Upgrade - 447: I've Got a Lot of Code in the Catalogue

Episode Date: February 20, 2023

With Jason on vacation, Myke is joined by David Smith. They discuss a potential delay for Apple's headset and what that may mean for WWDC. Also, David's AI-powered podcast transcription website, and the introduction of 'Ask Underscore'.

from relay fm this is upgrade episode 447 today's show is brought to you by squarespace and text expander my name is mike hurley and i am joined by a very special guest underscore david smith hi david hello so i called you david there i always call you dave that just tends to be how i refer to you i don't know if dave is a chosen like a chosen nickname for you but i do call you dave a lot is that okay yeah yeah totally fine most i would say most people call me dave i'm like formally david and then informally underscore yes or just under for sure yeah uh you have an honor today do you know what that honor is i do not know what that other than being on an upgrade which is in itself an honor

so jason's on vacation he's on vacation for the next two weeks and you are my first guest host of the show well that is quite is quite an honor a special feeling. Like there was an episode, I think last year, where Jason went away, but he was still on the episode. We recorded some stuff in advance and I had a bunch of guests and we did guests like segments. So you are the first person to fill in for Jason in full in the 447 episodes that we've been doing this show. Well, I will do my best. I bring the big guns, you know?

When I need a good guest host, I bring an underscore. And I have a Smith Talk question for you to start this week's episode of the show. Peter wants to know, how did you settle on the underscore for your online handle? Where did that come from? Sure.

So as you might imagine having a name david smith it is quite popular there are many of us in the world it has caused all manner of challenge and hilarity over the course of my life having a name that is i think by many measures perhaps the most popular name in the world certainly certainly in the English-speaking world. And it has definitely been interesting. So whenever I sign on to a new service, so in this case, where the story starts was on Twitter, I was signing up for an account there.

Obviously, I couldn't get at David Smith. That would have been, or even Dave Smith. I couldn't, you know, no version of my name would have been available. And I believe the the best i could do and i'm not a big fan of like throwing numbers at the end of my handle or anything like that like that's not gonna work so i did the david smith initially which was available and then like just like maybe two days later and this this isn't, I mean, I was no, nobody, I had no followers,

it was no anything. It was just like me on trying to get, see what Twitter was. And then like two days later, my like shyness kind of kicked in. And I was like, Ooh, that sounds kind of presumptuous. But like, I'm the, you know, the David Smith, like all the other David Smiths. No, no, no, that, you know, they're not the one you want. I'm the one. And so I was like, ooh, yeah. And I'm going to be a little, I'm too shy for that. So instead, I was like realizing that, oh, you can actually just like throw underscores into Twitter and that's a valid character. And so I just replaced the the with an underscore and it became underscore David Smith. And then I think the nickname itself kind of got started by Marco Arment,

underscore David Smith. And then I think the nickname itself kind of got started by Marco Arment, who started referring to me as just underscore on Build and Analyze. This is probably like, I don't know, whatever that is. It's ancient history now. Yeah, it's like 10, 12 years ago. And that's just what he referred to me as,

it's like underscore. And then it just, it stuck. Because rather than saying underscore David Smith, you just say underscore. And it works well as a nickname and has kind of stuck with me ever since. And I appreciate it because, I mean, even in Apple, like the Apple world, there are other David Smiths. Like there's an Apple, there's a David Smith who works at Apple, who works on Foundation. And that has caused challenges for both of us

over the last few years. So, you know, just being underscore, being that that's a great way to kind of differentiate between myself and the other David Smith smiths out there you didn't consider like david smith official or anything like that you know like you went for the david smithing but the real david smith no i yeah it's i mean for those are even those are even more presumptive i think so i really like that you went with the and then changed your mind i'm happy you changed your mind, but that is very fun. And I will say as well, if you would have been able to claim David Smith on Twitter, that wouldn't have been a good thing for you, right?

Because you wouldn't have been the same everywhere else, right? Like it was, you know, wouldn't have got that handle in other places. If you would like to send in a question to help us open a future episode of the show, I'll say in advance, next week's guest is the one and only Casey Liss. So if you would like to send in a Liss Talk question for next week's episode,

just go to upgradefeedback.com and you can submit that. Thank you to everybody that does send in these questions every single week. Saddle up, David Smith, because we're heading down for a rumor roundup. Yeehaw! Thank you. Someone finally gives me what I'm looking for. I have a report from Digitimes that the 15-inch MacBook Air is set to launch in Q2 of this year. As of right now, it is unclear if these machines will see an M3 chip or they'll be sticking with the m2 chips

