Big Technology Podcast - Apple's AI Play — With M.G. Siegler

Episode Date: April 24, 2024

M.G. Siegler is a venture partner at GV, formerly google ventures, and the author of Spyglass. He joins Big Technology Podcast to discuss what Apple may have cooking for its AI-themed WWDC event in June. We discuss what Apple may do on the consumer front, including how much of iOS it's willing to change in service of AI features. And then we dig into its potential AI play for developers, including on-device processing and its own foundational model. We also talk about its interest in robotics. Tune in for a deep dive into Apple's AI options as it gears up for big announcements in the coming months. ---- Enjoying Big Technology Podcast? Please rate us five stars ⭐⭐⭐⭐⭐ in your podcast app of choice. For weekly updates on the show, sign up for the pod newsletter on LinkedIn: https://www.linkedin.com/newsletters/6901970121829801984/ Want a discount for Big Technology Premium? Here’s 40% off for the first year: https://tinyurl.com/bigtechnology Questions? Feedback? Write to: [e-mail]

Apple's big AI announcement is fast approaching. So let's look into what may be coming with one of the most insightful Apple Watchers in the business. That's coming up right after this. Welcome to Big Technology Podcast, a show for cool-headed, nuanced conversation of the tech world and beyond. We have such a good show for you today because we're going to look deep into what's coming with Apple's big forthcoming AI announcement at WWDC. We don't have the full script or really any of it, but there's enough smoke out there that we can try. angulate what's actually going to come. And we have one of the best in the business here to do it with us. M.Gie Siegler's here. He's a venture partner at GV, formerly Google Ventures, and the author of SpyGlass,

which is a great, I guess you could call it a tech blog and you can find it at spyglass.org. MG., welcome to the show. Hey, thanks, Alex. It's good to be back. Yeah, it's great to have you back. I can't believe it. We did a show. You were like guest number four on the podcast. That was in 2020. So four years later, here we are. Yeah, I was in San Francisco at the time. Now I'm in London, so it feels like a whole different world. Yeah, we were both in San Francisco. And in fact, I think I recorded that show from a ranch outside of L.A. I had rented like a room there mid-COVID to escape. And now live in New York. So crazy times. We spoke last time about how Apple's at a crossroads, and that certainly hasn't changed. I mean, it's been quite a

roller coaster ride, perhaps that we couldn't predict since then, right? Which they just sold massive of amounts of phones and computers during COVID. And then they had the sort of the after effects over the last couple months or year plus where they've been, they've seen revenue declines in four, the last five quarters. They'll probably see another one in the quarter that just ended in March. And so rough time for the company, both in terms of sales and then also everybody's kind of scratching their head. What's going on with their AI play?

And we're about to find out in June at WWDC, which I'm looking forward to. hopefully I'll be there. And there's kind of two aspects with this forthcoming AI announcement with Apple. One is consumer and the other is sort of like the enterprise or the tools that they're going to give developers to build. So we're going to dig into both. But first, let's talk a little bit about consumers. So you've been writing about this extensively and you highlighted some reporting from Bloomberg that Mark German has unearthed. Yep. He gets most of it, obviously. Yeah, he is, he is very good. And he says Apple's artificial intelligence strategy will include a slew of new proactive features. Proactive is a very interesting word there

to assist users in their daily lives. And the company isn't planning to debut its own generative AI chatbot. Instead, it's holding talks with potential partners, like Alphabet's Googles and OpenAI to supply generative AI services. I mean, you wrote about this. Read between the lines. What does that mean? So, again, as you know, Obviously, no one knows for sure, and Apple is, of course, very secretive about it. German has great sources, you know, usually and almost for sure here, I'm sure. It seems like, you know, obviously the word out there in the markets is that Apple's behind in this regard, right, in AI. And I don't know if that's, like, correct in that it's still so early days, right, for all of this.

and everything you see launching on a, you know, weekly, if not daily basis, all sorts of new pieces of technology, new companies, you know, new models that are rolling out. And so I do think that Apple is behind in terms of what they've showcased publicly. Obviously, a lot of the other companies from Google on down have been doing a lot in that regard. But what would Apple actually do here when they ship? Obviously, they're famous for sort of sitting back and waiting. for a market sort of to be seated a little bit beforehand and then coming out. And it seems like there's going to be something similar here, but it's also a different world because a lot of this is obviously software dependent and backend dependent.

And so if you, again, read between those lines, it sort of feels like they might take a two-pronged approach here where they basically have their own models that they've been working on and that they would do on device. Obviously, Apple has an advantage here in that they do their own. hardware and they've done their own chips for a long, long time. And, you know, Tim Cook, of course, in more recent times has been touting how those chips are sort of AI ready. And I'm sure they are, but they're not like Nvidia chips. They're not, you know, necessarily built just for AI purposes. They're usually multipurpose chips. But they've been building some that seem like they're

going to be, you know, used for those purposes. And so again, I would guess, based off of, you know, all the reporting, German and a bunch of others, that they would do the two-pronged approach of having both AI that's going to be on device and they'll tout sort of the privacy aspect of that and also speed and some other, you know, advantages that they can take advantage of because they, again, they built the hardware. And then I think they would outsource. And this is where the reporting on, you know, talking to Google and I have no inside information or no knowledge about what's going on with that. But it does make sense at a high level of like that they would potentially partner with a third party. You look at the early days of the iPhone, right? They partnered

with, again, with Google on YouTube and Google Maps and Yahoo on things like stocks. And they still partner with Yahoo on things like stocks, right? So I think that they could pretty easily just slot in there a partner. And maybe it's Google, maybe it's open AI. Maybe it's a combination of all of those for things that they want to do in the cloud, some generative AI, you know, capabilities that maybe they can't do or feel like they don't need to do necessarily on device while on devices use for different types services. And so effectively, the net output of that is going to be two different things.

