The Vergecast - Phones are the ultimate AI gadget

Episode Date: April 30, 2024

Today on the flagship podcast of dedicated AI hardware:  The Verge’s David Pierce and Allison Johnson debate whether the emergence of standalone AI gadgets like the Humane Pin and the Rabbit R1 are better off as apps or should exist as its own hardware.  Humane AI Pin review: not even close  The Humane AI Pin worked better than I expected — until it didn’t  A morning with the Rabbit R1: a fun, funky, unfinished AI gadget Can Rabbit’s R1 outsmart the smartphone assistants? Let’s find out! The future of AI gadgets is just phones The Ray-Ban Meta smart glasses actually make the future look cool   The Verge’s Alex Heath joins the show to discuss Meta’s big move into AI with its multimodal AI smart glasses and a new AI model called Llama 3.  Q&A: Mark Zuckerberg on winning the AI race  Meta wants to be the Microsoft of headsets Zuckerberg says it will take Meta years to make money from generative AI Nilay Patel answers a question from The Vergecast Hotline about Microsoft and antitrust. Microsoft splits Teams from Office as antitrust pressure ramps up Microsoft and OpenAI deal may face anti-trust investigations in the EU.  Email us at [e-mail] or call us at 866-VERGE11, we love hearing from you. Learn more about your ad choices. Visit podcastchoices.com/adchoices

Welcome to the Vergecast, the flagship podcast of dedicated AI hardware. I'm your friend David Pierce, and I have a label maker. One of the things I've learned about myself is that I procrastinate by organizing, which mostly means I just like take a thing that's over here and put it sort of arbitrarily over here, although it improves nothing, but it like scratches my brain in a very helpful way that makes me feel like I'm accomplishing things, even when I'm mostly not. But now that I have a label maker, I'm actually going to get stuff done. There's the spice rack over there.

I'm going to label all the spices. So I stopped using paprika instead of cinnamon accidentally. I have a bunch of tea up here. I'm going to label all of that. So I stop drinking caffeinated tea at night and staying up all night. I'm going to label the baby bottles over here. So I know what he's supposed to be drinking out of and what we're not supposed to be drinking out of anymore.

I'm going to label the milk with the word milk, which will accomplish nothing, but will make me laugh every time I pull the milk out of the fridge. I'm just going to label everything. And then my wife is going to come home from work and have a lot of feelings about what happens. So we'll see. Anyway, we are not here to talk about label makers. We are here to talk tech. And we have a lot to do on the show today.

We're going to talk to Alison Johnson about AI gadgets. She's been testing some. I've been testing some. And she has a theory that maybe none of them need to exist and that maybe a phone is the answer. I have some thoughts. We're going to talk about that. And then we're going to talk about meta and AI. because meta has been on a huge run of new AI products,

and there's a big story in there somewhere. So Alex Heath is going to come on and help us figure it out. Then we're going to do the hotline, lots to talk about this week. All that is coming up in just a second, but first, I'm not kidding. I have a fresh roll of label tape. The battery is charged. I have work to do. This is the Vergecast.

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

Proms something like, build me a revenue dashboard on our Salesforce data. And Retool actually builds it on your company's data and your cloud with enterprise security built in. Go to Retool.com slash Vergecast. We all need to retool how we build software. Welcome back. All right. I believe I have labeled everything in my house that there is to be labeled, and then some. We'll see how my wife responds when she gets home. Let's get into it.

For the last few weeks, I've been immersing myself in AI gadgets. The humane AI pin, the Rabbit R1, the meta-smart glasses, this voice recorder I have called Plaud, a bunch of apps, everything I can get my hands on that is like a way to use AI. And the number one question in front of all of these devices in particular, I think, is the same thing. Why isn't this just an app on your phone? Actually, that might be the number two question. Number one is just, is this thing any good and a lot of them aren't. But let's focus on the phone thing for right now.

The Virgins Allison Johnson wrote a piece last week arguing essentially that a smartphone is the perfect AI gadget, that we don't need dedicated devices. We have the dedicated devices. And she ran an experiment to prove it. So, of course, I had to ask her to come on and tell us how it went. Alison, welcome back. Thank you. We usually talk about phones when you come on the show.

And this time we're going to talk about phones, but maybe in the weirdest way that we've ever talked about phones, which I'm very excited about. I'm here for it. So let's just start at the beginning here. Tell me about this wearable phone experiment that you devised for yourself. Oh, my gosh. I have never been so glad that I just work in my own house alone. I was just sort of ranting about how.

you know, the humane pin reviews came out and I was like, this thing is just an Android phone, like a mid-range Android phone. I was like, I have so many mid-range Android phones. So I just kind of set out to like recreate the capabilities with what I had. And it was surprisingly hard. I tried a flip phone. I had the Motorola Razor Plus and I had it like kind of clamped over the collar of my shirt, you know, like you do. And I got like so far with that. Like I could kind of talk to Google Assistant. So wait, so you you dangle it over with like what, the front screen facing out. Yeah.

And you just, and then you just sort of like, hey, gee it from there. Yeah. And I figure it, because then you have cameras facing forward. You could do Google lens. That was the thought. It's not very easy to do any of that. Kind of out of the box. Like the flip phones won't let you download the Gemini assistant, which is strange. So you kind of That's weird. Like side load it. If you try and try and. trigger the Google Assistant with the phone closed and the cover, just using the cover screen, it tells you to open up the phone. So it kind of like, it was already not off to a great start.

So did you ever find an actual solution with the phone itself? We're going to get to headphones in a minute. Yeah. Foreshadowing. But did you ever find a version of like a wearable phone that actually kind of worked? I did not really. I had, I just ran chat GPT in the kind of like, talking mode, the conversation mode. And I had that on my shirt and just, but it's just always listening and like waiting for you to say something. So you can't just like go about your day and we had a very pleasant conversation about like the weather or whatever.

But yeah, nothing really worked the way I thought it would. Well, this is why I wanted to talk about this because I think you and I have been talking about this for months now. And the question everybody brings up about every single AI gadget is like, why isn't this just an app on my phone? and my thoughts on the subject and why it shouldn't just be an app on your phone are actually becoming increasingly strong and intense. And I feel like I'm more right than I was before. But I am curious, like, at a broad level, do you buy that argument?

Like, are you also one of the people who is just like, this is all ridiculous? This should just be an app on my phone? I think I'm like 20% rooting for like a gadget that does these things. And 80% just feels like realism. like phones have already solved so many of these problems. Like these little gadgets get hot. They have to connect to the internet. They need cameras.

Like checkmark on all of those for phones. Like they did it. Battery life, you know, regular software updates. I got it. And I'm just like, I'm having flashbacks of covering compact digital cameras for years where it was like, you know, oh, these are so much better than your phone, and which is like not even true anymore. But it was like, no, the phone is good enough for most people. And people do not want to carry around two things.

That is what we learn from that experiment is if you can carry one thing instead of two things, you will carry one thing. I do agree with that. And that is, I think maybe the most convincing argument against the AI gadget is just it's the good enough hypothesis, right? that it's like I don't need it to be as good. I just don't want to carry and charge and worry about two things. But I think my most strident opinion on this subject is that I think we all have forgotten how annoying it is to use your phone. Like we've just been doing it for so long that I think it's why I believe in flip phones for the same reason. It's like our phones are these big, chunky things that you balance on top of your pinky to the point where most of us have like divots in our fingers now from trying to hold these things.

you have to sort of like worm your thumb around this gigantic screen in order to touch anything. Like from a pure in a vacuum user experience, it's not good. And if your hands are full, you can't do anything. And I spend like an alarming amount of my time with like dog and stroller in one arm and phone in the other. Like it's just bad. And I, this idea that, okay, every time I want to do anything, the right user experience is for me to dig into my purse, pull out my phone, unlock it. Swipe, swipe, swipe, swipe, open an app, tap on the app, log in, do a thing. Like, that's bad.