so dave i know that you were a big fan of the 12 inch macbook right i know there are stories told of you sliding one inside of an inside jacket pocket so i'm assuming a bigger macbook air is not necessarily of interest to you but i do wonder what you think about this product entering the lineup yeah i mean i i think this feel a it's just kind of interesting to transition away from air being small like initially i would say when the you know when the air first launched it was one of its big things that was you know small enough to fit in a envelope was like a you know that's how it was reversed revealed and i think air has now sort of just in some it's sort of instead been sort of basic or, you know, initial rather than having the sense of it being always the smallest. And so

introducing a big MacBook Air is kind of an interesting evolution of that name. And I think probably is representative of where this falls in the lineup now. Whereas I don't think, it's not necessarily related to its size anymore. I think people think of a MacBook Air as the entry-level MacBook. That if you want an Apple laptop and you don't want to spend a lot of money

or you don't have huge needs for it, that's what you get is you get a MacBook Air. And I think that could make some sense as to why they would you know introduce a bigger one as another way to kind of increase their margins uh you know presumably you know cost you know 100 or 200 or 300 more than uh the smaller size and i think the thing that really makes me curious with this is like, is it going to fall to the same fate as the, uh, the 14 plus this last iPhone cycle where they took the base and added a bigger version of it. And it has not been selling nearly as well as I think perhaps Apple would

have liked or expected it to that. Um, it continues, I think is being outsold by the just the iPhone 14, like three to one in my stats, like for my apps. And I can't imagine that's what they hoped for for this device. And I think it's interesting with this is like, are people buying the air and it's it's so popular because it's cheap, or because it's an air. And if they're buying because it's an air and they love the air, oh because it's an error, and they love the error, oh, it's an error, but bigger would be great. If really what they want is cheap, then it may not

actually be as compelling of a device. And so I think that's what I'm really curious to see with this sort of a larger error to see if it is actually successful, or if people are actually just they'll just be they'd rather get, you know, a basic MacBook Pro, if they want a bigger device because they want the power and the other capabilities of that. And this may actually not have great fit, but we'll have to wait and see. As you mentioned that, it does strike me that this is an interesting product

to add to the lineup, no matter how it falls, right? Where it's like either scenario A, that it's just not popular because people want to get the macbook pro in the same way that they do the pro iphone or option b people buy the 15 inch macbook air instead of a macbook pro so it's like you know like either way that it falls it's an interesting there is a potential interesting downside i guess apple is expecting that a 15 inch macbook air may bring more people

into the macbook Air line that would otherwise be elsewhere. I do wonder, it's an interesting thing that you pose about the iPhone, if it will stretch over. My gut feeling is that people buy the Pro iPhones because they want the best iPhone, where when it comes to laptops, I'm not sure if they're the same status symbol i think a lot of people would just want a bigger screen for their laptop and so they would get the macbook air but i don't know and like similarly is will this become the standard issue laptop from corporations because it seems to be that macbook airs macbook pros are or or as we found out when I got 700,000 people writing to me

that the 16-inch MacBook Pro tends to be a pretty standard issue. We'll find out. But either way, I am happy that Apple is continuing to expand the line, but it is going to be fascinating to see what fate

this product ends up having. Exactly, yeah. And 9to5Mac reports that they have obtained renders of the upcoming iPhone 15 Pro. They have like a CAD model that they were able to develop and show off. So I'll tell you what we see here.

So we see a USB-C port, we see, which we'd kind of expected, a thicker camera bump, which we'd expected, but it is worth noting, this is the Pro model, not the Pro Max, where we're expected to see even more changes from a periscope lens. I expect it will be even bigger. And then there are a couple of things that are newish here from some of the stuff we the sleep-wake button becoming capacitive with haptic feedback, kind of like the old Touch ID button rather than a clicky button.

And 9 to 5 say that these renders indicate to them this would be the case. And also the glass on the screen, so like on the front of the device, features more of a curve to the kind of frame of the phone, making it a smoother transition with the frame itself carrying more of that curve. So rather than these kind of very harsh, straight, flat sides and sharp corners, the iPhone might get a bit rounder and softer in its feel again. Does any of this intrigue you at all, Dave? I mean, I think it's such a tricky thing where i feel like

the physical shape and dimensions of iphones i don't think matter very much to anyone i know like i mean there i think there are people who care about that but i don't think the vast majority of people myself included really don't care too much. Like I buy an iPhone, it goes into a case. I never really see the shape of the phone again. Like it's the entire, my entire experience of it is the quality of the screen and the, I guess the physical dimensions, like width and, you know, sort of the width and height of the screen. But beyond that, like having a slightly thicker camera bump or having curves that are slightly more rounded, I just don't, it's, it's the kind of rumor or the kind of aspects of a device that doesn't particularly seem to matter. Because, I mean, there are people who use their phone without a case, but they are few and far between.

