Tricky Bits with Rob and PJ
Rob Wyatt and PJ McNerney discuss the latest and greatest news in the tech world and to figure out where things have been, where they are, and hopefully where they are going.
Tricky Bits with Rob and PJ
Apple and Generative AI
Enjoying the show? Hating the show? Want to let us know either way? Text us!
Could Apple be quietly gearing up to redefine the AI landscape? With all eyes on Siri's evolution and Apple's famed secrecy, we delve into the tech giant's delicate dance between user privacy and the cutting edge of AI innovations. Join Rob and PJ as we unravel Apple's enigmatic strategy, considering the possible hardware constraints and their steadfast commitment to a reliable assistant amidst the whirlwind of AI advancements.
Apple's Siri has been a household name, but it's not without its critics. In this episode, we dissect the challenges it faces, balancing user privacy with the need for more personal and proactive functionalities. We speculate on the potential for Apple to introduce more generative AI capabilities and how this might revolutionize user experience while sticking to their privacy-first mantra. Moreover, we explore the juxtaposition of Apple's hardware prowess with their approach to personalized AI, pondering the efficiency and practicality of running inference models on mobile devices.
Wrapping up our deep dive, we tackle the broad implications of AI in our daily tech interactions. From ethical concerns over data usage to the burgeoning role of edge computing, we question whether AI will be the next big revolution or a passing trend.
(And yes...this description WAS in fact generated from the Episode transcript and AI...wild times, right folks?
You know what what is really going to get your noodle later on? Did the AI write this footer? )
Welcome back everyone to Tricky Bits with Robin PJ. Right now, we're in the middle of an AI frenzy, from investment to all the big companies getting into it, but there's one company that we haven't heard a lot from on this front and, given that they're the second biggest biggest depends on the week company of the world, it's important to talk about. So where is Apple in all this? And, Rob, we've had Apple play in this sandbox to a certain extent in the past with Siri, but they are going nowhere near as nuts at the moment as Microsoft through open AI, Amazon, Alphabet, Google, Meta Facebook. So where is Apple in all of this?
Speaker 2:I think it's the typical Apple approach they're saying nothing, I mean they are hiring people.
Speaker 2:If you look at openings, there's a lot of AI positions. Some of these people might be in the infamous self-driving car division, but they are definitely hiring people for AI and typical type-lipped. They're not saying anything at all. They do also put out papers in the AI space, which is very rare for Apple. They don't put many papers out at all. So they're definitely working on things, but generally, I think they have to announce something soon.
Speaker 2:I don't suspect it'll be at WWDC that they'll have some updates. It's most likely, I think, going to be an update to Siri and a more generative form of Siri. Most likely it may just be a bunch of APIs that you integrate into your own things, and we'll have to wait and see. I do think overall it gives Apple kind of a dilemma as to what they want Siri to be. There was lots of infighting when Siri first came about as to what it was going to be and Apple went for the canned responses over being more generative even back then, due to the lack of control that you have if you go down the generative path and Apple, as we know, is all about that control.
Speaker 2:Over time, Apple have been moving their AI functionality to the neural engine on the local device and for things like Siri and simple voice recognition, that's a totally capable device to do that sort of modeling. But if you move to the full generative world, the devices where Siri runs are not powerful enough. Yeah, you could compress it, you could get it more powerful, but you're not going to compete with the likes of chat, GPT and Gemini. If you were restricted to what's available in a home pod just the same hardware as a watch yes, your phone and your Mac could do even better, but it's not going to be a consistent experience. So I think it brings Apple back to doing more work on the server side and not on the client side, which takes away their argument of where more secure, where not sending your data to the cloud, blah, blah, blah and everything they've told us over the years.
Speaker 2:And I think the big question comes about if they have these large language models that they're going to feed into Siri, where did that data come from? They've not been stealing our data over the years like Google have and Facebook have at least they say they know but then suddenly they come out with a GPT quality large language model of how did they train that? Where did that data come from? Have they been listening to our voice calls and text messages for the last 10 years, and in an Apple different sort of way? I don't know. I think that's a question that they'll never answer. Maybe they just bought the data set, I don't really know.
Speaker 1:So one of the things you've talked a lot about, rob, about Apple, is that they are very much an end user experience first company. They start with the end user experience and then craft backwards to figure out the slew of technologies to make that happen. Now I know you said that Siri has been handicapped a bit by virtue of its canned responses. I contrast this with what we're seeing with Google and ChatGPT and a few others, where there's this great slew of technology that's been created that have been producing some mixed bag I will say most generously of experiences for users. What do you think, or what do you suspect, is the core user experience that Apple's trying to go after with Siri and generative AI, other than a? Me Too, I got to do that.
Speaker 2:I think, from a consumer point of view of being a market leader, that's where the Me Too for Apple ends. They are, like you said, far more interested in designing a user experience and going all the way back to transistors if necessary to make that experience possible. That's exactly how things like the Vision Pro came to be. It's how every Apple product in the last decade came to be. It's not we have this cool technology. Can we monetize it? This is the experience and this is the technology we will use.
Speaker 2:And so the Me Too for falls away at their image image in the consumer view of we're a tech leader. We need to be in the AI space. I think, from Apple's point of view, they would want it to be like an obedient child of like, hey, siri, do this and it does it exactly the way you want to do it, but the problem with generative AI is it doesn't do that. I also think Apple want to avoid the whole race and political issues that generative AI tend to kick up, and they just want to avoid it completely. They don't want nothing to do with it. They want Siri to be, like it says, an obedient servant.