And the idea that you can shortcut all of this by just yelling a thing is good. I believe in that. And this is why I liked your experiment so much because it's like, okay, can I make my phone that thing? Which I think is where we get to earbuds. Yeah. The longer we do this, the more I am coming to the answer of like maybe headphones are the problem solver in all of this. Yeah. And this is something our colleague V-Song has written about is like,

With wearable tech, it's so important that it's just like something you want to put on your body and is comfortable and is kind of like socially accepted. And I just come back to earbuds. Like we're just people just walk around with them. I feel like such an old person being like, these kids, they just walk around with their earbuds all the time. But like you have a good transparency mode, you know, and those things could get smarter about the sounds they let in and the sounds they keep out. And then you have this connection to your phone that you can just like summon it when you want it and keep it in your pocket and not have to do the whole dance of like tapping on things. Because I think we've come. We're sort of blind to that.

Like we are just so conditioned to like, this is how you get things done on a phone. You take it out. You do all this. You get distracted. And yeah, once you start kind of like conceptualizing not having to do that, it's like, oh, yeah, I buy into that vision. And I, but I think we're just going to be able to do it without touching our phone. But it's still going to be phone. Yes.

And that turned out to be sort of the second half of your experience with this phone AI gadget thing, right? You basically went like full pixel. Yeah. I did like, I downloaded the Gemini assistant and switched into it. And there's still like weird workarounds you have to do because that you can have Gemini assistant running on the phone, the earbuds are not going to trigger Gemini because they don't talk yet. This is the most Google thing you've ever said. Yeah. Seriously. Well, like the Pixel Watch, you can't have Gemini on it yet. So you're like talking to one assistant here. You're talking to a different assistant on the phone. The Google home in my kitchen is constantly being like, what? So I had to like unplug that. And I kept Gemini like open and running on the phone, which is not an ideal user.

experience, but had the earbuds in. And my like light bulb moment was I took a picture of this recipe I was making and I was like, look at this and remember it. And I just walked around my kitchen doing things. And I would ask you questions like, how long do I put the fish in? How do I chop the vegetables? Just kind of out of order and like as I was doing things. And it got it right all the time. And it was like, oh, this is super helpful. I would be running back to this recipe with like, stuff on my fingers trying to figure it out. And it's like, man, those earbuds. That is maybe the single coolest use of AI I've ever heard, by the way, to take a picture

of a recipe and then just pepper the assistant with questions about that recipe as you go. That's very cool. I was super into it. And it's that kind of thing of like, I run out of things to talk about with chat, JBT. I'm like, there's only so many like business proposals you can brainstorm. But I'm like, if you give it a little data set, of like, I'm doing this thing, help me out. Like, it makes so much more sense to me that way.

Yeah. Are you a voice assistant person in your day to day? Not at all. I set timers and I ask the weather and, like, trying to get myself, I'm starting to, like, realize I can use it a little more, especially if I am using a phone with Gemini on it. Like, I ask, but you just, man, you got to keep an eye on it because it'll make up a word or, like, I was asking it if, the bridge they were working on in my neighborhood had reopened. And it like got the days backwards.

It was like, yes, it's open, but it's opening tomorrow. So you just have to keep an eye on it. But it's like converting me a little bit to more of a voice assistant person, I think. Okay. Yeah, I think one of the big questions I have about this space right now is, did voice assistants work because they were a bad idea? Or rather, did voice assistants not work because they were a bad idea? or did they not work because they weren't very good?

And I think they definitely aren't very good, right? And I think the Siri experience sort of proves that for a lot of people. And I think a lot of people's first experience with these things was Siri, and you ask it to do something and it just fails or it's like, here's what I found on the web. And you're like, well, this is useless. And then you never really try again. And so the first attempt we all had to build that habit,

it failed spectacularly. And there are a lot of, like, kids who grew up asking Alexa silly questions and getting silly responses, and that was the thing that worked. So I think the attach rate there has actually been better as a result. But then I come around to, okay, we all bet on this a decade ago. Like the tech industry was like, this is the thing. Conversational interfaces, they're going to change everything. And they super didn't.

And now to some extent, we're back in that exact same thing. And the promise is, no, it was the right idea with the wrong underlying technology. And at least my experience so far has been, I still see the theory. of it. But even as I use all these things, I'm not, I'm using it more. I agree with you. I've definitely found more reasons to do this stuff, particularly with like, I really like to use an AI voice notes app to like say my to do list out loud. And they're actually getting pretty good at turning that into just a structured list of stuff I have to do today. I feel like a doctor who's like saying all of my charts into my voice recorder and then like something magical happens. It's great.

But I still don't feel like I have hit the thing where it just, sort of that becomes the main way I use my phone. And what people say when I say that to them is, oh, it's just an interface question. It's because once you have your phone out, why not just use your phone the way that you're used to, which is why they're all excited about these new gadgets. And so I just, I don't know, I spin in circles about like, what is the actual sort of roadblock here to this actually becoming real? Because it seems like there are a lot of them. Yeah. There's just sort of like a pain versus reward, you know, ratio of like, I feel like it's starting to even out a little bit with the voice assistance, but you still get burned where you're

like, and there's nothing worse than like being at the grocery store and being like, I'm going to use the voice assistant. And then it's, it just fails and you have to like either keep talking to it or like take your phone out. I think once you had that experience, you're like, oh, just use my phone for this. Yeah. But yeah, I think, I think it's something we can adjust to. And it really has to like, that threshold of public embarrassment or just like frustration has to change. And there's like signals that that will happen, I think. But yeah, it's it's not proven, I wouldn't say. Yeah. One of my developing theories is that headphones are going to be where a lot of this goes for that reason. It's like it's a thing you can wear one of in the grocery store and no one will look

twice at you. Like we have solved the public weirdness problem of headphones. Sony did that for us 45 years ago. Yeah. Thank you, Sony. But what I wonder now is, like, should you have the AI system that is just purely baked into your headphones? Like, one of the conspiracy theories that I have no evidence for, but believe to my soul, is that the next AirPods maxes are going to be local AI devices. Because they're big and they're heavy and they have a big battery and they can do stuff. But, like, if you try to run some of that stuff on AirPods or the pixel buds or it's just the battery, the battery.

is going to die in one minute and there's just there's literally no place to put the AI stuff. Yeah. But maybe it's maybe that's okay. Maybe Bluetooth headphones and your phone in your pocket or purse is enough interface. And it does seem like in your experiment, that's kind of where you landed. Yeah. That that kind of proved out. And that didn't solve the camera thing, which is also like an intriguing related, you know, like feature, which the humane pen has the camera.

and we're both like want to look at something and ask a question about it or if you want a hands-free way to take pictures, which both are kind of intriguing to me. So, I mean, everybody keeps bringing up the meta ray bands and man, that kind of solves a couple of those problems and doesn't look stupid. True. Do you wear glasses normally? I wear contacts. So I'm so close to buying these things. I don't know. Really? Yeah, it would be so good for just like taking pictures, kid, running. running around and I don't know, but another gadget. Do I want another gadget? I don't know. I hear you. I've been debating this with some of the folks on our team for a while because I think

they're definitely a huge leap better than like having a headset on your face to have the meta smart glasses. And meta is leaning into the fact that we've also solved having glasses on your face is not weird. And I don't know what your experience has been, but when I wear the meta glasses, no one ever looks at me sideways. Yeah. Like I got more public notice, more sort of weird double takes wearing the humane pin than I ever have wearing the meta glasses by a mile. Yeah.

