Why Microsoft bet on Surface | The Vergecast - video Dailymotion

The Verge

Today on the flagship podcast of open smart home standards:  Microsoft’s Pavan Devaluri, leader for Windows and Devices, joins the show to discuss the future of the AI PC and what’s next for Microsoft’s hardware. The Verge’s Jen Tuohy and David Pierce discuss the latest updates in the smart home world in a segment called “Does Matter matter yet?” David answers a question from the Vergecast Hotline about AI-powered search engines.

Transcript

00:00:00Welcome to The Verge Cast, the flagship podcast of open smart home standards.

00:00:04I'm your friend David Pierce, and I am cooped up inside.

00:00:07It's a holiday weekend, but it's like hot and rainy all at the same time, and it's just like

00:00:13a nasty time to be outside. But the upside of that is it means I have an incredible,

00:00:18built-in excuse to just spend three days playing video games.

00:00:22That is essentially what has happened. I don't know if I've mentioned this on the show before,

00:00:27but the game over probably the last decade that has eaten up the most of my time

00:00:32is FIFA, the soccer game. It's not even called FIFA anymore. It's EA Sports FC now,

00:00:37but who cares? It's FIFA, and I play arguably too much FIFA. I'm pretty good at it in a way

00:00:43that I'm not good at a lot of the battle royale games. I just get crushed at Fortnite and Warzone

00:00:48and all of those, but I'm pretty good at FIFA, and I have become alarmingly into it, especially

00:00:54over the course of this year. And now it's May. It's almost June, which means it's the end of the

00:00:58soccer season, which means it's almost the end of the FIFA season, and I'm too deep in it at this

00:01:04point. I am in the subreddits. I'm in the trading discords. I spent 20 bucks on a guide to trading

00:01:10players inside of FIFA to get more coins to buy more players. It's a lot. I'm deep in it. It's a

00:01:16whole thing. I can tell you who all the leakers are. If you ever want to talk about FIFA and

00:01:20compare Ultimate Teams, get at me. It's the only thing I ever want to talk about.

00:01:25Luckily for most of you, that is not what we are here to talk about on this show. Today, we're

00:01:30going to do two things. One, we're going to talk more about what has gone on with Microsoft the

00:01:34last couple of weeks. It was Build, it was the Surface launch, and we have the person who runs

00:01:41Windows and Surface at Microsoft, Pavan Davuluri, on the show to talk about both the beginning and

00:01:48end of this project that Microsoft has been on for a long time, and what he sees coming for the

00:01:54next decade, too. Super fun, really interesting. Then, Gen 2 is going to come on, and we're going

00:01:58to do a bit of a round-robin of smart home news, because a bunch of stuff has happened, some of

00:02:03which I think is actually bigger than I realized, and we're just going to dig into all of it and see

00:02:08what's really going on. Then we're going to get to the hotline. We're going to talk about search

00:02:11engines. Search engines on everybody's mind right now. All that is coming up in just a minute,

00:02:16but first, FIFA. It's champions. If you know what that means, you know that's a big deal.

00:02:24If you don't know what that means, stay that way. It's for the best, but anyway, I got to go play a

00:02:28bunch of FIFA for some reason that feels very important to me right now. This is The Verge

00:02:33Cast. We'll be right back. Welcome back. I know we've talked a lot about Microsoft the last few

00:02:40episodes, but it's because I think this is a big moment for the company. It's pushing really hard

00:02:45on AI everywhere across the company, and especially across the products people use.

00:02:50It just rolled out these new Copilot Plus PCs that Microsoft seems to think might change the way we

00:02:55use our computers forever. And I mean, Microsoft is the most valuable company on earth for a reason,

00:03:01right? But what happened over the last week actually feels to me like both the beginning

00:03:06of a new AI era at Microsoft, and in some ways, the end of a decade-long project inside of the

00:03:12company. And I wanted to hear from the inside how that moment feels, and how we got here. So I

00:03:18called up someone who knows. My name is Pavan Devaluri. I am leader for Windows and Devices.

00:03:24Pavan currently runs all things Windows and Surface inside of Microsoft. He's been in that

00:03:29job for a few months, but he's actually been at the company for more than two decades,

00:03:33and as a result has seen just about everything in the Windows and Surface world in that time.

00:03:38So we started our conversation by going back in time, all the way to 2012, to the first ever

00:03:43Microsoft Surface and Surface RT. Back then, we weren't talking about AI. We weren't even really

00:03:49talking about laptops. That was the time when everybody kind of thought PCs were dead and phones

00:03:54were going to be the only thing. You remember that? Like, this was supposed to be the end of

00:03:57computers, not the beginning of some grand project for Microsoft. But you can draw a line from that

00:04:03moment in 2012 with the Surface RT, which was bad, all the way to this one without trying very hard.

00:04:10So 2012, let's go back. What was the big bet behind the Surface, particularly the Surface RT?

00:04:17So our core thesis, you know, at that point in time, you're right, we made an ARM bet. The ARM

00:04:22bet was around modernizing the platform at the end of the day and bringing modern architectures to

00:04:28Windows. ARM was a part of it, for sure. You know, instruction sets matter, obviously, for the core

00:04:33operating system and the app experiences. But really, the theory of the case for ARM was a modern

00:04:38system, you know, in its entirety. And so what we wanted was, yes, great performance, we wanted

00:04:45great battery life, we wanted security, all these other aspects of modern platforms in that window

00:04:50of time. And so ARM was a big bet. And as you recall, for sure, Surface RT did bring a new user

00:04:56experience sort of, you know, paradigm to the table, for sure. And, you know, in that window of

00:05:00time, we were seeing confluence certainly across what was happening in the PC space and the tablet

00:05:05space. And in mobile, it was early days of, you know, tablet iteration back then. Yeah, so a lot

00:05:11of those constructs, I think, you know, are still true today, for sure. And the way I think about

00:05:15kind of this kind of way I thought was going to bridge that from back then to today is the ARM

00:05:20journey, as an example, has been a multi-generation exercise for us, for sure. And we certainly

00:05:27learned many lessons, especially on the Surface team between the work we did in Surface Pro X and

00:05:31Surface Pro 9 5G, for instance. And now I think we're in a place where we brought the totality

00:05:37of what was needed to be just building great product for customers at the end of the day to

00:05:41the table. And so I feel much more confident now in terms of the work that we have done.

00:05:46Did you know 12 years ago how long that list of things you just described was? Because I think

00:05:52part of the reason I ask is it feels like, you know, the running joke is next year has been

00:05:56the year of Windows on ARM for a decade now. And it seems like, at the very least, you and

00:06:02Microsoft are substantially more confident that that is the case now than ever. But if you rewind

00:06:0812 years, do you think your sense of the project was actually as big as the project turned out to

00:06:15be? Yeah, this is a great example of the breadth and diversity of the Windows ecosystem. I think

00:06:21our superpower is the diversity of the ecosystem. Which is such a great euphemism in such a good

00:06:26and bad way. It makes everything wonderful and massively complicated all at the same time.

00:06:30Massive, exactly. So you kind of have to get your arms around enough of the massive components,

00:06:34I think, to be meaningful at the end of the day. And that has been, in fact, the reason why we took

00:06:40for this iteration of it, quite frankly, David, I think we could have done individual bits and

00:06:44pieces sooner and then kind of done them on an asynchronous basis. We could have done some of

00:06:48the DevPlat stuff kind of out of band and kind of got those pieces out. We could have done parts of

00:06:53the core OS and then release them as, you know, incremental updates in Windows. But we chose to go

00:06:57ahead and take the time needed to go do the entire thing. You know, one good example of that is our

00:07:02partnership on the Silicon platform. To me, the Silicon platform is a foundation for what, you

00:07:07know, the OSes and app experiences are and what they can, you know, take advantage of. And it

00:07:11takes time to go rebuild the entire Silicon. In this case, our partnership with Qualcomm

00:07:16took us to a place, and quite frankly, with Arm Holdings, to a place where we had to go back and

00:07:20address some fundamentals. And, you know, Silicon design lifecycle takes time when you are rebuilding

00:07:24an entire chip for the exercise. So I think we certainly appreciated the magnitude of it in the

00:07:29more recent years. And I think we decided to take the time to address the completion of it, perhaps,

00:07:34versus doing another sort of incremental take on it. Is it easy then in that process to sort of

00:07:41find the moment where you're, or like identify the moment where you're like, okay, we're there,

00:07:45we've done the thing. And instead of, because my sense was something like this is you could tinker

00:07:49with app compatibility until the heat death of the universe, and never actually solve every problem

00:07:54in every edge case. How do you pick the moment where you're like, okay, we have kind of over

00:08:00promised and under delivered a couple of times, we know we have to get this right. And we have

00:08:04like, what, how do you put that flag in the ground? Yeah, it's a great question. You know,

00:08:08I think, to me, there isn't a single one answer thing here, you I typically, as a team, we

00:08:13collectively look for a set of signals, I would say, in this instance, there were some things

00:08:18that we clearly learned by way of customer feedback, in fact, you know, feedback from folks

00:08:22like you, for sure, looking at our products. And so that, to me, there's some front and center

00:08:27things that, you know, we just had to make sure we were we were delivering on emulator performance

00:08:31was one in such example, it was, you know, kind of sort of apparent to us what the benchmark

00:08:36industry competition, etc. So we have some clear benchmarks that we had to go hit that were

00:08:41kind of markers that we set for ourselves. The second part of it, I think, was we had some

00:08:45expectations of, you know, when we take another iteration doing this, we do have to be kind of

00:08:50meaningful and world class in this context. And so that was kind of a moving target for us. And

00:08:55we had to go as you know, it's a function of time, obviously, and time to market and so on. So,

00:08:58so some of those, we just decided we're going to take a quantum leap kind of in that sense. And

00:09:03then you will have to kind of wait and watch and decide and see if that's a meaningful exercise or

00:09:07not. The other big component is we do, you know, get a chance to talk to customers, we do that for

00:09:12consumers, for sure, we talk to commercial customers, we get a chance to do, you know,

