The Vergecast - Do we really want Rosie the Robot?

Episode Date: August 19, 2025

This week on The Vergecast, the co-founder and former CEO of iRobot, Colin Angle, joins The Verge’s smart home reviewer, Jennifer Pattison Tuohy, to discuss what the ideal home robot is. Are we close to creating a Rosie the Robot — an all-in-one humanoid robot that can take care of our homes, or should we take an entirely different approach to home robotics? They dive into the advances in technology powering this shift and ponder what purpose robotics in the home should really serve. Then, Jen takes a journey back into smart home history to help us understand its future. Grant Erickson, Principal of Nuovations, a former Apple, Nest, and Google engineer who was part of the team that developed Thread, joins the show. He shares the story of how and why, back in 2011, the Nest team, led by Tony Faddell and Matt Rogers, decided to create a smart home protocol. It involves a thermostat, fragmented ecosystems, and one of the best smart home products ever made. They discuss how Thread became the foundation of the Matter smart home standard — an unprecedented industry collaboration with a herculean task — to make the smart home simpler. To close out the show, Grant sticks around to help answer a Vergecast hotline question (call 866-VERGE11 or email [e-mail]) about how Matter manages your data. Further reading: Maybe I don’t want a Rosey the Robot after all Amazon left Roomba with a huge mess to clean up Figure will start ‘alpha testing’ its humanoid robot in the home in 2025 Amazon Astro review: too much Alexa, not enough arms Samsung is finally releasing Ballie This Pixar-style dancing lamp hints at Apple’s future home robot iRobot’s founder is working on a new kind of home robot iRobot OS is the newest ‘brain’ for your Roomba Amazon bought iRobot to see inside your home I tested a robot vacuum with an arm, and my dog may never forgive me Inside the Nest: iPod creator Tony Fadell wants to reinvent the thermostat Nest CEO Tony Fadell on Google acquisition Fire drill: Can Tony Fadell and Nest build a better smoke detector? How big companies kill ideas — and how to fight back, with Tony Fadell Situation: there are too many competing smart home standards Matter’s plan to save the smart home Nest’s home security system costs $499 and comes with magnetic door sensors Google says Matter is still set to fix the biggest smart home frustrations Thread is Matter’s secret sauce for a better smart home Nanoleaf launches a smart switch after eight years of trying Thread count: Ikea is stitching together a smarter home Why Thread is Matter’s biggest problem right now The four changes in Thread 1.4 that could fix the protocol It could be 2026 before all your Thread border routers work together Matter will be better in 2025 — say the people who make it The Nest Learning Thermostat gets its biggest upgrade in over a decade killedbygoogle.com Google’s ADT partnership finally has a new home security product to show for it Google discontinues Nest Protect smoke alarm and Nest x Yale door lock Google discontinues its Google Nest Secure alarm system Appliance makers are teaming up to reduce your electricity usage — and save you cash Learn more about your ad choices. Visit podcastchoices.com/adchoices

Welcome to the Vergecast, the flagship podcast of Why Matter Matters, no matter what David Pearce has to say about it. I'm your friend, Jen Toey, the Verge's senior smart home reviewer, and today we're talking about the future. First, the founder of I-Robot joins me to talk about how close we are to an all-in-one robot that will take care of all our household needs. But more importantly, whether that's what we really want. Then we'll turn to the future of the smart home. Grant Erickson, a former Apple Nest and Google engineer, joins me to talk about how and really why they decided to build yet another smart home protocol thread.

This is the Vergecast. Stay tuned. Support for the show comes from Retool. Too many companies run critical operations on duct taped spreadsheets, slack workflows, and whatever else they could cobble together. Not because they want to, but because building internal tools means weeks of waiting on someone else's backlog.

That's where Retool comes in. Build custom internal tools just by describing what you need. Proms something like, build me a revenue dashboard on our Salesforce data. And Retool actually builds it on your company's data in your cloud

with enterprise security built in. Go to Retool.com slash Verchcast. We all need to retool how we build software. What's up y'all? I'm Skyler Diggins, seven-time WMBA All-Star, Olympic gold medalist, and mom. And I'm Cassidy Hubbard, host and reporter for nearly 20 years

covering the biggest names and stories in sports and mom. And this is Am Mom, a community for athletes, game changers, and moms of all kinds. Dropping May 14th. Tap in with us. Do you ever wonder what's in your lotion? If you look at the back of the bottle, it could contain more than a dozen ingredients. And they may not all be regulated. The threshold is so high that only 11 cosmetic ingredients have been restricted by the FDA since 1938.

This week on Explain It to Me, the chemicals lurking in your cosmetics. New episodes, Sundays, wherever you get your podcasts. Welcome back. Now, if you're too young to remember the Jetsons or didn't have a basic cable subscription to Cartoon Network when you were growing up, think of a futuristic version of the Flintstones set in 20. There were flying cars that folded up into briefcases, 3D food printers, and my favorite, a robot made named Rosie. It's time to blast off for Dreamland. Well, okay. If Rosie tucks me in, then tells me about the cow that de-gravitated over the moon. Miss, I got it here, too. Will you, Rosie? Yes, ma'am. Thank you, ma'am. Good night, ma'am.

Rosie was a big blue-metal robot that rolled around on wheels. She had dials for eyes and claws for hands and a pretty little apron. She cleaned their house, took care of the kids, and even functioned as a security system. Now, Rosie inspired a generation of roboticists to create home robots. Today, we have robots that vacuum your floor. We have ones that scoop your kitty litter, mow your lawn, even wash your windows. But we don't have a Rosie, at least not yet. So is a humanoid helper robot something we could actually achieve?

And maybe, more importantly, is it something we actually want? To answer these questions, I invited a leading expert on home robotics to join me. Colin Engel is the co-founder and former CEO of one of the world's first robotics companies, I-Robot. And while it became known best for its Rumba robot vacuums, I robot was originally founded to develop robots for all sorts of applications. From exploring Mars to building stair climbing robots for the military, essentially Angle and his co-founders, all MIT roboticists, laid the foundation for the robotic future we may or be living in one day.

Hello, Colin. Thank you so much for joining me. It's great to see you. I'm happy to be chatting once again with you. It's been a while since we talked. You've been under the radar a bit. since stepping down from iRobot last year following the sale to Amazon falling through. But I understand you're still working in robotics with a new startup called Familiar Machines and Magic. How much can you tell us right now about what you're working on?

Not very much. Darn. We're still in stealth mode, but I think that it'll be fun to be able to sort of pull back the veil, you know, as we get toward the end of this year into the next year. I think that if the name is a bit intriguing, what we're doing will be even more so. Ooh, good tease for the future. We'll definitely have you back on when we can talk more. But today we want to talk about what's going on now right today in the state of home robotics. So just to sort of kick us off, I kind of wanted to start by getting your thoughts on the current home robotics field. You know, what's good in the space today?

what's bad and what is the potential here to be really game-changing in home robotics? So, you know, a wily coyote, Acme robotics kit was just dropped from the heavens on us a year or two ago. And that obviously was generative AI and its remarkable ability to impact the entire field of robotics. But even before generative AI, I think that the industry writ large has been, you know, struggled with what is the role of robotics in our home, in our everyday lives? You know, back when we founded IROBIT, IROB, if he's okay, 35 years would go by. And the best we had was mainstream floor care. I think that I and everyone else who heard that answer would be up in arms in violent disappointment

that we couldn't do more rapidly. I think we're still learning. And even though we have great new tools, we shouldn't forget that we hadn't finished learning the first set of lessons even prior to getting these new tools. Yeah. And now obviously, generative AI seems like it's going to be a big game. changer for robotics. And I definitely want to dive a little deeper into that or a lot deeper into that. But just in terms of like the space today, where do you feel like most people have embraced

home robotics? I think a lot of people don't really realize how many robots they actually have in their home today. Yeah, it gets at what is your definition of a robot. Yes, exactly. What is yours? You know, I've gone through many different generations of answering of that question. You know, it started out with a robot as something that can perceive its environment, think about it has perceived, and then taken an action based on that thinking. But by that definition, virtually every device that we have in our world is a robot. And I was forced to say, okay, that doesn't quite work anymore, at least it doesn't fit with my,

gut definition and evolve to, you know, a robot is a machine you name as a imperfect but more accurate description of what a robot might mean, because I think that there's an aspect of my interpretation of what robot means that involves a human connection of some sort. that human connection is crucial in a class of machine which goes beyond pure function into support and relationship. And so by that definition, there's a lot fewer robots out there, and there may be some things that are technologically simple and yet fit the definition. This is something that I've been struggling with.

of the reasons I wanted to have you on to really dive into this. So I've spent a lot of time with the robots that you were originally describing, not the ones that I necessarily named. So like the vacuums, the robot lawnmowers, the robot kitty litter, things that have a purpose and have some intelligence because they know to do something at a certain time or when something happens rather than me having to direct it. Then we've, you know, anyone that's been in the smart home at all, the dream that's sort of been forced on us has always been this Rosie the robot. She was a character in a cartoon, The Jetsons, from the 60s, that has sort of long been held up, not so much as what we ultimately want in a robot, but sort of an inspirational device for what we may see in our

home, a multi-modal, humanoid robot that can do chores for us, that can do the laundry, that can wash the windows, that can make dinner, that can play with the kid, that can do these chores. And then as you say, and in the Jetsons, this is very, very much part of the character, becomes part of the family. And that's an area where it starts to obviously get to another level. But in terms of the actual technical achievement that we're looking at towards getting a humanoid robot in our home, a Rosie the Robot, inspired robot. Do you feel like we are close to that? And if yes, if not, what is the tech that's kind of put? pushing us to that. And do we want that? Well, I think that the do we want that or when do we want

that is maybe the more interesting question from the can we? Yeah. Because, you know, it is certainly within the realm of technology that exists today to do something rosy-like. Could we build such a thing? Yeah. Yeah, I could build that. Is there a demand for it, or is it what we really want today? No, not at all. Humanity has a long way to go before Rosie is what we really want. In fact, one of my journey in my journey at IRobot, the question of after vacuuming, what do people want?