One is going to be product changes that Apple will make, and then others will be products that Apple will enable others to build. The changes that Apple is going to make, to me, is really interesting. So they have Siri, of course, so you would imagine, let's talk about for a Siri, it would be somewhat souped up. But the other thing is, if you use, let's say they're partnering with these other companies to create AI experiences on the phone. that changes the way the operating system works. And last I checked, iOS is a pretty good operating system. I would call it the best operating system of all mobile operating systems.

So how fundamentally does Apple change or potentially how big of a risk that can they take in changing their operating system? Because that could potentially have them falling behind others in the space that they lead. It almost reminds me a little bit of like the Google dilemma of how much do you want to change search. Do you view it as kind of a similar trade-off question? Right. It's sort of standard innovators dilemma type stuff, right? Because like you say, Apple's in sort of a key position.

And certainly now the Department of Justice says they're in the key position, you know, with the high-end smartphones, however you want to define that as they do. But Apple's operating system is a huge part of that, right? And iOS is, you know, considered by many to be sort of the pinnacle of smartphone operating systems and maybe all of operating systems. and maybe all of operating systems. And so to your point, how much do they want to sort of monkey with that, right, and change it? And there's, you know, low-level stuff that they obviously will do things that, you know, even Android's done, you know, earlier with things like generative wallpapers. You can imagine that being like very low-hanging fruit, right?

They'll do, I'm sure they'll do something along those lines because it's sort of these table stakes type stuff that you need to do now and fun little things for the user experience. But there's much deeper stuff that they can do. And one of the things I actually look to and think about there, and I don't think there's been much reporting on this. Maybe German said something about it in passing. But I actually look to sort of what Microsoft is at least rumored to be doing. And I know Tom Warren over at The Verge does a lot of reporting on this,

which I think is pretty good and he seems insightful and to have good sources on this, that basically gearing up for the next version of Windows and Microsoft's going to go all in an AI, which is not surprising, obviously given their open AI partnership and everything else that they've been doing. you know, in recent quarters, but so what if Microsoft sort of reconstitutes Windows to be predicated around AI? And again, you talk about iOS being in a dominant position, you know, in terms of being a great product. Windows, you could, you know, it's been a long time since I use Windows regularly, but you could say, you know, it's sort of long in the tooth. I think it's been, you know, getting better in recent years as it has become more modern. And as Microsoft suggests less, you know,

built the entire company around it now in the Satya Nadella days, which is a good thing. But if they really are going to sort of base the entirety of windows around this, they might do something like you know, have something running in the background that's always sort of recording. And that sounds sort of nefarious. But, you know, it's something that a that an operating system can do is like, you know, sort of basically monitoring what you're doing on your computer to be able to recall at a moments notice, if anything, like web browsing, Word documents, all different sorts of things. So now take that to the Apple side. Imagine if Apple did that for iOS, again, they have an advantage there maybe because they have such the privacy angle and they've been talking about that

nonstop for many years at this point. So what if they basically touted like, look, we can, you know, obviously we, iOS runs the iPhone, we in a very secure way on the secure enclave or whatnot. We will record everything that you're doing and need to, if you need to recall anything on a moment's notice, we'll be able to do that. And they can do that with third-party apps. That will be a little bit complicated. And again, I don't know that they're going to do this, but you can imagine that would be a big, big thing if they did something like that. There've been a few startups who have talked about doing that. But the problem is it's very hard, obviously, again, from a privacy perspective, but also just a technological perspective to be running fully in the

background all the time. But you can do that if you, if you can control the OS. And the OS is are obviously iOS, Android, and Windows, as we're talking about. And so all of those players could potentially do something like that. And that's something that would fundamentally change the iPhone, obviously. Now, here's my perspective, reading between the lines of where they might go. I mean, German talks about how they're going to be proactive. And we just had Jack Clark from Anthropic here talking a little bit about how they are moving from sort of chatbots where like you give a command and you get information back to agents that can sort of start to have dialogue with you and ask you questions. And I wonder if Apple in

this release will skip the chatbot stage and go right to the agent stage because they do have, they have your health data. They have your, you know, basically all the data that a company could want to have on you. And they have this privacy thing going for them like you mentioned. So maybe they just go straight to agent. And their AI play is like, you know, almost like this super enhanced Siri that will end up like helping. you go about your daily life, like maybe the vision of like some of these assistants in the past, just borne out by Apple. What do you think about that?

I think that's a really interesting concept. I mean, I do think, you know, zooming back, obviously since chat GPT, sort of the chatbot have been, you know, getting a lot of buzz and hype and a lot of it's warranted. Obviously, they're super interesting. But I do think, I do wonder if we're not in a stage where that's sort of a more temporary thing, basically to get to a means to an end, right? Like, what are you actually trying to do? So we're talking to the chatbots because it's sort of a known way to interact with a computer that's not necessarily just doing, you know, a search prompt, and it's a more sort of natural way

to talk. But eventually, you would imagine that, yeah, you use a whole slew of things, including voice, including, you know, images and all different sorts of things again. And so does app sort of look at that and say, yeah, the chatbots are interesting, but we can probably, yeah, just jump ahead of that. And like you said, sort of do a proactive agent that doesn't necessarily need, you know, a one-on-one pop-up chat bot thing, chat box for you to talk to. That said, I do think that they have to have an answer to, you brought it up earlier about like what they do with Siri, right? And, and Siri gets, you know, dragged through the mud a lot, Larry David, you know, famous clip going around. in recent weeks about him just screaming at it on curb your enthusiasm. But the reality is like they were a first mover there, which is sort of rare for Apple, right? And it's been many, many years now. And it's just not where it needs to be in terms of where we are now in our AI world.