00:09:15iterations and, you know, trials and deployments with them. And so we learn through actual dialogue

00:09:20and data and telemetry from our commercial customers to see if we've addressed the core

00:09:25issues. You know, a great example of that is, you know, on the emulator itself, as much as we did

00:09:29tremendous work on the emulator on this iteration, there are some things that you can't emulate. And

00:09:34so, you know, kernel mode components in Windows and, you know, all the malware type stuff that

00:09:38and VPN applications that commercial customers do rely on, you have to get them native. So some of

00:09:43those things we do, we took time to go get the data, get customer feedback, and, you know, do

00:09:48the trials, do some deployments, get their feedback. And then I think you look at the culmination of

00:09:53them all and then see if you've done enough of the critical mass to decide that you've, you know,

00:09:57kind of met that expectation. The one thing that is kind of a gift for us in this context

00:10:02is the AI stuff that happened. I think the one real inflection point for us, David, was knowing

00:10:08that we could use this opportunity in a more general sense, certainly ARM is a big piece of it

00:10:13for us and the first wave of Copilot Plus PCs are on Qualcomm Series X components, but really using

00:10:19the AI point in time as an inflection for us to kind of go after a bigger vector for what, you

00:10:26know, value we could serve at the end of the day was another thing where, you know, timing figured

00:10:31into our decision making for both value and timing for the exercise. That makes sense. That's actually

00:10:36a good segue into the next thing I wanted to talk about, which is that it feels like to some extent

00:10:39in 2012, the Surface's job in particular was to sort of remind people that PCs matter. Like we

00:10:47were in the era when, you know, smartphones were going to kill everything, laptops are dead,

00:10:51who needs any of this stuff anymore? Like that's the old way. And that was, I think, obviously

00:10:56never the case, but that was, I think, part of what Surface existed to do was continue to, you

00:11:02know, push that along. It feels like the job now seems to have shifted a bit where you're trying

00:11:08to, and I even noticed this sort of as you're on stage talking about this stuff, like you have to

00:11:12make the case for these devices in a new way. And it's not just these things still matter because

00:11:17typing is annoying on your phone. You have to like describe a different way to think about my computer

00:11:24than it has. So I'm curious, give me sort of the big philosophical vision of like the job to be

00:11:29done of a PC in this era that we're just starting right now. Has it changed at all? It has. And I

00:11:35think to be honest with you on that exercise of my own learning and journey, I would say the last

00:11:41four years have been particularly instrumental because to your point, I think people go through

00:11:44that thought process periodically. They're like, you know, what is the job of the PC? And one thing

00:11:48we learned through the pandemic and coming out of the pandemic on the other side was how vital PCs

00:11:54were for people and just how fundamental Windows is in that context of serving the

00:11:59breadth of education customers, through commercial customers, through frontline workers, through,

00:12:04you know, consumers at large. And so I think it reinforced for people in a way that obviously

00:12:09we took a pandemic, but I think in a way it also forced us to kind of go through the introspection

00:12:14of how are we making sure we are thinking deeply about the promise of the platform, the ecosystem

00:12:20that Windows is on, the role for Surface in that context as we go forward. And I think for me,

00:12:26there's two constructs. One, to your point earlier, I think the deep-rooted value proposition

00:12:30of Windows, I think just it is a great reminder. It is kind of a humbling thing and an inspiring

00:12:35experience at the same time for how important and foundational and key Windows is across all

00:12:40of kind of work and play for people and life. As we think about the world going forward, I think

00:12:45for me, my sense and orientation in the space is this AI journey that the world is going through,

00:12:50our industry, at least for sure, is a tremendous gift for us. And, you know, I think it gives us

00:12:56a sense of agency for what is possible on a PC and to kind of reimagine what was possible from,

00:13:01you know, 10 years ago, you know, 2012, for sure, we were not thinking about this in a way that I

00:13:07think is somewhat of a tidal shift for us. To me, what I love about this is, especially with the

00:13:14ability to build an entire system stack, you know, kind of be thoughtful, and this is also

00:13:19part of why this project took, you know, kind of longer perhaps than others would have expected,

00:13:23is it needed us to kind of go through the deep thinking on how do we deliver this proposition

00:13:27in a way that is somewhat durable. I think we're at the infancy of this journey, David, I think

00:13:31there's going to be, you know, iteration and refinement and learning and evolution in that

00:13:35sense. And so, it required us to go think deeply about how do we go integrate AI into our products

00:13:40across the board. It took us down the path of making sure we were building the hardware

00:13:45platform and capability, and that's a statement across the entire Windows ecosystem. We have a

00:13:49whole modern generation of SoCs that I'm actually very excited about that are going to start showing

00:13:54up for Windows. It's going to be great across our partners, because I think that modern platform is

00:13:58what then we can build on top of, both from an operating system standpoint and as a device maker

00:14:04in Surface. But I think, I mean, just to that point, actually, one of the things I think you

00:14:08could have done, and you should tell me if you did do this, because I think it's very interesting,

00:14:12is say, you know, Microsoft's kind of lost in mobile, the AI thing is happening, this is our

00:14:18moment, we're going to relaunch the Surface Duo as an AI device and take over the universe. That

00:14:23is one, like, perfectly rational version of that thing. But you went another way, and you went all

00:14:29in on PCs. Like, even the fact that the Surface Pro itself is the same size as the last one,

00:14:35but is in such a meaningful way, like a fundamentally different device, I think is

00:14:39so interesting. And it's like, there's something about what happens when you put AI inside of a PC

00:14:45that I feel like consumers just have not seen yet. This stuff is so new. But you've had to, like,

00:14:50think ahead about what that will look like and how we'll use it. And I'm sure you've just used

00:14:54these devices longer than the rest of us. So, like, what is it about the size and shape and

00:14:59form factor of a PC that you can add AI to, and it changes it meaningfully?

00:15:03That's a really good question. As somebody who worked on Duo, you know, we definitely did think

00:15:07deeply about that, you know, in terms of vectors and options.

00:15:09I love Duo, by the way. I have a Duo right over there, and I love it to pieces.

00:15:13Warms my heart. I love hearing that.

00:15:17I think, you know, my sense on that topic, first, I think AI is going to show up in a variety of

00:15:22different forms. For us, you know, Microsoft, we're the co-pilot company, and we're going to,

00:15:27you know, deliver co-pilot as a set of services and experiences on a variety of different devices,

00:15:33a variety of different form factors, you know, different platforms.

00:15:35So you're seeing that from us. We talked about it on Monday as well.

00:15:39So that, I think, is certainly a thrust for the company, you know, across the board.

00:15:43I think we're seeing tremendous momentum with things like the M365 co-pilot, for example,

00:15:47in terms of embedding agentic capabilities, co-pilot capabilities, and productivity suite

00:15:52and applications and services. In terms of the device form factor,

00:15:55the thing that I'm particularly excited about with the PC orientation around it,

00:16:00by the way, one thing we should mention, mobile, you know, the one reason why,

00:16:04you know, the value of the neural engine and efficiency and just matrix math computation

00:16:08and performance per watt and all this stuff, we do inherit a lot of those capabilities of

00:16:13mobile platforms into these next generation mobile SSEs. Obviously, we have the sensibilities,

00:16:17because anyway, so I think that I think, you know, this, of course, and you're seeing this,

00:16:22it's happening in industry, you know, broadly, and we are in pole position when it comes to

00:16:26taking advantage of those IPs, those capabilities, and then bringing, you know, the largest possible

00:16:30neural engines at scale into Windows on the devices themselves. The thing for us when it

00:16:38comes to large screen form factors, you know, pro devices, laptops, and by the way, I think

00:16:43Surface is one great embodiment of it, but what was kind of magical for us with this co-pilot PC

00:16:47moment is it was not just Surface. We had, you know, lean in from our entire ecosystem partners

00:16:52and the signals we got from all of our OEM partners here was super strong. So it kind of

00:16:57gave us conviction on thinking about this, you know, broad and holistic and making sure where

00:17:02we think of it as a transition for a class of devices versus, you know, one particular form

00:17:06factor. The couple of things that make AI kind of meaningful for me in PCs, one of them is,

00:17:13by definition, Windows is a multi-platform, multi-app operating system. And one of the

00:17:18things that I think that make this AI experience powerful is the fact that these models can reason

00:17:23across things happening, you know, across your screen, across multiple applications,

00:17:27across workflows. And I think in a world where a lot of what we do is task flow across applications

00:17:33and so on, having the ability for having AI help you both in your own app, but also be able to help

00:17:39with workflows across apps is huge, I think. And for us, the idea of recall, the notion of search

00:17:46and search becoming semantic, I think that, you know, today, most people don't interact with their

00:17:51devices in a natural language sense, like I think intuitively we are able to do or would like to go

00:17:55do. And so, but for search to be semantic, you know, you do need to flatten what that information

00:18:01architecture and data scheme looks like, because, you know, if you, again, you don't want to be app

00:18:04specific, you want to search the way, you know, your memories would work, for instance, and how

00:18:08you recall things. And so, that way for us, PC, that was one powerful thing, the ability for us

00:18:13to be able to bring AI across applications. The second big thing is multimodality. On a PC,

00:18:19you're looking at content, you're typing on a screen, you have the opportunity for voice

00:18:22interactions, you have touch, you have ink. And so, these AI models are increasingly going to

00:18:28become multimodal for us. And so, the fact that we are a, you know, multimodal operating system,

00:18:33we are a multitasking operating system, it gives us a fertile ground to go, you know, bring these

00:18:38capabilities. And quite frankly, I think my sentiment is we have a point of view on where

00:18:42that value starts. The fact is people are going to do things with them that we start with as a

00:18:46point of view, and then they are going to bring their own ideas, and I think that's going to be

00:18:49powerful. The last thing I'll say, David, is, you know, the power of the Windows ecosystem is our

00:18:54app catalog and what developers do on top of it. And the beautiful thing with AI now is we are

00:18:59finding a strong pull from traditional, you know, Windows applications, as well as a set of folks

00:19:04who are traditionally sort of web apps, if you will, and build websites and stuff, who are very

