go on. You have to throw your phone into the sea. Yeah. Yeah. Something has gone horribly awry. I like that as sort of like a framework for where useful AI is like.
That's not doing something ostentatious. It's not changing the world. It is just being helpful in small, very precise ways that require a little bit more contextual awareness than I think perhaps traditional programming could have gotten you. Yeah, we've had ways to put things on calendars before that didn't involve generative AI. We've been trying to do that in emails and everything like that. This just feels that extra little bit smarter where it can connect dots a little bit better.
I did remember one time MagicQ came up with a conversation with my friend and I was like, oh, and she was mentioning things that we could do for day trips. And she mentioned some festival in Rhinebeck. And I was like, I don't know where Rhinebeck is. And then it popped up like a Maps MagicQ thingy. And it's like Rhinebeck. I was like, oh, I can see how far Rhinebeck is. And I was like, oh, I'm not driving two and a half hours. So now I know. But I was like, that was one time where the MagicQ was like, oh, OK. That's not so bad. You two tested a bunch of these sort of AI-ified features, right? What were the other interesting ones?
OK, so V helped me out on the day that she was technically not working, but it sounds like doing kind of a lot of work. Listen, it's a long seven hour wait. You got to figure out stuff to make the time pass. I was like, can I call you and you can speak in a different language to this phone? And she had a Pixel 10 as well. So we tried out the voice translate. And that one is like we've seen this from Samsung previously.
It just kind of chimes in and speaks the translation live on the phone while you're talking back and forth. So you can make a dinner reservation or whatever. This one mimics the speaker's voice both ways, which is a little wild. And we had a weird time with it. My friend was also there on the side. And my friend was just like, her eyes were just like, yo, what? Because at one point, Robot Allison, who was speaking in Japanese to me, just went like, she glitched.
She went, eh. It was wild. And we were screaming, like, just, like, going, what happened? And then, you know, we kept, like, talking through it. And so then we were confusing the robot at one point. And then on her end, Robot V went, eh. Yeah, it started, like, screaming back at me. This doesn't sound very successful.
It seems like worse than it was before. It was kind of half successful, I think. Like Allison coming through in Japanese was actually pretty good. It was one of the most... Until the screaming. Until the screaming. Until we broke it. But I actually think I was the one that like broke things because Allison was talking like a normal person in fluent English because that's her native language. And unless you're hiding something from me, Allison. But she's talking in her natural language. She's talking in full sentences that make sense. Me, in my Japanese, which is a little rusty, my sentences were a little bit broken. They weren't fully textbook appropriate, like, you know, like textbook grammar. Because when you talk in Japanese, it can be very...
You can just start leaving out pronouns and people will understand it because it's contextual. The only way I could explain it was that I was glitching in real time because I'm code switching between languages. So part of me would forget a word and I'd be like, ah, what would I call that? And it's translating all those little like, oh, how do I say that again? And me talking to myself a little bit while I'm talking to Allison. And that's what I think caused it to go. That's fascinating. Absolutely not. I broke the machine. And, you know, she's asking me about Shakespeare. I'm like, oh, I don't know the Japanese title for Twelfth Night. So I'm just going to katakanize it and go Twelfth Night. And it was just like, nope, not translating that. You ho. This is what I imagine Gemini is thinking in its head. But it was very much like, oh.
That kind of informs when you would use this, right? Because if you're making the reservation, like Allison's saying, you're talking in a measured tone. You're talking in full sentences, probably. And the person is also doing that. And you're talking in your own native languages. But if, you know, I think one of the demos, they were saying you could talk to family members who you maybe don't necessarily speak the same language as. And that's a situation in my actual family. And I was like, listen.
We speak Conglish in my family, in which people are non-sequiturs, just mashing grammar of two different languages together in a pidgin. And I think that would just trip up the robot. And so the solution would just be speaking in your native language. But after years and years of just talking to each other in this pidgin, I don't... Like, that would be really hard to just do because my aunt will be like, oh, did you get phone call? And it is just like this is going back and forth in between the English and plugging vocabulary in different ways. And I was like, oh, that would not work. That's fascinating, though. Like, it hasn't adapted to the way that a lot of people actually speak. Yeah. And that's my commentary on most translation tech. It's getting a lot better with AI. That's very exciting to me. But just the way that we actually speak is so varied and changes minute to minute.
And, I mean, you see the creative linguistics the kids are doing with skibbity toilet and all of that. So, you know, we should test that if it can translate skibbity toilet. Oh, I'm calling you right now. Yeah. It's going to scream again. It's going to go like a scream. What? I mean, it did sound like V at times. It was just so weird to hear, like, an imitation of your voice coming through the phone. There were times it sounded, like, kind of like you. And there were times it would just go in a different direction. I'm like, that is a different person. That's a different voice. But it tried. The little robot tried. So Allison speaking in Japanese is very deep-voiced.
She's just very deep voiced, which is like an interesting choice because like Japanese and Korean especially, those are languages where you tend to speak higher. So like when I speak in those languages, my voice pitch goes up because that's just like naturally a feature of the language. Everyone's pitch goes up. It still sounded like Allison to you? Sometimes. Like in the beginning, yes. And then it would just turn into a robot Japanese lady. The other question for me on these is like, what is the latency? Because it's like it has to, number one, do the translation. It then has to make it sound like the other person's voice. And if there's any latency, like this just breaks. Nobody's going to sit around for that. It sort of intentionally like lets you talk for a minute. And then there's a little like.
like it makes a little noise like something is happening and then the translator starts talking. So it's kind of nice because it gives you a minute to be like, okay, they're talking, you hear them talk for a second. It doesn't feel like a delay I'm annoyed about. It's just sort of, and from there it flows pretty well. I didn't think it was like super slow or anything. Just like that conversational rhythm that Allison was going into where she's pausing, stopping to think.
switching gears mid-sentence, that trips it up. It absolutely trips it up. Because, like, in the demos, when you get that, you're just speaking in full-form thoughts, and that's just not how conversation is. So, like, this is not necessarily a Google AI problem. This is an all-AI translators in my experience problem, where you just be like, oh, you know, I'm talking, and then, oh, wait a second, I have a different thought, and, uh, actually, yeah, you know, like, mm. So, anyway, and you do that. Very natural. You understand it in person. The robot just goes, ah! It's like, that's actually what the robot did. It did start screaming on both of our ends. I was like, oh, okay, yeah. You both have the phones in front of you, right?
I need to hear this. Let's give this a try. All right, I'm calling you V. Okay, yeah, yeah. The rest of the Vergecast will be conducted this way. Oh, God, this is going to be... Do you need to, like, activate a certain mode or just, like, starts? I know you do need to activate, like, a certain thing. Okay, here we go. Answer. Call assist. Voice translate. V.
What did you have for breakfast today? I'll definitely know, maybe. V, whose fault do we think this translation is?
normal. You understand it. The AI translated it in a funny little way. It's not wrong. The translation is not wrong. But because Allison is speaking in a native way, the Japanese that's coming across is more correct in a certain thing. And so that was something I noticed and was like, oh, it reminded me of this time when I was in college and I went to college in Tokyo. And one of the classes that they made us take, if you didn't test out of it, was conversational Japanese. And so the assignment was that you had to actually call Pizza Hut, the Japanese Pizza Hut version, and order a pizza.
And because they were like, you're going to live here for so many months. So you need to know how to be a functional human. And these are not necessarily things that you learn in a high school Japanese class. So we were doing this and my friend had a meltdown because she didn't realize that how to read a Japanese address and how to give a Japanese address. So the person on the phone was just like she was reading out numbers as you would in a like, you know, like.
something, something, something in Tokyo. And it's like, you have to actually say no in between instead of dash. That's how you would say it. And so she lost her mind. They lost their mind. And I, you know, I was thinking back to that. I was like, this would super have like come in crutch at that time. She could have just talked to them, ordered a pizza, given the address. And that would have been written. That would have been like translated in a way that's super helpful.