And so does Apple sort of, you know, come back with a vengeance with some sort of very souped up version of Siri? Do they even still use the name Siri? I mean, probably because it has such, you know, brand awareness. now, if not exactly great, uh, you know, a halo effect. But, uh, but, uh, I, I, I sort of wonder if they don't sort of also do that. So they basically have a version of Siri that's, you know, again, maybe not just chat bot full on like chat GBT, but it's sort of a voice based assistant as it is now, but running as effectively what is a, a chat bot behind

the scenes, but souped up. And I think that that is super compelling because that's not just potentially phone. I think if they do that on Apple Watch and they do that on AirPods, you know, I think that that becomes really, really interesting. And now here's my sort of guess on that front. I think they're going to bring in for the first time other companies' models into Siri. So we talked about like this licensing and I kept thinking like we're good licensing or partnering with Google Gemini or Open AI land because effectively like, okay, maybe it's part of like an Apple agent, but you know, if they end up like, you know, there's little

distinction between like that and an app unless you bake it into the operating system and the way to bake it into the operating system is with Siri. Do you see that as a possibility? Yeah, for sure. And I think there's also a bonus side effect of that, which I've heard a few people sort of mentioned in passing, but I do think you could see easily play out this way, whereas obviously a lot of these companies, big and small, big companies and startups, are getting dinged a lot for, you know, what gets returned from the prompts, right? And what if Apple can basically outsource that and say, like, it's not, you know, it's not our data that is being returned to you.

And obviously, they need to do some work there because they can't just have it, you know, spitting out things that they deem to be sort of inappropriate that they think would, only because they would be concerned, obviously, that it would ruin their brand, you know, and tarnish their brand in some way. But, again, you could see a world in which they use third parties. integrate third parties to get all of the functionality that sort of people come to expect from it, but the same time they can say like, yeah, that, you know, you're sort of using this at your own risk and you might get some data back that's, you know, not up to the Apple

standards, but we're willing to work with third parties here type thing. They'll obviously massage it in a much better way to say that, but along those lines. You're going to be so excited to access OpenAIs GPD5 to the Apple AI iPhone. yeah something like that you can so yeah so uh there's also so let's then talk about this is obviously happening at a developer event and that i think will be the main thrust of the news i think a lot of people will be like so my series getting smarter and the answer will probably be yes but what apple is really going to want to do is enable developers to build generative AI applications on their phone this idea of an AI app store has been something that's been circulated and

seems like inevitable because what we've been hearing about has been again the fact that you this up that Apple has a heck of a lot of compute that they can do on device and that's going to enable developers to build things that they haven't been able to build before because a lot of the compute right now is happening on the cloud and what if they can actually enable developers to build more powerful, faster, smarter AI applications using some of the compute power on the phone itself? Is that going to be the most important thing that they introduce and where do you think that's going to go? Yeah. Again, I could see a two-pronged approach here. where they can basically, you know, maybe strike a deal with one or many of the, you know,

LLM companies and say, you know, you can work with their APIs. We've made it super easy to do that to build your applications through, you know, our own, you know, development areas. Or we're going to offer our own on-device capabilities for certain types of AI. You know, again, not for everything necessarily, but, you know, they would lay out for X, Y, and Z. You don't necessarily need to go up to the cloud. It's going to be much faster, much more secure to keep it on device. And you should build these types of applications are the ones we know about.

But we love to see what you guys come up with and that type of thing, right? And you can imagine an AI kit and all different sorts of stuff that they launch in order to leverage that. Yeah, AI kit seems like it will definitely be something that manifests. Yeah, for sure. So talk a little bit about fair. you've written about this ferret research paper that Apple has released. Yeah, again, you know, going back to the notion that Apple is obviously super secretive and it doesn't want things out there before the big reveal.

AI is an interesting element to this, right? Because in some ways, they've been getting, Apple has been getting dinged because researchers famously want to be able to put their work out there, you know, to work with an open community and let it be known of like what they're doing. and being collaborative and Apple's, you know, obviously hesitant to do some of that work. But with AI, because of just the, this is the state, the normal state of being for that type of research, they have been putting some stuff out there. And one of it is this, you know, this supposed ferret project, which they published, I think,

a couple of papers on now, one of them recently, I forget where through, was it Northwestern or, I think it was Cornell, actually, they published it through. And so basically you can go in and look at this paper, which is public and it talks about, some of the AI work that's being done internal to Apple, which again, they're publishing in public. And so if you read about this ferret project, it basically sounds to boil it down very simply, they've created a system that is able to parse UI, so application, you know, user interface to be able to deduce what's going on, like what's in the image. And, you know, that sounds sort of obvious. But it's actually, from my understanding, which is a layman's, you know,