and even within the realm of floor care, what did it take to grow to the point we are today where robot vacuuming is generally accepted as a mainstream alternative to upright vacuuming? One of the things that I got most wrong in the early days of Roomba was imagining that the perfect Roomba was a Roomba you never saw, you never touched, you just, came home every day to a perfectly clean home. And that was completely wrong. The perfect Roomba was the Roomba that cleaned around your kitchen table did a great job, eventually earning your trust that it could actually do that job, at which point you would give it permission to clean the rest of the kitchen, at which point after proving that it could successfully clean the kitchen, you would give it

permission to do more. And so that people don't instantly trust technology. That's very true. And so that this fundamental disconnect between what is technologically easy and what do people want their technology to do creates real barriers to progressing down the path of customer acceptance. Yeah, I mean, your example there of saying, well, I'll try one thing. And if it does it well, I might let it do the next thing. I can see that logic. But we are seeing companies with these humanoid robots. In fact, just recently figure they're really all really pushing hard on their own generative AI models that use vision, language, action to create these humanoid robots. They have a video where gentlemen have told these two robots to put the groceries away. And we see these humanoid robots actually sorting the groceries, passing them back and forth to each other, putting them in, you know, the fridge or in the pantry.

If these humanoid robots could really do so much more for people in their homes, I think I could see where people might be drawn to them. Right now, we're not spending $10,000 for a robot to do this. But this kind of is the, you know, the future Rosie, the Rosie that, I guess when we're talking about, the science fiction of the humanoid robot. This is probably what people envision. What do you think about this type of robot? I mean, there's the Tesla optimist, there's Boston Dynamics Atlas. I mean, we've seen these being developed. And although mostly for industrial applications, you know, figure and Tesla have very much expressed their desire to push into the home. So, I mean, technologically, this is a marvel. I mean, this is incredibly exciting. But it is,

Is it a product that people want? Yeah. And if so, why? You know, that's a completely different question. And forgetting to get good answers to that question can lead you to, you know, making some mistakes as far as where you put your energy. Yeah. So, I mean, I think that is a magnificent video of robotic competence. I think it is irrelevant from the perspective of any kind of realistic vision for robots in the home.

Yeah. Because it's, is that really a problem? Right. Most of us could put our groceries away relatively easily. You know, at I robot, we would create a graph where on one axis of the graph it was activities, the frequencies, the frequencies, with which an activity was done in the home. And then on the other graph was, how much was that activity hated?

You know, so that if... Emptying the dishwasher. Well, I mean, so emptying the dishwasher, okay, there's something that's disliked. And done frequently. Okay. And, you know, vacuuming was done also frequently and disliked. Mopping was disliked so much that it wasn't done frequently,

and people would value the mopping robot far, far less than the vacuuming robot, simply because they had trained themselves to be okay without mopping every day, and so that they sort of mentally... Didn't have to be done. It didn't feel like it had to be done. And ultimately, you know, and ultimately mopping as a standalone robot, failed, it became, at best, differentiator. If you could buy a vacuum that mopped two, that was better than a vacuum that didn't mop. And so it was a secondary feature that had some

differentiating benefit. And, you know, you talk about emptying the dishwasher. I mean, I think that's interesting because it is done frequently and it is disliked. In which case, you say, okay, well, how much would you be willing to pay to not have to do that? $20, that's how much I pay my daughter a month. $20 a month to empty the digital. Not the same as buying a humanoid robot from Tesla. I mean, this is where it gets hard. Yeah.

And trying, you know, I think that it makes sense to me that so much energy is going into sort of B2B applications. Yeah. Because you can sit down with a spreadsheet and you can decide what needs to be true in order for your robot to be economically viable. Now, you may not like that math, but at least it's doable math. Yeah. You know, and when you start to say, okay, let's just vision cast into the home where

we can make some, holy cow, that's ridiculously awesome kinds of videos. But does that translate into people opening their wallet for it? Yeah. You know, that's a whole different world that the industry of physical AI is going to have to be very careful of, or else, you know, we're going to be sitting, you know, 10, 20 years from now, feeling like, gee, not much progress has been made because not many of these robots have been sold. And the other thing that is, you know, I've been on both sides of, but I sort of had to, based on brutal experience, except the fact that general purpose robotics doesn't translate well into product.

Yeah. Meaning that if you said, well, putting away the dishes is worth $20 a month. Folding the laundry is worth $30 a month. Pushing a vacuum cleaner is worth $5 a month. And so that the value of this robot is the sum of all those values doesn't actually work, at least not in traditional consumer behavior. And so that it's more like add up all the applications and if they're five of them, then divide by five.

or pick the one that the consumer is going to actually want, and then that's the value. And so if you had all these things, one of them was the $5 push a vacuum, then it could be the $5 is the perceived value. And people don't do summation of benefit math in their head very well. You know, you're talking here about how we should focus on the use case, but there is a lot of interest in some type of sentient robot in our homes? I mean, we have seen in the last five or six years, more recently, Samsung, LG, Amazon, all coming up with their own home robot that doesn't really do anything, but does have a position in the home. And what that position in the home is is

still very fuzzy. We've got Samsung's Barley and then Amazon's Astro, which I actually tested in my home, and it was very cute and it had a lot of personality to your point earlier about having something in your home that you have a connection with. Our former deputy editor, Dan Seaford, actually described that Astro is a love child between a Roomba and an Echo Show smart display, which I think was the best way of summarizing that creature except for it couldn't vacuum. It didn't have an actual obvious purpose. We've seen Apple has been working on as tabletop robot with sort of personality. They did a whole sort of white paper on investigating the value of having sort of a connection to the robot.

And when you were at iRobot, you and I talked about this a few times. You were developing something called an iRobot OS, which is still something that the robots, so Rumba's worked on for the last few years, an operating system. And it gives the robots, and these were your words at the time, I think, like a deeper understanding of your home and your habits so that they can clean harder and smarter. So, and at the time, I think when this happened, I surmised that this was one of the reasons that Amazon wanted to buy a robot because it was creating this kind of purpose for robotics in the home, like a central intelligence, I suppose, of your smart home, a rober.

a robot that knew what was happening in your home. So do you think that kind of central intelligent device in the home is kind of the key to unlocking that value of the domestic robot? Is that what LG and Samsung and Amazon are sort of working towards? And what's that going to look like when we get it? It's a great question. And I have many opinions. But to try to unpack the question down to, you know, what was IROBOS about what might the smart home be, you know, you brought up Rosie. And if you talk to anyone who has been living the smart home challenge, there's a first question in approach.

And that is, are you trying to build Rosie the robot or the Starship Enterprise? Yes. Well, and you know, that's what Amazon's trying to build with Alexa. It's like these two things coming together, right? Well, obviously, you know, Rosie the robot would be the general purpose robot that does everything in your home. Yeah. The Starship Enterprise is the home is the robot. And so, you know, the Rumba is a white blood cell. The thing that changes the thermostat is, you know, a control system in the body. And, you know, maybe. the eyelids are the opening and closing of the shades in your home. And if you can think about the home as the robot, then it leads you down a very interesting path, which to date, there's been one of these barrier tapes across after you take the first three steps. And the

barrier tape across the journey toward building the Starship Enterprise is complexity. overload. Right? I mean, I love this idea. I'm just going to go incrementally add things to my home, and when I have enough things, I'll be able to say, computer, make my housework, and everything will magically happen, and I'll feel like I'm living inside a caring dwelling that looks after me, except for it doesn't work. The maintenance of this home of the, as the robot, concept has been blocked by complexity. So ACB toolkit comes down. You need to have an architect willing to prioritize consumer experience over inclusiveness

and force some of the, you know, that usability magic that made the iPhone such a success. and we can start to see the Starship Enterprise take shape. And I think applying generative AI, the ability to unlock perception to extract user intent without burdening the user with complexity, that is a very interesting journey. And the cool thing about the home is the robot or the Starship Enterprise metaphor, is you already own the Starship Enterprise and so that you're just adding little things and so that some of the investment dollars required

can be minimized and you can go buy the function you want to add your home and it just can sort of incrementally do what you expect. And so I feel like, you know, I'm not sure these attempts that rolling something down the middle of Rosie and the Starship Enterprise are strategies that I would advocate. You know, once you have the Starship Enterprise,

maybe there's a Jeeves, the Butler, that has a role, but its role is to interact with people. And so go build it in a way that interacts with people, if that has emerged as something being sufficiently valuable that people would pay X,000 dollars to have that. And maybe that's the humanoid. Yeah.

Because then it's the humanoid being used for its core functionality. And I think that that's an very interesting direction that I think the smart home would go. But the common trip-up is everyone thinks they can run through that yellow barrier tape and it's just going to be okay. And they get tangled and nearly decapitated as a result of ignoring that being fundamental problem of the Starship Enterprise. The complexity. The complexity, yeah. Well, and now the Starship Enterprise was run by a central computer. So for this home as a computer to work, we need that central intelligence. I mean, and that's another benefit of the Starship Enterprise, because that could be up in the cloud and be very efficient, you know, the most expensive possible thing to do with that computer is put it on a wheeled platform or something. I mean, it's why we can do that. making it in, but what we do have today already are voice assistants. And we're seeing the slow,

painful shift of the voice assistant from the command and control device, like a robot, command and control to the generative AI intelligence infused voice assistants like Alexa Plus, which has recently launched to some. And Google Assistant and Amazon, Apple's series is supposedly undergoing a transformation. So is that really the rosy of the future? Like this sort of of sentient being in the home. If you buy into building the Starship Enterprise, then one of the best parts of the Acme toolkit is speaker-independent, natural language,

command and interface with the Enterprise. Yes. Because... That's what I want. Right, right? I mean, it's the... You know, we had voice connection with Roomba for a very, very long time.