Not helpful in the sense that she wouldn't have learned how to actually say it properly in Japanese. So in that sense, I was like, oh, it's actually also not great if you're trying to get better at the language, but might come in handy in a pinch when you're just like, I don't have the brain cells to attempt this today. I just want a pizza. To be fair, I don't like ordering a pizza on the phone in English. It makes me nervous. They also have AI for that, Allison. This is all pointless. They're just going to AI everything eventually. Yeah, it's true.
Okay, there is another big AI feature that I really want to make sure we talk about, which is the ProRes zoom, where on the camera, there's a 5x telephoto optically, and you can zoom in and after 30x, it starts using AI built into the camera app to enhance the photos. And last week, I was like, I was like, this looks incredible. This looks, this is like blowing my mind, because I just seen a few sample images out of. They're really hands-ons. And some more stuff has come out since then. You've had a chance to test it. And I would say some flaws are showing in places. I'm curious, Allison, you actually toted around another giant camera to test this against an actual 100x zoom. How is it? How do you feel now that you've tried it about having AI in your camera changing your photos? I feel weird about it.
Big picture take is like, it is pretty good if you stay in the 30 to maybe like 50 X range. It can, you know, I'll take a picture of a container ship from a bridge and it knows what to do with the, you know, you get kind of those jagged lines and digital zoom and things just look weird and crunchy. It will clean that up and it looks totally passable. I'm like, all right, this, it doesn't look amazing. It's not like a National Geographic photo of a container ship. Like it looks like a decent phone picture of a container ship. But I did carry around.
Jake's urging. So this was his fault. I'm sorry. I didn't know how much of a burden this would be. I was like, we should get a real 100x zoom. And it turns out that's heavy and large. I should have known too. But I did the math and I was like, there's exactly one camera that can do this. It is a Nikon Coolpix P1100. It is a gigantic beast. This thing is hilarious. The photos of it that they advertise, it's like the size of somebody's face. Yeah, it is. And there's just no like...
camera bag i have to put it in that makes sense i just had to like lug it around uh the first time i went out i was like i'm gonna take side-by-side photos i've got this phone this camera um was trying i was trying to hand hold the camera at 2400 millimeters that doesn't work you don't like i failed db review college or whatever um so I went out again with the tripod. My tripod is flimsy and crappy. And the camera, once you get zoomed out far enough, it starts sagging a little bit, even though I've got it locked in. So I did not end up with amazing side-by-side photos. I have a side-by-side photo in the review, which is basically.
A boulder with some trash next to it. It wasn't the most scenic photo. Like everything else in your review, lovely. It's like a boulder and like a stick. Yeah, yeah. There's like some garbage just sort of laying around. And I was like, well, this is what we've got. Because it was the only thing I could get framed up reliably. And very humbling experience. The Nikon, and I put the Pixel on the tripod too, to give it a fair shot. Yeah, as you would expect, the optical zoom does better. I was sort of surprised. I thought the...
The Pixel did okay, like, considering everything it was doing? Yeah, the 100X, like, that one clearly looks optical. And it's not a great photo, but it's significantly clearer. The Pixel photo is not bad, though, particularly when you compare it to the, like, pre-AI version. And this is, the thing that's been fascinating is, like, I think there seems to be certain subjects where this AI enhancement is great. And I think if you didn't tell me it was AI, I would be like, cool, they're just doing some fancy upscale and looks great. And it's, like, you know, particularly for photographs of buildings, architecture in general.
It seems to look great. It's just like, you can't tell that this has been upscaled in any way. I've done like, I look really closely side by side and I'm not seeing any weird AI wackiness. The thing that trips it up, as soon as there is a single letter of text, it goes, uh-oh, I'm AI. I have no freaking idea what I'm doing. And so there'll be this image that like, the thing that's wonderful is like the pixel gives you the before image and the after image so you can see what's happening. And so you'll have this image that by all accounts, everything else.
It's just like perfectly cleaned up, crisper, sharper, looks clearer. And then there'll just be these letters. They're just like, I don't know what language is. It's just like, it's just total garbage. And I'm like, what happened here? Like, it's weird because it's like, when are you going to do this? Right? Like, when are you going to zoom in a hundred times? For what purpose are you zooming in a hundred times on a thing? If it's, I mean, I know I did it on the Shakespeare line to be for shits and giggles because I was just bored. And so I went a hundred times zoom on my friend's eye and that was a nightmare. I hope that didn't enhance. That's not. Oh, that is a nightmare. That is her eye before the AI got involved.
That's the AI after. Oh. And I think that's just actually a really fun art project because what is this? This is a nightmare. This is like a black pit of despair with some like streaks of color around the outside. That looks like an eyelash maybe. That's something. And, you know, sure. Like, what are you trying to do? Are you trying to find some interloper's license plate number? Is that why you're doing 102? I mean, the idea is landscape kind of stuff, landmarks. You're on vacation. You're taking a picture of the Statue of Liberty. It's kind of far away. So I get that. There are all kinds of implications of...
what is a photo? Did you really take a photo of the Statue of Liberty if AI just kind of filled in the gaps for you? The writing thing is funny because it does get really nervous and doesn't know what to do. I found that really big lettering, it could do okay, but as soon as it was a little smaller in the image, it just kind of like, I don't know what that is. That's what kind of breaks my mind about it. Because again, I think if I took a Statue of Liberty photo with this, it would appear to be my photo by all accounts. But as soon as there is text in it,
I'm like, no, there's literally an AI rewriting this image. And that's very trippy. And I think last week I was like, man, maybe this is some new era of photography. Everything's getting upscaled. But it seems like they've actually been pretty precise here and pretty limiting in what it's applied to. It feels like it probably shouldn't be applied to text at this point in time, even though it would be cool to be able to zoom in and read some signs from far away. But it doesn't know what language is, so it's not going to happen. Yeah. And I didn't see anything.
too wacky or unexpected outside of the letters and signs. It gets a little confused by, I think, specular highlights. If there is, you know, like bright sunlight on the surface of a pond and there's a bunch of garbage and like duck feathers floating around in the pond. It didn't know what to do with those because they just look like points of white stuff. And you magnify that a hundred times. It just looks weirder and weirder. So it could kind of, it sort of took those things and went in a different direction. You're like, no, this isn't it. But for the most part, it was...
It was pretty impressive, closer to 30 times, and as long as you're not asking it to read English. This is going to give us, I think, material to debate the what is a photo question for years to come. I suspect this is also not going to be the end of AI and camera apps. Very interesting year for the Pixel. Not a lot of huge hardware changes, but the AI stuff is, I think, particularly interesting this year. Other big thing in phones happening this week, Apple announced the iPhone 17 event. It's happening September 9th. I think we're expecting that to be a pretty interesting one. But now that we've seen what Samsung has to offer this year, now the Pixels are out, what do we need to see from Apple to...
kind of keep up the momentum. I think it's going to be another situation where they kind of tap dance around the AI, you know, thing that's not happening or happening. It seems like we're going to get some different hardware. Camera bump goes sideways or something like that. But I think that the elephant in the room is going to be AI again. Yeah, I don't know. In a weird way, it feels like these phones are merging into each other. Because if the camera bump goes sideways, like some of the leaked renders I've seen, I'm like, it just kind of looks like this.
It's not a bad look. I've looked at these mock-ups and I'm like, it's a little funky, but... It's not a bad look. But then I'm just like, they're merging into each other in a way that's like, you know, Pixel comes out with a journal app. This comes out with a sideways sidebar. MagSafe is on both of them, which, you know, honestly, that was great. Having MagSafe on this was like really great. But I don't know. It's weird. It's bizarre. And Google coming out with such a strong opinion on AI and where it belongs in our devices and our lives. I'm just sort of like, ooh, well, they threw down the gauntlet. What's Apple going to do? A big design refresh is always a good way to...