understanding. It's a pretty complex thing to do. I've known some startups in the past that have been trying to work on this type of product, this type of product and problem. And again, it's not as straightforward as you might think it is because there's all different sorts of screen sizes and various permutations that you can do that change, you know, with, with UI. And so in Apple system, they, you know, presumably will be able to, you would say take a look, but it's obviously all happening behind the scenes. They would be able to know what is going. on within the UI. And that's important because you could imagine if they could open this up either to their own applications or eventually to third parties, you could sort of see just like

they do right now with the shortcuts application on every iOS device. They could allow this to sort of like in real language to say like, hey, send this to Instagram and send it as a direct message, you know, to my wife. And then they could let the ferret system go in there. And even even if Instagram didn't do anything specifically to sort of create to tap into APIs that are already in existence or build new APIs to allow developers to do that, they could potentially, Apple would allow a system to interact with a third party application that they don't really know that much about just by allowing the machine basically to use the AI, sorry, to use the UI as if a regular user would. Yeah. And just to think back towards the agent thing, I mean, you could totally again see. these proactive notifications. You take a picture, so maybe juiced up series, says, you know, tell me where you want to send it. You speak into the thing, you know, DM it to my family, post it on Instagram, send it to X with

this text, and then there you go, and it does it. Yep, yep. And again, I think the key to that is you can do some of that now, obviously, again, through shortcuts, but that involves APIs that have to be built by someone. And with this type of system, you know, some very one person, you know, shop could build an app and they don't need necessarily to build all these APIs and need all these developers to help them out and all these new capabilities. They could just give the app over to Apple and just, you know, allow this type of ferret system to be able to work with the app using AI. It's, you know, it sounds like almost too simple, right? It sounds like too obvious, but I think it could be pretty impactful if they're able to get it working, which sounds like they are.

Yeah. I mean, it is sometimes the simplest things in AI. You just like it's some of these solutions that we've been throwing like or the tech industry has been throwing like different technological fixes toward. But they're just too complicated with AI. You can just make it more simple. And that's sort of the magic of this stuff. The another thing that I've been wondering about Apple. And I'm curious if you have some insight into this is they've, I mean,

Whether they're behind or not, we're going to talk about that sort of as we end this segment. But they certainly don't seem to ship fast. And you do wonder, I've wondered whether they have the culture that's, you know, sort of been able to build these things. Obviously, they were unsuccessful with this self-driving car. To me, that was a cultural issue. And I wonder if they have the culture that's going to be able to build these products because it requires so much more collaboration, sharing of knowledge. Like you said, publishing papers, which they don't. usually do and now they're doing. So maybe they've shown some signs of adjusting. So there's a

culture and then there's also a talent question because they haven't really been stacking up the AI engineers the way that Open AI and Google have in the past. And I think, yeah, you mentioned I'm going to pull it up right here in your writing. The real, the war is for talent more than so for products at the moment. That's the real risk to Apple. So culture and talent, what do you think about that? Yeah. I do think that that's, you know, one of the key areas of downside to focus on with them for all the reasons that you mentioned. I do think, you know, like we just talked about, they're doing work that's maybe outside of their comfort zone, you know, to necessarily try to attract the talent with regard to publishing papers and, you know, bending to some of the

industry norms in that regard. But yeah, they, you know, for what, I don't know that it's totally fair or unfair, but they're not right now known as one of the players that's attracting. all of the top AI talent, right? I think they've made some acquisitions in order to get some of that talent in-house. You know, the reports, again, indicate that maybe a lot of the, not a lot, but maybe some of the car project people have come over from that. And, you know, there's tangential things, right? Like with computer vision and machine learning in that regard,

that obviously a lot of people were working on that on the Apple car project. And now they've come over potentially to work on other AI projects for Apple. So they definitely, obviously, of course, have talent. they're, you know, almost a $3 trillion company. They're going to be able to hire who they want. It's a very specific type of talent, yeah. It's a very specific type of talent, right. And there's a lot of other players, big and startups out there right now who are all vying

for that same talent. And so, I mean, I have not heard what the sales pitch is from Apple, obviously, in terms of trying to bring those, that type of talent over. You could say, I can imagine a world, though, where they say basically like, look, we have over a billion active devices in people's hands. in people's pockets. How many other people can say, how many other companies can say that? Maybe Google, but, you know, even then it's they're partnering with Android, with Samsung, and with, you know, other OEMs, you know, not making all of those devices by themselves,

whereas Apple does that. And so they have an advantage in that regard. And you can see a world in which they basically ship their AI, say with, you know, iOS 18. And again, they ship it to a billion plus devices overnight. And then all of a sudden, you have access to camera information. You have access to location information. You have access to videos. You have access to all different sorts of stuff that basically no one else can get to. When we talk about, you know, these companies, and I know we're going to talk a little bit about this working on hardware, why do they want to do that? I think because there's a wide range of data input that they would want that's not just text box based. Yeah, when Gruber was on,

he talked about when Apple has a high priority project, they tend to just swarm it. with talent and it does seem we were that he basically said that's they didn't do that with home pod which we and i talked about last time you were on and like now seems like a distant memory even though they've revived it in some disappointing way uh but it does seem to be like this is what they're doing now they know they have to execute here and they're swarming the talent and really directing it from places that it was somewhat like wasted before which is the self-driving car project yeah and And to your earlier point, like if they can get over the cultural hurdle that they're viewed as, you know, too insular, too, too close-minded to work on some of these, you know, newfangled AI projects, if they can get over that hurdle, I think that they can make a pretty compelling sales pitch again to a lot of would-be, you know, AI experts and researchers to say, no one else can offer you billions of devices sort of almost overnight, right? because so many people will update iOS to the new version when it comes out immediately.