Speaker 1:Is there actually any real pressure on Apple to get into this game? I mean, I could say that, look, the difference between the first Oculus coming out and the Vision Pro is close to it decade at that point in time. Is it actually sufficient at this point in time for Apple to effectively say you know what? We're not going to be occupying the generative AI space for another five to 10 years, because all this stuff is insane right now.
Speaker 2:I think they have to be. I think investors and the consumers want them to be, and everyone knows Siri is garbage, so it kind of has to be. And in your comparison I think the tables are turned Like from the first Oculus to the Vision Pro, apple's the one in the advanced seat when it comes to AI. You said it's a decade for VR. It's a decade for AI too. Siri's been around forever and that's a good point. They're still where they are. So Apple in this case is the simple model, and then you've got chat, gpt and Gemini and things like that at the cutting edge. So from a purely optics point of view, apple is far behind, and I think it's easy for people, even non-tech people, to put two and two together and realize how useful Siri would be if it was a more I wouldn't say necessarily a generative system, but a more flexible system. I've just been able to ask you're asking a question right now and she said oh, the answer's on your iPhone and she just did a web search. Why can't you just read the website to me? Why can't she just gather that and be more useful as a whole? I'm not saying that has to be fully generative, because I think there's a lot more to a user experience than generative AI. Consider something like oh as the rabbit one that came out, where it's the Android-based piece of hardware that can do actions instead of just predicting what the next word's gonna be, and as far as I know, they train that on knowing how to click buttons. So I think it can go to a website it hasn't seen before and still figure out how to navigate it for you, so you can say things to it like buy me a plane ticket and it will ask you a few questions, and then it'll go buy you a plane ticket. We'll also one of the questions it will ask is it's $300, are you sure? With that? It's very similar to saying to your kid go just buy me a plane ticket to LA and she'll ask you a few questions and she'll pick on a few things for you. That sort of end user experience, I think, is what Apple would go for. So it's not like a generative AI. Isn't that useful? It just spits out a wall of text or an image and you can tell it's done by an AI that already exists.
Speaker 2:Having Siri just make up a story of no use to anyone as a home assistant. Having it do things and trust that it does them properly, I think is where Apple will take it. They have no interest in being the next chat GPT and just have Siri talk to you, but having Siri say buy me a plane ticket to LA and she knows that you like window seats and you like the front end of coach, reasonably priced. Always check a bag, things like that or things that I think they'll learn about. But then it comes back to the privacy part of like. Does Apple really? Are they in a position to take that information in a way that's pleasing to people, or people willing to give them this amount of information? You're giving it to them anyway indirectly and they could mine it to find it. But Apple have always said we're not in that space. But they kind of have to be in the space to make a not even a generative AI, just an AI be useful, an action AI. It needs to know a lot about you.
Speaker 1:So, with where Siri is at today and I think, like my traditional experience with it has been as needed to be connected to the internet I think that may have changed. Why does it suck? Like it's been out there for what A decade-ish. At this point in time and, to your point earlier, like it's the one who's the simple model. It really hasn't advanced from my mind or in terms of utility. I use it for day in and day out, except for, maybe, speech to text, which still gets wrong a lot of the time. So what have been the factors in play here? Is it image, like Apple didn't wanna have a disobedient child? Is it technology? Apple didn't wanna be doing a lot of server-side stuff. They wanted to focus on the local Like. Why is it that we've had this in play for a while and it was ahead of Alexa, it was ahead of Google and it's still not really?
Speaker 2:useful. It isn't useful. I mean it's only recently. You can ask it to have two different timers. Give them names. I've, like, set one timer for the potatoes in the oven and set another timer for the bathtub that I'm filling and actually have them. Keep them separate. I think it was released because of Alexa and all those, and Cortina and the Google Assistant were just a thing at the time and Apple had one and it fit into the ecosystem and I think it's the ecosystem keeping Siri alive. If Siri didn't have the entire iOS, mac OS, apple ecosystem, it would have been gone years ago because it pretty much is useless.
Speaker 2:You can ask it to set an alarm and I always double check when I ask, because I'm never quite sure whether it actually set the alarm for the time I wanted. And if I'm catching a plane or got to be somewhere, then it's pretty important that your alarm is set. So, having to double check, I might as well just set it myself. She can't even play music very reliably. You say, play this and you might get it or you might get something related to it and things like that. It's hard to control and I think it's as far as Apple have wanted to go because of that control. They're very, very controlling company and if they can't dictate the experience you're gonna have, they won't do it, and I think that's held them back. That's why they are where they are today.
Speaker 2:So why is it bad? It's never really been a full AI. It has some AI bits, but it's lots of canned responses. It's really just oh, you said this word. It means this you can speak in very broken sentences to Siri and she'll still do the same operation as if you said the full thing, as if it's just like oh, this word and this word, and this word means that. So it's more like an AI just detecting shapes of audio waves and things like that than a true like I'm actually listening to you.
Speaker 1:From the Genesis standpoint, where it's like Cortana, I think, and then Alexa, google Assistant, like that's also a form of its own Me Too. Is it effectively the fact that Siri has always existed in kind of this Me Too? And now that AI generative AI is out there, you think that there's pressure on the company to actually transform it into something that's more useful and potentially less controlled than before? With the whole generative AI thing, I think there's limited use cases where I find it extremely useful, and then other ones where I find it just garbage. And so is there any kind of historical sense for, like, why Apple would wanna do it now, especially given the sort of problems that other companies are seeing and how you think they can do it better, and is there anything that prevents it culturally from being better?