I would say, distract from any other things. Like, you can do nothing else and you're like, completely brand new design. Like, I'll take it. I feel like there's such, right. Oh yeah, this is the thing. You've got to find something to pair with the lovely liquid glass redesign. You know, I think different gears are for different things, but also like, it's been a minute since the iPhone has changed in a pretty significant way. The current design is getting a little stale. So like, I don't hate it. The AI thing is really interesting because it's true. Apple sat down like a year and a half ago. I feel like we're going to be a major player in this year and a half later and we kind of know that they're skirting around it, skirting around it still.
I don't know. They could have some surprises in store. I don't know that it's going to happen. I think there's been ongoing rumors that they're going to get a Gemini partnership to go alongside the ChatGPT thing. It's not clear if that's actually going to happen in time. We'll see. We'll see what the one more thing is, if there is one. It's not going to be Google. I mean, honestly, I was just like, orange? Orange phone? Maybe? Orange? Because orange is not my color. But I've yelled for years that I would like an iPhone Pro.
model to actually have a color and not like a hint of a color. So I saw the renders and I was like, listen, that color's not for me. I'm not an orange person, but I appreciate it. This is what Apple needs to do this year. Make colors happen. Make colors happen. Drag that saturation slider way up, you guys. You know what? That is enough. I will be like, yay, colors. Drives me crazy. Okay, we got to take a break. Allison, thanks for joining us. I know you've got to run, do an interview for next week's Vergecast, which I'm very excited to hear. That's right. Tune in on Tuesday to hear more about that.
Awesome. All right, we've got to take a break. When we're back, we're going to have Richard Lawler. We'll be talking about DISH, Elon Musk, and Intel. Support for the show comes from Upwork. Business owners know that you kind of need to become a jack of all trades to grow. But as you do, situations arise where you end up wishing you could simply hire a jack or a jill of a very particular trade. In those cases, you'll be glad to have Upwork. Upwork gives your business fast access to highly skilled specialists with deep expertise. With their help, you can fill critical skill gaps, launch projects faster, and scale support up or down without committing to a full-time headcount. There's also Business Plus, which lets you access the top 1% of talent on Upwork. And when it comes to the operational stuff, such as contracts and payments, Upwork takes care of it all.
With over 170,000 freelancers holding five-star ratings, you can hire with confidence knowing you're getting someone with a verified track record of applied expertise. Visit Upwork.com right now and post your job for free. That's Upwork.com to connect with top talent ready to help your business grow. That's U-P-W-O-R-K.com. Upwork.com.
Support for the show comes from OutShift, Cisco's incubation engine for frontier technology. You can find multi-agent AI in every enterprise roadmap right now. But here's the problem everyone's facing. Agents can pass messages but can't think together. These systems silently underperform through cognitive failures, misreadings, unfairified claims, false consensus, and conventional monitoring can't detect them. OutShift by Cisco is building the fix with their new comprehensive resource, the Internet of Cognition. It's an open source guide that's a foundation for building multi-agent systems from design to production. The goal is to give agents shared context, shared memory, and guardrails.
all to drive the results we actually expect from AI. You can lean on this resource to understand agentic AI's evolution and how it's shifted from isolated agents to orchestrated superintelligence. You can gain insights for how to future-proof your business by tackling complex challenges with multi-agent systems. You can read the paper, experience a demo, grab the code from outshift.com. That's O-U-T-S-H-I-F-T dot com.
We'll see you next time. All right, we're back. Richard Lawler is here. Thanks for joining us. It's been a big week for disasters of companies and the U.S. government being involved with them.
I'm not happy about any of this. The first thing we have to talk about, and I think we're all in a very sad, dark place that Eli isn't here to commiserate with us about this one. I think, as you all know, Neelai, for a period of time this year, has run a very famous, well-respected, well-received segment internationally known as Brendan Carr is a dummy. This segment highlighted the ups and downs and mostly the downs of FCC Chairman Brendan Carr. Now, this segment will not be returning today because it is copyright, trademark, all rights reserved, Neelai Patel Enterprises, and I just couldn't get a license to it. But it is here in spirit because I cannot emphasize the disaster, the absolute disaster of what has happened with DISH.
Like, DISH was already a disaster. And I do not know how it could get worse. And now it did. It got worse. It's just a complete fail. Like, there's been nothing redeemable about DISH's 5G network. Is that correct? I'm sorry. I'm trying to think of something. I'm just coming up with nothing. I got nothing. I think Mitchell, this piece that Mitchell Clark wrote nearly three years ago now, I went to the Genesis point of DISH's 5G network, and all I got was disappointment.
just captures the entire experience, I think. And not only that, that was the high point of DISH's network. So, okay, what happened this week is that they no longer have a network. They just sold it. So here's what happened. In April, SpaceX, J.L.A. Muscones, complains to the FCC that DISH isn't making use of some of their spectrum. Now, in fairness to SpaceX,
They're right. DISH isn't doing anything. They should have. They should be doing something. They're on the hook to launch a nationwide wireless carrier. They didn't do it. So a few weeks after that, FCC Chairman Brendan Carr opened the investigation into DISH's 5G network.
This causes huge problems for Dish because Dish is a disaster. Dish, by the way, I don't even know how to get into this. Dish sold itself back to Echo Star, which was originally what Dish was called and then spun out Dish. It doesn't matter. The point is they're not in a great financial place. So the Wall Street Journal reports in June that Echo Star is considering filing for bankruptcy because Brendan Carr's crusade against them is making it impossible for them to do anything. They said that Carr's threats have, quote, effectively frozen our ability to make decisions, which in fairness, it was frozen anyway. They weren't doing anything, but it would be nice if they had done it.
Let's go back to 2019. Donald Trump was president. This is important because Donald Trump was president again. And in 2019, T-Mobile wanted to buy Sprint. And at the time, the Department of Justice, again, under Trump, was like, hey, it's a bad idea to reduce the number of major wireless carriers in this country. We had four, that would make three. And so they cut this deal. They cut a deal where T-Mobile says, fine, we will sell off Boost Mobile, Virgin Mobile, Sprint prepaid, and some Spectrum to DISH in order to buy Sprint.
And DISH will agree to actually launch a mobile network. They'll agree to become a 5G carrier. And DISH even agrees to a timeline. They work with the FCC. And by June 2022, they had to cover 20% of the U.S. population with this 5G network. By June 2023, they had to cover 70% of the population. They do it. They meet these requirements in some, you know, alleged capacity. And then they never actually launch a service. This is, Neil, I've been on about Project Gen5Sys, aka Project Genesis. I don't know what this thing is. This website hasn't been touched in years. They haven't made a social media post since 2022.
It's now the second Trump term, okay? The first Trump term, they were like, Dish, you have to be a wireless carrier. It's important that there are four wireless carriers. Okay, that's a big gamble because who's going to trust Dish to do this? But they have a point. That would be good. It would be good if Dish launched a wireless network. Now it is 2025. Trump is president again. I know I keep reiterating that, but it's important because this is the same administration.
And now, Brendan Carr, the FCC chairman, is out here saying stuff like, there isn't a, quote, magic number of carriers that the U.S. needs. He doesn't care. He's like, four carriers? Whatever. This all leads to this week, when Dish just comes out and they're like, okay, we just sold all our spectrum to AT&T. Like, basically all of it. They sold all the good 5G spectrum to AT&T. And Dish is like, yeah, we're kind of a carrier, but not really. It's three carriers again. We're just back to three carriers. Not only are we back to three carriers, but AT&T got this huge chunk of really good spectrum that's going to make them an even better carrier, which is going to make it even harder for other people to compete with them.