00:19:08interested in knowing what they can do with the AI capabilities local on device. And so, the pull

00:19:14from them of reimagining what apps can be, and that's full spectrum, quite frankly, from I want

00:19:19to change the UI because I don't need, you know, as many radial buttons, dials, and so on, because

00:19:23the agents can start understanding what's happening, all the way through I want to do net

00:19:28new things where I need local models, and I want those local models to have a set of attributes

00:19:32that we can, you know, kind of, from a Windows platform standpoint, build platform capabilities

00:19:36to serve them. That's another strong signal, which is why I think, you know, Windows devices

00:19:40are a great place because we have a rich catalog that exists at scale today. And they have a very

00:19:45broad surface area for taking advantage of AI. Yeah, you just brought up my favorite user experience

00:19:51thing to think about right now. And I can't stop thinking about it. And I'm curious how you're

00:19:54thinking about it, which is this sort of UI of AI also feels very uncertain, because a lot of what

00:20:00you just described is what I would call like infrastructural AI, it's sort of underlying

00:20:04technology in service of something that shouldn't really feel like AI as you do it, right? It just

00:20:08makes it faster or smarter or whatever else. When we talk about AI a lot, we talk about chatbots,

00:20:14right? Like that is I think, the sort of accidental synonym of AI in a way that I really

00:20:18hate and hope we get past. But I am curious, I mean, I think AI, it sounds like for you sort of

00:20:23runs that whole spectrum. But as you're thinking about like, what this stuff should look like,

00:20:28and where people should see AI and sort of how loudly it should scream AI at them when they see

00:20:33it. How are you thinking about that right now? You know that as a company, David, we are there's a

00:20:39variety of teams are looking broadly and deeply at that topic. Yeah, because nobody knows anything.

00:20:44It's so fun. It's all so new. I think honestly, that's kind of the beauty of it right now.

00:20:49It's a place for innovation. We are early in the life cycle. Like you said, I think

00:20:54my sense is people looking for a paradigm and for a period of time, at least I think there's going to

00:20:58be iteration, evolution, there's going to be different people with different points of view,

00:21:02the value is going to surface itself in different ways. And then I think perhaps there'll be some

00:21:07patterns that are more stickier in a certain set of use cases than others. Can you get away with

00:21:11that on Windows, though? Like Windows is not, I would say, famously a place where you can do a

00:21:17lot of experimentation without immediately infuriating millions of people around the world.

00:21:21Can you play like that in something as big and important as Windows? Yeah, let me actually let

00:21:27me try to complete response for the previous thing, and I'll come back to your iteration

00:21:29question. The question, where does it surface itself? We shared a little bit, we shared a

00:21:34glimpse of this at the Build Conference. One of our big learnings, at least in the canvas of

00:21:40Windows, is using AI for serving customers where they're at in their workflows. And so,

00:21:45we have a lot of very highly used surfaces, very high traffic surfaces in Windows right now.

00:21:51And one of the things we're thinking about is how do we have these agentic experiences just

00:21:55help people get to their intent faster, easier, quicker? And so, that's one sort of thought

00:22:01process where I think in terms of UI evolution, you'll see the UI evolve itself to a place where

00:22:06you'll see copilot showing up in more diffused but infused manner across the operating system.

00:22:11It's not going to be screaming at you, but I think it's the subtle ways of being able to do things

00:22:16that you perhaps do today, but it'll just get you through them faster, more efficient, in a more

00:22:20intentful way. There's another big thought process, which is around the idea of recall that

00:22:24we just introduced, features and preview. The notion that you can bring net new capabilities,

00:22:30I think is going to create new shell surfaces, is going to create new experiences, and they'll

00:22:33become a place where we will refine and iterate and add new value as well. Once you have a semantic

00:22:39index, to your point, that's sort of a structural notion, that is going to surface itself in a

00:22:43variety of different ways. And so, I think that's another way to think about it. And of course,

00:22:47then there's the app experiences, and how do those app experiences in themselves take advantage

00:22:51of copilot capabilities. We introduced the new copilot thinking, yes, on Monday as well.

00:22:57And then I think you'll see, we think of it as hybrid AI. There's going to be a pattern, I think,

00:23:01where you'll see these large models in the cloud working with models on the edge. The model on the

00:23:06edge will bring context and understanding, and these large models will do reasoning at a scale

00:23:11that's not possible. And I think that concert is going to open up some new experiences, both at

00:23:15the OS and app layer. Anyway, so I think that's kind of our three vectors of thought process right

00:23:20now, in terms of where UI affordances and constructs will take themselves. And then we'll

00:23:25see. A year from now, I'm sure if we talk, you'll see how we're going to be smarter across the board.

00:23:30I think the point on these interfaces can be minimalist over time, because of the reasoning

00:23:35capabilities is certainly a deep... I know you've talked about it in a couple of different places.

00:23:39I think that's kind of on our minds. Your experimentation iteration point is a great one,

00:23:44and I think that is a thing we are on a learning journey ourselves, quite frankly,

00:23:48and how do we go do that? We'll continue to use the Windows Insider program. It's been powerful

00:23:52for us to learn what works and what doesn't. What I love with the Insider community is we're

00:23:57able to do a bunch of different experiments. We're able to do them asynchronously and get

00:24:02signals at the level that matters for individual features in themselves.

00:24:06But we also have a variety of other learning tools and experimentation methods.

00:24:10We are able to talk to customers. We certainly get the chance to go experiment with commercial

00:24:14customers. And then I think from a Windows standpoint, we get to learn at the rate of

00:24:20the entire company. And so for me, the fact when it comes to the idea of responsible AI,

00:24:25for example, Microsoft is leading responsible AI in the cloud. We are learning a bunch there,

00:24:29and certainly we take those lessons and we apply them to customers in the Windows context,

00:24:33because they're the same person, they're the same corporation, same set of customers.

00:24:37So I find, in fact, in the AI world, I have more learning tools, because the way AI capabilities

00:24:43and services and even social experiments and how people are experiencing them is actually

00:24:47pretty broad-brush across the company than features in Windows in the past by themselves.

00:24:52Yeah, I've always enjoyed the pros and cons of being Microsoft in this ecosystem. On the one

00:24:57hand, you have this incredibly long history and a lot of people who are very accustomed

00:25:02to certain workflows, and the cost of change is very high. But on the upside,

00:25:07you can put that kind of stuff in front of huge numbers of different kinds of people

00:25:11pretty quickly. And that tension I feel like I see everywhere across Microsoft and is always

00:25:17really interesting to watch. It is a great balance. Yeah, I don't know if I have a recipe

00:25:21for an answer that has resulted to success. Like you said, it's a tension and we have to

00:25:26work the balance there for sure. Totally. So I know I have to let you go here in a minute.

00:25:30So one last question. We started 12 years ago. Fast forward 12 more years for me.

00:25:35Is the bet now the same one you feel like you were making with Surface 12 years ago,

00:25:41or has it shifted? Is this a sort of inflection point in the life of what you think these devices

00:25:47are and what you're pushing towards with them? Okay, it's a great question. Let me try to do

00:25:53this. The philosophy and our strategy and mission with Surface has been around, you know, driving,

00:26:01leading innovation in the Windows ecosystem in a way that we think allows Windows itself to move

00:26:07the needle for what is possible, but to do it in a way that is actually responsible for end

00:26:13customers and as a business within Microsoft. I think that aspect of Surface is going to be

00:26:19instrumental for us to continue to move the needle of what is possible, both in a technology

00:26:24sense, in a product making sense, in a sense of how do we deliver these capabilities to customers

00:26:28and to businesses in a way that, you know, actually is credible for them at the end of the day.

00:26:32I think that aspect of Surface will continue. It has been powerful for us. Quite frankly, I think

00:26:37the co-pilot wave, co-pilot plus wave of PCs was only possible with the deep work that happened in

00:26:42Surface over the last, you know, four or five years. And so that will continue to be instrumental

00:26:46for us, and I expect and believe that will continue to be great for us going forward.

00:26:50In terms of the devices themselves, you know, time will tell. I think, you know, we've grown

00:26:55our portfolio, we've played, we've added, we've tried, we've learned. I think that spirit will

00:26:59continue with the team. We will continue to drive innovation. We will learn through those experiments.

00:27:03I'm excited many of them are going to be successful, but I'm also, you know, clear-eyed.

00:27:06Some of those experiments and learnings may or may not result in, you know, in products

00:27:10in themselves. But in either case, I think they serve the outcome and the mission for what's

00:27:15critical for Surface, and quite frankly, Microsoft, I think, at the end of the day.

00:27:18Fair enough. All right. Well, I've made you late for a meeting, so I'll let you go. But someday,

00:27:22we're just going to talk about, like, hinges and stuff for several hours, and I'm very much

00:27:26looking forward to that. It's going to be great. Sounds lovely. Thank you.

00:27:29All right. We have to take a break, and then we're going to do a big smart home catch-up,

00:27:33because there's a lot to get to. We'll be right back.

00:27:41We're back. It's time now for my favorite occasional segment on the Verge cast,

00:27:45which is when I drag Jen Tuohy onto the show and then complain to her about the smart home for a

00:27:49half hour or so. I call this segment, Does Matter Matter? It's pretty good, right? But actually,

00:27:55there is some fun stuff going on in the smart home world right now, including some potentially

00:27:59very big news from Google and Amazon and some other companies that matter in this space,

00:28:04which means Jen and I actually have a bunch to catch up on. Jen Tuohy, welcome back. It's been

00:28:08a while. It has, David. It's always a pleasure to be back, though. I feel like we haven't had

00:28:13enough time to yell about matter on this show recently. It feels like it's time.

00:28:18I know. Well, I got to yell about matter with with Nilay on the other show.

00:28:21So we don't talk about the other show. How dare you? Okay, so I have come up with a little

00:28:28game for us, and I didn't tell you about it because I'm a very cruel and unhelpful podcast host.

00:28:32I have no idea what's about to happen.

00:28:34So I have five topics for you. They're all based on stories you've written recently,

00:28:41and I'm going to throw them at you one by one, and you're going to have to tell me whether

00:28:45this thing is a big deal, a medium deal or a small deal, and then we're going to talk about it.