And so that's a pretty compelling pitch if you're an AI researcher who wants to do something that is unique and can be done at a scale that's no one else can match. There's one more thing I want to get to. I don't want to miss on this conversation, which is that Apple is rumored to be developing its own foundational model called Ajax. Maybe it's not there yet. It seems like no one's really talking about it. And it's interesting that they're doing that in conjunction with bringing in

partners. Is that something that just doesn't seem ready for prime time or any thoughts about, you know, why that's not the main thrust of this? Hard to say. Again, this is all based on reporting, of course. And you're right that it hasn't been, you know, as at the forefront as some of the other reporting about their AI projects, including partnering with, you know, other, other players. And you would think that maybe those two are related in some way. And so that's why it's on the back burner. But Apple, you know, occasionally does surprise, right? Still. And And maybe they do have more behind the scenes and they're sort of letting on or have let, you know, got out there in some capacity. And so while this project has, yeah, been rumored about it could be as simple as like maybe it is farther along than we know about or that's been reported on.

And maybe they are ready to go with it. Maybe it's also related to the on device elements. And maybe that's what it, they'll launch with specifically. As you know, many of these LLM projects, you know, are now splintering off where they have, you know, the premium version, the big version, or they have the smaller on-device, right, that are meant to be on-device versions. So maybe Apple launches with the on-device version of Ajax, you know, as their sort of first foray into LLMs. And maybe they have a bigger version that they're working on, which won't come out now. You could see a bunch of different ways that could go. But again, there might be able to surprise

with something at, they'll obviously surprise with something at WBC, but is it going to be, yeah, that big LLM? We'll see. So let's just take, before we end the segment, your gut on where Apple is behind ahead, sort of with the pack right now, and where do you think they're going to be after WWDC? So my general question, but go forward. But my general take on this, not with any, you know, just trying to read a bunch of different tea leaves, you know, and bringing them all together to form a point of view is essentially I don't view, I don't worry so much that Apple is very far behind.

again, we hit on this earlier. I think things are moving so fast, so rapidly, with new things shipping all the time, including LLM updates, that I'm not so sure that Apple's sort of missing the boat here. If anything, I'm not convinced that it's still not too early to launch some of this stuff, except that I think there's no way that Apple could get away with not launching something. There's obviously benefits to being in public and having people know what you're doing with regard to AI because, again, for attracting developers and attracting talent to the company again,

as we just talked about. But I'm also, again, not convinced that it's still not too early for all of this. And I think that that matters from a product perspective. I think we've seen a lot of fun stuff launching, you know, from chat chip-t on forward. But again, it's all a moving target. And so Apple, as we talked about, as you well know,

like they like to sort of come into a market that's, you know, sort of been meandering and really firmly establish it. And I'm not sure that that's the case here. I think that it's still going to be meandering for a while, just again, by the nature of how fast things are moving. And so again, Apple needs to do something.

Certainly Wall Street believes that they need to do something. If you look at Apple stock prices, what Microsoft stock prices and all that. And so they're going to need to do it right now, but I'm not sure that they absolutely would do it, all things being equal. And I think that what they're, want to get out there is ultimately the end product and what can they do there that's unique

compared to what's in the market. There's so much more Apple that I want to talk about, including a surprising play in robotics, the stock price, which you just mentioned, and declining iPhone sales. So why don't we do that after the break? We'll be back right after this. Hey, everyone. Let me tell you about The Hustle Daily Show, a podcast filled with business, tech news, and original stories to keep you in the loop on what's trending. More than two million professionals read The Hustle's daily email for its irreverent and informative takes on business and tech news. Now, they have a daily podcast called The Hustle Daily Show, where their team of writers break down the biggest business headlines in 15 minutes or less

and explain why you should care about them. So, search for The Hustle Daily Show and your favorite podcast app, like the one you're using right now. And we're back here on Big Technology Podcast with M.G. Seagler Venture Partner, a GV, which is formerly Google Ventures. He also writes Spyglass, which is a great tech site. You can get it at Spyglass.org. Very quickly, let's talk about this robotics thing. While I was doing research for this interview, you have been on this, right?

There's been a tremendous amount of smoke, shall we say, when it comes to the robotics play that Apple might be getting into. And it sort of breaks down into two different categories. One is like an actual home robot that Apple has been talking about developing. and two is there's like a almost like a FaceTime robot with a robotic arm that will feel like when you're speaking with somebody that it's like kind of moving along with you and it seems like Apple has a lot of this technology already developed but isn't quite sure if it's good enough or Appley enough or or going to bring in enough money but it is interesting that they are so

interested in pursuing robotics we also know that Amazon is interested in pursuing robotics what's going on there? So to me, this is a very obvious extension of Apple searching for the quote-unquote next big thing, right? We talked about the car project not being shut down. That was at various times over the past decade going to be their next big thing, right? It's one of the automotive business and car sales are one of the few things at volume that can sort of move and move the needle for a company like Apple, right? Because the iPhone is so massive, such a success that there's. very few products that they can launch, including the recent launch of Vision Pro, no matter how well