And as soon as you put it in a case, like having a slightly more curved transition, like if anything, it makes me wonder, it's like, like oh is this going to make it harder to put on a screen protector or those types of things where you know if you have the more rounded you make things the harder those types of uh things become and when i think of the capacitive the buttons potentially going capacitive it's like i just hope they work like especially i mean the the volume ones i'm less worried about i feel like i use those and those are used much less frequently than the sleep wake button, which is, you know, I probably am pushing that hundreds of times in a day to sleep my phone when I'm about to put it in my pocket. And if it's not 100% reliable, like if that's 99% reliable, it's going to be annoying because

that 1% is, you know, many times in a day. And so it's like to be annoying because that one percent is you know many times in a day and so it's like it sounds reasonable that's interesting that it doesn't seem like it's a dramatic year it's still i feel like we've been on the same general design language for these phones for this will be what the third ish year third or fourth year like it was the first one that had this kind of design on it yeah and it's like it's a slight change but it feels much more like we've had i think maybe this was the seven where there was a series this there was a period where it was like they just kept they started with a basic design and they just kept rounding things off yeah um that was the change with the six yeah and so like that seems to where we're going but it's like it's fine like i i mean the reality is it's like i i'm always

excited for new iphones because the the shape of it is not the interesting part it's like it's fine like i i mean the reality is it's like i i'm always excited for new iphones because the the shape of it is not the interesting part it's what's going inside that shape and what you know it's like if you make the camera bump thicker that's exciting because the camera's going to be better and that's always exciting for me but you know the seeing the physical dimensions of it is just kind of like it doesn't really matter um and i mean i'm always excited when like and make that camera bump as thick as you want like just go wild with it because it's you know i'm i'm the kind of person who has an iphone in a card case so the you know my case is super thick anyway so it doesn't even matter to me it's like make that as you know you put it you can put a you know like a you know

like a like a like a thickness camera like a camera of the thickness of a like a camera in there and i'd be happy like i don't care i just you know give like a Leica thickness camera, like a camera of the thickness of a Leica camera in there. And I'd be happy. Like, I don't care. I just, you know, give me awesome pictures. I had two thoughts on this. One, well, actually I have three thoughts now. I just had a different one. One, I hope that they don't get rid of the switch for the mute.

Like if everything's going to be like capacitive or like, I guess, kind of solid state, right? Like there's no physical movement. I'm worried about the mute switch because that is a very important like just you know it's muted right you can just put your hand in your pocket and feel that the phone is mute is on mute i would be sad if they got rid of that uh too i wouldn't mind this at the moment because right now on my my current iphone 14 pro max the sleep

wake button has become a little spongy i don't know what's happened like maybe i spilled something on my phone i don't know what's going on but it's now not working very reliably so i wouldn't mind it if nothing could get in there sure um and then this is more of kind of like an existential thing i was thinking i was thinking about this the other. I don't use a case on my phone, right? Yeah. And I enjoy that feeling. But really, everybody, by and large, does use a case on their phone, right?

Like, I feel like I am in a tiny percentage. Why do they make these things out of glass still? Like, if everybody's just going to put a case on it, and we all know that and apple knows that and they want you to buy cases like well obviously i want you to buy cases but like why did we continue to make these products at least especially the back of them out of a breakable material if everyone's gonna put a case on it to stop their phone from breaking if they drop it just an interesting thought that i had yeah and i feel like I feel like it's almost like they're trying to optimize

for that experience of going into an Apple store, picking the one up off the desk and holding it in your hand. And that being feeling premium, that feeling beautiful in a way that is not actually representative of your actual use of it. It's almost like if the display model was glass and then the one that you could buy came with the plastic back like i'd be perfectly happy with that like i'd be i'd love it if my phone was a little bit lighter and was much more durable like that sounds a great great trade-off to me but i could understand not wanting to have people pick up the phone in the apple store and have it feel less premium to

have it feel like it's plastic like this is so strange right because like everyone's popping off in the live discord right now telling me that it's for signal it's for wireless charging but all of these things work through plastic cases right so that's obviously not the reason and it is this strange thing of like i could i believe you like what you're saying i believe that is the case right that they do it because you want the thing to feel premium but then if you're going to put a case on it what does it matter you know just like i i would i would like it if apple started to experiment with this stuff but in a way that made sense because a couple of years ago samsung put a plastic back on one of their like mainline phones people lost their mind and i was disappointed