And it was unfortunately disappointing how rarely that got used. Yeah. It had to be so precise. You had to remember the exact phrasing. And I remember, I mean, Roomba had like, what, like 600 commands you could use, but no one can remember all of those.

It was actually much, much, much larger than that because we tried to imagine every possible way that you could say, go clean the kitchen. And we thought we did a pretty good job, but 98% accurate isn't good enough. Yeah. And it only takes a couple of failures before you never do it again. Yeah. And so this idea that with generative AI, we can start having natural language interfaces

that actually work is going to unlock usability. Yeah. Break through that yellow tape of complexity. I mean, the difference between 98% and 99.99% sounds little, but it's transformational. Especially in a home environment. You know, I think that all of the lessons on voice should be thrown out and we should start again with the power of actual natural language comprehension. Yeah. Because there's so many, well, it didn't work.

well, now they will. And a lot of conviction around the limitations of voice as interface needs to be completely rethought as it comes to how do humans and machines interact. I think that's one of the exciting things to see. And then translate that over to machine vision. And the amount of perception that is capable, machines are capable now, is a little night and day as well. Yeah. Okay. Now go find the right application.

And this comes back to the application. And it's a very good point. And I guess I have two more questions around this for you before we let you go. So there's one that I want to wrap up with. But then there's one that I don't think any of the listeners or my readers would forgive me if I didn't ask you about is appendages. Robots with arms and legs. We've already seen a robot vacuum manufacturers come out with an arm to pick things up. I know you've told me in the past that you guys, and you did actually develop a stair climbing robot, not for the home, but for military applications. So whether we're talking about a humanoid robot or a computer-like Star Trek Enterprise, we're still going to want our robots to do a bit more than they do today. So what's the next step there? Do you think that we're going to be able to get more utility out of current robots without getting to the humanoid, necessarily going to the humanoid level? Yeah. I think generalizing appendages is a mistake because I think legs and arms are completely different.

Yeah, yeah. I think arms with our favorite Gen A.I. Acme Toolkit are about to become spectacularly useful. Interesting. And so that, you know, what you can do with the single arm, dual arm manipulators, I think that the costs of high performance dual arm manipulators are coming down very quickly. And so putting an arm on something is suddenly wildly more useful than it used to be because we can grasp organically shape objects. We can recognize where they are.

We understand this visual perception that I was alluding to earlier is. suddenly made the fact that once I'm holding this, putting it away now is something that I can do because I actually understand where stuff can go. And I can, you know, I can, you know, go get me a beer. Suddenly makes sense because I know where the kitchen is. I know where the refrigerator. I can open it up.

And so now I can pick up a beer. And those are all really, really, really, really hard problems prior to, you know, Gen. I, you know, Gen. I think that the idea, if, you know, If I was building a warehouse robot, I would put arms on it immediately, but I wouldn't put legs on it because warehouses are flat. Wheels work really, really good on flat. So Rosie had it right. I mean, she was arms and no legs.

Yeah, I think Rosie did have it right. Honestly, and if she wanted to go upstairs, she just hovered, which was like the, you know, just like R2D2 could magically make itself upstairs until they put rockets on it, which really wouldn't work very well in a home environment. But I think legs are hard, you know, the amount of force that needs to be stored in the knee of a leg in order for it to do the types of things, lifting moves. And if that amount of energy in the knee is not necessarily something you want people near. and so that there's some physics that need to be figured out a little bit as we think about the realities of thoughtful Rosie the robot design so that we can get all of the good minimize cost and maximize

safety. And that will require letting go of some of the preconceptions of how things need to be. No, that makes a lot of sense. I see an arm being much more useful, although to date, arm that I've tried in my house has not really been very useful. But the potential's there. The potential is there. And that's exciting. It is exciting. But just to sort of wrap up and to bring us sort of full circle to when you first talked at the beginning of the chat about where you feel, I think you mentioned, like companionship and bringing like robots that have a connection with people in the home. This is something that I've been struggling with in my mind. It's like the idea of anthropomorphizing appliances is sort of brings with it like a whole host of complicated challenges

to me around like the nature of consciousness and the boundaries of humanity. And we don't have to dive into all of that now. But I would love to just sort of ask you if you think if you if as you said, generative AI and this sort of more true intelligence in robots is going to become very much part of the future of home robots. And if we're building something that we're looking to have a connection with that is perhaps part of our life or our family, not just a machine. What is that? You know, like everything in robotics, it's complicated. Right.

You know, I think that what is true is humanity is desperately isolated and lonely. And, you know, 50% of Americans would self-identify. as feeling isolated and lonely. And the health impact of feeling self-isolated and lonely has been well documented as, you know, a U.S. Surgeon General number of years ago said it was the same as smoking 15 cigarettes a day in its negative impact.

And so this is a epidemic level challenge that society is facing. Where do robots come into this? Yeah. And I would argue that building, whether it be Rosie the Robot or the Starship Enterprise, without considering the fact that what you're doing is creating an opportunity for real human connection is a mistake. Now, once you've realized and once you've committed yourself to the fact that human connection is an important part of, of what the world needs, then there's a question of,

okay, where are the challenges that we need to make sure that we're very aware of, what metaphors can be healthfully imposed, is it recognized that the goal of your smart home, it's just occasionally motivate you to get out of the home and go do something that would drive

additional human connection? Or how do you become part of the solution, as opposed to an increasing part of the problem. But it begins with a recognition that this is a important challenge, and generative AI without taking on, at a very fundamental level, the opportunity and challenge of human connection is not going to succeed in the long term. Yeah, yeah, wow. Yeah, it's interesting.

I mean, it's there. It's happening. do, how do we make sure we do it right? And part of doing it right is having it be just a wonderful and healthy experience and not just a wonderful experience. Well, and I know you can't tell us much, but I'm just guessing from the name of your new business, familiar machines and magic, that maybe this is something we might be hearing more about from you later. And I'm excited one day, hopefully, to hear more. But I really appreciate your time chatting with me today. walking me through this and helping me sort of figure out what I actually want from my home robotics

has been very, very illuminating and really interesting. And I very much appreciate you sharing your time with us today, Colin. Thank you. My pleasure. Thank you very much. And we're going to take another ad break. And when we come back, we're going to be talking about the red. No, not meta-social network, but the thread protocol and how one goes about creating a smart home protocol.

We'll be right back. Support for this show comes from Shopify. Starting something new isn't just hard. It can be really scary, too. So much work goes into this thing that you're not entirely sure will even work. But here's a better thought. What if it did all work? What if your instincts were actually right all along?

Shopify wants to help you get there. They're the commerce platform behind millions of businesses worldwide and nearly 10% of all e-commerce in the U.S. from established brands like Allbirds and Hines to companies just getting started. Their design tools make it simple to create the exact online presence you're envisioning, with hundreds of ready-to-use templates available. And with built-in marketing tools, you can launch full email and social campaigns in just a few clicks. So you can connect with customers wherever they are. It's time to turn those what-ifs into with Shopify today.

You can sign up for your $1 per month trial today at Shopify dot com slash vergecast. You can go to Shopify. com slash vergecast. That's Shopify. dot com slash vergecast. Support for the show comes from LinkedIn. If you're a small business owner,

you know that every hire counts, but time and resources are limited. Finding, connecting with, and screening the right candidates takes up valuable time you could be giving to your customers. That's where LinkedIn Hiring Pro comes in. It's built.

to be your hiring partner, helping you find the right candidates faster. That way you can hire with confidence without turning it into another full-time job. Hiring Pro streamlines the entire process from drafting your job to shortlisting candidates and conducting AI-powered interviews for initial screenings. Its updated conversational interface lets you describe what you need in plain language. Nearly 60% of hirers find a candidate to interview within a week. With Hiring Pro, you spend less time searching and more time connecting with the right talent. And instead of getting buried in resumes, you get a focus shortlist that actually moves your hiring forward.

Join the 2.7 million small businesses using LinkedIn to hire. Get started by posting your job for free at LinkedIn.com slash track. Terms and conditions apply. Support for the show comes from LinkedIn. If you're a small business owner, you know, know that every hire counts, but time and resources are limited. Finding, connecting with, and screening the right candidates takes up valuable time you could be giving to your customers. That's where LinkedIn HiringPro comes in. It's built to be your hiring partner, helping you

find the right candidates faster. That way you can hire with confidence without turning it into another full-time job. Hiring Pro streamlines the entire process from drafting your job to short-listing and conducting AI-powered interviews for initial screenings. Its updated conversational interface lets you describe what you need in plain language. Nearly 60% of hirers find a candidate to interview within a week. With Hiring Pro, you spend less time searching and more time connecting with the right talent. And instead of getting buried in resumes, you get a focus shortlist that actually moves your hiring forward. Join the 2.7 million small businesses using LinkedIn to hire. Get started by posting your job for free at LinkedIn.com slash track.

Terms and conditions apply. Welcome back to the Vergecast. Putting together a smart home can be complicated, especially if you want to buy devices from different companies. That's because as manufacturers have made new devices for the Internet of Things, the IOT, they've all used different protocols. and created different walled gardens. Zigby, Z wave, home kit, Bluetooth, Wi-Fi, thread, weave, and more I could go on and on. It's been a frustrating experience for consumers. Do I need a hub? Will it work with my Android and my spouse's iPhone?