00:28:50Okay.

00:28:50I have theses, but I want to know what you think about all of these.

00:28:53Does that sound good? We're just going to dive in. It's going to be great.

00:28:56Sounds good.

00:28:57The first one, big deal, medium deal, small deal. The Dyson Wash G1, Dyson's first mop.

00:29:04People are very excited about this thing, Jen. Big deal, medium deal, small deal. What do we think?

00:29:09From a smart home perspective, teeny tiny deal.

00:29:13Okay.

00:29:14Because it's not a smart thing. It's, well, it depends on your definition of smart,

00:29:19but it's not internet connected in any way. It is not an IoT device.

00:29:23From the world of floor washing, however, I would say it's kind of a big deal,

00:29:29because, you know, it's very expensive, but Dyson, love them or hate them, they are innovative.

00:29:36They do come up with some interesting solutions for common problems in our home.

00:29:41And whilst there are lots of great ways to mop your floor, this does actually seem like it has

00:29:46an interesting construct that's going to take a lot of the hard work out of mopping.

00:29:51I mean, frankly, I have never mopped my floors in years because I use robot mops,

00:29:56but that's part of my job. But when I used to have to pull the mop out with the bucket...

00:30:00Oh, it's the worst.

00:30:01It's the worst. And this does actually seem like a good solution.

00:30:04And I have used some of the automated versions. So it's a mop that does not have any kind of

00:30:10motor in it or any kind of fan. So it's very different from any of the products that Dyson's

00:30:15brought out prior. You know, it's known for its fans, it's known for its hair dryers,

00:30:19it's known for its vacuum cleaners. This does not suck or blow.

00:30:22Oh, interesting.

00:30:25Although it sucks in a different way.

00:30:29And that's the kind of neat engineering trick here, which I think is interesting

00:30:33in terms of cleaning your floors. But people love Dyson products and people are prepared to

00:30:38pay a lot of money for them. I do think it's ridiculously overengineered and ridiculously

00:30:42overpriced, but I'm quite confident it will work well. So that makes it a big deal, I suppose.

00:30:48Okay.

00:30:48What do you think about this?

00:30:50I feel exactly the same way. Every Dyson thing I've ever tried or owned,

00:30:54I think is wonderful and preposterous at the same time. Like,

00:30:58I bought my wife an Airwrap for Christmas a couple of years ago.

00:31:02Oh, you wonderful husband.

00:31:04Oh, it is hands down the most successful gift I have ever bought my wife. She loves it.

00:31:08And it's an awesome thing, but it is also, like, ludicrously expensive and it involves a lot of,

00:31:13like, let's be clear, you're not getting any more gifts for several years after this. Like,

00:31:16this is, it was like your engagement ring and this, like, these are the two big gifts you're

00:31:20ever getting from me. But the stuff they make tends to be great. And I think part of the reason

00:31:24I'm fascinated by the mop thing is I think, as someone who has had a robot vacuum running around

00:31:30for a long time, the robot mop thing seems like it ought to be the next thing. But at least from

00:31:36the little bit of experience I have with them, they're not nearly as good as the robot vacuums.

00:31:42And you're shaking your head like you agree with me on that.

00:31:44I agree. Yes, they're not. They've got better, but you have to pay an awful lot for the really

00:31:49good ones. And yeah, it's just nowhere near the same as actually mopping your floor. And this

00:31:55thing will actually mop your floor because it uses water, like the water is constantly flowing

00:32:01onto the mop and to a big mop, most of the mops on robot mops are small and maybe, you know,

00:32:08have a very small surface area. So you're not really getting a lot of clean for your motion.

00:32:13Whereas this uses like a two large roller mops that constantly spray clean water onto them and

00:32:21then sucks up with agitation the dirt. So it doesn't, like I said, it doesn't have a motor,

00:32:27so it's not a vacuum, but it uses agitation to kind of get up the dirt. So yeah, you know,

00:32:31the pictures they sent and the demo they showed me, I mean, they're very excited about this.

00:32:35This was big enough that they even rolled out, you know, Sir James Dyson himself to the press.

00:32:39Oh yeah, which is pretty rare these days. Yeah.

00:32:41So they were very excited by it. But $700 per mop!

00:32:46I just, I'm sort of in the position of like, when in the early days of the Roomba,

00:32:50it was like, yes, this is a good idea. I hope everybody decides to work on this.

00:32:53And they kind of did, and robot vacuums got to be pretty good. I'm hoping that's where we are

00:32:58with mops now, because it feels like if I could solve that problem robotically, it would be

00:33:03amazing. Even if I bought this Dyson one, I would use it like twice and then be like,

00:33:06mopping is a lot of work. I don't, I don't care. But I'm hoping this is like a spark for

00:33:10people to work on this stuff. Yes, I think that's true. I'm just

00:33:13surprised they haven't come out with a robot mop. You know, they came out with a robot vacuum,

00:33:18but they've yet to come up with the robot mop, which seems like a natural progression for them.

00:33:23I don't know, maybe this was their sort of shot to say, actually, we don't think the robot version

00:33:27is ever going to work. You know, you still need to do the manual part, but we've taken a lot of

00:33:33the hard work out. I mean, the main, the good thing about this compared to, so there are other

00:33:37options out there from companies like Roborock that make robot vacuums that are handheld wet,

00:33:42dry vacuums. So they can mop and vacuum as you go, but they're huge and bulky. And you know,

00:33:48it's like a workout pushing them around the floor, whereas this is much more lightweight.

00:33:54And apparently kind of, you know, it has the motion. It's doing the hard work. You're just

00:33:59guiding it. Eventually, you know, we'll have the full size Android in the house,

00:34:04Android robot that can come around and push it for you. Maybe that's their plan.

00:34:09Sold. That will upgrade to big deal. But for now we'll leave it at small deal. I think that's fair.

00:34:14All right. Next thing on the list. You tested out Amazon's Mattercasting. A,

00:34:19I want to know about your experience and B, Mattercasting in general, big deal, medium,

00:34:24small deal. What do you think? I think it's a big deal. I do. But it's, it's stymied at the

00:34:30moment because I just don't see that Apple or Google are going to embrace it because they

00:34:36already have their own proprietary solutions for casting content to your TV. And Mattercasting is

00:34:42like, it's like Amazon's working on it the most right now, it seems like. But in theory,

00:34:46it's like an open standard for sending content around between devices, basically, right?

00:34:51Yes. So it is. It's part of the Matter standard. It actually came out with the first

00:34:55iteration of Matter, so 1.0, but it's been very sort of under the radar. I have written about it,

00:35:01but it's not, not many, there's not much adoption. It has been driven by Amazon. So the way Matter

00:35:06works, there's working groups and you have sort of a lead of each, a head of each working group

00:35:11and the head of that working group is an Amazon engineer and he's pushing this. And he's working

00:35:17with, they're not allowed to tell me exactly who's in the group, but he implied it's mainly

00:35:22TV manufacturers and app makers. So like Hulu or, and I don't know that Hulu's there, but that kind

00:35:29of, those kinds of people are working on developing it. So, but yes, just to clarify, so Mattercasting

00:35:34is a way of casting content from one device to another at its core. But what it's been used for

00:35:41here is casting content from your phone or tablet to a screen, i.e. your TV, but it could be to any

00:35:49screen in your home. So Amazon's using it as a way to cast prime video content to the Fire TV stick

00:35:56or an Echo Show 15. Mattercasting has to be enabled on both ends. And if you're familiar

00:36:03with Fire TVs, right now they do not have a native way of casting content. So this is a,

00:36:09this is really the first time Amazon's had something like this. And it's been kind of a

00:36:13frustrating experience for people that use Fire TVs not to have, you can mirror, but mirroring's

00:36:17never great. But right now it only works on prime video, though they say they have a number of app

00:36:21makers who are going to enable it, including Plex and Sling. So they're really pushing it hard.

00:36:27The thing that's different about it versus Google Cast or Apple Airplay is, and this is

00:36:34where the app makers come in, is you actually, it's an app to app communication. So you're

00:36:39replicating the app on your phone and on your TV. So you can control everything you can do in the

00:36:45prime video app or the Hulu app or which other Netflix app mirrors exactly on your TV. So you're

00:36:50not kind of like, when I use Apple Airplay, I'm like, I want to fast forward, but I can't, like

00:36:55the controls aren't intuitive and app makers are like, we've made these great experiences for you

00:37:00on our phone apps. Let's just make that what you use on the TV. So it's simplifying the process

00:37:06basically. And I think it works really well. My experience was just limited to prime video

00:37:10because that's the only one that works, but it worked great. It was much easier for me

00:37:14than other forms of casting. So I think it's going to be a big deal for Fire TV at the moment.

00:37:21But in the smart home in general, Mattercasting means any device can cast to any device. So your

00:37:29washing machine can cast to your TV. I know that's the stuff of nightmares for some people,

00:37:34but I like getting an alert on my TV when my washing's done. You could also, once Matter

00:37:40brings cameras in theory to the standard, you know, you could have a non-proprietary way of

00:37:47viewing your video doorbell on your TV. So you wouldn't have to have a certain TV. You wouldn't

00:37:52have to have a certain video doorbell for that to work, which is the way it is today.