it does out of the gate, it's not going to be able to move the needle versus what they ship in terms of revenue with the iPhone. And so, again, the car project now being gone and Vision Pro being out there, they have to move on to see what is next. What else can they do that moves the needle from a revenue perspective? We're sort of running out of things. There's banking right? And they are doing stuff with Apple pay that I think is going pretty well. But in terms of, again, it's Apple. Their DNA is to ship products, not be financial services, services, businesses itself is massive for Apple. And I think that that's sort of interesting. But again, they want to ship products. And so what's the new product that they could do? Robotics would be one sort of more obvious element. Obviously, it's a big, all-encompassing type of situation. It could mean a lot of things. hit on two of them. And I do think that in some ways, I had written about this for a long time because I felt like, and this came from, you know, we kicked off talking about when we last spoke in the pandemic. I was thinking about that a lot during the pandemic, that it was weird

that Apple didn't have a device in the home that was great for FaceTime, right? You still had to use your phone, which was great, obviously, and an iPad is also great. But it just felt like Apple had all of these different elements, including, you know, Apple TV. Back in the day, they had routers right which they've discontinued but they had like all the pieces in place to build a really great living room device and that it should have included face time now they've added face time on the software side to apple tv since then and you can use it with your your iPhone and it's pretty well done but you could have just imagined just selling some sort of unit that's basically the quote unquote face pod right where you basically put a put a device in your living room and it's

hooks up to your TV and it's probably an Apple TV as well. It's a home hub. It's Apple TV. And it has a small camera. And others, of course, including Amazon, as you mentioned, but even Facebook slash meta at one point was doing this, right? And they had some level of success. Even though they were probably the least trusted company at the time to be able to launch something like that, I think that they were in the right place, right time for a little bit of time during the pandemic. So that's one element of what they could launch. The other stuff, like a more generalized home robot, Some of the reporting, which again, I think was Mark German talking about it. It almost sounds more like Rosie the Robot Jetsons type, you know, housekeeper type thing.

Obviously, Amazon tried to buy iRobot, the Roomba Maker, and, you know, had to move on from that after, you know, regulatory concerns there. But they're doing other stuff. They have their own robot in the world. And so you can see a world in which Apple wants to go into this space. I think that they have a lot of, a lot of the pieces, again, sort of that are not brought together, at least forward facing for us right now, but they have a lot of the pieces, including some players from, you know, potentially that car project, again, which could be brought in to do some sort of home device. And again, if you just think about, you know, 10 years out, what should Apple be

doing to move the needle and ship interesting products? I think it's pretty compelling to have the idea of some sort of home robot, you know, that's running an iOS like operating system and interacts with all of your other devices, you know, beyond cleaning, doing different sorts of stuff. Yeah, I'm with you on robots and I think it's going to, robots is an area that's going to get a lot of attention, both from investors and from the big tech companies, maybe especially the big tech companies, because these large language models, they basically, you know, because you can converse with computers now and they can, they're multimodal. they offer much better functionality than you could have had previously with a robot.

And the big glaring hole for them is that they don't really understand the real world. So it'll be like a research frontier that they're going to. So it's sort of like both it will help them get better by being out in the real world and they're better because they can rely on this technology. And so we see Amazon and, you know, I'm sure meta has something. Apple, you know, this is going to be an area to watch. Yes. And that sort of speaks to, you know, the notion in AI itself where,

you know, there's some school of thought that believes that you would, either you'll never get to AGI, obviously, many people believe that, or that you won't get there until you do some sort of personified or in the real world version of a robot like we're talking about, where it's able to not just, again, get data inputs from people inputting stuff manually on a computer, but it's out there in the real world interacting with it in order to learn different types of stuff that you can't do just from, you know, being on a computer. And we've talked so much today about how the, next step for Apple or Apple getting into the next big thing is so crucial for its future

prospects. But then you kind of look at the present and you're like, oh, is the house on fire? I mean, that's a little dramatic, but it doesn't seem good in terms of Apple's core businesses. So we're talking just now, you know, when I, there's been a report where Apple's had a 9.6% year on year drop in iPhone sales and it's losing ground in China in a big way where there's band there's you know ostensible bands or you know wink wink don't use the iPhone in certain workplaces and Huawei is getting better and meanwhile like you look at the core numbers for Apple and it's it's not good I mean they're their Mac and iPad are struggling to sell the iPhone isn't growing the revenue has been decreasing and we talked earlier about like maybe

this is like comparisons to COVID, but we're in 2024, like you have to start growing on your 2023 base at this point, and it just seemed rough for the company. So what do you think about the core business of Apple? It's pretty concerning, I would imagine. Yeah, I mean, so the way I think about this is tied into a lot of what we're seeing with, you know, some of the backlash on the developer side and obviously with the regulatory environment and various different bodies all around the world that are sort of hitting Apple pretty hard right now, right? Because they view them as, you know, both dominant in some capacity, but also not willing to play ball with, you know, the various different regulations that, again, various governments and entities are trying to put

in place. And I think to me, a lot of that goes back to exactly what you're sort of talking about in that and what we had talked about previously with the iPhone was such a great business and remains such a great business. One of the best businesses ever. from, you know, revenue from profit, it might be the best business ever. And how do you top that? The answer increasingly has been that you can't, but instead you have to come up with other ways to sort of grow the pie, grow the business. And I think a lot of that, when I look at Apple these days, the only thing that's been

growing the past, you know, several quarters now has been services reliably, I should say. as you note the other businesses that are big you know are running into just the law of large numbers where it's very hard to grow beyond the the extreme end of where they are in terms of shipping number of iPhones and again as you note the pressure from China and these other areas where they could potentially grow a bit more but there's there's other types of pressures on that so when I see Apple pushing back against all this regulation and be and you know many of us looking at them just like, why don't they just like throw a bone to some of these folks, you know, and just make some tweaks to the app store that, you know, would, would maybe ding revenue a little bit, but would, you know, garner a lot of good well with developers and, again, with regulatory bodies. But again, they look at, I think they look at services and say, this is the only thing we have where it's, it's a unique business because it's sort of compounds versus their other businesses, right?