Am I locked into one manufacturer ecosystem if I just want to turn my lights on with a voice command? So to fix all of this, about five years ago, Apple, Google, Amazon and Samsung, and a bunch of manufacturers, but not all, rallied together to create a solution. Matter. Think of Matter as a shared language for your smart home devices so that they can talk to each other, no matter where they came from. But for them to talk, they need a way to do that. They need like telephone lines or a network infrastructure.

And Matter works with two. One is Wi-Fi, and it's hardwired sibling. Ethernet and the other one you may be unfamiliar with. It's called Thread. Thread is a low power, low bandwidth, mesh networking protocol that, unlike all those other protocols, was specifically designed for smart home devices. It's super fast, works locally and doesn't need an internet connection to turn on your lights. Today we're going to do a deep dive into Thread, how it came to be, what makes it stand out, and then explore why matter and thread haven't yet seen the widespread adoption that some were hoping for. To do this, I've invited someone who's probably forgotten more

about thread than most people know. Grant Erickson is an engineer who actually developed Thread while working on Nest's early smart home products. He then helped grow the protocol as president of the Thread group and was instrumental in bringing together Google, Apple and Amazon to form the Matter Smart Home Standard. Hi, Grant. Welcome, and thank you so much for joining me. Good morning, Jennifer. It's a pleasure to be here and a pleasure to finally be able to chat live. Wonderful. Great. Yes, we've communicated over the years over email and this is our first time face to face, so it's exciting for me. I can't wait to get some more insight here.

And just to kick us off, can you just give us kind of a brief overview of what is thread? Absolutely. So in order to really orienture yourself to what thread is and what where it fits, I think the best peer that you can think of is Wi-Fi. We love Wi-Fi. It's on our phones. It's on our tablets. It's on our laptops. It's on our computers at home. It allows us to do things like stream Netflix or look at Facebook or do all the things that sort of the killer app of the internet and the web have become. But Wi-Fi is, in addition to providing a lot of bandwidth, it needs a lot of power to do what it does. And for constrained devices that either need to be small and tiny or need to run on a very tiny battery, Wi-Fi.

can be made to work, but it's not ideal because you're often going to replace batteries very frequently. You're going to have to figure out how to plug the thing into a wall. And thread is really the low power sort of constrained wireless network peer to Wi-Fi. It allows you to sort of bring internet connectivity to things for which it was either previously impossible or expensive, or as people in the biz love to acknowledge required gateways and coordinators and bridges and all these other types of technology that consumers really hubs, exactly. consumers don't care about or don't want, but there's sort of little necessary evils that we have to bring in when the technology doesn't get us to the internet end to end. So in summary, really think

of thread as a super low power complement to Wi-Fi for bringing devices that previously couldn't connect to the internet to connect to the internet and other wireless devices or wired devices. Excellent. But one of the main things I wanted to talk to you about, because I think it's fascinating that you created a new smart home protocol. So we're going to do. go back in time and do a bit of smart home law here. And this is for the smart home geeks along here. I hope you guys will enjoy this. But thread began way back in 2011 with the original Nest learning thermostat. And you were involved in developing that, having worked with Tony Fidel and Matt Rogers at Apple before joining them when they founded Nest. So can you tell us

why you guys decided to create your very own smart home protocol, considering there were already a few of them. Why create thread? So back when I got the call for Matt Rogers to join getting the band back together was with my wife at the Space Needle in Seattle. And even at the moment that he had called, there was already a prototype. And even before I joined of the Nest Learning Thermostat, and it had what we call some dark silicon in it. And I think it was the TIC-C-something or another 802-154 radio. For those in the audience that aren't aware, your thermostat fundamentally controls one of the largest energy loads in your home. And the Zygby Alliance, of course, at the time had a standard called Smart Energy Profile. And the idea was that the Nest Learning Thermostat was going to participate

with your energy meter in your home through the Smart Energy profile. And by time I joined, I was sort of helping lead the charge with some other engineers on how we were going to make this work. And we found something very interesting. And that is, even though there was a smart energy profile and standard, as we approached, for example, Pacific Gas and Electric or Con Ed, all the different energy companies, even though there was a standard, at the top of that standard, each energy company had a special sauce in terms of onboarding and security in terms of how the device would get involved on their platform, on their network, and participate. And what we quickly realized is, wait, we're a small startup. We only have, you know, less than 50 employees. There are 20 energy, major energy companies. We're going to have to have effectively 20 different, firmware loads for our thermostat to deploy this across the country. We're like, disaster,

that's not going to work. That's not scalable. We just simply can't support this. And so we basically said, let's stop, let's retrench, let's re-figure this out. And what we ultimately did, which I think was the right choice, is of course, all these energy providers had cloud platforms. And what we realized is we could integrate at the cloud level where software and resources are much more fungible and easily changed and redeployed. And so we did cloud-to-cloud integrations with all the energy providers. And so what that meant is we now had this radio that we thought we knew what we wanted to use it for, but suddenly we didn't have a use for it that we thought we did. So that kind of brings us to the next part of the journey. And that is, it's important to understand that the company

was called NEST. It wasn't called the NEST Learning Thermostat Company. It wasn't called the NEST security company. There was always a vision for an ecosystem of connected products and services, not just within the walls of NEST, but outside to other partners and products and services outside of NEST. And nest, presumably, being like the home, like nesting in your home. Exactly, exactly. And so there was always this vision for more than just the thermostat. And there was always a vision for connectivity, both outside the home. You're outside the home.

You want to have awareness of what's going on in your home. How do you connect through back to your products and back to your home? There's a notion of, I'm in the home. I want to control my thermostat from the bed. Maybe not going all the way up through the cloud, as a lot of early smart home integrations were. What would this need to look like? And in fact, even before the fact, we were, of course, many of us Apple alumni, but we knew that Apple was not the start in the end of mobile platforms and tablets.

And so we were thinking about, again, Android and Google and other ecosystems. And so we asked ourselves, what technologies do we need to bring to bear to bring this vision to life? Hey, we've got this 15-4 radio. What could we do with that? And the real key moment for us was, and this sort of Wayne Gretzky moment, skate to where the puck is going, not where the puck is. we bet on IP. We said, yeah, there's all these protocols. Yeah, they've got some market traction. But ultimately, if we want people to start with a particular product in our ecosystem,

we don't want them to have to start with that product. Plus, oh, you need to buy a gateway or a hub. Or, oh, I have the product, but I want connectivity. I can't do it until I get this hub. And so we really wanted to make the ability for consumers to choose the product in our ecosystem they wanted to start with and be incrementally grow from there without having to be networking engineers or without having to buy additional pieces of just sort of network-only hardware to make the whole ecosystem work. And ultimately, we decided we need to build something ourselves. But building yourself, I mean, you mentioned earlier. You were a 50-person company, I think you said, right?

And that was an already taking on a large project launching essentially the first smart thermostat. I think E-Cabee was that just before you guys. But what made you think, okay, let's create a whole new smart home protocol. And how did you then decide to develop thread? So the actual genesis for thread didn't actually start with a thermostat. The thermostat was the thing that kind of triggered the vision. But by time we released the thermostat, the first generation, we already had our second product in the pipeline and engineering development, and that was Nest Protect.

And NessProtect had this interesting problem of, from a regulatory perspective, having to be interconnected. Just to clarify for everyone, that is the amazing, smoke. alarm. That was the first smart smoke alarm, which unfortunately has recently been discontinued, but I still have two in my home. I'm hanging on. I think mine expire this month. Ah, okay. Yeah, it's really a fantastic product, one of the best I've worked on in my career. So yes, that is the smoke and CO alarm. And of course, it's not just enough to detect smoke and CO where you detect it. You have to broadcast that alarm to every other connected alarm that's in the home. And so we, we have to, you know,

had a challenge of, okay, we have these connectivity requirements for people, for end users, for the cloud, but we also need these smoke detectors to connect to one another. Hmm, okay, what does that need to look like? What do we need to design to make that happen? And that was really where the genesis of Thread started to come to the fore. Again, we looked at all sorts of different technologies, proprietary radios from different vendors that we could have done, other network types. And ultimately, we said, okay, we could do this with this particular technology. Like, okay, well, now we need to have that radio,

plus we need to have connectivity to the cloud that increases our product costs. What can we do with just one radio, one network stack, be an interconnected alarm, and do all of this, and still achieve the vision of, okay, let's say people just start in their home with NestProtect and want to grow from there. What would that look like?

And ultimately, that was sort of the seeds to get us to thread. And again, we went back to that North Star, which is this has all got to be IP, it's got to be mobile connected, it's got to be cloud connected. These things have to be connected to each other. We need something that can carry IP, but can also talk to each other when the internet is down, broadcast these alarms, all that,

and that really was the genesis of, okay, we need an IP-based wireless mesh network that's low power because, yes, some homes have connected, powered smoke and CO detectors. They've got line voltage in the box behind them, but some don't. Some just have to run on batteries, and those batteries have to last for a couple of, of years. And so how do we do that? How do we meet that multi-year AA battery lifetime with an

interconnected wireless alarm and meet those product costs and technology constraints? And so Thread was born and that was the genesis. And what was the vision then? Like in 2012, I think you said, 2012, 2013. So what was the vision then? What was coming next? I mean, again, at that point, we already knew that there was the thermostat, even with the first generation, the second generation was already on the drawing boards. We had Ness Protect already in flight. Ness Guard was already on the drawing board. We already knew we were going to be building eventually a security system.