So that's a huge victory. But I also don't wear glasses normally. And I don't want to start wearing glasses. Like I wear sunglasses when I'm outside and I have mostly replaced my sunglasses with the meta glasses, which I like a lot. But in terms of a sort of always on assistant, like I'm not going to get to work and keep wearing glasses I don't need just to use my assistant. It's a different vibe. Yeah, you have to be ready for that shift. Yeah. Like, if we can normalize sunglasses indoors just in order to make

AI happen, I'm here for it. This works. We'll work on that. But yeah, I do think I have definitely come around to the idea in testing the humane pin and now the rabbit and some of the other stuff that's out there now that actually what we need to figure out is how to put AI into other socially acceptable things that already exist before we try and invent some wholly new thing. Yeah. Like the humane pin kind of tries to hide it, but it's sort of like, look at this. This is kind of a cool gadget. I appreciate that the rabbit is just like, this is a thing.

It is bright orange. It like practically glows in the dark. I don't know if you have this experience where I walk into a dark room and I'm like, there's the rabbit art. It really does. It sort of screams at you from. wherever it is. What do you think about the rabbit? Speaking of experiments you've been doing, you've been testing rabbits AI stuff with some of the other stuff that's out there. How's that

gone? It is, boy. That's a tough. That sounds right. It's a tough one. The battery life is just god-awful, which, you know, as it turns out, is important because I don't, if I just have it in my bag and I go somewhere, I'm like, I'm going to leave it off because the minute I turn it on, and it's going to be like draining the battery. So like, A, if there's no battery, it doesn't help you do anything. It's just a cool orange thing. I've seen a couple of moments when I'm like, okay, this is maybe doing a little more than a Gemini. I took a picture of my plant on my desk and it did the typical thing of like, this is a pothos plant, blah, blah, blah.

And it kind of like went into the care instructions and it was like, your plant looks happy in its current pot. I'm like, you know what? I don't get that assessment. from Gemini, it just sort of spits out a bunch of stuff at you, like, don't overwater it, don't underwater it. So I don't know. There's like something that is a little more like could tune itself to like what you're actually doing or looking at that feels like not just Googling something. But I don't want to give this thing too much credit. It is not done that very often. And it kind of feels that basic stuff. Yeah, it's bad. Like I think we're going to have,

We're going to have more to say about it, both on the site and on this podcast, but the, the, the, the, the, the, the, the, the rabbit R one is not good is, is a pretty easy takeaway. But I'm curious, you've been testing it against chat GPT in particular pretty strenuously, right? Are they, my assumption would be that most of these things should come out relatively the same, because it's all kind of pulling from the same backend infrastructure. Have you noticed anything interesting poking at all this stuff? Well, I think that something you mentioned, I think, is that it's running perplexity. So you get a little more like real time info than a chat GPT or the free chat GPT I use that was trained whenever I go. So I can ask it like, is that bridge open that was closed over the weekend and I get a real response. So that's interesting, but it's it's such a strange little, strange little crazy. It's like very adorable. I think there's an adorable factor. It perks up its little ears when it's listening to you. But TBD on that. Does that do anything for you? Actually, because I think I have, I've definitely had the same experience that I like that the little rabbit on the screen sort of bounces waiting for you to talk. And then you ask it a question and it perks up its ears and stops bouncing. And when you're playing music, it wears headphones. And there's a lot of sort of charming little things about it, I think. And I can't decide if that's, what I want or is totally the opposite of what I want from a device like this. Because you use

Gemini or ChatGPT or Siri or whatever and it's all very matter of fact, right? Like they are tools to do a job. And actually, every time we've seen these things exhibit personality, it's been in really bad, ugly, problematic ways, mostly. And so they've all learned to just sort of shut all that down and just be there to execute whatever they are required to execute and then move on. And part of me is sort of endeared to Rabbit for trying something. else. And part of me is also like, well, maybe if it worked but was less cute, I would like it better. And maybe that's what actually matters here. Yeah. Like there's a cute little animation of the rabbit in a hamster wheel or something when you need to charge it. I'm like, this is adorable,

but I'm annoyed that I need to charge this thing again. Like cute only gets you so far, I think. Yeah. Do you think there's going to be ways to put more of this stuff onto phones in interesting ways? Like is the next round of flip phones and foldable phones and smart phones going to have some of the stuff we're seeing in these gadgets? Or are we just going to keep getting phones? Because I just keep thinking about the razor as you're talking about it. And it's like, what if the external screen of the razor could run an assistant and could have that outer facing camera? Like it kind of has all the hardware it needs to do that. Are we going to get that, do you think?

I would like it. I mean, it would be an ideal world where you could kind of choose from your AI assistant. and you don't just have to use the one that your operating system baked in. Fair. But yeah, then you get into like the tech company is going to let the app makers in, you know, like get into the system the way it would have to where you don't run into a wall. Like chat GPT can't change settings on my phone or like put something on my calendar, you know. So I feel like there is a pretty like firm wall on that right now.

And that would be interesting. I think that's kind of like. Like rabbits, you know, ethos is like, how do we get around that? And the answer is a weird system where you log into stuff and on, I don't know, a virtual computer. Yeah, the answer is largely you can't right now. It turns out. That seems to be not working also. But yeah, it's an interesting question.

And my sense would be that's only happening if regulators make it happen. Right. It seems to some extent, like, more powerful than setting your default browser, which you can do, even more powerful than setting your default search engine, which has only become a thing because of regulatory pressure. And essentially, you could kind of argue it's like allowing you to install another operating system on top of your operating system. Like, if you were allowed to let chat GPT run your phone, that's a pretty big thing for any of these companies to be able to do. I kind of agree that that's how it should shake out if this tech is going to get better in the way that everybody says it's going to get. better. It'll be cool to have options and not just be forced to use better Siri or better Gemini or better Chatchipti or whatever. I don't know that I would bet on that ever being a possibility.

Yeah, that does seem like a big question mark. And then you get back to just like, well, what if you could just talk to your operating system while you're walking around the kitchen? It's like not the ideal future that I want. I think I want a future where there's weird Or you can download a virtual assistant that can actually do things for you. Yeah, I think that's why I'm on like team phones because it's the realist of me is like, oh, they're not going to let this stuff happen. I think that's probably right. But I do also remember somebody, I forget who it was, somebody at one of the phone makers years ago

was like, we're going to get to a point where the phone in your pocket is basically just a cellular modem. And all the accessories around are going to be. connected to that in some way. Your wearable is going to use that to connect to the internet. Your headphones are you going to use that to connect to the internet. Your laptop is going to use that to connect to the internet. But it is, it's going to be less the device you use for everything and more the device that just lets you use everything everywhere. Yeah. And I think that's a really interesting version of the future. And it's kind of what you discovered with this is like, I can just use my

phone as a go between to all these services that exist. And with headphones that I already own, I actually don't need to invent a completely new pipeline for that stuff. I just need a new way to kind of get at that pipeline. Yeah. And I've become just an addict for the smart watch. Like I, over the past couple years, like, I never used one before. And now I cannot live without one. It's just like, it's the right kind of light interface for when I need to do stuff.

Like, I'm out on a walk and I need to pause music or just check a quick notification or a text. And it's like you run into the wall very quickly of like you can't really respond to a text very well or you can thumbs up emoji something. Right. But boy, has that been like that is the point where I'm like, I am so willing to charge this extra thing and it has its own special charger that I have to bring along when I travel and like it crosses that threshold of like, yes, I will do all of that because I find it so useful.