Like as the user base keeps growing, they also are adding things to the services layer. And so they have all sorts of new things that they can just keep piling on to keep growing. And so eventually, I would imagine services is, you know, by far their second largest business after iPhone. And so that's the way that they're growing it now. And then we talked about all the other things that they could potentially do, banking and, you know, growing that more. And then robotics and different types of devices. but it's going to be the answer is going to be it's got to be a bunch of those things together and yeah at doing that while facing all the headwinds against device sales um and it's it's a big

challenge i mean yeah they the the max the iPads they're launching new versions you know there was a report about the new m4 chips already even though we just launched the m3 chips and you know trying to get the word out there but when you look at the way that you know you have to run as a public company, especially a company like Apple where the stock is so heavily owned by mutual funds and all of these big entities and it's such an important stock for so much of the economy that they need to keep growing. How's Microsoft doing it? They're doing it through their AI narrative and that they're forward thinking, right, that they're a growth company. And the way that Wall Street seems to be looking at Apple right now is that they're not a growth company. They're an old school sort of

becoming a dividend, right, company and just like they're, they need to change that narrative. And And I think that's why all we're hearing about it, WWDC is AI, AI, AI, AI, AI. They want to show that they're going to go back into growth mode. And I think that that's a lot easier said than done for all the reasons that we just talked about. That's right. We have a few minutes left. I want to hit a couple more topics if we can. So first of all, you main pin.

We're not going to go, you know, sort of too deep into it because at this point, anyone who's listening to this show has probably read the reviews, listened to our Friday show about it, March watched Marquez Brownlee, you know, rip it a new one on YouTube. But I also want to talk to you about your view about whether the new AI era sort of warrants, and you write this, warrants some sort of new hardware paradigm. Do you think that the new, I mean, it's been a lot of chatbots. We're seeing agents. Do we necessarily need new hardware for this AI era? Or is that just something where like people are like, ah, AI, hardware. Yes. You know, throw money at it and see it happens i think it's i think it's obvious why uh various different players are trying to do that

including humane um you know there's a new one that launched today called limitless which seems interesting ninety nine dollar pin a pendant type device um which you just reordered yeah which i just preordered because it does look it looks compelling it's a good price uh humane's little expensive as has been talked about a lot um but then of course there's what's rumored with open a i with Sam Altman and Johnny Ive potentially, you know, teaming up to work on some sort of hardware. My view is that I'm not sure that we're anywhere near the point where we need a new piece of hardware yet. I think that, again, all this stuff is moving so fast. All this new stuff is launching on a weekly basis where it's just there's too much, too many moving pieces to know exactly what you should be building for.

And I think, frankly, that's part of what Humane's problem has been, right? They started a long time ago before chatGBT sort of blew up. And so they built like the projector and they built the camera and they built all this other system when I'm not so sure that what they shouldn't have just built. And obviously this is easy for me to say in hindsight. But I do think like there's there's something compelling to be built there. It might just be like a super simple, you know, pin that only does chat interaction. And that doesn't try to do all this other stuff.

You bring the price down a lot. You bring the complexity down a lot. It tethers to your phone. It doesn't need its own cell service. You launch that in the success state when you have, you know, a second or third iteration. You sort of launch that stuff. You get why they launched what they launched with. They come from Apple.

They wanted to build an all in, you know, all in one, all encompassing type of device. And it's obviously a noble effort to try to do that. But I just don't think that that type of device is going to work out of the gates. I think that, again, everything is moving so fast. I think that what you would see in these early days is if anything takes off, it's going to probably be from one of the larger players, Apple being a prime example of that, right? Again, we talked about this a little bit already, but if they can get some of these AI models working on the phone and then eventually working on the AirPods or on the Apple Watch, and they can do that all on device, they have stuff

already that's working on device on the Apple Watch right now with Siri. Again, Siri's not very good, but if they eventually launch a better version and different capabilities, you could see where that type of thing gets supercharged by all of this. And so, Again, I think that there's probably an avenue down the road for new pieces of hardware that are, you know, using AI. But I come at this, again, from that flip side that we had hit on a little bit earlier. But it's like, why are some of these companies trying to do that? I think because they know the advantage that someone like Apple has in having billions of devices out there. When they launch their AI products, they're going to have such an advantage in having these in people's pockets at such a scale that no one else can hope to match.

And then they have to hope to like they can work with Apple on some of that, right? And maybe that's why they're doing it on their side, right? Because they want something from Apple to be able to work on the device scale that they can offer up. But I don't know if any of these, you know, the rabbit device also is, you know, coming soon, I think. And that got a lot of buzz out of the gate. I have a hard time seeing any of these things taking off in a major way early on, again, just because it's so early. So maybe in a few years we see something that's a compelling use case. but right now it's going to be sort of phone and or watch and or, you know, AirPods dependent, I think.