We had, you know, ideas of additional sensors. We had the Nest Yale Lock was already in flight at that point. So we already had a first-party ecosystem that we had a roadmap for, and we had different points on that roadmap for, here's when we're going to do this and here's what we're going to do that. So as we were building that first product, we already had, you know, points on the horizon that we knew. we needed to get to. So as we were making decisions, again, we already kind of had needles that

we needed to thread to make sure, oh, if we make this decision, we need to make sure that we're enabling that. And even at that time, we also had the Works with Nest program that was getting off the ground. And that was, of course, a cloud-based integration. But we always had it in mind that in addition to cloud-based integrations, we wanted to have local integrations. We wanted to bring weave and thread to other first-party, second-party, third-party partners for doing integrations within the home outside the cloud. So again, all this stuff was even as early as 2011, 2012, stuff that we knew we needed to get done and we needed to thread that needle correctly to make sure we got there. Well, and then something happened. Google bought Nest

in 2014. What specifically happened to the technology of threat when that happened? Yeah, it's an excellent question. And I'll do an unofficial plug for those in the audience that haven't read it, Tony Fidel's excellent book build really covers a lot of detail that we probably won't cover today about that relationship and that acquisition. But there sort of was an interesting, I don't want to say cultural clash, but an interesting clash of sort of agendas and goals. Of course, Nest at the time we were required was probably 175, 150 employees very small, still laser focused on what we were doing. Google, of course, at that time was a multi-billion dollar international company with many divisions, many different initiatives, and of course, Google's famous

for sort of controlled chaos and not top-down decision-making. So one of the interesting things that happened, of course, is IoT and Smart Home was sort of getting off the ground, but had not been acquired, and that was sort of a match in the haystack for igniting interest in that space. But it was interesting. The day we were acquired, one of the founders of Google sort of said, you know, to the extent that we are doing other things related to IoT within Google. Today, we are going to stop doing those and we're focused on Nest. And that was kind of sort of true. There was an initiative within Google called Android at Home.

And within a week of that acquisition, Android at Home was shut down. There was an opportunity for those teams to come together and work together with Nest. But what was interesting and what we learned shortly thereafter is, I'd say, within a month of that acquisition, many of us would start getting meeting requests for say, I'm working on this little IOT project, and how can Nest help me in my IOT project? And there literally was tens, if not hundreds, of IoT initiatives within Google. And so how do we make thread and weave move forward? Of course, it was abundantly clear that we needed to do what we needed to do to achieve our roadmap.

But there was a lot of, I would say, competing interests within Google that were not aligned and convinced that what we were doing was the right choice. And so a lot of us spent a lot of time and energy trying to convince Google what we were doing was the right choice. and that wasn't obvious to them, even though it maybe seemed apparent to us based on what we were doing in our vision. But ultimately, you know, the people that needed to be were supportive and we kept moving forward

and kept doing what we were doing. But certainly there was, I think, a lot of confusing competing initiatives within Google with their partner ecosystem and what they were doing with IoT that made it beyond the challenges that we've talked about already. There was a lot of internal challenges

that, again, just made that journey a lot more difficult than probably otherwise could have been or should have been. Well, and then in 2016, Google open-sourced thread, correct? Yeah, that was an interesting moment. Open thread is really another interesting part of this journey. What you typically see a lot historically in the world of standards is sort of what I call leading with paper.

You sort of have multiple companies coming together with an idea and the ideas that we're going to write a specification, we're going to write a standard. And eventually, to the extent that standard and that document generates excitement, you get multiple takers, multiple comers on that say, we're going to write a software to that standard. And different standards, bodies have different thresholds. Some require one implementers. Some require two.

Some require three. But the idea is we're going to get some number of implementers, and we're going to have a race to have them all implement this. And then we're going to do an interop fest and make sure they all work together. And when we get sort of the right level of convergence and interoperability, we're going to laugh and clap and say the standard is complete. we have this many implementations and it's done. And that is an okay way of working,

but we needed to move faster. As you knew, as you heard earlier, we had a lot of products, we had a lot of ambition, we had a lot of vision, we needed to get done, and we needed to get done quickly. And so as that process was happening

within the thread group and was not necessarily happening at the speed and the quality that we wanted it to happen, we had a unique opportunity. We had a brilliant hire in the industry, Jonathan Way. We picked up from Cisco,

Jonathan had a ton of experience in this space, and he and the rest of the team suggested, hey, I think I could write a thread stack, and we could open source it and really unlock and scale this market in a way that's different from the way this industry works in the past, and Open Thread was born, and we approached some partners, and we said, hey, we're doing this, and they're like, oh, my God, this has never been done before.

Like, this is fantastic. Can you really do this? I'm like, yep, we're doing it. And Jonathan got it done, and we went and did it, and it really, again, took thread. from sort of in terms of silicon vendors and stacks and availability from very small number. So we literally had 15 partners overnight with Open Thread. It was truly a watershed moment for thread.

And Jonathan is Google, correct? I mean, he's still. Jonathan is still at Google. He's actually also VP of Technology of Thread Group as well. Right. Now, so you mentioned Thread Group. Now, when did that develop? Because that came in.

So Nest was bought by Google and then Open Source Thread. Somewhere in the point here, this becomes, thread becomes a small industry collaboration with thread group. I think Nest, you mentioned Yale, I believe, Silicon Labs. How did all these guys come together? And what was the sort of impetus of creating thread group? So we talked about winter 2011 kicking off the NessProtect program. Again, when you sort of think about things from a supply chain perspective, thinking about Apple levels of scale of producing something,

one of the worst things you can do in the industry is have single source suppliers. Like, oh, wait, we've got a production problem. We can't move forward because we're locked into the single supplier. And so, as I mentioned, choosing 802154 was a key choice to make sure, one, we didn't have to spin up that silicon program. But two, there was already a wide range of silicon vendors that were all making 802154 solutions. And so as we launched the Nest Protect program, we were out looking for who is going to be our supplier for 15-4 solutions in the NEST Protect. And we went and talked to all the major

players. And we talked about our vision. We talked about the problems. And they said, hey, we actually believe in this. We have been hearing from our customers, a lot of the same things. We think you, Nest, have the right vision. We're on board with this. And so we started to sort of think to ourselves, maybe the right way to do this is not just a sort of a vendor-to-vender relationship. Maybe there's an industry collaboration here. And so we had a meeting even before the Ness Protect program was fully kicked off. And Tony and Matt said, no, we are going to make this an industry initiative. And so we got all the different silicon suppliers we were talking to together, NXP, Silicon Labs, etc., Texas Instruments, so on and so forth, and said, hey, we're going to create

an industry. Do you guys want to be a part of this? And they're like, yep, absolutely. And at that point, my predecessor as the thread group president, Chris Boros, was actually our first partnership lead. He was already working with the likes of big ass fans and Yale, et cetera, on the the Nest lock and other programs. And we got them together and said, hey, we're thinking about doing this thing. And here's the vision and here's the idea. And they're like, yes, absolutely, how do we be part of this? And so we got sort of this core group of, I think it was 11 or 13 companies together up and down the supply chain. And thread group was born even before we started on the Ness Protect. So sounds very familiar to something that's happened more recently. It's very much, I mean,

I went back and read the first press release from Thread Group. And it basically sounds exactly what, like, the idea was what Matter eventually became. I think the quote that caught me was, the thread protocol takes existing technologies and combines the best parts of each to provide a better way to connect products in the home. But then this was 20, what is this, 2013, 2012? Now, that's a long time ago. matter came in 2019, just for some of our listeners who maybe aren't super keyed in to the smart home space, what is matter? Absolutely. So we talked about with Wi-Fi that there's a killer app, and the killer app for Wi-Fi ultimately became the web and all the different web

properties and applications. Ultimately, matter, like its peers' home kit or OpenWeave or Zigby, is an IoT application layer. But what really was critical for Matter is matter is based on the internet protocol or IP, and because thread is IP, just like Wi-Fi is IP, and Ethernet is IP, you suddenly have a singular application layer that can connect devices that are on any of these IP together to connect, to interoperate with one or another, and work together to provide features and services in the home. Right. So thread is like the plumbing and matter is like the fun stuff, the showers and the taps, and not a great analogy, I'm sorry. Exactly. But for the smart home, It's helping run, do the fun things that we see, and thread is working behind the scenes to make sure everything connects with each other.

Exactly. To borrow a popular metaphor from Linus Torvald's, thread is the plumbing, matter is the porcelain. Right. So the main, the key here is that interoperability, right? Exactly. Any device can work with any platform. And that's something that's really been a struggle in the smart home to date silos and wall. gardens and now everyone's coming together, all the big companies, everyone's involved in matter, right? Apple, Google, Samsung, most of, the small manufacturers, the chip makers, everyone's come together and said, like, this is what we need. And thread, along with Wi-Fi, are the sort of underpinnings of matter today. Exactly. To have, you know, a home that you want to invite people into,

a hotel you want to invite people into, like they're not looking under the cabinets, they don't necessarily care what the plumbing is, the engineers do, the people making the products do, the people building the building do, but what people interact with is the porcelain. And so an application layer, that killer app that has the right ingredients, the right special sauce to bring all this stuff together to bring the visual eye was what we ultimately needed for this whole industry to really take off. I can imagine our colleagues and peers over at Apple thought they were doing the same thing with HomeKit and truth told. I think Apple was really doing a fantastic job with building out as there want to do with made for iPod, made for iPhone, a really rich

vibrant partner ecosystem with HomeKit. Of course, the Zigby Alliance have been doing this long ago since roughly, you know, the early aughts. And so we're all kind of doing our thing in our silos to try to bring this vision to life. But ultimately, to make the plumbing valuable, people needed a killer app. And, you know, let's be candid. Most companies aren't as insane as Ness to say, here's this super ambitious vision. And we're going to go do this with a limited number of people. Most companies are fast followers. They just say, you know, tell me what to do. tell me what to adopt, and then we'll go make product based around that. They're not necessarily technology creation companies. They're product creation companies. So we're all kind of doing our

thing in our silos trying to figure out a way to be successful in this market. Yeah. Well, and you mentioned Apple there. They did adopt thread, right, in 2018. So how did that happen? So we talk about plumbing and choosing the right plumbing for the smart home or the smart building. And one of the sort of elbows, pipe fittings, etc. sitting on the table is Bluetooth. And you might think, hey, Bluetooth is everywhere. It's in my car. It's in my phone. Maybe you think, I'm going to build an IOT solution out of Bluetooth.