Speaker 2:I don't think they'll do it better. From a purely technological point of view, they don't do anything better. Nothing they do is the best. Never has been. Apple's not a leader when it comes to cutting edge technology. They'll look at oh, this is what we need and it's the experience, and I think this is kind of a dilemma for them, because it's hard to curate the experience with something that's so generative at the back end.
Speaker 2:I do think they have to do something, this pressure from investors. I just read this morning that a bunch of UK activist investors are pressuring them to do something in this space, and now something. I really think they have to do something. It's just what are they gonna do? So I think, where, as in the dark, is everybody else? I don't know anything that they're doing that's groundbreaking, but they really have to. They need to come out and knock this out of the park and be like you can control generative AI. You can have a system which still gives you suitable results in a more controlled environment, but that opens up a whole can of worms in itself.
Speaker 2:Yes, they don't have to go face the press as to why the AI is racist, but who gets to pick? What's acceptable in a controlled environment. Is it gonna be another John Stuart where he can't talk about this because it's a fence of Chinese and therefore you'll only hear what Apple want you to hear? It's like that's kind of a little of a bad world there too. So I think they're stuck between a rock and a hard place on multiple fronts, and I don't think they're gonna get out of it cleanly. They're not gonna come out and make this obedient child slave AI that knows everything about you and can do actions on your behalf the first time around. I think that's where they'd like to go. I don't think they're gonna get it right. I don't think anyone's getting it right.
Speaker 1:I actually wouldn't be surprised, I mean, if I was Tim Cook, which I am not. I'd probably say back to the investors like look this generative AI space. Yes, there's a bunch of money relative to the rest of the market being invested into it, but gesprochen夠. The experiences that are being generated right now are a mess and, furthermore, apple's at a particular disadvantage. It's like you mentioned earlier Facebook collects all your data, Amazon collects all your data, google collects all your data and Microsoft probably collects some data or they're scouring it with open AI on the web. Apple has had a very strong stance on privacy. Historically speaking, that has made them very adverse to mining any data in the cloud or locally from you. So where is Apple getting the data that it used to train Siri in the first place, and where would it get the data from here on out to try to craft some better experience for AI for its user?
Speaker 2:So I've always had this dream that AI would be trained as you use it for you, and I think that does give you a good experience. For example, let's take maps. I always drive through Boulder, boulder, colorado. In the same way. If I'm going to South Boulder, it isn't the most efficient way that I go. If I ate it in the maps, all it does is bitch at me for the next seven turns. So you turn here, turn here, turn here.
Speaker 2:Why can't it learn that I always drive this way? And the reason I drive that way is the quick way goes through the campus and it's just a pain in the ass. Sure, it's not a pain in the ass for any metric you'd measure in driving scenario. It's just a pain in the ass because there's students walking everywhere and it's just a pain. So why can't the AI learn that I do this? So when I ask directions to somewhere I've been before, it goes oh, I know you go this way. I'll just tell you to go that way, not be annoying while I'm driving.
Speaker 2:And I think this is a very unexplored area of AI, of like AI that's trained on your own behaviors. I says I buy a plane ticket and it knows I always sit at the front of coach and it knows I always check a bag. It learns things about you the same way as a child would learn things about you. I think that's when generative and assistant type AI becomes way more useful. And no one's doing that, everything's just bland. Basically, we go to what it read on the internet and we format in it and that's literally what the generative AI is. I mean, technically, it's predicting what the next word is going to be Right, picking the one that it thinks is going to be most likely. And yes, there's a lot of tech behind that, but that's really all it's doing and it's read a lot about everything on the internet and it's scouted everything and that's why it's generative.
Speaker 2:It's same for generative pictures of like yeah well, I think I'll just this color picture will be this, and if I'm drawing a picture of a forest and colors, most likely going to be green, they are, in theory, quite simple and it's hard to see how they work. But that's training on massive bulk data sets. And the question you ask is valid when did Apple get that data? If they did it locally trained, based on your own behaviors, then they could get it from you and they could be open about getting it from you and they could say, yeah, we collect this and we store it and we don't sell it to marketers and all things like that, and I think that's how you get a usable experience. Is your theory. He'll be very different to my theory due to yours was trained on you and mine was trained on me. As far as I know, nobody's doing that right now, whether it be for driving directions or whether it be for a home assistant.
Speaker 1:Well, this is interesting and I want to. Let's zoom out for one second. Then I want to zoom back in. Right now, Nvidia is making a killing by effectively selling all the H100s that can to everybody.
Speaker 1:Like they are selling the machinery for building out these data centers, for doing the AI generation, for doing the training of it. I'm really curious. Again, apple is in a very unique position because it is so vertically integrated. It will go all the way down to the transistor. Do you think that Apple Silicon actually provides an advantage here for being able to tailor their chips or tailor the devices or tailor whatever part of the stack to be able to do that local generation? And it's not something that Microsoft or Google even, or any of these other companies can do because they don't own the stack? Like, is there something there that there's a play here that they can effectively get away from Nvidia for needing, like all of this high end hardware because they control everything down to the silicon?
Speaker 2:No, okay, I don't think so. Well, maybe there is obviously something you could do locally. I don't think what I suggest. It just would be done locally. I think they gather the data locally and then feed it back into a private AI that's yours on the back end. They do it on the server. I don't think any of the devices where Siri runs and it's got to be consistent too, of course. So just fact that Siri heard you say something on your watch doesn't mean that your Mac didn't know about it. Doing it all to the back end keeps that a lot more consistent, as there's no question about where it's stored. So I think everyone gets their own little bucket of AI state that's done on the server and uses server class hardware.