So when Marissa Meyer was on, I spoke with her about your post, about how consumer is basically done and don't bother trying anything consumer. But it's interesting to hear you talk about the hardware thing because it's almost like you say you view that there's more opportunity to do stuff in hardware than there is like an apps and software. Is that accurate or is that kind of, anyway, you go ahead. they're all they're both hard right consumer in general is very hard for the reasons yeah as as I've talked about as many people talk about it I mean it's just like it ultimately boils down to time and a day and we all are so busy we all have the services that we use already and so things not only have to replace those they have to be so much better in order to replace those and that's why like you know things were sort of interesting in a weird way during COVID times because it unlocked new time throughout the day right where people would try new stuff and things like club house and all of these types of services that sort of blew up because people had more different pockets of time that they weren't used to having. And so when I, again, going back to the hardware thing, we hear a lot that people are waiting for the new platform, right, the next platform and waiting for what Apple's going to launch.

But the reality is that even with the Apple Watch, which is at a pretty impressive scale, and even with AirPods, like these things, and with Alexa, which is at an immense scale, These things have not become platforms for, you know, new services built by third parties to launch on. And again, that to me boils down to just time and, you know, replacing other things that you're already doing in a day. So you're not an air chatter. I do have air chatter right now. I've been trying it out. Yeah.

I think there's something compelling about it, but I would be skeptical that it sort of gets to a scale, you know, over an extended period of time. Lots of things blow up and are big for, you know, a few weeks, and then they sort of dwindle away for, you know, very obvious and inherent reasons. But it makes consumer both software and hardware very, very, very hard. But I think you're right. Like hardware is compelling to a lot of folks because it does potentially unlock a new time variable if you can get enough people buying it and using it. And there's the sunk cost fallacy where you buy something and so you want to use it, right? And so that's natural. Okay. And then briefly on Amazon, you mentioned the echo device. You also wrote about how they're stopping to pay rewards to third party developers. Basically, it seems like they're throwing up the flag on third party apps and skills on the echo device. Do they have a strategy there anymore? Or is that kind of like remnants of Bezos that is just going to die a slow death over time?

To me, it's similar to the Syria discussion we were having. I think that they're definitely going to reboot it. They're very, reports about that. I think, you know, that's a big question of if it's, if it's Alexa itself or if it's something new, they have something called Rufus, which is sort of their shopping AI bot that they've, that they've sort of put out there. And so it'll be interesting if they try to do something new, or if they just sort of rejigger Alexa behind the scenes to work. But I think you're right.

Like the third party developer ecosystem stuff, as we were just talking about, doesn't seem to have worked except in certain pockets, gaming. there's some you know things that are working there but in terms of yeah just like it being a device where there's going to be multi-billion dollar businesses built on top of that's obviously not the case and can they get there uh that moment is probably passed them by is there something new to be built with AI and we talked about going back to robots and and what they're doing there like maybe there's there's something that they can do um that's compelling sort of bringing all these these different things together um but yeah it feels like Alexa as it's

stands right now and I don't want to trigger. I may have just triggered. I always avoid it. Yeah, it's rough. It will be changed just like Siri has been. And so it's a question of how much of it is behind the scenes versus does it require new devices and all that. So I want to take a minute to plug Spyglass because I've been enjoying reading your work there.

You had 500-ish or 500 words. And that was on medium. I think you had your own site. Now you've started something new. So talk a little bit about the impetus there. and what it's been like you're I mean you're super high output I'm looking at your archive and sometimes there's like five six short posts a day it's super fun and interesting what's your thought yeah uh thanks for asking about it I mean it's something so as

you know I come from a world where I was a reporter for a number of years before sort of becoming a VC um over a decade ago at this point and during the intervening time as you mentioned with 500-ish you know a site that I was running on medium for a long time is still up, obviously, always have just kept writing in my, you know, bag because it's something that I've just long done and enjoyed doing. But it's also just something that helps me. It helps me sort of formulate my own thoughts, helps me, you know, get obviously thoughts out into the world. But doing Spyglass now for, you know, the past few months, I've sort of built it up and talk about it as sort of three things. It's a pretty standard link blog, linking

to a lot of things, as you note, a fairly standard sort of newsletter, you know, once a week, summarizing what's both been linked to before and also, you know, linking to some new things. And then a column. That is sort of more of a work in progress that I'm still sort of toiling away with behind the scenes a little bit more. But I basically wanted to be those three things. None of them are new, of course. You do them.

Everyone does them. But for me, it's just as much about doing it for myself as it has been about sort of doing externally. And I sort of, you know, challenge myself to see if I could come up with a new type of site type of publication, you know, starting from scratch, not wanting to do what I'd been doing for the past several years. And honestly, I'm loving it so far because, again, it sort of feels like it's unclogged my brain in ways that I had missed dating back to the reporter days where it's like, all this stuff is just in my head and I'm thinking about it, but it all just jumbles up in there. And it's not till I get down on the page what I want to say

about it that I really form my own opinion about it. And that's helped a lot in conversations on the VC side, on conversations with entrepreneurs, on conversations with other investors, just with technologists in general. It's just like in dinner party conversations. It helps because it unlocks, again, my own thinking, which is very helpful. So yeah. Yeah. And it's always a fun read. And like we spoke about the last time, I like went through your archives when I was writing the Apple chapter of my book and just love the stuff there and love keeping up with your work. So the website is spyglass.org. MG, thanks so much for coming on. Thank you for having me. I hope we talk again soon. Me too. All right, everybody. Thank you so much for listening. We'll be back on Friday to break down

the week's headlines and we'll see you next time on big technology podcast.