Bluetooth is great. Love Bluetooth. But ultimately, the best way to think about Bluetooth is sort of a wireless USB. What are the applications that really made Bluetooth prevalent? It was telephony. It was headsets and audio. It was human input devices, keyboards, mice, etc. If you put a USB cable between any of those two devices, you're like, oh, okay, it's a point-to-point connection.

It's networking, but it's not a network in and of itself. And so, you know, you translate that. And again, I'm speculating here. I don't have an official line since I wasn't involved in their product development plans. The Apple guy's still in you there. Not speaking for Apple. But you can imagine a Bluetooth home kit door lock on your front door, right? And now I happen to be upstairs in the bedroom, and I want to unlock my front door because my kid has come home late, somebody's come, the dog walkers come, whatever it is, and I want to open the front door.

Oh, but wait, guess what? The bedroom and the front door are too far apart. Too far away. They can't talk to each other. Yeah. And now the question is, how do I solve that? Well, I can move closer to the door lock. But then I may as well open it manually.

Exactly. But because Bluetooth isn't a network, there is no repeater, there is no router, there is no router, there is. is no additional devices or things I can put between me and the front door to make that experience better. And so my suspicion is that, like many companies that were building their IoT strategy around Bluetooth was, we need to do something different to make this experience better for customers. We need a network. And I think, you know, I love Apple and they're great because ultimately, as we were at Ness,

they were believers in IP of, again, Gretzky, skating towards. that puck is going, and I think they reached the conclusion, like many companies in the ecosystem did, we need something beyond Bluetooth to make this in-home networking experience for IoT better. We need to be able to have a home pod, for example, sitting between me and the bedroom on my iPhone and the door lock, and hey, now they're all thread, they're all meshed together, my request to open the door, jumps to the home pod, the home pod jumps to the front door. Great. I've got a NesperTech sitting somewhere between those. I jump from my phone to the NesperTech to the home pod to the front door. Again, it's just sort of a network that grows as you add devices. Customers

don't have to think about it. And so I again think thread was a natural way for them to expand HomeKit and sort of solve some of those customer experience problems that I imagine they were having. And again, HomeKit already worked over IP because they had a cloud component of HomeKit. And so for them to adapt that HomeKit over the cloud component to thread was, I suspect, a pretty easy exercise for them to do. Okay, right. So we've we've now, got thread quite established, right? And we've got HomeKit using it. We've got We've using it. We've got Google and Apple. I think Samsung was also involved in the thread group here at some point. So they were interested. They were also very early. Yeah. So, but we still have these ecosystems. We still have

these silos of the smart home. The smart home is stalling out to some extent at this point, because people are confused tearing their hair out, trying to make things work. I've spoken to a few people about this in my time reporting on matter, but I understand there was a moment. Michelle Turner and yourself from Google, Apple, Amazon counterparts, and the Zygby Alliance all came together to figure this out. Was this kind of a watershed moment like Apple, Amazon, Google, Zygby all sitting down together and being like, we're going to work together on this? I think so. I think all the players had reached a point where we sort of had figured out what we need to do to get to the vision of where we need to go. We think we can make a bigger market. We think we

can make it better for consumers. We think we can improve all this. And so with that, the team started working together, figuring out what the best of each of these things was and how we were going to bring it all together. And as somebody who's been using it, certainly it's got warts, but I think relative to where it was, say, 10 years ago, 20 years ago, it's getting there. It's getting to the vision. Why was it thread and Wi-Fi? And why not Zigby or Z-Wave? or, I mean, Bluetooth mesh was an option that some companies used. There were other things out there. Was this a political move to some extent?

I mean, I know with protocols, there's a lot of politics. Or was it very much based on the technology? Like, was it a mix of the two? No, it was absolutely technology. We really, again, were putting the consumer first. And in order to have the consumer experience that those of us who were in product wanted to have, which is, you know, you start with this product, you start with this product, you start with this product, and add incrementally over time, and things just evolve and get better as you add those products to the mix. We, again, had a very fine line on the horizon that said, we don't want people to have to acquire gateways, converters, bridges, etc., just for the sake of converting one protocol to another.

But there was another practical reason for all that. Another North Star was, you know, at Nest we had the Nest detect window and door sensor. Oh, I love that thing so much. I'm so sad. It's gone. The resources on that were mind-bogglingly small. And the fact that we packed Bluetooth and thread into that thing, and it could connect directly to the Ness Cloud, it could connect directly to the security system, and to do all that in the form factor that we did and to have it last as long as it did in a battery. Exactly. It was incredible. And so what you learn from an engineering perspective when you're building these products is every protocol you add is call it another

128k of RAM and flash you need. And so to add this thing that is an IP, great, there's another chunk of space we need to find. And now you need to add this thing, another chunk of space you need to find. And so we already had IP and Bluetooth in there. There just was no room for anything else. And so between the customer vision of all this stuff connected without gateways, bridges, and converters, and the engineering constraints of this all has to fit in window and door sensors, it really had to be as simple as possible. And again, IP was we knew where we needed to go because we knew we needed to connect to the cloud,

we knew we needed to connect to mobiles, we knew we needed to connect to other devices. That was key. And so we didn't say no, you couldn't have these other protocols, but ultimately, the way matter brings those into the four is you sort of virtualize them into the matter ecosystem. You have hubs or things that sort of say,

oh, the Zigby light bulb over here is actually a matter light, and so you interact with it from the matter perspective, and then that thing reflects that off into the Zigibu world or the Z wave world or whatever it happens to be. Okay, well, you brought it up. So I love the story. This is wonderful. It's great to have. And it was an really interesting and exciting moment, it sounds like, for everyone in the industry to see this sort of North Star.

Okay, we can get there. But I'm going to kind of bring us slightly crashing back down to Earth because you were going, you were mentioning there that this big selling point here is we don't need proprietary. HUB's, bridges and gateways, you know, to this point, anyone in the smart home has, maybe a Hugh Bridge, a Lutron bridge, an Apple Home Hub, maybe an Echo device, maybe a Smart Things Hub. There's also so many hubs. My, if I showed you my Hub closet, you might have a heart attack. So this is a big selling point, but it really wasn't true, was it? Because Matter does need a hub, and Thread needs a hub, thread border router. Okay, maybe Nomenclature is a little, you know, some marketing speak, the word hub kind of disappeared.

But can you dive in a little bit there just briefly? Like, what is it about matter and thread that makes it different from what the hub world we were coming from to today? Yeah, absolutely. So from a thread perspective, you have different roles in the network, just as you have in the world of IP or Wi-Fi, in fact, you have access points and you've got stations and you've got repeaters. etc. You've got routers and switches and hubs, etc. So from a thread perspective, from this plumbing perspective, these different roles are networky plumbing bits. But the cool thing with from a thread perspective is their application agnostic. So if I've got a thread border router or I've got a thread router, I could be running HomeKit over it. Doesn't matter, doesn't care. I could be

running matter over it, doesn't know, doesn't care. So these aren't necessarily a sort of proprietary ecosystem vertical silo specific bits of networking, they're just functionalities that you can build into other products. They could be dedicated if they need to be. But again, going back to the Ness Protect, we had Wi-Fi in the Ness Protect, and we had thread in the Ness-Protect. So we already knew A priori that if you had a house full of NessProtects, you already had two, three, four, five border routers and thread routers in your home, and that you didn't have to think about it. You weren't thinking about, I need to buy these things. Additionally, you just suddenly built out these nests smoke and CO detectors, and suddenly you had a thread network that you didn't even

know you had. And so that's really the key to understanding thread, it's just networking, it's just plumbing, it's just infrastructure, again, no different from the stuff that you have with your Wi-Fi network. You need to have that access point. Right, and that's the thread border router. The thread border, of course, is a thing that gets you from the thread network to other networks like Wi-Fi or Ethernet or Doxus or DSL or whatever it happens to be. And then, of course, the thread router is the thing within the network that gets you, hop across the mesh. Right.

So you have the threadboarder routers, which can be anything, which is kind of cool. And then Matter, you need a hub, which would be from your ecosystem. But when this all rolled out, I mean, everyone was excited. I think the smart home community users, not just, you know, as well as manufacturers and companies were excited, were very sort of bullish on thread because on paper it checks all the boxes. And some of the first thread devices I tested were super responsive and really fast. But then when Matter did launch,

eventually, which took a while. I know in the scheme of things maybe not that long, but it felt like a while for us as we were waiting for it. There were problems. Like only a few companies had thread products available. The infrastructure wasn't really there to support the protocol, at least not in the way that perhaps matter promised. There were only a handful of threadboarder routers. Amazon and Samsung were kind of slow to adopt that protocol, even though Samsung had been involved with thread from the early days. And it really was Apple and Google. And it really was Apple and Google that seemed to be pushing it. We ended up with these weird scenarios where companies had to print like long lists on the sides of their boxes of what you could use to use this product

in your home. And it felt like, tearing my hair out, we're just back to what it was like before. So do you think there should have maybe been a bit more focus on infrastructure rollout before the launch of matter and thread? Was that a big misstep? Or am I over-egging it? I think you can look at it both ways. I think there's sort of, you know, how would have you done it retrospectively from a perfect execution perspective? But I think given the state of the market, perfect was the enemy of good. And to me, proving things out in the market is the best feedback that you can get about what actually works and what doesn't. So I think there was value in sort of having the vision right, sort of stepping and iterating your way to the product.