Speaker 2:The big Nvidia AI chips versus Apple Silicon. There's no comparison there. They're like Nvidia will destroy them on performance and any sort of inference that's done local has to be simple. The local hardware is mostly done for the final stage, the inference stage of like I can take the big model data that was trained elsewhere and I can run it in a very small space efficiently. And don't forget, a lot of these are still mobile devices, so we can't be burning through power. Generating those models is very, very difficult, and that's where I think your data will go to the server. It will get compartmentized and kept secure. Train your AI on Nvidia hardware and then send the results back.
Speaker 1:So your local Siri is now running a different model than someone else's local Siri, if that makes sense, no, it makes total sense, I mean, but it does imply that Apple needs to do a significant build out on the data center side to actually be doing this kind of training right, Meaning that they can't then ignore purchasing H100s. They actually have to either acquire them from Nvidia or again they have some secret project for a beefy Apple Silicon chip that we have no idea about.
Speaker 2:I'm not sure that's a viable thing because there's no market for them. Everyone's just going to buy Nvidia hardware. That's why Nvidia is a two trillion dollar company. Now Apple come out with just the GPU. Well, first of all, they never make just the GPU, so it would be a huge step for Apple to even go there of like, ok, we have this biggest chip ever made and it's just compute units and it's or it's just the a massive neural engine.
Speaker 2:The Nvidia hardware is interesting because it's not just a neural engine. The neural engine in the Apple, silicon, is basically a giant matrix multiplier. It can do the neural node math quite easily. It can do this times, that plus that times that. It can do basically big dot products, big matrices, which is kind of how the basis of a neural net and video. It's all computer. They have the tensor calls, which kind of the same thing, very simple, very, very fast multiply and ads and it's all connected to the compute cause.
Speaker 2:And so the H 100 isn't just a big neural engine, it's a massive compute box that can run CUDA and all the libraries that go with it. So I think Nvidia have got themselves in an excellent position that they're basically untouchable at this point. Apple could make Silicon, but they'd have to make the libraries in the ecosystem, and all that open source code that goes along with it doesn't exist for Apple. So even if they made hardware, they'd only use it themselves because no one else would want to use it. And could they do that? Yes, but would they? I doubt it.
Speaker 1:I think they just swallowed the pride and go buy Nvidia hardware which is hilariously ironic, because you've not been able to configure a Mac with Nvidia hardware for I don't know about a decade or so.
Speaker 2:Yeah, but it wouldn't be a Mac, it'll just be a server Linux box like everybody else's and just ruin a stack of h100s in the server, and they do that anyway. I mean it's not like every single thing they do is Apple based. Yeah, they have Windows machines and Linux machines and servers are Linux and I Think somewhere like Apple music is not even running on their own servers. It's like formed out of a zoo or somebody like that, or at least some of it is. So they're not against using other people's hardware, they're against you using it. They want you to be using their hardware and they'll do whatever they need to do to support that hardware. So They've already gone down that road.
Speaker 2:I don't see them making hardware just for themselves and I don't see them making hardware in a way that would be sellable to anybody else, and they're a consumer company. They make consumer hardware. None of what Nvidia is making right now is really consumer hardware. It's almost like Nvidia's put the GPUs on the back burner. It's like, yeah, they're okay, but this is the thing we do now. We make these giant.
Speaker 1:AI chips. So, to recap, from Apple's standpoint, you know, let's say they go down this route for a bespoke personalized AI for you and me and Everyone else is using an, you know, an Apple device. Let's say that they have the sort of private security, but it still means that they have a need for a set of hardware To do the initial model generation, which either they have to build out to themselves or they will have to run out to someone else, like like Azure, to actually do that. And they potentially also have this other general data problem, right? Because if they're isolating each of our data, they can't use that in a general pool. So presumably they would still need to get a general pool of data from somewhere.
Speaker 2:Yeah, that's why it's not been done, I think. I think you have to start with a generally trained AI and then tweak that AI based on your data. But your data stays with you and doesn't get shared. Could, can they aggregate it? Blah, blah, blah, and then we share it. Yes, technically, I mean, that's what Google have done for decades. Well, apple have always said they won't do that, and I even think this needs to be done server side. I don't think a lot of this can be done client side.
Speaker 1:Right, right, right the model generation still needs to be done.
Speaker 2:The model generation service side. But I think the model in furrances a lot of that service side to maybe some of its client side but it's so it's going away From where Apple have been going.
Speaker 2:Maybe Apple just went down the wrong path of doing it all client side and they've dug themselves a hole that they now they now need to get out of, because the client isn't powerful enough to do a lot of these things, even if you throw the whole box at it and that Apple have always been like, well, we'll do it on an u-roll engine and it's it's, yes, great, it's not that powerful, it's Doesn't have enough storage, doesn't have enough memory, and a Lot of this pushes back. And but then Apple like we've said multiple times now, apple have made a big point of not sending all your data back to an anonymous server in the background. It's done locally, it's secure, your data doesn't leave your device. We've heard this many times and I don't think that Vibes in a world, especially a mobile world, where you need these heavy AI models to do anything useful.
Speaker 2:I think there's a power problem and I think there's a Genial compute problem, so they have to send some of this data back now. Maybe they they could split it in a way. No one else is thought of some of its back end anonymous and some of its local and Unique to you, assuming they go down that sort of path but they have to to make it useful, then Maybe that's where they are, indeed dollars ago, and I have no idea. But it's a very hard problem for them to get out of the hole they're in now in the AI space and Maintain their privacy statements that they've made for a decade now right, there's.