Dish, everybody. Brendan Carr, everyone. The Trump and Mr. This is how that has gone. Did I miss anything? My brain hurts. I think you covered everything. In the middle of that, I went back and pulled up our visual history of the whole timeline that got to DirecTV and Dish coming together. There's even more mess when you consider the media side. Another nail I favorite. But yeah, that is about it, I think. Has Dish done literally anything other than like...
split apart and reassemble its assets over and over again for like 20 years? Nope. That's what that timeline shows, right? Like, that's what it is. I don't, I don't, I don't understand what Dish is now. I don't understand what it ever was. They did some innovative DVR stuff for a while there. Yeah. You know, I'm realizing that I've heard of Dish, but I don't think I've ever known truly what Dish is. If you're still rocking a Dish DVR, vergecastwithverge.com. We want to hear from you. How's that going for you? If you ever use Project Gen 5, sis, please.
I do not know what's happening. Brendan Carr's argument here is he's like, oh, the MVNOs are really big now. And so we don't need more major carriers that have their own networks because there's competition. This does not seem like a good argument to me. Right. The MVNOs, they operate on the carriers. The carriers can do whatever they want with them. Competition that is the same, that is someone in the same outfit wearing a different shirt is not really competition. That is like precisely it. This is such a disaster and no one cares. And I think for the FCC in particular, it's just a complete abdication of their role. And I think particularly like this is Carr's doing. He was just like, hey, we gave him a deadline. They didn't really exactly follow through on it. So just just quit.
Instead of being like, hey, you're on the hook for being a major wireless carrier, a thing that we regulate. Why don't you go do that? He was like, hey, get out of here. He just wanted to wash himself of the dishes, wash the dishes. Listen, I've been trying to think of that one for a solid two minutes. Listen, nobody wants to spend time with Dish. That's the lesson we've all learned. They're Echo Star now. They are Echo Star now. A fact that confuses me every time I see it. And you know what? A fact that won't matter because it will never show up on our cell phones.
But I mean, as you mentioned, that really is the problem. Because if you live in an area that isn't well covered by the cell towers, if the prices are too high for the kind of services that you want, if you have fewer real options, there's no one who has a reason to try and get your business. And oh, and MVNO, we've seen the MVNOs. We have any number of them. Maybe some of you have signed up for Trump Mobile or tried to. Let us know if you get those phones. But I don't look at any of those. I've tried some of them. I don't look at any of them as competitors for the major carriers. They just aren't. They aren't on the same level. The thing that gets me is like what, say one of them became like, you know, a major player. What leverage do they have, right? Like what's going to stop the major carriers from being like, hey.
Pay us more money. Okay, now they have to raise their prices. Like now this all goes away. They're only working because it wasn't necessarily a huge thing. The carriers are giving them discounts. The carriers can make their service worse anytime they want. The carriers can charge them more anytime they want. And when you're on one of these more budget services, oftentimes you're dealing with things like the kind of prioritization or these caps or, you know, kind of buying details. And I think there should be more options and plans, but there needs to be more competition. And you just don't have that if you don't have carriers with their own towers. I think you can look back like the decade leading up to T-Mobile buying Sprint. I think because of T-Mobile in particular, there was actually a lot of competition in the mobile carrier space.
They were making noise. They had to get scrappy. They had to fight pretty hard to be on par with AT&T and Verizon. I mean, that's why I got on T-Mobile because, you know, like they had that guy, I forget his name. John Ledger? Yes, him. He was in the pink shirt and he was like acting up. But at the same time, he was acting up all the time. But at the same time, I was like, oh, you know, that's more affordable. I'm poor. I can do T-Mobile and the service is actually pretty good. But I went down to my spouse's ancestral homeland of Tyaskin, Maryland recently. You know, and everyone there has Verizon because it's the only network that works.
And we're going to put works in scare quotes because on his Pop-Pops farm, there's only one tree where you can get one bar of service. So to make a phone call before Wi-Fi, you always had to go to the one tree on the nut farm to make a phone call. So to your point, maybe Dish could have helped him out there if he had lived up to the promises because that tree is not close to the farmhouse. It's quite far to have to get the one bar of Verizon. But maybe now the government has more reason to give SpaceX a lot of money for Starlink. This is the weirdest part. The thing that instigated all of this, SpaceX complaining.
Echo star slash dish slash whatever it's called now still owns the spectrum that SpaceX wants. They're not using it. They're not doing anything with it, as far as I know, or as far as SpaceX claims. So in the end, everybody loses. Yay. That's just the story of America in 2025. Okay, it's now time to talk about a...
slightly more successful company at times. And that is Intel. Which... If you average it out over the years... You know, I think slightly is a little harsh. You know, people know what Intel is and what Intel does. Those stickers went a really long way. They really did. You know, bump, bump, bump, bump. That's true. What is the current name of Intel's processor series? Core what? Pentium. Ten Lake Moon.
Something. I don't do laptops no more. Yeah, yeah, yeah. No, if you average it out, it's right. Intel has been a bit of a rough streak, which has led up to, this happened on last Friday, so it was after we recorded the last Vergecast.
Intel gave a 10% stake of the company to the U.S. government in exchange for $8.9 billion. This is weird for a lot of reasons, including that the U.S. government does not typically take ownership stakes in private companies. But I want to talk about why this is a just bizarre and questionable deal for Intel for a minute. The first being that they were already getting this money. This money was already awarded to them, primarily from the CHIPS Act. What happened is that Trump was like, what if we just don't give it to them? Now, I don't know if that's legal.
Might not be. The CHIPS Act, I will remind everyone, passed the Senate and the House on a bipartisan basis. Everybody was like, hey, it would be a good idea to prop up the U.S. semiconductor industry. That's an important industry. We realized that during COVID. It has a lot of national security implications. Let's make sure that's doing well. Intel, they're not doing great. Let's give them some cash. Trump didn't like it because it was associated with Biden. And so he's like, I'm just going to, I'm not going to give him the cash. And so he, so this is the timeline of what happens. Early August, Trump demands that Intel's new CEO, Liputan, resign. He's just like, I don't like this guy.
He has some time to China to get him out of here. Tan goes and visits the White House, as CEOs do these days, and suddenly they're good pals. And I assume that has to do something with the fact that, you know, a week or so later, he gives 10% of the company to the United States in exchange for his money that the United States was already paying them. What a deal. What a deal. Art of the deal. This is, I mean, this is bizarre on all counts. Number one, Tan, in some ways, gave away 10% of intel to protect his job, which I think is a questionable decision.
Number two, what is really happening here is that Intel's is foundry business. Foundry business is not doing great, right? It's not making the best chips. Companies don't want to make a deal with Intel's foundry business because they don't even know if Intel's foundry business is going to be there long enough to keep making the chips. And so the U.S. government is now getting involved. And I think the reason that Intel wants to do this is that they're like, hey, Trump, maybe I'll just pressure Apple to make some chips with us. Everyone's leaving Intel. And they're looking to Trump to put some pressure on them to get customers for Intel.
There's some real galaxy brain math happening here. Like, I feel like my brain is leaving my skull, yeeting itself to Jupiter and doing some mental gymnastics to try and see the logic behind this. And it's finding none. It's finding none. The logic, as far as I can tell, is like crony capitalism, right? The logic here is the U.S. government will get involved on Intel's behalf and Trump is going to put pressure on people and the folks he likes are going to win, the folks he doesn't like are going to lose. And right now, at this moment, he likes Intel and he's going to try to get some people to be Intel's customers.
And hopefully Intel can make some good chips for them. But we don't know if they can. Which has been basically Intel's whole problem for years now, that they've been supposed to be making better chips and sort of haven't. And I guess we'll see what happens with these next generation of chips that they're building. But they missed out on really the AI boom. TSMC has become massive on their watch. They missed out on this technology that they developed. Just purely from a tech perspective, I don't see how this solves the problem of Intel messing up, which has been the story of their last decade or so. From a business perspective, I don't know what's going on. And from a political perspective, it's scary.