And again, I think missteps have happened. experiences from a customer perspective, maybe not been quite as ideal as they could have been, not quite to where the vision was. But again, I think getting that market feedback early and often was good in terms of providing real-time feedback to make things better. And of course, you can always do that within one company in the lab internally. But when you're collaborating across the industry, it's kind of hard to keep all that as behind closed doors as you might like. And so it happens maybe a little bit more out in the open than we would ideally like it to be. Well, and then And the other issue I've seen here since the rollout is you mentioned at the beginning that Wi-Fi and Thread are kind of, you know, partners. They do similar things, but for different applications. So Wi-Fi is good for high-bandwidth, high-power use, low-power, low-bandwidth is great for thread. And Matter works over both Wi-Fi and thread. And everyone has Wi-Fi. So the infrastructure really wasn't an issue there for people when a company came out with a Matter product that works over Wi-Fi. They don't have to worry if you're not.

there's a thread border router in the house. So one of the things we started to see, I think, is a lot of companies launching matter over Wi-Fi products that potentially would make more sense as a matter of a thread product, like light switches, some, you know, light bulbs. And we also started to sort of see some companies start sort of to pull away from thread because of some of this confusion and issues early on. Nanoleaf is one company that was very bullish on thread, but then recently launched his own protocol called Lightwave. A Cara is one company that now offers dual thread and Zigby devices because it's easier for the, I think, one of the reasons they've said to me

is it's easier for their consumers. If they don't end up having a thread border router, they're going to be able to work with their Zigby hub. So I guess it sort of seems a little bit like the future that some of these manufacturers were building for didn't necessarily arrive when they were, as you pointed out, you know, it was a work in progress. But why do you think that that has been the adoption of thread among manufacturers has been slower perhaps or a struggle for some than perhaps

you would have wanted it to be? I think as you stand in the home and you look out various windows, you see different views of what are ultimately the same environment. So I think you can sort of stand on one side of the house and look out one window and you see that, oh, wow, There's now 220 thread group member companies. There's now over 600 certified products and solutions. IKEA just came out and said going forward, they're thread only. So I think different companies sort of are at different places and positions in their journey.

But ultimately, I think if you've either got a greenfield and you don't have to think about the past and legacy or legacy customers, there's sort of one way to approach and engage with the market and bet on the market. But as you move across the house and look out a different window in the environment, you still ultimately are seeing the same landscape, but a different part of it with a different view. And I think as you look at some players that do have legacy, do have existing markets, existing customers to consider there's a different set of constraints. And I know one of those constraints in particular is if you had a huge Zigby installed base and you had one 802-154 radio in your product, but you suddenly want to make that product not only Zigby compatible, but throughout. and matter compatible, you now have this challenge in that to share that 15-4 radio across those two protocols is actually a bit more of an engineering trick than it would initially seem to be. And if you say that you want to support both those at the same time, that's a really challenging

engineering problem to solve. And ultimately, I think they're going to solve those. But those are tough things to make work. And I think that maybe explains some of the challenges that those players have had. So what do you think needs to happen in the first? near-term future to make thread a success for everyone, to make it easy for everyone to jump onto that green plane field? I mean, I think really we've got the key components today. I think there's one issue that probably warrants a bit more standardization and refining to really make things as good as they can be. I think Apple and Google have done some really fantastic work on their mobile platforms, but thread credential sharing is, of course, a challenge that comes up a lot

in your content with your audience, others. Will you explain just briefly what that issue is? Yeah, absolutely. So like any secure wireless network, ultimately, there's a number of different security credentials for securing actors on that network to make sure that they have permission to be there, that they have permission to act on that network.

And one of those is sort of the thread administrative credential. And it's just sort of like your Wi-Fi password that you share or if you're an iOS user and or next to another iOS user, like, hey, would you like to share the Wi-Fi network password with this other person? It's sort of synonymous with the thread administrative credential. In the early days of thread,

that part of the spec was probably retrospectively underspecified. But how do we share that credential was not something, again, we standardized and we didn't think about how that would happen. But fast forward to where we are today and matter and everything else, The question then becomes, well, I created this thread network in my home or my business, and now I want to add other devices, and maybe they're not in my application ecosystem.

How do we give people permission to add devices and share this credential in the way that we do with Wi-Fi? And so that's ultimately the problem is how do you share in a trustworthy, secure manner, this Wi-Fi equivalent to the Wi-Fi password, the Thread Network administrative password, to other entities, other people, other ecosystems? So Apple's sharing with Amazon and Samsung sharing with. Google and this comes back to the whole, all these guys are working together, but how is that really going? Yeah.

So again, it's going as well as it could. And again, kudos to Apple and Google for really saying the first place most people discover this pain point is the mobile to mobile. Like I've got iOS in my family. I've got Android. How does that happen across those two platforms? And again, Apple and Google did a great job of collaborating to make that experience great. But now we need to say, okay, it's not a mobile platform.

it's not Android, it's not iOS, how do we again standardize that? So it could be something else, another device, another controller device, a cloud service, whatever that might be. How do we share that? Yeah, and I mean, I remember when Matter launched in 2022 at the launch event, you could already see this could be an issue because the issue that it's created is that people end up with multiple thread networks in their home. In theory, that shouldn't be a huge problem for someone who's new to the smart home,

but it is something I've already seen from readers having issues with this, like so one device can't necessarily communicate to the other device. It can get messy. And I asked all the ecosystems right at the launch, like, how are you going to solve this? And they all said, we're working on it. But like two years later, you said Apple and Google work together, but everyone needs to work together.

And I guess one of the things that sort of exemplifies for me this issue, even though it feels like perhaps a small niche one that wouldn't necessarily affect a lot of users, but it required everyone to work together to solve this solution. And they weren't able to do it. In the end, the thread group had to come out and say, this is how you're going to do it. So I guess one of my concerns, and I think a lot of concerns from people in the industry and users, is this is an industry collaboration. Matter and thread is supposed to be about these major ecosystems and manufacturers all playing nicely together. But are they really?

Is this sustainable? or could this all still implode from the inside? Is it sustainable? I think absolutely yes. Could it implode from the inside? I think if you look at any one company, the answer is for that specific company, yes. Companies are constantly shifting, their priorities where they're investing.

I think if you looked back in the heydays of Nest and Google and IoT, and you said, what's the next big thing? What have we predicted AI? And the answer is probably not. wouldn't have predicted that watershed moment when Chad GPT came out and suddenly all this energy and investment and money came flooding into AI. And so it may be that some companies, one of these companies, decides that, you know what, smart home is not where it's at. We're going to double down on AI and that's where we're pumping all of our engineering and our money and our cash, and we're going to

deemphasize. But I think the use cases for the smart building space, the use cases for smart home remain. You talk to builders. Builders still see a tremendous amount of interest in people that want to buy new homes that have some level of smart home technology in it. As again, things are shifting with post-pandemic and return to office, and so it's hard to say what the office of the future exactly looks like. But regardless, people that are owning those properties still want to optimize resource usage, energy usage, occupancy usage, etc. Ultimately, that demands technology to help make those decisions and run those resources and optimize those resources. So I think, again, short answer, it continues. It has momentum on its own, but individual kind of

companies may make decisions to say, you know, this isn't as important as we thought it was, where we're out, that certainly may happen, and I suspect will happen. Well, yes, speaking of companies that might want to devalue the smart home side and focus their efforts much more on AI, not to make too obvious of a jump here, but I would like to close things out by coming back to Google and Nest and get your thoughts on how Google is implementing thread in its nest hardware today. I was very vocal about how sad I was when the fourth gen Ness Learning Therostat was launched with no sign of thread. So would you be able to share just a few of your thoughts slash hopes on what Google might be doing with its nest hardware and where you see

the future there when it comes to thread? Yeah, absolutely. I'm not affiliated with any particular partner or company anymore, including Google, who was my happy employer for many, many, many years. so I can only speculate and project. But I think to give you a short, concise, immediate answer, I think the Google product graveyard, killed by Google.com, is a very popular, well-known property. And over the years for one reason or another, Google has a lot of initiatives that they decide

are no longer important for one reason or another. So that's sort of one way that we can maybe approach and answer to the question and say, maybe, hard to say, things change. But I think if you sort of look at some of the organizational and product choices that have been made, certainly would seem to me that there is emphasis doubling down on platform. So I know that the people at Google that I still have relationships with doing thread are still very active, very well supported. The people at Google doing matter are still very active, very well supported. So what I see happening is Google's sort of saying, we're a real believer in the platform and the infrastructure,

And that's the thing that we can do best. But as it concerns product, maybe ecosystems and partners. So you saw that there was a transition of security from Ness Guard and NEST Detect off to ADT. And so now as you see ADT trucks rolling around, you see ADT powered by Google. And so that was a partnership relationship. You saw with the sun setting, sad as it was, of Ness Protect, that actually Google invested effort to make sure that first alert and the, I think what is the SC5 follow-on product actually is interoperable with the NessProtect. They didn't need to do that, but that I think was a really important investment in the customer

and making sure that there was a seamless transition there. But again, the signal is, we're not doubling down in that product category. That's partner. As far as I can see, speakers, displays, thermostats, cameras still seem to be invested in those. Will that remain forever? Hard to say? It seems like speakers with the locus of AI and an AI interaction surface, it seems reasonable that they'll continue to do that.