Yeah, I think smartwatches probably land in the earbuds category of that kind of light input system. that really works. Like when I wrote, when I wrote the Humane Pin review, the overwhelming response was what you're describing is just a better Apple Watch. And I think other than the camera stuff, which is left to be solved, but right now is only kind of somewhat useful anyway. That is true. Like my big regret right now is that I didn't buy a cell connected Apple Watch. Right. And I wish I had because then I could do all that sort of lightweight stuff you're talking about without meeting my phone in my pocket. Yeah. Yeah, you just leave the house. Can you imagine? I mean, first of all, no. But also, it sounds wonderful. The idea of your phone as a thing that lives in your pocket that it lets everything else connect to the internet strikes me as much more of the sort of AI gadget future than like these all encompassing new ideas about gadgets. Yeah, I know. It's kind of depressing, but it's also a little bit. Yeah. And it means like all these other companies, humane, rabbit, everybody. else is like desperately trying to get around your phone, just like you said. And it's like, oh,

this maybe keeps the world the way that it is, if that's how we do it more than I would like, but it also just might be inevitable. Yeah. And it's an interesting moment at least. Boy, is it not dull right now. I know. And I'm still thinking about the camera comparison. And cameras never came back. No. They just like mid-range. We found a lot of interesting stuff to do at like the high end. And I think about like the Fuji series that people really love. Like, I feel like if you want like a good camera, your options have never been better. But if you want a pretty good camera, they don't exist because you have a smartphone. Yeah.

You already have it. Yeah. And I wonder maybe that's where we're going to go with AI stuff. And we're a ways away from the AI gadgets being any good. So maybe we're going to be in a real sort of valley with that stuff for a little while. It feels like. Yeah. We're going to find out.

But luckily phones are getting better. So I'll take that. Yeah. Yeah. Phones are sticking around. I think. And you've given me yet another reason to buy a flip phone, which is all I really needed from you was more reasons.

I took that Motorola out of the closet and I was like, this is so nice. Just shutting it, just shutting it feels good. I'm going to start buying T-shirts with the pocket on the breast, and I'm going to fold it out and just stick it right there and chat with it all day. It's going to be amazing. Yeah, it looks super cool. Yeah, I believe that. I bet you made a lot of friends and everybody thought you were super normal and cool all the time.

I was taking pictures of it for the article and my husband walked into the kitchen and just slowly backed away. I don't understand what's happening here and I'm not going to question it. That's the correct way to react to AI of all sorts. Yeah, probably. Just back away. All right. Allison, thank you as always. Thank you.

We got to take a break and then we're going to talk about. We'll be right back. Support for this show comes from Shopify. Starting something new isn't just hard. It can be really scary too. So much work goes into this thing that you're not entirely sure will even work. But here's a better thought. What if it did all work?

What if your instincts were actually right all along? Shopify wants to help you get there. They're the commerce platform behind millions of businesses worldwide and nearly 10% of all e-commerce in the U.S. From established brands like Allbirds and Heinz to companies just getting started. Their design tools make it simple to create the exact online presence you're envisioning, with hundreds of ready-to-use templates available. And with built-in marketing tools, you can launch full email and social campaigns

in just a few clicks. So you can connect with customers wherever they are. It's time to turn those what-ifs into with Shopify today. You can sign up for your $1 per month trial today at Shopify.com slash vergecast. You can go to Shopify.com slash vergecast. That's Shopify.com slash vergecast.

We're back. While we're on the subject of AI, there's one company that we don't talk about much as an AI powerhouse, and we probably should. That company is meta, and it's been on a tear with new AI products recently. It added multimodal AI to the smart glasses that it makes with Rayban. It launched a new model called Lama 3 that it says can keep up with Gemini and ChatGPT and the rest. It added meta-a-I to like every one of its apps. If you use any meta product, and statistically, if you are a person on Earth, you do.

you basically can't avoid meta's AI anymore. There's a big company-wide move happening here, and it feels like meta, which was all in on the metaverse, not that long ago, might be all in on something else now. Alex Heath on our team knows this situation better than anybody, and so he's here to help make sense of all of it for us. Alex Heath, welcome back. It's good to be back on the People's podcast, David.

Yes. Who cares what the judges think, Alex? I don't care about them. That's a Webby reference. So there's a lot of stuff going on at meta, and basically you've been reporting out a bunch of puzzle pieces over the last few weeks of what's going on there. And I want to see if we can make sense of them all together. But let's start with last week with the big Horizon OS announcement. Explain to me what meta is up to with Horizon OS and trying to open up the quest universe to other people. Yeah. So meta's been doing headsets since before it was meta for a very long time, 10 years actually, when they bought the, Oculus startup that then became Quest later. And what they did recently, which I think it actually makes a lot of sense in the bigger picture of Meadow, which I know we'll get into, is that they decided that the OS that they've built for the Quest, which they've now called Horizon, and we

are approaching Google-level naming confusion here, David, so asked me to explain if it's confusing at any point. But they are taking the OS that they've built over time for the Quest. and starting to license it out to OEMs. So basically replicating the Android playbook that Google used to become the number one player by user base in mobile. And they think a similar thing could happen with headsets. It's very early.

So all they've announced is that Lenovo is going to work on a custom headset using their Horizon OS. And Asus is also doing something. And then there's going to be less custom, but I think more just special addition with some, you know, bundled controllers Xbox-themed Quest. And I think what they've decided is that they want to be the general purpose

headset maker, so making kind of the state-of-the-art thing that could be used for a lot of different tasks on your face. And they want a wide network of OEMs to basically invent all of these more niche, narrow headset use cases, whether that's something for work, something for gaming, something for, who knows, construction, whatever. The idea is just let a million flowers bloom here. That's really interesting. So yeah, my next question is going to be,

what do you think this means for meta's hardware plans? But that's a really interesting way to think about it, that actually this is meta saying, we want to build this sort of all-purpose headset, but then let lots of people build lots of other kinds of headsets, which I think actually makes sense. Like, we certainly have not arrived at the perfect hardware for a headset. So you might as well let lots of people try lots of things.

No, and the quest is still pretty small. I mean, it's approaching console level yearly. volume in terms of sales. I think they've sold a little over 20 million quests to date. Now, how many people actually use those quests on a regular basis is a whole other story, right? They were giving them away basically for a little while. But yeah, I think even with this market being so young, they see the potential for others to come in and help them grow it. And I think crucially for meta here with the OS is that it comes with Horizon the social network. So yes, Horizon OS also has what has been called Horizon the 3D kind of Sims meets Roblox social network that we have relentlessly made fun

of here on the verge. I was going to say, which sucks. To be fair, they have squashed a lot of the bugs and the avatars look better. But yeah, it's still pretty rough. But that comes bundled with this OS. It kind of reminds me of when Google was licensing Android. I mean, that's still how they do it. That really to get the best experience, you've got to have Google's search.

based into Android, and this has actually gotten them into some anti-competitive, you know, regulation trouble over the years with how they bundled things. But meta's kind of doing the same thing here. They're bundling what is their actual way that they see monetizing all this metaverse stuff over time, which is their social network, which will be monetized with ads, of course, and commerce, assuming that enough people eventually start using it. Interesting. So I have been trying to think what is in this for meta, because on the one hand, I think meta's assumption would be that the more people use any kind of immersive headset stuff, the better, right?