00:37:55And it's a universal, you know, standard open protocol. Anyone can use it. You know,

00:38:01that's a good thing in the smart home. We like these open standards as opposed to the closed

00:38:05infrastructure of, you know, Airplay or Google cast. You know, there's, there's a lot of potential

00:38:12here. Yeah. Right now though, it's, yeah, it's all Amazon and no one else. I've spoken to Apple

00:38:18and I've spoken to Google about it and they're all like, nope, not saying anything. We're not

00:38:23using this yet. And this is one of the problems with Matter. Not all platforms are required to

00:38:29use or implement every part of the Matter standard. And for example, one of the problems right now,

00:38:35I love the idea of being able to cast to my TV from any Matter device, but Samsung,

00:38:41one of the major TV manufacturers, even though they're fully on board with Matter,

00:38:45they have not committed to their TVs or any of their appliances being Matter enabled. So we're

00:38:52going back to this issue of, you know, there's these great open standards and open protocols

00:38:57that you can use, but not everyone's adopting them. So that kind of, you know, kneecaps the

00:39:02whole initiative from day one. Right. Well, and with Mattercasting in particular, it seems like

00:39:08if I'm Google or Apple, I'm looking at it saying, okay, at least Matter in general

00:39:13solves problems, right? Like you can understand why eventually you're going to want to be part of

00:39:18this broader ecosystem that you don't have to control all of by yourself. Whether they want

00:39:24to now or not, a spoiler alert, we're going to get to in a minute. But the question of like,

00:39:29is it a long-term good idea? It sort of makes sense. My sense is for casting, if I'm running

00:39:34Airplay or Google cast, I'm like, eh, we've kind of solved this problem. And like, sure,

00:39:39it might be nice to have a bigger, broader thing to do, but what most people want to do is they

00:39:44want to tap a thing on their phone and have it play on their television. And we've done that.

00:39:49So to me, it's like, I'm curious to see if anybody can sort of build a use case that is so much more

00:39:55useful for Mattercasting than that. And then I think you might start to see everybody get on

00:40:00board. But until then, my sense is Apple looks at Airplay and is like, yeah, we did that. We want

00:40:05to get it on more screens, like in hotels and whatever, but Airplay works. It's fine. We're

00:40:10not going to invest more resources. But the advantage there, the hotel TV screens is a

00:40:15great example. Hotel TVs would not need all new hardware to work with Mattercasting. All they

00:40:21need is an update to an app. So instead of waiting till, what, 2050, when we can go into a Hilton

00:40:27and Airplay to our LG TV or whichever, I forget who it is that made the deal with, but you'd be

00:40:34able to do it tomorrow. It would just, all you need to do is update the software on the app on

00:40:39the TV and the software on the app on your phone. I mean, you can have it built into the TV itself,

00:40:45but you don't have to have a Matter-enabled TV for this to work necessarily. So it would

00:40:51be an easier solution. And I think where the momentum might shift is if Amazon is successful

00:40:57in getting the app manufacturers on board here and the TV manufacturers, because yes, the TV can

00:41:03have Airplay and Google Cast, but if it also has Mattercasting, it would be a more seamless

00:41:08experience for the user because it's very, you don't have to figure out if it supports this or

00:41:14that. It will just work, which is the whole promise of Matter. It will just work. Big air quips.

00:41:21Yeah, exactly. All right. So while we're on the subject of Matter, Mattercasting,

00:41:25big deal. I agree. I think long-term big deal, short-term.

00:41:29Not so much. Yeah.

00:41:30Big deal for Prime Video Fire TV owning customers. Sure.

00:41:33Yes. Matter 1.3 came out a couple of weeks ago.

00:41:38You wrote a story about it. Matter 1.3, big deal, medium deal, small deal?

00:41:42Big deal. Okay.

00:41:44Big deal. A lot of momentum. I was very not excited by Matter 1.1. Matter 1.2 was a little

00:41:52bit more interesting because there was a lot more device types added, but Matter 1.3 just brought

00:41:57a whole load of more potential with it because more device types, the device types weren't that

00:42:03exciting. It was kind of what we expected. Like we got washing machines last time,

00:42:06so now we had tumble dryers. Like, woo.

00:42:09Sure.

00:42:11There were a few interesting additions there, but I think the big deal here, well,

00:42:16Mattercasting came, updates to Mattercasting came with 1.3, which I was excited about. That was the

00:42:21ability to cast from device to device, not just the app to app. But the big deal, I think, is the

00:42:27energy reporting, energy management features, which we've been waiting for. We knew it was coming,

00:42:33but we've been waiting for this. And this is, to me, energy management in the smart home

00:42:38really takes the smart home from sort of niche to almost necessary. Because today,

00:42:46conserving resources, using less energy in our homes, saving money are all things that many

00:42:53people are worried about, concerned about trying to do. And the smart home can make it so much

00:42:58easier to do all of this stuff automatically to one of the main sort of showcase pieces that came

00:43:05out with Matter 1.3 that exemplifies this is electric vehicle management. So EVSE charges,

00:43:11so electrical vehicle supply equipment is now a device type in Matter, as is the ability to

00:43:19control and create management of the way it charges. So for example, you could use your

00:43:26Matter app or Matter-enabled app to tell your EV, I want to have 80 miles of range by 2pm this

00:43:35evening or this afternoon. And I want you to use the least expensive energy to get there between

00:43:41now and then. And you just, you know, hit button and it will do that for you as opposed to you

00:43:45having to go in and like, check, you know, is there clean energy now? And some electrical

00:43:51vehicle supply equipment can do this today, but it's proprietary. So you have to have, you know,

00:43:56the Tesla charger in the Tesla app or the Kia charger in the Kia app, whereas this will open

00:44:01this ability to any smart home app that uses Matter. So in theory, your Apple Home app or your

00:44:08Samsung SmartThings app could do this kind of management for your EV for you. In fact,

00:44:13Samsung already does have some electric vehicle management, but again, it's specific brands and

00:44:18specific cars. And so you have to have the right equipment, whereas bringing this type of management

00:44:24to Matter will open it up so that it should work with anything that can work with Matter.

00:44:29And it works locally, obviously, which is a bonus, but it will need to use the cloud,

00:44:33obviously, to get things like energy prices. But beyond EVs, you know, you could see this applying

00:44:39to using your tumble dryer, your washing machine, making sure your fridge and freezer uses defrost,

00:44:45does a defrost cycle at a, you know, at a time when energy's low. Like if you connected everything

00:44:50in your home and appliances, almost all appliances are now part of Matter, which is what

00:44:55came with 1.3, we're mainly just waiting on heat pumps, which they say are coming next. And that's

00:45:00obviously a big part. And there was one other appliance I can't remember off the top of my head.

00:45:05It was water heaters, right?

00:45:06Water heaters. So both of those obviously are big energy consumers, although heat pumps less

00:45:11than traditional HVAC. So, but still being able to manage the energy use, energy reporting. And

00:45:18this is where I think it's going to get interesting with the platforms like Apple Home,

00:45:22Google Home, because this gives them the opportunity to start differentiating, right?

00:45:28So to date, a lot of people have said, well, Matter just commodifies everything. Everything

00:45:33works with everything. So why is there a benefit for these platforms and for anyone that wants to

00:45:39create something in the smart home and how you manage the energy, what kind of service you can

00:45:44provide on top of this type of reporting that the system is now providing is going to help

00:45:52differentiate. So if Apple Home comes up with a great way of managing energy throughout your home

00:45:57with your connected appliances, you might be more inclined to use their app. Samsung already has a

00:46:02pretty good Samsung energy system in place. So I think it's got the headstart there and Google Home,

00:46:08we'll see. Alexa already has some energy management built in if your device supports it.

00:46:14So I think in the next few months, and I've already seen a few come out, we're going to see

00:46:19a lot more devices adopting Matter in order to, because this is a reason to put Matter in your

00:46:25device. Having energy management is a huge sort of push for the smart home. It's not the sexy,

00:46:31fun stuff, but it's really important. I think back to like the days of the Nest

00:46:37thermostat. And to some extent, that thing that it was just like, we will help you lower your

00:46:43heating and cooling bill for your house is maybe still to this day, the single simplest, most

00:46:48compelling smart home use case for most people. And I think this just adds to that in really

00:46:54interesting ways, but also seems like it sets up this cool flywheel potentially where it's like,

00:47:00okay, not only are you able to control more and more of your stuff, but the stuff that you're

00:47:05controlling can feed more data back to the system, which means like you're saying, the system

00:47:10can start to do better. And the people building these systems can do more stuff with that

00:47:14information. And you can actually start to like, improve upon itself over time, as opposed to just

00:47:20a lot of what we've had so far is just sort of increasingly elaborate controllers. And that's

00:47:25fine as far as it goes. But like the job of the smart home should be to do most of this

00:47:30on its own. And we're slowly getting to the point of it feels like the puzzle pieces for that are

00:47:36starting to get into place, which is very exciting. It is. And this is something I've been talking to

00:47:40companies about for years, and it has taken them years to get the partnerships, get the electric

00:47:46companies on board. It's a huge project, whereas this should make it a lot easier to push us

00:47:53forward. And the really exciting part, I think, and this is also something I wrote about in the

00:47:58last couple weeks, is that it could help shift not just your home, but energy use in the entire

00:48:06country or world by promoting and helping push forward the idea of virtual power plants.

00:48:13And this is something so Nest Renew, you mentioned Nest, so Nest Renew has turned into a new program

00:48:18now called Renew Home. But Google's still involved, but Google kind of shifted its Nest Renew,

00:48:24which is the service that helps you save energy when you use your Nest thermostat, to a new sort

00:48:29of platform that merged with a company that had already been doing demand response programs in

00:48:34select states in the country. But now it's working towards sort of creating virtual power plants,

00:48:40using every smart house and every house that has smart appliances as virtual power plants. And

00:48:46that's something that could really help reduce the strain on the grid, you know, and create a better

00:48:51experience for everyone. So it's for the greater good.

00:48:56Totally. I think that's really exciting. You mentioned heat pumps and water heaters,

00:49:00are there other huge categories still missing in Matter? What's kind of still on the to-do list

00:49:07there?

00:49:07Cameras is a big one, although they said that's definitely coming. Although it will come,

00:49:13from what I understand, in a very limited fashion. It'll just be that you can view your stream and

00:49:18control your stream in any platform. But if you want recorded video, you want the added things like

00:49:24AI identification of people, animals, that's all going to be from the manufacturer still.

00:49:30So it'll just be there that you'll be able to, in theory, view your video anywhere or use it for,

00:49:36hopefully, for triggering automations like motion detection, or if you spot an animal,

00:49:41turn the siren on. This is for my chicken coop. So I think people want cameras there because you

00:49:48want one app to see everything. But you're still going, and this is a theme with Matter,

00:49:53you're still going to need to use manufacturer apps if you want the added benefits, if you want

00:49:59the added features. But if you just want the basic use case of smart home, that's one that's really

00:50:05requested. People really want cameras. And then the other, which I don't think we're ever going

00:50:08to get, is security systems.