Speaker 1:The technological hoops are Understood. You can buy hardware, you can aggregate data, like that is that's what Google is doing. That's what everyone else is doing, to your point. Apple has this philosophical stance that makes it more difficult and creates more hoops for actually like doing this, because, in order to not break their own promise, they have to effectively figure out some clever way of Dicing the data up, storing it and maybe not even being able to send it to someplace like Azure, like maybe they do need to build their own data centers, simply to say we keep it.
Speaker 1:You know lock that for that.
Speaker 2:I do that. I mean they have data centers. They'll definitely do that. But don't forget, the example I gave of making an AI customized for you was just an example. Sure, it's not. It's not where they're going with it. They won't, definitely won't be doing that rev one. But I think it fits the model off. To be useful, it has to do these things. No one's there right now. No one's making an AI that we trains itself based on your own behaviors, and At best it's a generic model with a few tweaks based on your rate. This on one to five, and I'll I'll bias the output stage of a new one, that based on the weights like you give me, and that's kind of where we're at technically today. Nobody's Feeding the data. You give them back into the training to get a better initial answer right, and I think, for my point of view, that's the only way an AI is going to be useful is if it becomes basically ironman.
Speaker 1:Sure, but it also goes back to the point that Apple, from a different philosophical standpoint, has always wanted that end user experience solid and Not just have it be a tech demo. So either it has to create something that is useful and maybe it isn't you know the ironman Jarvis program but it's either has to be useful or it becomes a tech demo, and we've seen where all the tech demos are at right now. It's basically what's out there again.
Speaker 2:It's the, it's the whole Apple are in right now with investors and people, and everybody is expecting something and all, in some cases, demanding something. And and what can it be that says this is an Apple product? It's everything out there is way too off the rails or goes off the rails pretty damn quickly. Who was it? Was it Microsoft who had the bot that would learn from things it gave it, and it became instantly racist. Yep, yep that was.
Speaker 1:They shut that one down pretty quick.
Speaker 2:Yeah, exactly. So it's like I don't think you can make it learn from public data. You can base, train it from what's out there and Then it has to learn based on you, and if it's racist, it's because you're racist. That's fine, you look, you'll get along with it great. You can't have person a's data Gentrified with person B's data, because you end up with that Microsoft bot. So I'm pretty convinced you have to make it unique and you have to make it Train itself on what you want. That's the only way you ever gonna agree with it and you know, the way you ever gonna trust it is if it was trained by you about you. But it's the Apple problem of how do they get out of this privacy hole and Do they just buy data from Google and be like, oh okay, we'll just buy it from somebody else already has your data. We're not breaking any of our promises. But if that's a ridiculous assumption, of course. But even if they did that, it only works right now. It works once it's at some point.
Speaker 2:How your data is used for an AI is something that we need to address as a society. You either do the Google Facebook thing of just say that that's the way we are. All your data is our data and we'll use it for anything we want. And the other side of the coin, you've got Apple, where they try to use any of your data At least say so but then you can't do these new R&D things, things we didn't even know about. When the ideas for privacy as we know it today were created, no one thought of all. If we had access to everyone's data, we could train these big models. Without everyone's data, that model is not trainable. The companies that have your data got this inherent advantage.
Speaker 1:Well, given Google's troubles at trying to create a Generative AI tool, it might actually be in their best interest really just to become a data broker at that point in time.
Speaker 2:I mean, google definitely have access to a lot of your data and they never said they won't sell it. It's not gonna happen because that is who Google is. Their data on you is Google, so selling it would basically be selling out, and Now they don't have anything that no one else can do. The fact that they have so much data on you it's same for meta. They're not gonna sell that.
Speaker 1:That is the crown jewels the data they have from the standpoint of why it would be good for Apple's business if they were able to thread this needle and actually get us private, local fuzziness of AI generation for us. Do you think that is sufficient cause to effectively Juice the sales of Apple's hardware? And to put this in context, I've been trying to think about this from every big company that's getting into AI, which is how is it gonna improve their existing businesses or how it will create new businesses? So for Apple, is it selling more hardware?
Speaker 2:That's the only way, only reason they would do it. Get more people into the Existing ecosystem and keep them there is how Apple makes money. Then they can sell the services and all of that To them, and then I think a smaller maybe it's not smaller Selling more things to the existing people in the ecosystem. I don't know which is any more beneficial to them in the grand scheme of things. I assume getting people in there, because once they're in there they can then sell more things to them. They were consumer product. They've got to stay at this cutting edge of Consumerism and if they don't, they fall out of grace pretty quickly. I think there's lots of people who've been in this same position and never kept it.
Speaker 1:Do you think that Apple has an opportunity here To change the conversation on AI like avoid basically the generative AI mess that Google, for example, is in and say, look, we're gonna focus on AI and how it empowers AR? I know we talked about this in a prior episode, but could this actually be a better route for them to go down to say look, the Apple vision Pro. You know we have AI that's gonna be used for seeing, understanding and we can make AR even better because of the investment we've done here? Like, does that help them escape some of these privacy problems too?
Speaker 2:No, it don't. It's a totally separate thing. That's AI as a technology, and this is AI, is a user experience.
Speaker 1:Okay, they have to do both.