Is there another part that I forgot? They also missed out on the mobile boom, which is like if you talk to Qualcomm for any second, because Qualcomm has always had a chip on their shoulder about how everyone knows about Intel and not them, even though they're in every Android phone and they're all over the place, because you literally talk to Qualcomm for five seconds and they're like, we're in mobile, we're in everything, we're Qualcomm, Snapdragon, yay. And nobody knows, like outside of tech nerddom, no one knows what Qualcomm is, but everyone knows Intel. But yeah, they missed the mobile train, they missed the AI boom, and now they're like... I don't know. Intel's just sad, man. And to Richard's point, this doesn't solve Intel's recent inability to make the correct business decisions. The tech isn't there. The business decisions have been off. Forcing people to be Intel's customers, it keeps them afloat, which I would like Intel to survive. I would like its foundry business to succeed. I think that is a good thing for the industry. But they need to make good products, right? Fundamental rule of business is make a good product. So, yeah, they kind of need to do that.
I don't know, man. Their marketing team is still great. That might be the only team at Intel that's still good. But they've been losing ground to AMD and now Windows on ARM is really good suddenly, apparently. Qualcomm and everyone is coming for them there. NVIDIA is reportedly making a chip. There's a lot of crossover and a lot of...
A lot of overlap, I think, in these companies and what they're doing. I still don't see the position for Intel to get themselves into to make things better. There's also a bizarre national security component to this. Part of this is great. The United States wants to make sure that there are foundries in the U.S. that can make good ships, that the supply chain is here. This makes sense. On the other hand, TSMC was a really valuable national security contribution to Taiwan, right? And if the U.S. is trying to remove reliance on TSMC, then that's bad news for TSMC. It's bad news for Taiwan.
Which is bad news for the tech industry in general, because again, most of the chips are coming from TSMC. TSMC has the best stuff. And so this just gets messy on basically every level. Trump sees what he's doing as just getting 10% of this company. But if you continue to follow out the next steps of this, what has to happen in order for this to make any sense, it is not good for anybody in the industry except perhaps Intel. And maybe not even Intel. I don't specifically know what the common argument is, but I guess if you take a look, how different is this from something like the auto bailout? Like in the treasury, like kind of taking over GM in these bankruptcies. I feel like that was an emergency from outside. This is an emergency that Trump created. And so it's not the same. No, I think that makes sense. And then I don't...
The U.S. didn't take a stake in those companies, right? They were giving them money, right? For a time, I believe the Treasury owned GM. Oh, interesting. Okay, yeah. But part of that was just that the U.S. bailed out the auto industry just because of the jobs and the industry, how much it would devastate the local economy. This kind of just devastates Intel. And yes, there are jobs that are associated with Intel and whatnot. It's just, I feel like there were bigger implications with, I don't know, the entire U.S. auto industry going.
versus one semiconductor company that happens to be American. If their problem had started with the president of the United States claiming that their CEO might be an agent for a foreign government, we'd be discussing it very differently. That too. Yeah, I think the auto bailout also goes to show the government has a lot of levers it can pull from the outside without going so far as taking an ownership stake in a company indefinitely. And I think that is where it gets concerning, right? Intel has the backing of the U.S. government. That just necessarily gives Intel a very fascinating position and advantage over...
everyone else in the industry, and it puts pressures on everybody else in the industry, whether they like it or not. Okay, one more thing we have to talk about today. A few weeks ago, we maybe made fun of Elon Musk a little bit. Did we call it a shenanigan? We did. I think we said this was a shenanigan. Elon threatened to sue Apple over not featuring Grok in the App Store. And honestly, I think we all thought it was a joke, right? Because Elon threatens lawsuits a lot, does not always file them. He did, in fact, file a lawsuit.
I don't know whether to say yay. You know, like, oh, you finally followed through on a thing. Yay for you. Or to be like, oh my God, you followed through on this thing. Not yay for me. It's not a great lawsuit, but he did do it. So XAI suing OpenAI and Apple.
And they're alleging anti-competitive conduct over a bunch of different issues. They have beef with OpenAI getting an exclusive deal to put ChatGPT into Siri. They claim that Apple is, quote, manipulating App Store rankings. And they say that Apple is delaying App Store reviews for Grok. Now, I think we can just say off the bat, famously, all the App Store reviews take forever. So that's not unique. But the premise here is basically, this deal is unfair because it's not open to everybody. And we're getting screwed in the App Store. And they undercut.
that argument right away. Because in this lawsuit, they highlight the fact that X is ranked number one as the free app for news and Grok is ranked number two as the free apps for productivity. So in this lawsuit, they're like, hey, we have these great apps. They're at the top of the app store, but they're also getting screwed by the app store algorithm. And it's like, what is this case? What case are you making here? They're just mad that it's not in the must have apps section. Don't argue against yourself. You're doing great. You just said you're doing great. Your argument is we're doing great and so we should also get featured in this other area. Having tested Annie.
Grok's not a must-have app, I must say. And Elon has been obsessed with app store rankings. As someone who has notifications on for his tweets, he was always tweeting about X or XAI or Grok or whatever being number one in the app store in some country that you forgot existed. He gets his feet of data and must share the news with everyone all the time. Richard, that is extremely unhealthy. I have to tell you. It's the healthiest. I know that I have not gone into active psychosis because I still can't understand what Elon is saying. And I get a reminder of this every 10 minutes or so. It's great. It really is. It's really that often. It's horrible. You are right, though. Just the other day, he retweeted somebody who's like, which again, cuts against the argument. He's like, we're featured in the must-haves, one of those editorial sections. They said they were featured in the United Arab Emirates in one of those editorial sections. And it's like, okay, so they are getting featured. I do not know.
that they have much of a case here. This is just like not a good thing to argue when at the same time you're saying that actually we're doing quite well. The other part of this case I think is actually very interesting. And they're saying, hey, we actually can't compete with ChatGPT because they're getting so much data and so many queries via Siri, which is this really interesting thing where they're kind of acknowledging, oh, we have to farm how everyone uses this thing or we can't improve it. Which is one of those things I think we all know, but it's like another thing to see it written down. I mean, if you want Grok baked into Siri.
maybe don't make sex bots. Like, I don't know how else to say that part, but I don't want to talk to Siri and have bad Rudy, the little angry panda, say that I'm lame because he's... if you insult that panda, all he says is like, I'm not lame, you're lame. And that's just, I don't know what he expects when he acts out the way that he does. And Apple, Apple has not been like hiding how they are this entire time and what they want out of the apps in their app store. They haven't hidden that. Well, and it's like, right, you have the chatbot that number one, recently went on a tear and called itself mega Hitler.
for no clear reason that you have now baked several like pre-made bots into it they've updated annie the the you know physical digital person incarnation of grok um with some new outfits so she can uh like strip down into like a bathing suit or something now which is just very embarrassing um but again it's like apple is like pretty you know for better or worse they're fairly like prudish with the app store right they do that thing they always say that they're family friendly that annie is not family friendly like it's just not right it's i think there's like again there's another world where i think if this was a random small company Like they're getting kicked out of the app store. Like I think Apple has plenty of reasons to not feature this app. And Elon is still asking for more. And I get it. Like I understand why you'd be pissed about not getting a special deal. But actually one thing that's really interesting, Tom Warren on our team, he writes the Notepad newsletter, reports on Microsoft. He reported, I think it was this month, that Microsoft has been testing, you know, the latest version of Grok and they don't want to deploy it to Azure because they're worried about the safety of it, right?
This is not just an Apple thing. This is Elon's product. It is not ready for this. It is not a thing other companies want to be associated with. You know, I don't think any of us are surprised when Elon has a tantrum. But this is the equivalent of him having a tantrum. I'm not the best in every single thing. So I'm going to sue you for not letting me be the best in every single thing. And it's not my fault. I have not done anything in this product that makes it not the best. You clearly, you know, it's just like very juvenile behavior and external blaming of...