Like, we're still very much in a sort of rising tide lifts all boats world. Then if you're meta, you're like, okay, do we want to be the ones who run the store? And we just take a commission on everything. Everybody buys on every headset and that's something. Do we want to be the ones who like charge, do we want to charge a licensing fee and go like the windows way and make all of our money that way? But you think it's actually the big galaxy brain play here is. actually a social networking play? I just think that's how they're going to make money off of it.

They've decided that they won't be making money on a hardware margin like Apple. And, you know, they're leaning into what they're inherently good at, which is like scaled ad-based attention services. And I think they think that they can translate that to headsets over time. And so that's why Horizon, the social network is in Horizon, the OS. I'm sure they're going to get paid also just for someone to license the OS, but I don't think they see that as a key. And so, moneymaker. And I know I've been mentioning Android and Google already a lot, but I would say that if you were to talk to Mark Zuckerberg, like I have, he would really rather be me making a comparison to Microsoft because Microsoft, when they, you know, began licensing Windows, OEMs really wanted it. It was actually the leading desktop PC software platform because Microsoft had invented the most useful productivity software at the time, right? Whereas Android was very clearly a reaction to the iPhone. It wasn't so much something that, uh, OEMs were seeking out because of how good it was, it was more of a defensive play. And to Meta's credit, they invested in this category before anyone else at the level that they have.

It hasn't gotten them a lot yet. But I think with this opening up play that they're doing, it positions them better long term, especially with Apple getting into the picture with the Vision Pro. I mean, it is a funny time because there was just this report that Apple is scaling back its Vision Pro plans because it hasn't been selling the way that it hoped. it kind of feels like everything you just described makes quite a bit of sense to me, right? Actually, for like both what meta is good at and also where this space is headed. Building an operating system makes a lot of sense and meta has done this really well.

But I feel like it all still assumes that eventually there will be a real market for headsets. And I don't know that I'm less confident in that thesis than I would have been a couple of years ago, but I certainly don't feel like I'm more confident in that thesis. But you've been talking to Mark, where do you feel like his head is with how this, market is going in general right now. I think they still see these headsets that fully immerse yourself in virtual environments as akin to the future PC market in terms of volume and scale. So you're not talking hundreds of millions of devices a year. You're probably just looking at, you know, low to mid tens of millions. And this is over a period of the next several years. I think

the industry is in single digit millions right now. The AR side, so where the meta rayband glasses are now, but eventually where they want to get to with displays in them, I think they see that by the end of this decade, early 2030, as something that could be approaching mobile in terms of its just widespread usage and adoption. So there's kind of two parts of the same play, and a lot of the headset technology eventually makes its way into the AR technology.

It goes back and forth. And, you know, we saw that with the Quest 3, right? It has the mixed reality component that you're also seeing in a lot of AR devices. Yeah. And I guess that also makes the social emphasis make sense, because if you're trying to build sort of from both directions, one thing that will flow kind of across that whole spectrum all the way up from smart glasses to like really high-end headsets is just people want to talk to each other and hang out. And so if meta can build that stuff for that whole spectrum, it can sort of win no matter how long it takes us to get to the final thing. Yeah, I mean, they see this hardware naturally as like an inherently social medium.

I think that's a bit of a dichotomy when you think about strapping something to your face. There's something inherently antisocial and putting something on your face between you and other people. But it's a pretty different approach to what we've seen at least Apple do to date with the Vision Pro, right? They're really treating it as more of a general purpose computing platform. They're not leaning into social. The only social thing they have, the FaceTime personas are pretty creepy based on my experience. They're so creepy. So I think meta recognizes where it can play.

to its strengths, and that's in building social experiences at scale and being this kind of quasi open player that is the Microsoft to Apple's Apple. Yeah, let's talk about the open player thing, because there was also, you know, this push to Lama 3, the new large language model, and their meta's trying to be open about how it approaches a lot of the big language model stuff. It's very strange to me that the company that built Facebook, which you had to be logged into in order to see anything and was so famously like its own walled garden and made a ton of money by being that walled garden that people came into and spent all of their time into, has now made

this grand pivot into being like a good citizen of the internet and an open player working with everybody. And they're talking about this with federating threads. They're talking about this with the AI stuff. They're talking about it with the horizon stuff. Like what happened here? Is it like Mark Zuckerberg got like really into fighting and was just like, I love the world now. Like what what happened here? I don't know if fighting would get you into being like, Embracing open source. That's an interesting theory. He just has somewhere else to put all of his rage. So now he's happy at work. I don't know.

I would challenge a little bit of your history retelling there. I mean, you remember the Farmville era. Sure. They actually made a huge bet very early to become a developer platform. So yeah, they've been a walled garden in the sense that they've pretty tightly controlled the walls around their apps. So it's not like, you know, the code to Instagram is open source and it never will be. But I would say they have historically actually, even on the hardware side, thinking about very early, you know, Oculus Rifts were done with companies like Xiaomi, Samsung, Lenovo. So they have a history of this. I really think that meta is becoming a kind of new Microsoft. And it makes sense because Zuckerberg has seen Bill Gates as a kind of a business idol and mentor personally for a very long time. And it makes sense.

I've heard him talk about Microsoft's influence on his thinking over the years, and he's talked about it with me recently. In what sense? Like what about Microsoft do you feel like sticks out to him is appealing? I think the way they leveraged their expertise in software and then licensing that software across different players to really just build a true platform. There's this old Bill Gates quote, and I'm going to butcher it, but essentially he said, you know, you know you've built a true platform. when the total value of what has been built on top of your platform is greater than the platform itself, right? And I think that's what Meta's hoping for on the headset side. And especially on the AI side, you know, with Lama, for example, I just interviewed Mark for

the Lama 3 release. It's interesting what they're doing because they're spending billions of dollars on compute to build these cutting-edge models that compete with OpenAI, Anthropic, and others. And they're just giving them away. And that seems counterintuitive for a company like META to do. But when you think about where they're kind of strategically positioned relative to the rest of the players here, they don't have their own cloud business, right? So they're not selling access to their servers. And their business is building consumer social products that monetize attention.

And rather than boil the ocean, I think they're realizing that's where they can play to their strengths. And the history with open source is if you get a bunch of people to build on your framework, that you've open sourced, guess what? It makes the technology you build more pervasive and value flows back to you in the sense that the community helps improve it, right? And Zuckerberg said something on a podcast recently where if the developers using Lama 3 can help them lower their costs on inference, which is the cost of like running these models over time, even just slightly, it will make up for the entirety that they've spent

on training all future versions of Lama just in terms of the cost of what they need to run this stuff in the cloud in their cloud. So it's a pure infrastructure play that I think Mark has correctly identified they're in a unique position to make. And it creates more competition because Lama 3 can be used by other developers. It forces OpenA and others, I think, to do more in open source, which I think we'll see over the course of this year. And, you know, META has a long history of doing open source, actually, on the, kind of just regular coding side. That's true. It makes sense for them.

And it's, again, I just think we're seeing kind of like the rise of perhaps the next Microsoft in terms of how they're attacking each of these new technologies. Yeah. It seems smart too because right now, most of those other companies you just named are all desperately trying to be the next Apple. Like Open AI is very much trying to sort of build a universe inside of Open AI. They like license some of the stuff out, but like it's clear that the thing that they want to do is like they want to build.

They want to make custom GPs and have you use their stuff inside of their stuff. And so many, and like Google is desperately trying to like pull the whole ecosystem in around itself with AI. And you get the sense everybody sees this as a chance to sort of own the whole experience. And meta is going the other way being like, this is a chance for us to like own the whole experience everywhere. Which actually strikes me as very smart in this moment. Yeah. And, you know, it's, yeah, I mean, meta should maybe be called Open AI and Open AI should probably find a new name at this point. I mean, Open AIsn't open at all, really.