Speaker 2:Okay, people are expected to be more useful than she is. It's basically where we're heading with this and how they do that Isn't an easy problem. The fact that your vision Pro now understands that your kitchen and my kitchen are both kitchens, even though they're very different, isn't an a Direct thing that the consumer sees. That's gonna get fed back through some experience, whether it be a game or a Movie playback of like cooking here and put this here and cook that and do that. Whatever it may be, it's a different experience and it may not come from Apple, even though the technology is apples. Just because you're using a rocket Doesn't mean everyone associates your AR experience with the AI that's built into a rocket. They associate it with you and your own company. That's why there's an app store. Otherwise, everything would be Apple.
Speaker 2:Individual uses of AI, which are everywhere now, are just going to become libraries and pre-trained models that you can download, including detecting the haystack. To start with, it's like that's a very simple thing to do these days. You've trained on a backend, got data nice and concise and you can run it on a small microcontroller that's always powered and that's a totally different AI space than the AI assistant or the generative AI space, which are far more user-facing. Using AI as a core technology to solve another problem, I think, is AI's best use.
Speaker 2:I think all the generative AI's they're off the rails. They'll do whatever the hell they want to do. They hallucinate and make things up. We're having an AI that just looks for handwriting recognition or voice recognition or phrase recognition of what it is. They're becoming more solved problems. We solve them more efficiently all the time. But I think that's the AI that's going to be useful for the foreseeable future of it just makes other tasks easier or more reliable and looking for signals and noise and things like that, and that's what an AI is good at like classify these objects or whatever. That's a more rigid world of AI and it's where we've been until these new generative models started to show up. But I do think that's two separate pieces of technology and you have to be in both.
Speaker 1:That's fair.
Speaker 2:Your framework to the back end are going to expect these AI assistant functions to be built in, and the front end is the user experience, the consumer facing AI's, which is for Apple. They are just different texts by different teams doing different things.
Speaker 1:From a usability standpoint. I really like your vision of the local AI that Siri connects into to make it more personalized to me. I do think that AI as a term has gotten amazingly bloated. It's sort of all things to all people. A lot of it is this generative stuff which is getting companies into trouble.
Speaker 1:There's four uses that I have come across that seem directly useful to me so far. One is I've done the AI image generation, which has been actually cool. I use a small app called Wonder. There's another one which is like an AI image conversion, where it can take my face and make it anime. That's another app called Glam. There's the AI that I use when I'm editing this podcast, studio Sound, which can help take out echoes. And then there's the AI thing, which I haven't tried yet, but I'm really interested in that you sent me about upscaling. Videos Like these are to me, for my use cases are specific, but I find them really interesting and intriguing. Beyond the maps example, do you think that there is a near term use case for Apple to go after with Siri at this point in time, other than doing image generation locally?
Speaker 2:Like I said, everything you mentioned is an applied use of AI. It's not necessarily the wide open wild west generative AI, right, I think an assistant has to be customizable in some way, whether that's retraining or whether it's just tweaking a few things to you. I think Siri's problem is everyone gets the same canned answer, and it's not what I was looking for. If they knew it weren't what you're looking for, even if they gave you a different canned answer, it would still be at least better.
Speaker 1:It would at least be something that's maybe more relevant to me. I wonder if Apple is afraid of getting into the scenario you talked about earlier, which is like, if I'm a racist and I train my Siri to be racist, is that a mark against Apple Because it let it happen?
Speaker 2:I mean, who gets to blame when someone goes to that racist house and asks their Siri a question? Siri already knows it's not the same person, so he could default back to the bland answer or say I'm not answering it for you. I mean it says that all the time If my girlfriend tries to ask Siri for my schedule for the day, it'll be like no, I can't tell you that. You've got to unlock his phone or something. There are some checks and balances in there, but it's just an optics of it. Obviously, no one would blame Apple probably for having that Siri now become a racist Siri. But it's the optics of someone now has a recording of someone famous as Siri answering questions to somebody else, and I'm just like whoa and I think they want to avoid the whole thing. I think they just want.
Speaker 2:That's the same problem as generative text. Today, if you ask Gemini to draw you a white person or a black person or something like that, it'll be like, oh, we're working on that. I think it says it will let you know when it's ready, when it's ready. And they had it and they took it out. Of course, again, it's the optics of it. It's not that technology is inherently racist or not racist, it's the optics that the PR, the company, has to deal with when they realize the training data may not have been optimal for the case they're pushing.
Speaker 1:Do you think anyone's doing this well right now? I mean beyond the applied stuff.
Speaker 2:I think the rabbit one has some interest and stuff, but there's no reason it can't just be an app on an iPhone and I think if Siri couldn't become that, I think they'd disappear overnight.
Speaker 1:Okay.
Speaker 2:It's a custom Android device. It's something you've got to upkeep and everything like that. Phones have been taken away hardware devices for years. Yeah, they take away the GPS, take away the separate camera, take away the flashlight. A bunch of people used to carry a bunch of devices. Now it's one device. People are not going to go back to carrying more devices and I think they only have their own hardware because they needed something to run their AIs on. There's no reason what rabbit is doing that can't be an app for a phone. Maybe integration with Apple would be difficult because of the access and things like that, but there's no reason Apple couldn't do what they do. So if Apple did that, they would disappear overnight, guaranteed.
Speaker 1:Maybe Apple should just buy Rabbit AI and then just incorporate into Siri.
Speaker 2:Again, I think, although they showed the idea, I don't think there's any inherent value there.
Speaker 1:So effectively, then nobody is doing this well.
Speaker 2:Being first means you look at us, we did this. Does it mean that you're the leader or doing it the best? You just did it first and you get someone else the idea it's potentially someone with a lot more resources, or we're applying those resources to a different path. And then we're like, oh, that's where we need to be. And all they do is just point the ship 10 degrees left. And now they estimate you.