00:50:10Oh, sure.

00:50:11So, you know, your smart home security system. I think those may bridge into Matter, because most

00:50:18of them are Z-Wave based, but there's just too much around the UL. So all security systems,

00:50:24most, you know, if you have a certificate, most of them are UL certified. And that requires a lot

00:50:31of certification. It's a tough certification to get, it's an important one. It ensures your

00:50:37security system is going to work in the way you expect it to. And I just don't see much,

00:50:42I think the meshing between security systems and Matter is just not likely to happen anytime soon.

00:50:48But I think we may get to the point where you can bridge it so you can still control

00:50:51everything in one app. But those are the main things that we're at now.

00:50:55That feels like an okay outcome to me with both of those, actually, to say, you know,

00:50:58you can sort of take what's happening inside of those systems and use it in the broader system.

00:51:05But the idea that like, in theory, everything in my Matter home should have perfect access to

00:51:11everything I've ever recorded on my security camera, actually feels kind of weird. So like,

00:51:17maybe at the very least, taking that one slightly slowly makes some sense to me. So I'm actually,

00:51:23for once, I will give the Matter folks some credit for it going slowly here. That feels good.

00:51:29And then there is one more though, sorry. Adaptive lighting, we're still waiting for

00:51:34better control of lighting. They say that's coming soon.

00:51:37But again, this is going to be one of the things you need to use your apps for.

00:51:40But I still I want to be able to use it. I want every light in my house,

00:51:44no matter which manufacturer it's from, to be able to sync to the scene that I've chosen.

00:51:48And they say that's coming. So hopefully in the next update, but sorry, yes, I could keep going.

00:51:53Got it. All right, two more. So one, I think I know where both of these will fall. So we'll do

00:52:00the less exciting one first. You wrote about Brilliant, this smart home company that was,

00:52:05I would say, once sort of high-flying and now seems to be in very rough shape. Brilliant,

00:52:11kind of collapsing, for lack of a better term, big deal, medium deal, or small deal?

00:52:17I can sense you wanting to be nice to Brilliant here.

00:52:21It's not surprising. I kind of saw it coming. It was a great idea. I have written a lot about

00:52:27wanting better controllers in my home, and it worked very well as that type of controller I

00:52:32had been looking for, but only with certain products. And it's almost like a symbol of the

00:52:40problems the smart home have had to date. They've been around a long time. They launched almost 10

00:52:46years ago, I think. They've been around a long time. It's a good product, good hardware. So

00:52:53they've basically gone out of business. I mean, they're still keeping the lights on,

00:52:57so the system still works. I have one in my house, and it's not stopped working.

00:53:01They're looking for a buyer. I did actually check in with the CEO just a few days ago,

00:53:07and he said no news yet, but it's looking promising. And he's hoping that users of the

00:53:12system, which is a smart light switch that's installed wired to your house, but can also be

00:53:17used as a smart home controller to control your Sonos system, view of your video doorbell,

00:53:22control your lights, your thermostat. It's sort of an all-in-one smart home controller,

00:53:26but it had very limited integrations. And the company said to me over the years when I'd asked

00:53:32them about this, it's like, it's not that we don't want these integrations, it's that companies

00:53:36won't work with us. And this is one of the reasons that Matter came about, because of this issue with

00:53:41interoperability and having to make individual integrations with every device out there.

00:53:46And they've been trying to do this for years, but they'd hardly added any new integrations in the

00:53:51last few years, and they weren't adopting Matter. And I had asked them, they said they were,

00:53:57but they hadn't done it yet, and didn't really feel like it was coming down the pipeline.

00:54:02Aaron, the CEO, told me that they did have a new generation ready to roll,

00:54:08better hardware, better integration, not better integrations yet, but they had sort of

00:54:14more there that was going to make on-device processing, AI, it was going to be a much

00:54:19improved device and less expensive, because, oh my gosh, I forgot to mention how expensive these

00:54:23devices were. Starting at $400 for a light switch was just too much, and that was always their

00:54:29problem, too expensive. But when I asked, it doesn't sound like they'd added Thread or Matter yet,

00:54:34even to this next generation. And I think there's some potential for a buyer who does, if someone

00:54:39does come and take this product and sort of keep it as it is, but improve upon it, that's really

00:54:46where I think it has a lot of potential to be a decent smart home controller, if it can work with

00:54:51everything. But yeah, it's kind of a medium deal. It is a symbol, I think, of the churn of

00:54:56the smart home. We're a decade or so into this sort of DIY smart home now, and we're seeing

00:55:05some of the early pioneers kind of drop off. They've either been bought, or they've sort of

00:55:11become somewhat boring, like Ring. They're not really innovating anymore, but they're,

00:55:16you know, they have a core customer group that they have a great business, but it's not exciting

00:55:21anymore. Brilliant is an example, I think, you know, they were an independent company and they

00:55:25kept independent until the very end, but it's hard now for these startups. And I'm guessing,

00:55:29I don't know if this next question is going to be about a startup, because I think it might be.

00:55:33It's not, actually.

00:55:35Oh, okay. Well, Quilt was the startup there.

00:55:38Oh, yeah.

00:55:39The other one that I wrote about, which is a heat pump company. But yeah, it's not an easy

00:55:43space right now for startups, and it hasn't been for the last few years. And Brilliant was kind of

00:55:49sadly a victim of that.

00:55:51Yeah, Brilliant is always going to have a sort of special place in my heart. I tried a very

00:55:56early version of it that just sucked. I mean, it was so broken and bad. I put it in my wall,

00:56:03and then the light on the balcony of my apartment wouldn't turn off until I took it out of the

00:56:09wall. But eventually, they fixed it, and I got a working version of it. That was like an early

00:56:12prototype that I just thought was hilarious. But eventually, I got a thing that worked. It was

00:56:16really great. They got a lot of things right, but it always felt like a touchscreen on your wall

00:56:21can't possibly be the correct answer to this problem. And it does seem like the thing that

00:56:26Brilliant was, was always going to be sort of a transition into how the smart home is actually

00:56:31supposed to work. And it feels like good for us, but bad for Brilliant. That is the transition

00:56:36that's happening right now.

00:56:37Yeah, yeah. No, I think you're right. The smart home needs to be smarter.

00:56:41Yeah, agreed. All right, last one. Google launched a bunch of new home APIs at Google I O. And I

00:56:49confess I'm torn between either this is a huge deal that might actually change a lot of things.

00:56:54And this is nothing and no one will adopt these APIs and no one cares. So where are you big deal,

00:57:00medium deal, small deal?

00:57:01Oh, I think it's a huge deal.

00:57:03Okay, good.

00:57:04I think it's huge. I hope it's not. I hope it's not Google just saying,

00:57:09we have had enough of this. Someone else deal with it.

00:57:13Yeah. So what is sort of the idea behind the APIs that they're launching here?

00:57:17So they have launched their home APIs, which is giving access to developers, device makers,

00:57:24app makers, to any device connected to Google Home, including Google's devices. So thermostats,

00:57:32presumably smoke alarms, cameras, they've said not necessarily initially, but that may be coming.

00:57:38But one of the key parts is it's also giving access to the automation engine.

00:57:42So now essentially, so with these APIs, a device maker can create an app from scratch

00:57:50for their device or for any device that connects to Google Home and create a smart home experience

00:57:56for you. So you wouldn't have to use Google Home, but you would that the Google Home app,

00:58:00but you could use any automation system that the Google Home app provides and any devices that

00:58:08connect to it. And with Matter, there are many devices to connect to it. So both works with

00:58:12Google and Matter devices will be accessible through these APIs. And it basically is similar

00:58:19in some ways to how Apple Home's original HomeKit worked. If you've ever used like an

00:58:24Eve app or Nanoleaf app, you could access every device that was connected to Apple HomeKit

00:58:29in the Eve app or the Nanoleaf app on iOS. This Google Home API will work on both iOS

00:58:35or Android and could basically mean people can create an entire smart home platform

00:58:41from scratch on an app that you could use that isn't Google Home, but uses Google's

00:58:48smarts in the background, which are good. They just have been stymied because I feel like

00:58:55perhaps not enough resources are going into developing Google Home at Google. I mean,

00:59:02Google's got its fingers in many pies and it's always felt like Google Home is a bit of an

00:59:06afterthought. I think it's really only still here because of the hardware. So I think this is big.

00:59:11I think it's also really big because open APIs are really the sort of lifeblood of the smart home.

00:59:18When things are closed and locked down, like Brilliant, it shuts down, it no longer works.

00:59:25But if Brilliant had a local API, like Google Home's opening its APIs, then even if they go

00:59:33out of business, the device is still going to work in your home. It's like the lifeblood of

00:59:37the smart home APIs and opening this to anyone to be able to use, I think is really exciting.

00:59:44But you are right. It does depend on people actually picking up and using it. So far,

00:59:49though, there are a number of big companies that are using it. ADT is using the facial

00:59:54recognition on Google Nest cameras. The facial recognition is going to let your system disarm

01:00:01when it recognises someone that you've authorised to come into your home. It's called Trusted

01:00:06Neighbour. It's a new service that they're rolling out later this year, and that's all

01:00:10using Google Home APIs. Again, going back to what I mentioned before, what this gives

01:00:15is the potential for companies to build better experiences than, say, Google Home has been doing

01:00:22today, but using the tools that are already there rather than having to reinvent the wheel

01:00:27or develop individual partnerships with all of these companies to have access to their devices.

01:00:32But Google is like Alexa in that it works with pretty much everything, but just in a limited

01:00:37way. And now you're going to have, I think, you're going to see more innovation from

01:00:42manufacturers and device makers. And that is exciting, if it happens.