Yeah, it's interesting. I agree with you. I think for a while Meadow was trying to have its cake and eat it too on the reality labs quest side. And I kept pushing them on this that, like, they were trying to be too much like Apple in the sense that they were doing all of the hardware, fully vertically integrated and the software. And I was like, when are you going to, if you want to be the open player, you've got to start licensing some of this stuff. Because Apple, you can't compete with Apple at its own game, right? Like Apple is the best at this.

And I think they finally realize that and are embracing it. And it's what Microsoft did. And it's smart. Yeah. No, I think that's right. And it also seems like it's given meta real momentum. Like my sense is this, the Lama 3 launch and just the sort of meta AI percolation around all of meta's different apps really worked very quickly.

Like it kind of, it became a big player pretty fast in this space. I think in part because it has been. sort of out in the ecosystem much more rather than just saying, you know, here is an AI you can talk to inside of Instagram, which I think everybody would have kind of just sniffed at and moved on. It feels like meta is able to make these big plays by playing this game differently. They are. I have a framework for evaluating companies, which is like, are you in the show me or tell me phase? And I think meta's been doing a lot of tell me stuff with AI, like the assistant is going to be everywhere, right? That was the recent news. They're putting it in the search box of Instagram. Which I hate, by the way. Yeah. So

that's where I'm going. Like, they're putting it everywhere and now the real test is do people actually want this? Does it make sense to have a chat GPT-like experience in your Instagram or your WhatsApp where you're used to just talking with people? And Zuckerberg really thinks we're all going to be interacting with synthetic AI type personas like people in conversations with people. You know, like you can add the meta AI to a WhatsApp thread with other people to give recommendations or something. So they've got to prove that people actually want that. The running joke inside meta before they put the AI everywhere was that like no one was using this. It was something you had to kind of find. It was pretty buried and not even like people at Meta were using it. So now that it's got

Lama 3 under it, which is pretty competitive as a model, and they've got Google results in there and Bing results in there, which are the only chat bot that does real-time search from both those. It's got a good shot at capturing a lot of the attention from people who have maybe never tried chat GPT or Anthropics Claude. Because, you know, for people who listen to the show, they know all these names, but you've got to think that the boomers sharing, you know, spaghetti Jesus viral generative AI memes on Facebook, they've probably never used a chat bot before. There's millions, hundreds of millions of people who have never used one of these things. And I think Zuckerberg's bet is just like he did with stories and Snapchat and reels with

TikTok. He's going to introduce this new format, this new way of interacting with the internet to more people than anyone else possibly could because his true inherent lever. and competitive edge is over 3 billion daily users. Yeah, I think that all strategically makes a lot of sense to me. And I think as a product, it works in some places. Like, I actually think putting the generative AI stuff inside of Facebook has terrifying implications. But as a product makes some sense. I hate that they have replaced the Instagram search bar with meta AI.

I just want to find a video. I don't want to talk to the AI. I want to search for a video. Like, it's this idea that we have decided AI and search are the same thing. drives me insane. And I wish Mehta would stop doing it. They might. You know, if there's enough uproar, they might. They did this like overnight. Meta's notorious. They invented the kind of ruthless A-B testing in Silicon Valley. So in a normal situation, they would have tested this with some, you know, in different markets for weeks on end, months on end, figured out if people

actually wanted it. They just flipped it on for everyone overnight, which means it's a Zuck level, top-down. We are betting the farm on this as a modality. And now he's got to be proven right or wrong. So that was actually one of the things I wanted to ask you about is like connect the AI push to that sort of big picture meta idea for me because I think I saw AI for a while at meta as kind of an end around metaverse play. Like AI got really cool right as the metaverse stopped being cool and they could build AI stuff that was actually metaverse stuff but they just had to call it AI stuff. But now it seems like there is this idea that some part of that, like, engagement-based community stuff going on inside of META's products,

they also seem to think AI as part of that in a way that I can't quite wrap my head around. Like, as you talk to people like Zuckerberg, like, what is the sort of galaxy brand vision for AI right now? It's this kind of scary vision that I, frankly, I can't decide how dystopian it is, but it's this idea that there are just going to be thousands, if not millions of AIs that we chat and interact with in our feeds and our messaging threads that are indistinguishable from people and that creators that Kylie Jenner on Instagram will have an AI version of herself, that she's trained with certain parameters that means that all her fans can interact with

quote-unquote Kylie Jenner, right? And they're doing some, I think, kind of clever early engagement hacks with meta-AI. There was an example in, I think, 404 media where it was really cringe, but it was an interesting way to kind of juice human engagement on Facebook. There was this group full of parents with special needs kids, and there's this new feature that Meta's trying in a group where if someone doesn't comment under a group post for, I think, like, more than an hour, meta-a-I will comment with its own just prompt to try to spark conversation with the actual human beings. And in this example, Medai, made up that it had a kid or a couple of kids with special needs and it was pretty creepy and hallucinating and all that. But you look at the

thread, it got a lot of humans commenting underneath it, right? So it's like you're using AI as the beginning flame to get a bunch of people to actually talk to each other. I like the idea that the AI's job is to just say some wild, insane stuff on every post on the internet just to get the fire. And we can all just rage at it. And yeah, I think that's part of it. But yeah, it's all of these things. And it's also the visual generation stuff. There's going to be multi-modality.

So generating video, you can already generate images. They upgraded it where it like generates the image as you type. Have you played with this? It's really cool. Yeah, it's cool. And, you know, obviously they're still figuring out the like things that will and won't do that are cringe, you know, the Gemini type diversity scandal stuff. But I think we're still in the very early innings of this. It's the, and now it's the show me, like I was saying.

They've got to show that people actually want all this stuff. That's a very broad way of thinking about it. Because like the way you were describing it earlier, right, is that there is that sort of unifying sense of what meta is very good at, which is basically building products that people engage with at scale, right? Like you can boil a lot of the things that meta does down to that. And the AI seems like it's partly that, but it could also be lots of stuff. And I get the sense that meta is maybe just going to explain.

that stuff to see what happens, but also I think, like, a lot of companies is going to be prone to being distracted by whatever shiny thing you can do with a language model and just shove it inside of apps and think people will use it. And I feel like meta more than most might have a chance to, like, really ruin some of its products by overextending AI into it. The thing is they're a large, mature publicly traded company. They won't torch everything for this. Like, if they see that it's tanking key metrics that will impact, you know, the next earnings. they'll roll the stuff back. I mean, I would point out they torched a lot of things in the name of the Metaverse for a while

there. Like, this wouldn't be totally unheard of. Well, kind of. I mean, it's all still separate, right? Like, yeah, so it's a lot of sunken money into it. But nothing they've done on the quest has, like, impacted your experience as an average Instagram user. That's fair.

They're not fighting Facebook. Yeah, that's what's different about MetaI is that they're literally putting it everywhere. So they're in the, like, like you said, shove it everywhere and find out phase. and we're all about to find out. Yeah. So what's your sense from talking to Mark about this Google stuff, by the way? This is the small detail of your most recent story with him and this whole meta-AI rollout that I found the most fascinating.