Speaker 1:My general statement on where AI is at right now is it's a very exciting thing that doesn't have a whole lot of real use cases with it yet. Like it's a great buzzword, it's a great term for getting like companies funded, but I don't think that it's actually transformed anything yet and I'm curious whether we're actually going to see a transformation in the near future. Like is one of the reasons Apple isn't doing anything is because there's not a real thing there yet.
Speaker 2:I mean there are some incredibly good uses of AI. I mean everything I've seen that's good is all in the embedded AI. The other side of the coin that we talked about briefly a few minutes ago it isn't that Genitive AI is inherently off the rails or it's inherently good or bad. It's the integration of that Genitive AI into something. I think Photoshop have done a reasonably decent job of having AI aware editing tools so you can say highlight a person and say delete and it just fills the background in. Very useful thing to do. Saves people a lot of time.
Speaker 2:I don't think AI is going to replace artists. I think artists with AI will replace artists. Without AI and through tools like this, if you refuse to use that auto fill, even though it may not do a perfect job, it gets you 90% of the way there. You do 10% of the work versus doing 100% of the work. If you did it without AI and it's kind of great. You can be like mark a section of an image and say put a street post there or sign or a neon light and it does what you say. I think when you start taking the out of Photoshop and just doing generic image generation, it's when it becomes off the rails of like what the hell is it doing?
Speaker 1:So it goes back to your point earlier about the applied AI.
Speaker 2:Applied AI is where it is. Ai is just technology. It needs to be applied to something Like the fact that you have the best A, b or C doesn't mean it's useful. It means it's a good piece of tech, and I think everyone out there who's doing it have always traditionally been the companies who push tech first, like we have. This tech will find a use case for it and I think that's Gemini. The old Bard chat GPT is just tech looking for a solution and that's why they let it go off the rails. Where tech that's useful, that you can actually sell to somebody and make their life more productive, make their job more productive, needs to be strained to the environment, it's supposed to be used in.
Speaker 1:And I think that's going to be one of the challenges, and I honestly think that Microsoft is a really good position to well, microsoft slash an AI, like to channel that through the tools. I'll admit I have not used co-pilot yet, but I could definitely see saying, hey, here's a bunch of data, although I'll also say that if the AI is able to do that, it sort of. What makes me wonder are those performance reviews actually work?
Speaker 2:Yeah, but then you factor in again local learning. Make a performance review, as I would have done it. It'd be very different to a performance review that I would have done, and I think all of these integrations of AI all benefit from the AI knowing something about you, even the Photoshop one. You tend to use this style, you tend to do it like this, so I'll just start there and that'll be the first option I give. You will be one that I've biased towards you, rather than just being like, okay, that's what I did. You go finish the last 10%. I think it becomes more efficient, more useful, if we can get into a world where every little AI that's integrated anywhere is tweaked towards the end user.
Speaker 1:But the data problem and Retraining problem remains so maybe that is the next frontier, like how do we have effectively all these Billions really of bespoke models that we want to create for each of the individual users that are going to be there because they agree, we want to have the Iron man suit, we want it augmenting what's already there, not being the can response?
Speaker 2:and again, if it's augmented in a way that's more useful to you, then it's a more useful product, and the way it becomes more useful to you is to know something about you, or Know how you would have done this if you did it yourself, and I think that's when AI gets more dangerous than it is today because all of a sudden it's gonna just try and Predict everything about going to do now.
Speaker 2:I think it just gets I don't say more dangerous, I mean more, that was probably a bad word it gets into position where it can do your job better, because that last 10% you were doing was critical and now that last 10% isn't necessary. But again it's your data, so does it stay with you? There's lots of legal questions here too, like if I Working for a company and I'm just running through this in my head right now, so this might be complete rubbish, but I work in for a company and I I've been doing this same Photoshop operation over and over and over again and my AI that's me Helped me now can basically do exactly what I ask of it without me doing anything if I leave that company, is that my AI or is that there AI? Can they keep using that without me, even though it's mine? It's the whole thing of.
Speaker 2:It's the same question of if you have a AI generated Tom Cruise from 20 years ago, do you own it? Tom Cruise own it in the? In that case, tom Cruise definitely owns it. But if, if I've got an AI that's been trained through my actions like your workplace, is that yours or is it mine? I think we're not in a position to answer any of these questions yet. But that's the problem with the locally trained model of like. Who actually owns that data?
Speaker 1:The lawyers are gonna field days with this stuff, and I'm sure that's actually One of the things that Apple, especially, is very sensitive about. Looking forward, rob, do you have a prediction of what we might see from Apple in this year? I'll give mine in a second. All right, I'll go go, then you go first. I actually think we will see. At max, apple will say we can generate images locally on your phone. That's it. I actually don't think they're going to do anything major Because I don't think they're ready.
Speaker 2:I Think we'll get something along those lines. Some look at those. We're doing AI, whatever it may be and I do think we'll get it upgraded Siri. How upgraded, I don't know, but I Think they have to do something more than that because investors are starting to complain. So I do think we'll get an upgraded Siri. Again, I've no idea how far it would. It would go, because they've also got a factor in. There's a lot of apps that are Built-in or have hooks into Siri so you can control all the apps. If they upgrade Siri completely, all that breaks. So they have to walk this line super carefully. But I do think we'll see. Yeah.