I think if he were to listen, he would hear some very fair feedback and criticism for all of his products. And despite that, it's still, like you said, it's still ranked very highly in various sections, just not the section he wants to be the highest in. Richard, have you spent any quality time chatting with Grok? I have not. That's where you draw the line. You will get a constant feed of Elon, but you won't talk. We've got Grok at home, okay? He's already getting the Elon feed. He don't need the Grok feed on top of that. That's real. But I think the thing about this lawsuit that struck me is that we're about as receptive of an audience as there is. I think two criticisms of the App Store and the way Apple maintains it. We've been talking about this for years. Anti-competitive, monopolistic practices, the decisions that they make, the way that they do that. Other companies have made these arguments before.
But as Vy put it, there are reasons why Grok might not be featured. There are really good reasons. And I think they can make those arguments. But it's pretty consistent with Elon's overall approach that if he's not very visibly winning and celebrated for winning, then it's someone else's fault that he's losing. And he's going to make it everyone's problem.
That's really true. And I do think there is some degree to which like, OK, Apple didn't pander to him. And so, OK, he's just going to try to fight them a little bit. Naturally, the lawsuit is filed in Texas where he can potentially find a friendlier judge. So we'll see what happens. Maybe he can cause some problems for a little bit. I'm not convinced this is a good case. Oh, one last very quick thing. Elon is also suing OpenAI over the fact that they're trying to become a public company, whereas they were. In the meantime, XAI drops its status as a public benefit corporation. This is...
I don't know. I don't know what's happening over there. Other than that, Elon is obviously only looking at it for himself and using lawsuits as a way to create problems for other companies. And with that, we've got to take a break. When we get back, the Thunder Round. Stay tuned.
Support for the show comes from IDERA by Palo Alto Networks. Every new cloud workload or AI agent your team sends out is another identity that gets permanent access to your systems. And in 2026, those non-human identities with access outnumber actual human identities 109 to 1. You need a tool built to manage them all. Human, machine, and AI. That's exactly what IDERA by Palo Alto Networks was designed to do. It's one identity platform for all, and it outshines legacy tools that were only built to manage human employees. How does it work?
Well, IDERA replaces permanent permissions with dynamic access, so you can lock down every identity without slowing down your own business. With IDERA, you get unified control by replacing siloed IAM or identity and access management, PAM, modern privilege access management, and governance tools, all with a single control plane. And it does that across human, machine, and AI identities. Secure every identity at your organization with IDERA by Palo Alto Networks. Learn more at paloaltonetworks.com slash I-D-I-R-A. Again, that's paloaltonetworks.com slash I-D-I-R-A.
Socks, underwear, tees, and more. Bombas. Everyday comfort worth obsessing over. Visit bombas.com slash audio and use code audio for 20% off your first purchase. That's B-O-M-B-A-S dot com slash audio. One purchase equals one donated. Always.
Running a business shouldn't feel like surviving a software group project. One app for accounting, another for inventory, another for sales, and somehow none of them talk to each other. That's where Odoo comes in. An all-in-one business management software that brings every part of your business together. From sales and accounting to inventory and marketing, all-in-one powerful platform. No messy integrations, no bouncing between tabs, and best of all, no spreadsheets. Stop managing software and start managing your business with one unified system. Try for free today at odoo.com slash vox. That's odoo.com slash vox.
All right, we're back. Time for the Thunder Round. Richard, you don't know this yet. Nilay obviously is not around. And so I have gone on a power trip. You know how Nilay feels with the Lightning Round. He thinks it's a fantastic part of the show. I have beef with it because there's a bunch of stories we want to get to. We never get to all the stories. And so I have destroyed the Lightning Round. It's gone. It's done. It's never coming back until maybe Nilay gets back. We'll find out. In place, we have declared this the Thunder Round.
This is the thunder round. Instead of the lightning round, we're going to do five stories, five minutes each. I've given Eric Gomez, our producer, the power of thunder. If we go too long, we're going to hear a little rolling thunder playing us off. And if we need to move on, thunder will crash, strike. There will be thunder. There will be a loud noise telling us it is time to move on. All right, let's get into it. V, what have you got for us? So my new title at TheVerge.com is Senior K-Pop Demon Hunters Reporter. Congratulations on the promotion. I love, you know.
A long time coming. I was a senior reviewer. Then I was a senior Cursed Tech reviewer. And now I'm a senior K-pop Demon Hunters reporter. And that's because K-pop Demon Hunters is officially the most ever watched movie on Netflix. Most popular movie of all time. And people actually like it. It's a great movie. This is the crazy. I cannot think of that for real theaters. Real theaters. This is going to the Oscars. Like, mark my words. I think I read that they are actually going to be nominating the soundtrack for the Oscars.
As they should, it is a banger of a soundtrack. And I'm not at all biased as a K-pop stan, but it is a banger of a soundtrack. And, like, the movie is actually, it's a spreezy 90 minutes. It's very good. This is legitimately the first Netflix movie that I've heard people talk about after the movie came out. And continue to talk about it. It's always, right, they'll create buzz in the lead up to a movie. The movie comes out, you don't hear a thing.
This is the only one that I didn't hear anything about this before it came out and nonstop for the past like two months. There was like almost no promotion there. I think there was one trailer for it. They did not promote it at all. It came out. All of a sudden, all my TikTok things were, and I actually didn't want to watch it at first because I like K-pop and because I'm Korean. I was like, oh, K-pop demon hunters. Are you kidding me? Pandering. And then I just, the feed was way too much. And then all of the people were like, oh my God, it's based on this band. And I was like, all right, all right, let me go watch this movie. And then I watched it three times.
That's one of my least favorite things, when everybody's right. It's just a good time. The singers who voice the characters, they're actual K-pop performers. So EJ, they're not. One of them is Audrey Nuna, and she's kind of a rapper. And then the other one is Rae Ami. But the main one, EJ, she is a bona fide. You could call her the Benny Blanco of Korea. Because, you know, Benny Blanco has written all these. How is she on a podcast? She speaks English, so she's probably very good. That's okay. But basically, she's written.
all of these major bangers for these huge bands. Hence the Benny Blanco thing. I found out that she wrote the song Psycho by Red Velvet and I went, what? That's not on my playlist. Holy crap, girl. And she was a trainee at one of the major K-pop studios, SM, known for putting out a lot of insanely talented vocalists, which she obviously is. And she wrote all of the songs for the movie and performed in them. And so it has that very authentic, genuine thing. But one of the most insane parts of K-pop Demon Hunter's insane success to me.
is that Sony, which made the thing, they basically gave it to Netflix because they didn't think it was going to do numbers. They just had no idea what they had on their hands. This movie took nine years to make is the other thing. Oh my gosh. It's like a passion project. They put their whole heart and soul into like representing Korea, representing like this whole culture, which is why I think it's resonated a lot with K-pop fans because there's just so much love put into every single part of it and it's authentic. But it's still just like a good story. And like the visuals are very like unique and fun and kind of not.
It's not Pixar, right? It kind of has the same vibe as the Spider-Man animated movies. The animation looks beautiful. It's really fun, and it's got a distinctive style. And one of the things I see people saying online that I agree with is that, for once, it's original. It's like an original story. It's original characters. It's not a remake of a remake of a remake. It's just something new that's out there, and it's going to be done to death in the same view as Squid Games. But yeah, just the fact that Sony needs a hit, and they're like, I don't know what to do with this here at Netflix. You take it. And then Netflix is like... This explains a lot. I want to ask something really quick of both of you.
Can you both try to name one other movie in Netflix's top 10 of their original films? One other movie that Netflix has made that was successful. Bird Box? Okay, okay. That's one of them. That's one of them. That's number six. Richard, do you have any Netflix original movies? Only because I saw the list do I know. The movie that no one admits to having watched. I've never heard of it. I constantly am like, it's like Gray Notice. Because they have Gray Man and Red Notice. I can't remember. I'm like, that's all I can remember as far as Netflix originals go. But this is it. This is like, nobody talks about them.