at them because they didn't change the price right like it was still the same price as it would have been if it had glass on it is how it appeared so that would be the thing right if they were going to do that it would have to be some kind of balance which is probably why they'll never do it but still it was just like a funny thought to me like we they put all this time and effort money and we pay all this money on these expensive materials but then you just put a plastic case around it and never think about it yeah exactly this device is also you missed you skipped over it's also seems to confirm that it's usbc which i think is also a pretty like nice confirmation that that's actually seems to be coming this year rather than at some point in the future like after you know if

apple fights the eu or whatever like it seems like nope it seems this is going to be the year that it's going to go usbc and lightning will be behind us at least at least on the pro phones i expect that it will be usbc on the pro phones this year and on everything next year i reckon that's how they'll do it just to try and get the last like the last advantage they can get from usbc right to be like oh it's it's an upsell for just one year and then it will trickle out to everything i have a question for you actually as a developer do you think and do you want the regular phones the the regular iphone 15 to get the dynamic island this year i think i mean yes i think is is certainly the answer to that just in so far the dynamic island is such a weird feature that it i think a lot of it's it's really hard to explain or to be excited

about it unless you use a phone that has it like Like it's a very, and this is just something I found from, I mean, I've been doing a lot of work recently with adding Dynamic Island support to one of my apps. And like when I'm explaining the feature to people, they kind of don't really get it unless they hold it in their hands and can actually see what's happening. That when you explain, well, you know, like while you're going to other apps, the small tiny version of your app will be visible in the top of the screen and you can long press on that to see it or tap it to go to that app. And there's sort of a complexity to that

that I think is really hard to explain. And so expanding the number of phones that can support that, that can do that, it's like if this is a feature that Apple wants broad and wide developer support for, making that a wide feature is just, I think, is the only way that that's going to happen. And I feel like the dynamic island, I mean, we're, what are we, four or five months on since it was first introduced.

And I wouldn't say that, you know, as a developer feature, it's sort of really caught on in a big, broad way. There's a couple of use cases for it here and there. There's some apps support it, but it doesn't seem to have had that. And I think part of that is, it sort of hasn't had a big moment. It wasn't like, you know, when it first came out,

it was very exciting and trending and all those kinds of features. And I think that's partly because it's only on the high-end phone of the current generation. And so it's going to be a long time. And if they keep it in the pro phones, it's going to be. And so it's going to be a long time. And if they keep it in the pro phones, it's going to be even longer before it's going to get there because none of the sort of hand-me-down,

or I guess like whatever you would call it, the second-tier phones next year, where at some point the iPhone 14 is going to be $100 less and move down in the line, it's not going to have it. And so if they don't get it into those kind of base models that stick around forever, it's just never going to be anywhere. And then at some point, it kind of turns into a touch bar or something where it's this feature that is technically cool and interesting in some ways, but just never really catches on. And that would be, I think, a sad fate for the dynamic island. I agree. This episode is brought to you by our friends over at TextExpander.

When you work in a small team, every single moment counts because you don't want to be wasting your time finding video conferencing details sent to a new client. You don't want to have to spend a ton of time tracking down the same questions to answer for the FAQs that are on your company website. In fact, just today, Dave sent me a message and was like, are we going to use Zoom for this? And I said yes, and I typed ZZOO on my keyboard, and it just pre-filled the message with my Zoom link because I just have that saved,

so I can very easily send someone my Zoom link when I need them to join a Zoom call with me. These are the kinds of things that you want at your fingertips so you can get your work done faster and easier. That is why you need TextExpander. With TextExpander, you can access what you type the most with just a few keystrokes, allowing you to work faster and eliminate repetition so you can focus on what matters most or get back to what it was you were doing. TextExpander's powerful shortcuts and abbreviations streamline the work of your team as well.

All you have to do is type that short abbreviation and let TextExpander do the rest. You can all just build and collect your most commonly used phrases, messages, URLs, and more right within TextExpander. Choose your abbreviations and they'll be with you and your team wherever you're typing. You can even customize the snippets to make them feel more personal. You can add in dates, fill in the blank fields, timestamps, and more. TextExpander is available on any in the blank fields, timestamps, and more. TextExpander is available on any device that you use across any app that you use on the Mac, Windows, Chrome, and iOS. If repetitive typing is getting you down, you need TextExpander. Check out TextExpander today at textexpander.com upgrade and you can get 20% off your first year.