Speaker 1:So I I will take the I will take the negative side on this, on this bet I don't think we'll see an upgraded Siri because I think for them it's gonna be too dangerous a user experience they're either will like. Like if they end up in the middle where they break a whole lot of shit but not provide a lot of value, that's bad. If they end up in the same place Gemini's at, which is like crazy shit coming out of it, that's bad for them too. So I still think they'll actually avoid doing anything with Siri and basically just have you know oh, this is your image generator, which may or may not have any any hook into Siri whatsoever. So we'll figure out.
Speaker 2:We'll figure out, we'll see. Wwc is when May, june time, and then there's an announcement coming up in a couple weeks and there is the big announcement, obviously in the fall when they tend to do the iPhones, we'll see what comes out, something in one of those three It'll be. I think there'll be a new Siri. I think it'll be a wwc. There'll be a whole bunch of New series stuff and a whole bunch of new framework AI based things for all the developers. I'll believe AI libraries.
Speaker 1:I would believe that it's the series stuff that's I'm I'm still gonna go negative on I.
Speaker 2:Think they can make it work the same way for the people who need it to work the same way and work better. For when it can? I don't think it's a or B, black and white thing, it's. It is a path that people can go down and those apps with integration can keep what they have without breaking and they can Progress to these newer integration libraries built in AI libraries, things like that to make their integration with Siri better. I think that's the only way they're gonna get from where they are to somewhere else. They have to take that first step and I think this will be that first step. May not be much of a step, but it's all right step in that direction.
Speaker 2:Otherwise, with what they're gonna do, stay where they are forever, make entirely new thing. I Don't see them getting rid of see when we're placing it with see we to call or what you want. This has to be a path from here to there and they have to start on that path and this is how they're gonna do it.
Speaker 1:I don't know, man. Everyone seemed fine with changing Google Hangouts to Google Hangouts plus to alo, to duo, to meet, to whatever the fuck it is. I'm kidding.
Speaker 2:Did they go?
Speaker 1:no, no.
Speaker 2:It's like no, it's like changes hard for people and, I think, gradual change. I do use series, which is a hate. Using Siri, I do actually use it. I'll yell out in the morning like what time is it? Set me alarm and play music and things like that. I don't have any own integration so I don't use like hey, sir, I'm home, turn the lights on. But I know a lot people who do and it seems to Work quite well. That could only answer questions from my own Case in. It's fairly minimal, but everyone's use cases different. So I think if you ask somebody else how they use it and what they wanted, a bit different answer.
Speaker 1:So I think take my well but I think this is a great illustration of your earlier point, which is that what everyone wants to use Siri for may be different, and Having something that's locally trained as a useful tool for you or for me has a lot of leverage to it. So, crafting something that is completely canned, you know you you'll effectively use whatever segment of that canned aspect you can, whether it's turning on lights or Opening your lock or whatever. But I think I think you're you're hitting the nail on the head, which is like you want Siri to be something that's Rob's, I want something that's gonna be PJ's, and I think I'm aware of this.
Speaker 2:If you look, there's been a lot of turmoil on the, on the Siri teams. Internally, people are leaving. It's not going in the direction some of the engineers want. It's Again, it's the can versus the non-can response, management's input and the direction of all that. They're aware of all of this, a lot of this, when I was Apple. They are aware of everything I've mentioned and everything everyone else is mentioned. It's just comes down to, I think, management and the path they want to go. I think the engineers are more than capable of doing Everything we said. It's just what do management want? And, again, apple very control and this is a very hard thing to control.
Speaker 1:And this is where I'm really curious to see if management actually is going to make a big change based on investor pressure. I think it's actually a really good litmus test for the company, like would they be willing to Put out a substandard product in order to Quell investor pressure?
Speaker 2:or is this apples downfall? Is this the same as Microsoft missing mobile and laughing that it's not Important and Apple just go well, we can't control it, so we're not gonna do it, and then, ultimately, that's the end of Apple in a decade? Hmm, it all traces back to this point. It's no different to the comments Steve Ballmer made about mobile a decade earlier.
Speaker 1:This will be interesting. I don't know if we'll be able to like run out a prediction for the next decade on this one, but it is a really good question of. Is AI as impactful and is it as revolutionary as we all think? Or, like another Nvidia fueled product? Is it possible that in that AI as we know, it becomes sort of you know, maybe slightly more useful than Bitcoin was or is?
Speaker 2:yeah, it's all unknown and I I think again. I go back to the integrated AI's, which are incredibly useful.
Speaker 1:Yes, that I 100% agree. I think that has. That is the hallmark, where it's like integrated with tools. It's this pure generation Skynet, c3po stuff that I get Much more Bearish about. It's the cutting edge.
Speaker 2:It's just people look at us, we can do this and it's the next big model. And Are they themselves useful? Not that we're useful, in my eyes, it's. It's taken pieces of that and the pieces that make a given tool for a given use case more Usable is, I think, our AI, ai. Ai is going to be used for the next couple of years at least. And Whether that be having AI based audio processing in the, the mix, as we're using to record these podcasts, so we've moved background noise or we cleaned your voice up, or we take the buzz of the air conditioner at the background All things that AI is traditionally been useful for the past few years.
Speaker 2:I think you'll start to see that into different devices. It won't just be a post-process tool on a Windows PC. It'll be built into the actual source and you have AI at the edge Doing things that are very useful at that position, less daily transfer, things like that. I think there's a lot of places AI can go to be more useful over than large language model. Being racist, I Agree. It's all questions that we don't know, which are kind of a brave new world we're entering it really is, and it can go in Both negative and positive ways.