The only thing that matters is K-pop demon hunters, but at least they've got to hit. All right, Richard, what have you got for us? A pretty big controversy going on in the YouTube world, surprisingly. Never happens. Several creators are upset because they noticed and they've started talking about how some of their YouTube shorts videos look different. Rhett Scholl, I think, had made a video about this. His report on the BBC and other places. And YouTube responded. Basically, what they noticed, particularly comparing the YouTube shorts version, I think, to Instagram reels, is that the versions of their videos that were on YouTube had these extra details and looked really...
absurdly sharp, and they felt like it looked like an AI deepfake, is how they kind of felt like their videos looked. And YouTube's Rene Ritchie, he responded. He said that it was an experiment that YouTube was running. As far as I can tell, they hadn't told anyone about this beforehand. And that they are using the kind of machine learning you experience with computational photography on smartphones, not generative AI. This is kind of the push and pull that's going back and forth between people saying it's AI or not AI. So if you've taken a look at the evidence, what do you guys think? AI or just upscaling?
It looks like nasty upscaling to me. This is just a failure on YouTube's part in general. You know what they look like? They look like ripped videos on TikTok that somebody has put an ad on top of. I've seen this look, and it is when somebody is boosting your video and trying to hide that they have downloaded it off something else. I had to go onto a Zoom call earlier today on the AI phone, and I don't know if it was Zoom or something else, but I looked at what I looked like in the camera. I'm like, oh, no. That's...
too much like me. It's just sharpened to the point where I was like, it's uncanny. It doesn't look right. And so I turned my camera off. I do think there's this very funny, weird thing happening where people are like, I don't like it. It must be AI. And the thing is, this is actually just bad. It doesn't matter what it is. It's just bad. They just shouldn't do this. It didn't look good. They didn't tell people that this was happening to their videos, the select number of videos they use this on. But yeah, I've been sitting in demos at trade shows with companies showing off their HD upscaling that would make your 480p into 1080p for years. And this is what it looks like.
did that and turned the dial up. And they all claim, every TV claimed that it has an AI processor in it for the last 10 years. So is it AI? Is it machine learning? What is the difference? I do not know. But it just, it looks way over sharp. They turned it up too much. And I think that the other part is how all of the tests and all of the features that YouTube is adding. We saw some backlash. We saw things pop up. There's one who was kind of upset because she noticed on Instagram reels that her voice translation, it translates it into another language, but also simulates your voice, which I think people were not ready for. So.
Whenever you see something and it's altered, of course, for many people, the first question is going to be, oh, it's AI. It's just where we are now. No, that's a good point, though. And that did happen. And Instagram really was doing that. And it was kind of impressive. But it's probably good to know that that's going to happen before you see yourself talking in another language. I was watching those. I'm like, is it even sinking the lips? It's trippy. It's really trippy. And super impressive. I'm sure if you're a creator, it seems good if you want more reach. But you want to know what's happening with your videos.
And there's this other element here where like YouTube is acknowledging that it's machine learning. And it's like, oh, God, where is the difference between machine learning and AI? And which is the one that we're mad at and which is the one that we and like my whole thing is like, what is the outcome here? And the outcome here is they made weird, crunchy videos and didn't tell anybody that they're going to make them weird and crunchy. And everybody was just mad that they looked ugly.
I mean, that's the real crux of it. It's just like, if you're going to do this, don't make me look ugly. We have vain creatures. If I look better, I'd be like, oh, thank you, AI. I look crunchy. You have these creators who worked really hard to make their videos look the way that they look. They have expectations. They know how YouTube works. They know how they'll be delivered on different platforms. And they've gone to a great degree to get it to look this way. And then suddenly you change it. Yeah, at first when I heard they were secretly AI-ing videos, I'm like, oh my God, did they facetune people? Because that's creepy. That's messed up. No, they took the clarity slider and dragged it all the way to the top. Don't do that. Okay, next story.
Google announced this week that they're going to start requiring verification for developers who want to publish apps on Android that live outside the Play Store. This, I think, doesn't sound monumental, but it kind of is. This is essentially how Apple operates the Mac. Anybody can publish a Mac app, but it's secretly a lockdown platform, which makes sense. That's how Apple has always kept its gates tight. But if you want to publish outside of the App Store and the Mac, as people traditionally have, you still have to have an Apple developer account. Apple's always required these developer accounts.
If you go awry and do something illegitimate, they can just shut down your app. This makes a little bit of sense to me on the Mac, where Apple has always had this level of control. Android, on the other hand, is ostensibly this open platform. It's ostensibly a platform that is open source, that nobody controls. Now Google is saying...
It's an open source platform. However, we would like every developer to register with us and we will therefore be able to shut off your app if it's a problem. Now, on one hand, okay, great. That's a good security feature, right? I see the reason for it. I see how that makes sense. But to me, the other thing I see here is Google continuing to tighten its grip on Android and make it feel even more and more like a product that belongs solely to Google. And that's sort of been the entire trajectory of Android, I feel like, over the past decade plus.
I think that's only going to get more true the more that they try and tie Android to Gemini. I think as that happens, as Gemini becomes another vector in which Google tries to assert itself into products, I think... I think you're going to see more of this going forward. Yeah, and I don't think this is the reason why everyone who uses Android uses Android, but there was in the past a reason to use Android because you could have more control of your device. You could install apps from anywhere. You could do what you wanted. And that is just, it's becoming a bit less true. And we saw a kind of example of what this could become just, I think, in the last day. Apple pulled a Torrent app for the iPhone that was published in the Alt Store in Europe.
But they can do that. It's not clear why. And we can ask them, but they don't have to answer. They can still just enforce their arbitrary rules. And look, is that app being used for piracy? 100%. 100%. Should Apple be in charge of deciding whether you can pirate stuff or not? I don't think that's the right move here. And this is sort of the direction this takes, where Google now is going to have that authority.
And it's also different if you, from the get-go, like Apple from the get-go is like, no, we have the control. But Google has historically been like, we love open source. We like being that kind of more interoperable company. And to Richard's point, who among us doesn't know the Android stand? And I say this lovingly, who's just like, oh, I just side-loaded onto this. Like, I can just do whatever I want with this. Versus you iPhone people enjoy your prison in the walled garden or whatnot. And so I think that when you position yourself one way and then you pivot and people start noticing that pivot, that's when people get mad. It's like when you're...
a free app and then all of a sudden you're like, hey guys, subscription. Even if it's a baby step, even if it's like a small thing in a direction, you just get a lot of backlash because people are like, that's not what you said you were. That's not what you said. That's not what we signed up for and now you're going the other way. So it's slippery slope. I think you're completely right. And I've talked about this before, but it's fascinating. If you look at Google's contracts with other smartphone makers.
They do everything to lock down Android to make sure it's a Google product, even though it's open source. They say, if you want the Play Store, if you want our best apps, you've got to put them in this folder, on this screen, in this spot. They're super specific about all of this. And so I think we don't quite see it all the time, but there are just a lot of different ways that Google very clearly keeps its hands around Android to make sure that it is a Google product, even though it remains technically open source. So you'll still be able to sideload. This is still going to be an option. I do think, yes, this is a good security feature, but it has some very, I think, strange implications for the future of Android as an open platform.
All right, V, what's next? So I really love this one, but Taco Bell, an executive there, is just like, so you know our plan to put AI in the drive-thrus? Maybe I have second thoughts about doing that because there's a lot of people complaining on social media and a lot of people trying to troll the AI on social media by asking for, I believe, 18,000 cups of water. And so he tells the Wall Street Journal, we're learning a lot. I'm going to be honest with you. I think like everybody sometimes, it lets me down, but sometimes it really surprises me, which is, you know, I don't want to go to a drive-thru and talk to an AI and get a surprise.