01:00:47Yeah. Well, part of what I wonder about the Home API thing is if it means we can get a bunch of

01:00:53new kinds of devices with new capabilities. The ADT thing you're talking about is a really good

01:00:58example of that. And again, it goes back to this idea of making the system work with itself in

01:01:04really cool ways, which I think is super exciting. But then I just keep thinking about one of the

01:01:07examples that they gave where they're like, DoorDash can make it so that when the delivery

01:01:13person comes to your house, it automatically unlocks for them and turns the front light on

01:01:18or whatever. And on the one hand, I think there are probably really cool, exciting versions of

01:01:24that that exist in the world. On the other hand, I think that might be the most solution in search

01:01:29of a problem thing I've ever heard in my entire life. So do you think there's anything to that

01:01:34idea that maybe this is how the smart home sort of invades the rest of the online ecosystem,

01:01:39or is this just a smart home thing? Yeah. So that was the way they were

01:01:43selling it. In fact, it was very not smart home. They were like, now any developer with any app can

01:01:49use smart home devices. And I get where they were coming from there. I do think that to date,

01:01:55the smart home has kind of felt niche and closed, and you need all of these devices to get any

01:02:00benefit out of it. And this, whilst it does seem like a solution in search of a problem,

01:02:07the DoorDash idea is, you know, you don't have to be a smart home device maker to take advantage

01:02:13of what you can do with these APIs. And I think that was an important message to get out there,

01:02:18just to show, to maybe get people thinking about what they could do with this type of technology,

01:02:24because it shouldn't just be limited to, you know, good night routines or welcome home routines.

01:02:30There is so much more that we could be doing with our homes. I think that was a slightly

01:02:34uninspired example. But I think that people that are going to be taking advantage of this really

01:02:41will be the smart home device makers. I'd be interested to see if, you know, Uber or Airbnb,

01:02:47I could see potentially taking some real advantage of this, and companies that are

01:02:52sort of home related, if not necessarily smart home related. So I think there's a lot of potential.

01:02:59But like all of these developer conferences, you know, these announcements we get,

01:03:02it really is going to depend on how much this has inspired developers to go out and do something.

01:03:08So we will have to wait and see.

01:03:10It feels like a fun moment in that sense, that there's actually a lot of sort of new,

01:03:15fun toys for anyone working in the smart home to play with. Between all the new stuff coming in

01:03:21Matter and what Google's up to, and obviously Amazon is continuing to invest in Alexa,

01:03:26there's just sort of more stuff to work with than there has been. Which I guess means we're probably

01:03:32due for some more chaos in the near term. But also maybe means a lot of this stuff that has

01:03:38just been sort of theory waiting for infrastructure is starting to become real, which I think is really

01:03:45cool.

01:03:45Yeah, I think you're right. And it is an exciting time. And you know, we mentioned that this was

01:03:49sort of the end of a cycle for one lot of startups in the smart home, but it is the

01:03:53potential beginning of a cycle. I mean, there's a couple smart home startups I've been talking with

01:03:58over the last few months, who have some great products that they're ready to launch, but

01:04:02they're still just been waiting on this type of infrastructure to come along to make it possible.

01:04:07And whilst the signs are there, and this is one thing that I was quite disappointed about with

01:04:13Matter 1.3, is that we still don't have the support from platforms. I mean, Google's opening

01:04:19up its APIs was a nice step in the right direction, but they still don't support half of the new

01:04:24device types that are in Matter 1.3. Apple hasn't gone beyond Matter 1.0. I mean, if we're lucky,

01:04:31we may get 1.2 devices, which are robot vacuums, washing machines, fridges in Apple Home with iOS

01:04:3818, if we're lucky. But that means another whole year until we get tumble dryers. So I just I

01:04:44really hope that the platforms and Amazon's been very slow here, Samsung's been a little faster.

01:04:50And I get why they're being slow, to some extent, they don't want to break people's homes. But if

01:04:55we don't start getting this out there, and the consumers get to use it and get excited about it,

01:05:01then we're not going to get the developers coming on board and making these good experiences. You

01:05:06know, we need we need both sides of the solution here to get people excited. So I really want to

01:05:13see Apple, Google, Amazon, Samsung, just go all in on what this this platform, this protocol, this

01:05:20smart home standard that they've developed to solve this problem. They can't just leave it there.

01:05:27They brought it to life, they've really got to put the work in and see this through. And so far,

01:05:33it's been pretty limited. And that's been quite disappointing. So I'm hoping that we'll hear

01:05:37something exciting at WWC next month. I'm not holding my breath. I think it's just gonna be all

01:05:43about AI. But yeah, I may be wrong. I suspect you're right. Prove me wrong, Apple.

01:05:50Fingers crossed. All right, before we go, you've been playing with tons of new smart home gadgets.

01:05:55What give me give me your favorite, coolest thing you've been playing with recently?

01:05:59Oh, well, actually, I have I have to go with my smart chicken coop.

01:06:02Okay, do tell.

01:06:03Yes, I've been testing a new smart chicken coop. It's called coop. And it is a startup based out

01:06:11in California. And it has is very well designed chicken coop. But the main kind of smart features,

01:06:18it has two cameras in with really impressive AI capabilities. So it can tell me if the door of

01:06:24the coop is open. It can tell me if there's an egg laid. It can count how many chickens

01:06:30in my coop so in case one's gone missing. And it will also tell me if it spots like a fox or coyote,

01:06:35or if it spots a cat or a bobcat, like different predators, different nuisance animals like it has

01:06:42some really interesting AI capabilities, as well as the automatic door that will open and close

01:06:48at the right time based on sunrise or sunset so that the chicks go in and out. I have baby chicks

01:06:54right now. I got new ones to test out this coop because my my ladies were too big to fit in this

01:06:59coop. It's quite small. But it's been a lot of fun. It's just I mean, baby chicks. I mean,

01:07:04it's that's fun.

01:07:05But you can't you can't beat it.

01:07:07You can't beat that. It's very expensive. I think I'm gonna say it's like $2,500.

01:07:11But having built my own chicken coop with wood, well, having watched my husband build our own

01:07:17chicken coop and helped buy the stuff at Lowe's, they're not cheap. You know, it's an expensive

01:07:24product, whether you buy one or build one. So it's interesting to see this kind of innovation,

01:07:28smart home innovation, sort of moving outside of the home and into the garden and, you know,

01:07:34homesteading with my chickens. So yeah, that's been really fun. I've been enjoying that.

01:07:39I love it. Baby chicks for journalism. That's the dream right there.

01:07:43All right, Jen, thank you, as always. Super fun having you.

01:07:46Thank you. We won't leave it so long next time.

01:07:50All right, we're gonna take one more break,

01:07:51and then we'll take a question from the Vergecast hotline. We'll be right back.

01:08:03All right, we're back. Let's get to the hotline. As always, the number is 866-VERGE-11. The email

01:08:08is vergecast.com. We love all your questions, and we try to answer at least one on the show

01:08:13every week. This week, we have a question about search engines, which actually, we got a lot of

01:08:17questions about search engines, but I was particularly taken with this one from Nathan.

01:08:22Hey, guys, this is Nathan from Milwaukee, Wisconsin. Listen, I was just thinking about

01:08:26this whole generated response from Google when you go to Google something, and I'm just wondering,

01:08:32is the big deal here that it's Google's doing it? Because they're not the first to be doing this.

01:08:37I'm sure, as you guys know, the Arc browser is AI-generating whole kind of web pages every time

01:08:42you search something on your phone, and then I'm one of the 12 people that uses Edge at work,

01:08:47and I actually use Bing to search because then it will auto-generate a response with Copilot first,

01:08:53and then kind of give me all the web results. So when Google was telling Eli that that's actually

01:08:58something they're experiencing, that people have higher click-through rates from those

01:09:00generated responses, I mean, I can really back that up because that's exactly what I do.

01:09:05And so I'm wondering, is the problem just that Google's doing it? Is the problem that

01:09:09they're so large that if they're doing this, this impacts the way that the search world works?

01:09:15Because obviously, they're the top dog, and that when the Edge browser and when the Arc browser

01:09:20are doing this, that doesn't really matter because they're just not the dominant search.

01:09:24I feel like that's maybe the argument here. So I was just wondering what your guys' thoughts

01:09:27were on that, if it's just a matter of scale, not a matter of what they're doing,

01:09:31or if there's more to this. Thanks, guys. Okay. I think the answer to this question

01:09:36is mostly yes, but kind of yes and no. I think on the one hand, the scale of Google is the thing.

01:09:44You just can't overstate the extent to which Google is the center of the internet. For anyone

01:09:50who makes a website, whether it's a cooking blog or a news site or just a place to show off your

01:09:55photography, Google is the internet. It is the main discovery tool of the internet. So what you've

01:10:01seen is everything over the last, I don't know, 15 years, really, has reshaped in order to work

01:10:09the way that Google has asked it to work. You see cooking blogs, just to keep harping on that,

01:10:14they have changed the way that they make their pages in order to suit Google. The thing where

01:10:20at the bottom, you have the recipe, but at the top, you have 2000 words of stories about the

01:10:24person. And there's a bunch of photos, and there's a bunch of sub headlines that are all kind of

01:10:28questions that don't really fit, but are questions people might be Googling. All of that

01:10:34is designed for Google. The SEO, search engine optimization, has completely reshaped the web.

01:10:39This is a thing we've talked a lot about on the show. This is a thing we've written a lot about

01:10:43on the site. I'll put some links in the show notes. But Google has fundamentally reshaped the web

01:10:48because it is so important. Because when people want to find things on the internet,

01:10:52they don't go other places. And I think the ARC search and Bing are instructive examples here,

01:10:58actually, because it's true that Bing has been doing this longer than Google. You could already

01:11:04AI your way through the internet with Bing. And it's not that no one noticed, but no one really

01:11:09cared because Bing is not meaningful to the traffic and revenue and existence of websites

01:11:15in the way that Google is. It just does not matter what everybody else does cumulative

01:11:22in the same way that it does what Google does. Google has something like 90% of the search on

01:11:28the internet. That's a crazy number. That is one of the things that has brought Google into the

01:11:33antitrust scrutiny that it has, because it is so powerful, because it has no real meaningful

01:11:39competitor. The flip side of that is that I think the things that we're talking about apply just as

01:11:44much to these other things you're talking about. When ARC, for instance, rolled out the browse for

01:11:49me feature, where you type in a search query, you hit browse for me, and instead of giving you a

01:11:54bunch of search results, it sort of compiles a web page for you, which is a lot like what the AI

01:11:58overviews on Google are doing. It gives you the summary and then some links and some multimedia

01:12:02stuff. People had the same kind of visceral reaction, which is that it felt like this

01:12:08exchange of value that people who make websites have with discovery tools was gone, right?