I suspect cameras and the awareness service model that goes with cameras are probably a big revenue driver. So maybe cameras will continue because that's a lot of high margin service revenue. But as you and I know, in the actual hardware-only business in which there isn't a service backing that, you're either a premium product or it's a race to the bottom and there's not a whole lot in between. So what happens with thermostats? Hard to say. I mean, again, I know just from how big that market is, literally. Hundreds of millions of thermostats have been shipped around the world. That's a big business.

That's a big installed base. My gut tells me they'll continue to invest in that, but it's hard to say. Excellent. Well, that's really interesting Insight Grant. Can I get you to stay around? Just for a few minutes, because I have a really interesting Vergecast hotline question that I think you would be, is right up your alley. Would you be willing to help me out with that one? Happy to take a look. No promises, but happy to take a look. Wonderful. Okay. So we're going to take one more break, and then we'll come back with that question. Fantastic. Support for the show comes from Anthropic.

Not every question has an easy answer. And the ones that are really worth asking usually come with a healthy mix of inspiration and backpedaling, aha moments, and quiet meditation. When you're working through one of those problems, you want a partner to bounce ideas off of and figure out where the deeper issue lies. That's where Claude can help. Claude is the AI for minds that don't stop at good enough. It's the collaborator that actually understands your entire workflow and thinks with you,

whether you're debugging code at midnight or strategizing your next business move. Claude extends your thinking to tackle the problems that matter. Plus, Claude's research capabilities go deeper than basic search. It can have comprehensive, reliable analysis with proper citations, turning hours of research into minutes. Ready to tackle bigger problems? Get started with Claude today. at cloud.aI.com.

That's clod.aI. slash vergecast. And check out Claude Pro, which includes access to all the features mentioned in today's episode. Claude.aI. slash vergecast. Support for the show comes from LinkedIn.

If you're a small business owner, you know that every hire counts, but time and resources are limited. Finding, connecting with, and screening the right candidates takes up valuable time you could be giving to your customers. That's where LinkedIn Hiring Pro comes in. It's built to be your hiring partner,

helping you find the right candidates faster. That way you can hire with confidence without turning it into another full-time job. Hiring Pro streamlines the entire process from drafting your job to shortlisting candidates and conducting AI-powered interviews for initial screenings. It's updated conversational interface lets you describe what you need in plain language.

Nearly 60% of hirers find a candidate to interview within a week. With Hiring Pro, you spend less time searching and more time connecting with the right talent. And instead of getting buried in resumes, you get a focus shortlist that actually moves your hiring forward. Join the 2.7 million small businesses using LinkedIn to hire. Get started by posting your job for free at LinkedIn.com slash track. Terms and conditions apply. Complex and unprecedented, the Spanish authorities are calling it. Asymptomachers.

Passengers who'd been stuck aboard the Hanta or maybe Hanta virus-stricken Dutch cruise ship disembarked in the Canary Islands this weekend, prompting the highest stakes game of where are they now since maybe COVID. Some of the evacuees, American and French, have since tested positive for the virus. And yet public health officials seem remarkably calm. We do have one individual who was a... taken to the biocontainment unit early, early this morning. And we assessed that individual.

They are doing well. Possibly because this is not the one to freak out over. Today, Explain drops every weekday afternoon. All right, we are back. Let's get to a question from the Vergecast hotline. As always, the number is 866, Verge 1-1. And the email is Vergecast at theverge.com. We love all of your questions, and we try to answer at least one on the show every week.

This week, it's a question from John from Massachusetts, who had a surprisingly great experience with a Matter compatible device that he bought, having previously had no idea what that matter was even a thing. And it got him to thinking about a very important question in the smart home, who gets your data? Let's have a listen. Hey, Virch Gas team. This is John from Massachusetts. Since last night, I unboxed a new air conditioner, and I immediately opened up my phone and had a matter notification saying that this device was ready to set up with matter. I just scanned the code on the side of the air conditioner, and it was in my Google home instantly. Who knew that was even a thing? Air conditioners with matter support. I was surprised by this because there's this little symbol on the box that says it works with the smart home.

app. There was literally an app called Smart Home, how original. This got me thinking, when I'm setting up with Matter, does the manufacturer get any data, like any user data, any usage data? Is there a reason why a company might not want to do matter, like how they don't like it when you use CarPlay in a car because they're not getting data they can work off of? And is that why Matter adoption is slowing? I'm kind of curious on this. Thanks. Excellent. Thank you, John.

Yeah, this is a great question. It's actually something Grant and I have talked about in the past because there have been some companies that have pulled back from matter. We don't necessarily know why. There is actually a competing organization or competing slash potentially symbiotic organization called the Home Connectivity Alliance, which is specifically for appliance manufacturers. And it uses all cloud-based integrations, which on the surface of it allows each. manufacturer to keep much more control over the customer's data when using their apps. And in theory, with Matter, you can set everything up in Google Home or Apple Home and never touch the manufacturer

app. So it does seem to most people that that means that I no longer have a relationship directly with that manufacturer and perhaps that may be turning some companies off from going to matter. So Grant, how would you answer that question? Is that a reason for that? Is that a reason companies not to go with matter, or is there a way for them to continue that data sharing if they want to? It isn't fundamentally a limitation. One of the precepts sort of North Stars about creating matter from the get-go was local only. And in particular, everyone is very sensitive about data and data sharing for that reason, as well as performance and making sure that everything still works when the Internet's down, all those kind of things. So fundamentally, matter sets forth the precept that your data is

important, your security is important, and keeping that all in the home by default is important. Now, that's sort of the default answer, but we talked to the NEST detect window and door sensor. We talked about how that sensor, in addition to being connected on the local thread network, also had the capability through a border router to talk directly to Ness Cloud Services. So you talked about the instance in which you've got this smart home. I'm using Apple's HomeKit app. I'm using Google Smart Home app. I'm interacting in John's case, the air conditioner.

Now, let's say that air conditioner has Wi-Fi. Could that air conditioner talk directly through the Wi-Fi network to the air conditioner's manufacturer? Absolutely. Matter has no prohibition against that. That's absolutely allowed. So if the vendor is saying the only way I can sort of get insight into customer interactions is through a mobile app, okay, that's a particular design and engineering product choice, but there's no reason to say that that air conditioner can't, again, still be connected to the vendor's cloud, sending telemetry, diagnostic data, how many cycles? what was the average temperature,

any of that kind of stuff, directly through the cloud, through the air conditioner. Let's say that air conditioner also implemented thread, because, again, thread is IP connected, just like the Ness Detect Window and Door sensor. It could also connect through a border router to the provider's cloud.

Again, share that data. So data sharing and matter are sort of orthogonal issues. They're not absolutely related, but matter says by default, within matter itself, data stays in the home, interactions stay within the home. But again, let's say you're controlling,

surface is a Amazon Alexa or a Google speaker device, there's nothing to say that those can't interact with matter within the home, but then couldn't share those bits of data up to the clouds of those providers, and there couldn't be relationships to say Google and works with Google cloud interactions with whoever the provider is. So there's all sorts of different opportunities for data sharing to still occur, and matter neither prohibits it nor enables it. It's just it's not part of what matter is about this data sharing. Is that? answer the question, hopefully? Well, yeah, and I guess, so especially for our listeners, a lot of them are interested in a

smart home that does not share their data with anyone. But with matter and thread and even Wi-Fi, like it can all be working locally in your home without a dependent internet connection, right? Absolutely. Because it's all working over your local network. Absolutely. And you can control that, as you say, you mentioned from ERO, or you can actually control that point of connection with the internet. but each device within the matter ecosystem is only going to communicate with the matter ecosystem you choose,

be that Apple Home or Home Assistant or Alexa or Google or any of the other companies that are starting to sort of, we've seen quite a few platforms now that work with Matter. So there's a lot of choice here. So I guess the answer really is your data is going to be shared with whichever platform you use and look at what their data sharing policies are in terms of. terms of what might leave your home and what can stay locally. And if you're very advanced, you can make sure no one can talk to the internet at all. Yep, exactly. But of course, there are benefits of connecting to the web, to the IP. Yep, absolutely. Yeah, you really brought

home an important point, which is with Matter, you as the consumer, get to choose, and it's your conscious choice that you are choosing what gets shared and what doesn't. And so if you're not comfortable with that sharing, then you get to say, I don't want to do that. Of course, as we said, that means things can happen on the side channel, but that is outside of matter. Right. So then it's about which companies, which air conditioner you bought. We don't know which brand you went with, so we can't answer that one. But yes, that's where you have to keep an eye on if you're concerned about data sharing. So hopefully that answered your question, John. Thank you, Grant, again, for your extra time hanging out with us at the end here. And I appreciate it.

And hopefully we'll talk soon. Absolutely. The pleasure's been to all mine. Such a pleasure, Jen, to be here this morning. That's it for The Vergecast and that's the end of the gen era for now. To read more about today's topics, check out the links in the show notes or go to Theverge.com and why you're there,

why not buy a paid subscription? This show is produced by Eric Gomez, Brandon Kiefer, Travis Laarchuk and Andrew Marino. The Vergecast is a Verge production and part of the Vox Media Podcast network. Jake Castanakis and the team will be here Friday to talk about the news

And next Tuesday, senior smartphone reviewer Alison Johnson takes the guest host chair. Thanks for listening.