So I just, this tickles me because it's just, the other thing he says is that he's discovered that using AI exclusively in certain situations, like a drive-thru for, quote, super busy restaurants with long lines might not be such a good idea after all, which I don't know, you think? Yeah, Richard, drive-thru AI, thumbs up, thumbs down. Could we just have buttons? You offer like 10 things. Just let me press the button and then we can do it. But yeah, I think it's not that hard to imagine that a bored teenager in many cases, or basically anyone just happening to do this job, is better equipped to deal with someone driving up to the window and saying, yeah, I want 18,000 cups of water than an AI is. It's just not going to work. Also, they might fight you.
Underrated elements of having human importance. Would you imagine? You're not going to do that at Waffle House. You're not doing that at Waffle House. Waffle House, have you ever seen that video of the lady at the Waffle House and they throw the chair at her and she's like, not on my watch. And she just pulls the chair out of the air. Yeah, you're not going to do that at Waffle House. This is such a fascinating use case because on one hand, couldn't AI handle a drive-thru order? Probably with some frequency, it could do that correctly. But this is like a business that serves customers and needs to do things correctly like all the time. Weekly too. Yes.
And it's one of those things where like, okay, great. If AI screws up a response to me one in 10 times, like it's not, it's not, it's annoying. It's not great, but it's not a big deal. It's just talking to me. I see that it's messed up. I can deal with it. If it's in a drive-through dealing with thousands of customers a day, like you can't have that error rate. Like AI's error rate is not ready for this. This is like precisely one of those situations where a company is like, automate, we can make a ton of money. CEO, I'm gonna sound cool. My earnings call saying AI, AI, AI. And then you put it into use and it's like, oh, like even if this works most of the time, like it needs to work.
all of it. Sometimes when Big Tech talks about things that way, I'm just like, dude, they seem so pure inside because I was like, did you give it to a troll? Did you troll test this for, because it only takes one jerk to break the whole thing. They're like, it'll be great. It'll do this. The ideal situation will happen and we're helping everyone. And it's like, did you give it to a 13-year-old who really lives for the shits and giggles? Because if it can't survive that, it's just not going to survive the real world. And if you're Taco...
The only people I know who go to Taco Bell are probably having the munchies and or really hungry. And so if your AI is not working, they have the munchies, they're in a drive-thru line, they have road rage, and they're hangry. This is not the ideal situation for an AI. And I kind of feel bad for that. Not that AI has feelings, but I kind of feel bad for it in one sense. And then my other spiel is that, you know, whenever I call an insurance company and they put me in the AI robot tree and I have to listen to the dumb thing and I'm just going, let me talk to a human. You know, how is that not the same thing?
Right. Can you just yell zero at the AI? I would just be like, human, human, just, I want a taco. Mild sauce, please. Chalupa. Like, it's not, it's not, I just want a chalupa. I, it makes a lot of sense that Taco Bell's customer base immediately made them regret this. Yeah. And I appreciate, they did good work here. Live must. You know, live must. Unless you're the AI at Taco Bell, in which case, live menos, I think. And that is the limit of my Spanish. Richard, wrap us up. What have you got?
This one came from Andrew Hawkins. We have Robomart. They have introduced a new delivery robot that aims to shake up autonomous delivery. It is essentially... Remember when Amazon introduced those lockers where you could get something delivered? Like if you didn't want it to be delivered to your house and sit outside all day, you can maybe have it delivered to the locker, get a code or something, and you just go pick it up. So what if the locker drove to your house? No! I...
I can't tell if this is brilliant or the dumbest thing I've ever heard of. It's both. That is exactly what I said. It's one of those two things that I don't know. It might be both of them all at once. It's both. I think the future is just lockers everywhere for everything. People love the lockers. I love the lockers. I'm like, ooh, a locker. I could pick something. I don't have anything to pick up. It could have something. It's the size of a shuttle bus and can carry up 500 pounds. What? You've got to look at this. No, no, no, no. Oh, my God. Sorry. I thought this was a small robot, not a full.
Bus-sized vehicle. Oh, my God. It's like a sprinter van, you know, just rolling around full of whatever. Does this make any sense? No. I don't. But the problem is this requires... What happens if the person... The thing with the locker, the locker sits there. It waits for you. If I'm not at home and the locker shows up, is this... Well, the locker could come to your office. But is it going to like... No, it's not useful for my groceries. ...parallel park and try to like wait for me?
Like, what happens to the person who's like 10 lockers in? What I love was one of the quotes on their website from Mars Unattended Retail. This partnership helps solve the industry's biggest issue. Ice cream melting before reaching the home when purchased in-store or through other delivery methods. What? Just get a cooler. Because it's refrigerated. Oh, it's refrigerated. I mean, this is a really expensive locker. This is the least efficient way to ship ice cream. How is this an issue? I don't understand this. I like...
I've tried grocery delivery before. I would say it has mostly been a bad experience. I've not had that bad of an experience with grocery delivery, even with ice cream. Have you tried Amazon Whole Foods, like grocery delivery? Okay, I don't use... They were not great. So I tried this because I got mad at my local grocery stores. You know what? They should carry better produce, all right? You cut open too many peppers that are bad inside, you get pissed. The problem I kept running into with Amazon is I'd be like, all right, I'm going to order some food, and then they would just be out of stock of everything. There's some weird online thing, but just it's not...
I was never upset. Like, I'm not going to order ice cream. It's fine. I don't need to order ice cream for delivery. You've never dreamed of sitting at home wanting an ice cream sandwich and having a bus rolled up to deliver you one ice cream sandwich. That's an abusive power. Richard, you have invented a business known as an ice cream truck. This exists. This is famously a very successful industry. In New York, those things are like rolling around at all hours of the night. I can walk outside at any time and get an ice cream sandwich. I don't need a giant truck with a personalized locker to give me eight of them at once.
They're solving an already solved problem. We have delivery. We have refrigerated trucks. I don't know what this is for. I'm sorry. You know when you put it that way? Yeah. You know, I understand why a business person, I understand why like an Amazon would look at this and go, this is genius. This is on the level of AI. It will revolutionize our business. And then as soon as you figure out, think about like, how am I actually going to put this to use?
Just have literally no idea. I'm also afraid of this on the road. I already have seen a TikTok of someone just like filming one of those, like the small ones that deliver drinks around. And they're just like, don't hit him. Oh my God, it's like stronger. He's going to get across the street and we're going to do this. I'm afraid of this. It's got a bit of a tank vibe. It's got it. It's like, yeah, this thing is hefty. It's hefty. Depends how much ice cream it's carrying. Can you just imagine it getting into a collision and then just the ice cream exploding out of the lockers and then it melts that way? You know what? That's the dream. Yes, I can. That is going immediately to the top of the verge.com.
All right, that's it for The Verge Cast. Stay tuned. Allison will be hosting on Tuesday. If you like what we do here, the best way to support us is to get a subscription to The Verge at theverge.com. We publish a lot of great stuff every single day. We'd love to hear your questions and feedback. Let us know what you want us to talk about in the show. Email us, vergecast at theverge.com, or give us a call, 866-VERGE-11. The Verge Cast is a production of The Verge and Vox Media Podcast Network. Our show is produced by Eric Gomez, Brandon Kiefer, Travis Larchuk, and Andrew Marino. See you next week.
Running a business shouldn't feel like surviving a software group project. One app for accounting, another for inventory, another for sales, and somehow none of them talk to each other. That's where Odoo comes in. An all-in-one business management software that brings every part of your business together. From sales and accounting to inventory and marketing, all-in-one powerful platform. No messy integrations, no bouncing between tabs, and best of all, no spreadsheets. Stop managing software and start managing your business with one unified system. Try for free today at odoo.com slash vox. That's odoo.com slash vox.