01:12:15The idea for two decades has been that I am going to make something and I'm going to let Google

01:12:22access it and index it and cache it on its servers and make ads when people try to find it

01:12:26in exchange for sending traffic to my website. The idea that I could monetize or in other ways

01:12:33benefit from the traffic to my website, that was the trade. Google gets to have and have access to

01:12:40and in some ways sort of intermediate how people find my website because it sends people to my

01:12:44website. That is the trade. That is the trade people have made on the internet for 20 years.

01:12:50That feels different when it's AI. And I think the way people reacted to ARC, there have been people

01:12:55who said this with Bing. There have been people who worried about this with Perplexity. All of

01:13:00these tools run into the same thing, which is they shortcut that. They completely end around that

01:13:05exchange of value and just say, don't worry, I went and visited all those web pages for you

01:13:11and brought back the information that you need. Never mind that the crawler on that web page

01:13:16doesn't accrue me any advertising money. It's not going to click my affiliate links. It's not going

01:13:21to be impressed with my stuff and read more articles. It's not going to make the cookies

01:13:27that I suggested and then save it and tell friends about it and bring more traffic. It doesn't do any

01:13:32of that. It just pulled it into a database and spat out the information with essentially no link

01:13:39back to the original source of the information. And certainly no reason to go find the source of

01:13:44that information because I've already given you everything you need right here in the summary.

01:13:48On the one hand, cool user experience for lots of things. On the other hand, totally breaks the

01:13:53value exchange of the internet. So I think the thesis is not that different between what you're

01:13:58seeing from all of these AI companies and what you're seeing from Google. The difference is just

01:14:03the scale. And Google has been existentially important to websites for two decades in a way

01:14:11that I think has been problematic for a really long time. There is this sense that Google won

01:14:16in some way that Google became the arbiter of the internet and that the only thing to do

01:14:21as a media company or as a cooking blog or as a person with a website of any kind was to play

01:14:28Google's game or try to find a way to get money from Google directly. And so you've seen a lot of

01:14:33these fights over the years of companies like News Corp picking fights with Google and saying,

01:14:39you have to pay us for our content or we're going to disappear. And I think what has been true

01:14:45forever is that these websites have needed Google more than Google has needed these websites. And

01:14:50that has been a challenge. And one thing you've heard us talk about, and Nilay in particular

01:14:54talks about this, is that Google doesn't actually owe these websites anything. And I think Google

01:15:00would tell you that it cares about them and believes in the web and wants to preserve the

01:15:04open web. That's all really easy to say out loud. And it's also not particularly meaningful when

01:15:11Google is able to do the things that it's doing now, which is say, well, our responsibility is

01:15:16actually to get users to the information that they want as quickly as possible. And again,

01:15:21that's actually an interesting product argument. It subverts a huge amount of the way that the

01:15:25internet has worked, in part because all of these websites are so reliant on Google and what Google

01:15:30has meant to them for so long. This is what we went through with Facebook when Facebook made

01:15:34everybody pivot to video with the promises of huge amounts of traffic and tons of money, and then that

01:15:39all sort of dried up. And everybody kind of got used to Facebook's wishy-washiness and back and

01:15:44forth on supporting websites and caring about traffic outside, and eventually just realized,

01:15:49well, okay, we're going to treat Facebook as kind of a value-add, right? If we get traffic from

01:15:54Facebook, terrific, but I'm not going to bend over backwards in order to get traffic from Facebook.

01:15:59And that hurt a lot of websites' traffic. There are a lot of websites that went away

01:16:03when Facebook stopped sending them traffic. And the same has never really happened with Google,

01:16:09in part because Google more or less remained a good citizen of that value exchange we're

01:16:14talking about until now. And there's a lot left to be seen. I think it's possible, as Nathan says,

01:16:21that really great content is going to be more prioritized. And the question will be,

01:16:27what happens to the folks who write this sort of commodity stuff? Like, I think a lot about the

01:16:32articles years ago that every time John Oliver would post anything, every website on the internet

01:16:37would be like, John Oliver roasts whoever, right? Like, he gets mad about hot dogs. And lots of

01:16:43people cared about what John Oliver did. So they would Google John Oliver on a Monday morning after

01:16:46the Sunday night show. And whoever was at the top of search results made a ton of money and got a

01:16:52lot of traffic. That was always kind of a weird game. And I think you could reasonably argue that

01:16:56the best outcome is actually just that you get to John Oliver much more quickly instead of a website

01:17:01talking about John Oliver. But what if there's really great, unique context added, or lots of

01:17:07interesting new information, or somebody did good reporting on top of the reporting that John Oliver

01:17:12did? This question of that sort of sliding scale between what is commodity information that

01:17:18everyone should know, and no one has particular claim to like, when was Abraham Lincoln born?

01:17:23That is just a question. It doesn't matter who has the answer. There are not particularly

01:17:27interesting reasons to go to one place or another for the answer. I just want to know

01:17:31when Abraham Lincoln was born, all the way down to, like, truly great original stuff,

01:17:37whether it's a new recipe, or new art that somebody made, or a new essay from somebody

01:17:42smart that you like, or original reporting in the news business. Google is now in a position

01:17:47of having to decide where one of those things ends and the other begins, and is also just making this

01:17:54bet, or at least this promise to publishers that, oh, we're going to get so much better at giving

01:17:58people information that they're actually going to be more curious, and they're going to want to know

01:18:02more stuff. I'm very suspicious of that argument, to be totally honest. Google is like, if we just

01:18:07surface better stuff to more people, more people will click on it. As somebody who makes their

01:18:12living making things on the internet, I sure hope that's true. I'm not confident that that's true.

01:18:17But it does put Google back in this position of, once again, being the arbiter of all of this

01:18:23stuff. And I think you'll see us talk a lot more in the near future about different kinds of search

01:18:28engines, because I think search engines are great, right? What Google was for a really long time

01:18:34was really valuable to a lot of people. And I think Google has shifted away from that to something

01:18:38else. That new thing might be great. It might be different. It might be terrible. Who knows?

01:18:43But I think there's going to be interesting competition in this space for a long time,

01:18:47because Google is kind of running away from the thing that people loved about it. And that

01:18:53is weird to me, to be totally honest, that there are a lot of companies that you can understand

01:18:57why they pivot, Google is taking this beloved, wildly successful business and kind of pivoting

01:19:03away from it. And I think you're going to see a lot of folks come in and say, Well, actually,

01:19:08search was both a good product and a good business. And maybe what we need to do is be part of that,

01:19:14because Google is going to just leave that gap open in the market again. So I think all of this

01:19:18is going to get more interesting and more competitive. But also the immediate threat

01:19:22to publishers and bloggers and anyone who makes a website on the internet is that all of a sudden,

01:19:28it could just dry up traffic. Google has made changes before that people have turned around

01:19:33and said, Oh, Google made a small algorithm change. And my website essentially disappeared

01:19:38from the internet. That's a real thing that happens. And it happens to bad websites that

01:19:42probably should have their spammy stuff removed from Google search. And it applies to really great

01:19:48websites that don't deserve that and just get caught up in Google's changes for whatever reason.

01:19:52And we're already seeing with these AI overviews, lots of unintended weirdness. There's there's

01:19:58always people out there posting stuff about, I think there was one that it was like, add glue

01:20:03to your pizza sauce to make it stickier. Like, don't do that. But that's the stuff that's out

01:20:08there now. So we're at the beginning of a thing that Google has just turned on and is making a

01:20:13gigantic bet on and is shifting its whole core product towards and the ramifications of that are

01:20:20just barely beginning to show up. And it's going to take a while. And I think it's going to be

01:20:24scarier to a lot of people than Google is going to be willing to acknowledge. So all of that is to

01:20:32say, yes or no, Nathan, right? I think that the same principles apply. But Google is doing it in

01:20:38a way that is so much bigger, and more important to the health of the open internet and the economy

01:20:44of how people make money from websites, that it feels bigger, because it is. And I have no idea

01:20:50where it's all going to go. But we're going to talk about it an awful lot on this show for the

01:20:54rest of the year, I suspect. All right, that is enough for now. That's it for the verge cast today.

01:20:58Thank you to everybody who was on the show. And thank you, as always, for listening. There's lots

01:21:02more, as always, on everything we talked about at the verge.com. I'll put some links to all of our

01:21:07SEO coverage from last year, we covered this a ton, we're still covering it a ton, super fun stuff,

01:21:11lots of Microsoft news, lots of smart home news, we'll put it on the show notes. But you know,

01:21:15read the verge.com. If you come to our website, we don't have to worry about Google. That's kind of

01:21:20the whole point of what we're trying to do. It's a good website, we do our best. As always, if you

01:21:25have thoughts, questions, feelings, surface devices, you would like to sell me, you can always email us

01:21:30at verge cast at the verge.com or call the hotline 866-VERGE11. We love hearing from you. Like I've

01:21:36mentioned before, my favorite slack channel at the verge is the one that just puts everybody's

01:21:41voicemails into a place where we can listen to them all day and I love it. And people keep asking

01:21:45to be asked, let into that room so they can hear the hotline voicemails. It's the best. I love it.

01:21:50This show is produced by Andrew Marino, Liam James and Will Poore. The Verge cast is a verge

01:21:54production and part of the Vox Media Podcast Network. Nilay, Alex and I will be back on Friday

01:21:58to talk about presumably more open AI shenanigans, all the stuff that's to come at WWDC and lots more.

01:22:05See you then. Rock and roll.

Why Microsoft bet on Surface | The Vergecast

Category

Transcript

Recommended