They're very excited about having Google integrated into meta-AI search results. Yeah. Which just seems bizarre to me that that is a thing that exists. What do you know about what's going on there? I thought it was interesting, too. Reading between the lines, you know, I asked him, What is the deal here? Because Google has not licensed its real-time search results, any of these chatbots yet. And I was like, are you paying them? Are they paying you? He basically said that meta is paying Google, but it's not, quote, a ton. I don't know if a ton is billionaire who owns half of Kauai a ton or normal a ton. I'm sure that will come out. But yeah, it's smart because that's one of the main problems with these chatbots, right? Is their recency isn't very good?

make up events. I was using meta AI without the Google integration. And it like used last year's Coachella lineup when I asked it who was just performing at Coachella last weekend. So you've got to fix that. Real-time search can do that. Zuckerberg hinted to me that Google's building a model. And I don't mean AI model, but I think a business model around this, around licensing search to these chatbots. So it makes sense because that is Google's bread and butter still. I thought that Google would keep that for Gemini, that that would make their chatbot experience, you know, more unique and their leverage there. But I think this means that we may start seeing them license it elsewhere as well. That's fascinating. A, sucks for perplexity. You had a good run. Perplexity. Great job. But B, the tension there just for Google, this is a total aside, but the tension there for Google between like search has made it all of the money in the known universe and also they're betting a lot of that search money on AI working.

But then you look at this and there is just a giant money faucet licensing this data to chatbots, all of whom would want it and pay for it. Fascinating. The like the business machinations between those two things are going to be really interesting to watch in the next couple of years. I agree. I'm eager to hear Google maybe talk about that more at I.O. soon. So big picture. As as the person who broke the story of Facebook renaming to meta, I'm curious how you feel about meta as a metaverse company at this moment. time. Like, there is part of me that is like rewind whatever five years and knowing what he knows now,

Mark would have renamed the company like artificial or something instead of meta. Like, is the metaverse still the thing when you talk to meta and Mark? No, I mean, it's not the thing they want to talk about the most. But I don't know, man. I think you can walk and chew gum at the same time. And I still think 10 years out they see more of their optionality on the future being on the metaverse stuff because they've hit a play. plateau with Facebook, it's not getting young people back eventually, and they talk about this pretty frankly internally, literally everyone's just going to die, right? Because like if you're not getting new users on your platform, your platform's in terminal decline. Instagram is their next Facebook,

right? It will be, we're all going to be on there in probably 10 years and it's going to feel just like Facebook does now to the younger generation. Whether they can buy the next thing to keep that whole cycle going 40 years from now, TBD, unlikely. But this whole push to control the next computing platform or at least have more say in it so that Apple just doesn't invent the whole next, you know, headset modality. I think it still makes sense. I mean, I don't know, I've been playing with the Rayban smart classes with multimodal AI on them. You have it too, right? And, you know, it's not great. It's wrong half of the time. I will say I'm impressed with the speed of it, but, you know, it just makes stuff up still a lot. But you can see when they iron out

these kinks, like, oh, the basis of this makes sense. Like, I actually do see over time as the tech gets figured out a lot of people wanting to use technology more through their eyes and through their kind of egocentric view of the world. And you're covering this a lot in the wearable AI side with, you know, humane and rabbit and all that. But this is all kind of converging over time. This generative AI meets wearables into what I think is a pretty compelling next

computing platform. And if anything, I think meta feels like AI has. only kind of made that more apparent and actually pushed off the need to have the whole display stuff that we always hear about that we're going to have all these holograms around us and it just keeps getting pushed out and out and out. And now maybe they think like, oh, maybe all we really need for the next five-ish years or so is compelling AI in a wearable form factor. So yeah, I mean, all these big tech companies, they overrotate on the topic de jour. And that's generative AI right now. I don't think they're changing the company name again. I think they're still investing a lot of

money in reality labs and we'll have a pretty wild product next year that I've scooped, which is the Raybans with a display and a bracelet that does EMG kind of neural interface control. So imagine like typing and clicking and pointing with, I guess, just basically thinking. So we're about to hit a lot of really interesting new stuff there. It's definitely like they over rotated on it like way too early, but that's like the story of meta. They spent, you know, whatever they spent on Oculus 10 years ago and where has it really gotten them in terms of ROI even now? So them like going too hard, too early is nothing new.

That is fair. That is what they do. And I will say to their credit, they both like built the way too early thing and the kind of right on time thing. Like the, I think the smart glasses were the correct move for now. The big push after the headset in the metaverse is like way too early, which is a classic Microsoft. move. So good job again, suck. But they're doing a surprisingly good job of both sort of running those in parallel and pushing them towards each other. And I think that is like, if you can keep that up and you can afford to keep that up until they actually hit each other, that's it. That ends up

being a pretty powerful place to be, as opposed to somebody like Apple, who now has this massively overpowered thing that it's going to have to figure out how to like pull back into actual mainstream interest over time. And I think that's harder than it gets credit for. Oh yeah, for sure. It's hard, but it's coming from different ends. Totally. All right, we've got to take a break. Alex, thank you as always. Good to have you back.

Yeah, thanks for having me again. Support for the show comes from Upwork. The days of doing it all, all by yourself, are over. There's no romance and burning out while you're trying to scale. Instead, you can check out Upwork. Upwork helps grow your business by giving you fast access to specialize talent across more than 125 categories, so you can fill skill gaps,

launch projects faster and scale without committing to full-time headcount. And finding the right talent is easy. You can browse profiles, review past work, and get help scoping the role so you can get started quickly. Seriously, you could connect with the right freelancer in just a few hours, especially when you sign up with Business Plus. Their AI-powered shortlisting pairs you with the top 1% of talent in under six hours. No endless searcher required. You can visit upwork.com right now to post your job for free.

That's upwork.com to connect with top talent ready to help your business grow. That's upw-w-rk.com. Upwork.com. All right, we're back. Let's get to the hotline. As always, the number is 866 Verge11. The email is Vergecast at theverge.com. We love all your questions.

We try to answer at least one on the show every week. This week, we have a question from Josh in Denver. Hey, Virchim. This is Josh from Denver. I have a question that runs the risk of sending Eli on a legal rant, but I have to ask it anyway. It seems like Microsoft is back on the path that it was in the 90s. They got it in a lot of trouble, specifically,

that a lot of anti-competitive stuff is making it into Windows. They're making it very hard to switch browsers again. They're making it so that the default edge browser is kind of like popping up constantly, even if you've selected it, not as your default browser. Same thing with the cloud storage. And I wanted to know, number one, why do they think they can get away with doing the exact same thing

that got them in trouble in the 90s? I mean, I understand the market share for their browser has dropped, but it's still the same anti-competitive practice. And two, why aren't we putting a stop to it legally? Anyway, thanks. Hope we can answer the question. All right, you ask for a Nelai rant.

You get a Nelai rant. Neil Pettel, hello. Hi, how's it going? So we've talked sort of obliquely about Microsoft's as part of all of these big antitrust things. So I figured this is useful and to just hit head on. What are your thoughts? What do you make of this question?

I don't know if you're going to get so much illegal rant for me is just a rant about the fact that nothing can displace Chrome. Like governments around the world have been trying to displace Chrome for 10 years more. And nothing they try works. And I think they're kind of at a point where they're like, I don't know, maybe Microsoft can do some stuff. And it doesn't work. And I think that specifically with the edge browser and the prompts to make edge the default and all that stuff, I think the market is very clear that it hates it. People don't like it.

Whenever we write about it, people are like, I hate this, this is stupid. And then it doesn't work anyway. And I suspect that regulators in Europe and the United States are kind of looking at that, being like, fine, it's just not working. Right. If you want to destroy your own reputation trying to make this work, like have at it. When I interviewed Jonathan Cantor on Decoder, the head of antitrust at the Department of Justice, he said his framework for decisions was hips. He's called it hands on hips.

High impact programmatically significant. And I think you just look at that equation, you're like, this is low impact because it's not working anyway. That said, Microsoft is not getting away with it totally. They rolled out teams. They bundled teams into office that basically put a ton of pressure on Slack. It forced Slack into selling itself into Salesforce. The European regulators looked at that very sternly, and Microsoft unbundled teams from office. And they have now done that globally because I think they saw the pressure was rising globally.

So I think there is, I think there are some places where Microsoft is responding to the pressure, but specifically when it comes to irritating everyone around Edge. And the regulators are like, yeah, that shit doesn't work. Right. Yeah. My read on this was that there are interesting Microsoft antitrucks questions, probably around Windows and probably around Office. And especially as it shoves co-pilot AI stuff into all of that, there might be interesting questions to come up there. But you just cannot make the case that Edge is powerful. Like, you just can't.

And I think you're right that everyone has been desperate for there to be a meaningful competitor to Chrome for a really long time. Mozilla has been running around saying, you know, Firefox is the thing. We're going to do everything differently. All these browsers now even run Chromium. Like, Mozilla is one of the few companies that isn't using the blink engine with the Chromium project underneath all of the browser stuff. Like, Edge is just another skin on the same stuff that Chrome runs. And so it's all really mealy and weird.

But for everything that Microsoft is doing, Edge still has like teeny tiny market share. And I feel like, if I'm a regulator, I'm like, aw, you're trying so hard. This would be sketchy if only it were working. Yeah, I mean, the regulators in Europe basically imposed the same thing that Microsoft is doing, which is like, would you like to try another browser? And everyone was like, no. And so Microsoft is welcome to try to decrease the market share of Chrome, which is the problem that those regulators have been focused on.

They're just not doing a good job. The other thing that I think is interesting and kind of virtualasty is the reason Microsoft was terrified of NetScape back then was they were very worried that the application, model for desktop computing would leave the Wintel monopoly. So they were doing anti-competitive stuff to protect that monopoly, protect the dominance of Windows, make sure Windows was the place where all applications were deployed. And they saw correctly that deploying applications to the web would break that model. And that 100% happened.

Yeah. So the fact that Edge runs Chromium is because if you want web apps to run well, it is better to standardize the browser engines. so they all work in the same place in the same way. And that's how all apps are deployed now. And so it's just weird that we've come full circle, but actually the thing Microsoft was the most worried about absolutely happened. And what the browser Chrome is,

like the way the browser looks and works and what default search service it uses versus what is the browser engine, like that war is lost. Yeah, and I think, I mean, it's a weird place to be in because my next question for you was going to be, should we be talking more about Microsoft in an antitrust way? because again, it's either the first or second biggest company on Earth. It is clearly as powerful as anybody. It's just so much more diversified in that way

that it doesn't have kind of the one thing that feels like its thing the same way. But I guess it would be Windows, but Windows has been completely commoditized. Like I think desktop operating systems in general have been made totally irrelevant by web browsers, essentially, in the internet. Like most people, you buy a laptop, you download a browser,

and that is your portal to everything else. And so maybe if I'm an antitrust regulator, there's just no upside and caring about desktop operating systems anymore, even though Microsoft is like by a mile the dominant player. But there's a handful of things that are worth considering. But even there, you can see there's enough competition. So the one I would just throw right back at your example is games. Sure. If you want to play video games on a PC, you're going to buy a Windows PC. And you are really reliant on Windows and all of its graphics support and the ecosystem of graphics cards available for Windows.

There are like three people with Steam running on their Macs who are like really upset that you said that. Yeah. And they're running it at very low frameworks compared to any Windows PC. Right. And so there's a little market there that feels very tightly controlled. But then you look around and you're like, oh, there's all these like cool handhelds that run Linux and Steam. And you can just see, okay, there's like meaningful, like the game developers themselves know they don't want to be totally stuck here forever. So they're creating new markets and there's new distribution.

And Windows itself is not even Microsoft some particular advantage in gaming. Xbox, its other big consumer gaming brand is not the winner. Like Phil Spencer will happily tell you they're not the winner all day and all night. So I think it's hard for a regulator to point at some big consumer harm in gaming. They couldn't even make the case to keep Microsoft from buying Activision. So I think that's challenging and plus that space does feel competitive. And then everything else kind of doesn't touch consumers that Microsoft does. Slack and Teams and it got the scrutiny and they had to stop bundling teams.

What else is there? Everything else, all of Microsoft's other business is a bunch of enterprise business. It's Azure. It's cloud computing services. It's security to some shaky extent. Like that's competitive now. I think it's hard to put together the political case. Like here's a real problem that people are complaining about unless it's something a consumer can see

or lots of people can see. In the way that blue and green bubbles is, like, it has that, like, visceral thing that you can point to. Microsoft doesn't have any of those. Yeah, and then when it does it, it's so blatantly annoying. Like, would you like to try Edge that, like, the market just sort of corrects for it by being like, no, I will now I will never do this. And that's fine.

Yeah. So you don't necessarily think any of this is even necessarily coming for Microsoft. Like, I think the two I've been waiting for are YouTube scrutiny and Microsoft scrutiny. And I think YouTube scrutiny is probably still coming. I'll take the opposite. I'll take the other end of that bet. Oh, okay.

YouTube is just a blind spot for everyone. We treat it like the ocean. Like, there it is. Forever changing and always the same powerful. It will be there long after we're all dead. Yeah. Yeah, it's like there lies the ocean. It's like, who knows what vast mysteries exist in the deep.

It's just an absolute blind spot for everyone. No one treats like a business. Microsoft, I think they deal with OpenAI, how they will integrate those products. the exclusivity around that deal, whether they sort of acquired Open AI without acquiring it, I think there's a lot of scrutiny left to come there, particularly as that market heats up. But I don't know what shape it will take, but I think people will understand that that's weird and there's money and power and characters over there before anyone even tries to understand YouTube.

That's fair. And that is, to your point, that, like, directly regular human relevant thing. Like, if you want to start talking about what happens to chat GPT, like now you're back squarely in regular consumer territory, which is where that goes. Yes. Did Microsoft illegally invest in whatever, blah, blah, blah with Open AI to make these guys rich? Sure, you can tell that story. I think it's very hard to be like, we're going to mess with YouTube. It's just the ocean. That might be the best metaphor for YouTube I've ever heard.

I very much enjoy that. That's not what we came here to do, but I'm glad that's really well. One day, someone's going to be like, huh, YouTube. seems a company. And we'll see what happens after that. Yeah, fair enough. All right, Josh, I hope that helps. Nil, thank you. Thanks. All right, that's it for the Vergecast today. Thanks to everybody who came on the show. And thank you, as always, for listening. There's lots more on everything we talked about, from Rabbit to Allison's experiment, to Alex's coverage of all things going on at Meta at theverge.com. We'll put lots of links in the show notes,

but as always, read theverge.com. It's good website. If you've thoughts, questions, feelings, or other AI gadgets you want me to try. You can always email us at Vergecast at the verge.com. Call the hotline 866, verge 1-1. Thanks again to everybody who voted for us for the webbies. We're still riding high on the victory there. Thrilled that you all listen to and like this show. It means the world to us. Send us all your thoughts and questions, ideas, everything. We want to hear them all. Call the hotline. Send an email. We love hearing from you. This show is produced by Andrew Marino, Liam James, and Will Pore. The Vergecast is Verge production. Part of the Vox Media podcast network. Nelai, Alex, and I will be back on Friday to talk about presumably more AI gadgets.

some iPads coming up next week and everything else going on in tech. We'll see you then. Rock and roll.