We also use TextExpander here at RelayFM across the team and make sure that everybody has access to the information that they need and to keep our responses similar across multiple people. It's super awesome. That is TextExpander.com upgrade. Go there today to find out how you can get 20% off your first year and to say goodbye to repetitive typing. Our thanks to TextExpander for their support of this show and RelayFM. Thanks to TextExpander for their support of this show and RelayFM. According to Mark Gurman at Bloomberg, Apple has delayed the launch of their headset to WWDC this year. Here's a quote from Mark Gurman. Apple made the decision to delay the launch earlier this month after product testing showed that both hardware and software issues

still need to be ironed out. Apple has been working to fix issues with sensors on the device to enable the hand and eye control mechanism. It's also trying to strike a balance between battery life and performance. I would say on this one, considering this was supposed to be a March, is what the original rumor was, it feels like they're cutting this one really close if they're still making these decisions about the hardware maybe not being where they want it to be right yeah i mean i mean even wdc is

close i mean i think i was i looked on a calendar and based on like the most likely day you know if wdc is the first monday um in june that's like 105 days from now i mean if for if the hardware isn't final like it's not going to be you, you're not going to be mass producing these in 105 days if you haven't locked in the hardware at this point. So, I mean, in general, it feels super like down to the wire. And I mean, they keep moving this forward in a way that I mean, obviously, they haven't they've never announced anything. So it's always just like entirely speculative. But it seems externally based on the rumors that it continuously is getting pushed forward and forward and forward. But it seems externally based on the rumors that it continuously is getting pushed forward and forward and forward.

And it's like at some point you just kind of have this feeling of like, man, like how are they how are they ever going to get across the line if they just, you know, it's like every time they get close, it just gets kind of pushed out another three and three months, another six months, whatever it is. I can imagine that genuinely the reason they've chosen WWDC, it is the very end point. Because whatever it is, they're not going to be shipping these things in July, right? They will start shipping towards the end of the year. But if the plan is for this product to make sense, they have to be able to show developers what the operating system is at WWDC this year. If they're not doing that, you might as well wait another year.

And I don't think they have that ability anymore. Because I feel like, I mean, you could correct me if I'm wrong. I feel like as we are leading up towards this, thinking about what potentially could be on the roadmap this year, what all of the rumors seem to be pointing towards, this is the focus of 2023,

especially for developers. So if they don't have it, there won't be a lot for people to get busy with yeah and i mean that's always so hard to tell like i mean i would say the the software teams at apple are so are by far the quietest most secretive groups and so like there could be other amazing things down the road i think there is certainly a lot of indication that i mean even just from a conceptual perspective think there is certainly a lot of indication that, even just from a conceptual perspective, if Apple is launching a whole new platform,

that whole new platform is almost certainly going to be a focus of WDC. Because there's the most to educate developers about. And WDC is functionally an education event. It is about them communicating to developers what they should be aware of, how they can actually use it. And that is the venue for them. And they have other ways they can do that.

Like if they didn't make WWDC and they instead launched this in September, they could do tech talks and they can do videos, especially since WWDC is much more virtual now. I mean, I think the big thing they would miss out on is presumably if they have an event at WWDC with this announced, is the developers who are there in person

and the invited guests and all of that will be able to try it out, to experiment with it. You could imagine having on-site labs where developers could use this before it actually is available. It's the kind of device where I'm really curious if they're going to do a development version of this where developers can buy it ahead of time,

like they have done with the Apple Silicon Macs or some of the other devices. In order to use this, you have to be on-site at Apple, and they may have on-site labs throughout the year. But having it ready by WWDC gets rid of a lot of the initial complexity of that because you're going to bring a thousand developers into one place, and so you can show it to them, you can have them try it on. And it seems the kind of device that, no matter how good your simulator is

that you can run on your Mac, it is not going to give you nearly the impression of how the device will actually work in practice because the whole point of it is that it's kind of revolutionary as a display technology and so you know it just doesn't really work that way and it does feel like the longer it is until they announce it the less likely i believe that there would be developer kits like if there's not showing it until this time until june i don't imagine they would be like hey come here and get a developer kit and it'd be like a hardware unit i mean it would be incredibly helpful if apple can do that but i don't know how that's going to shake out and i would imagine that for something like this especially a

development kit would be really helpful for you right yeah i mean i think it's the only it's going to be a really difficult device to to develop for without something like that i think you can do the basics but actually experiencing like the experiencing the device is going to be so important. I mean, obviously, and they could do things where you can run it with a Quest or something. You could use other kinds of VR devices to get some impression of it, but that seems very un-Apple-like too, that the developer kit is here, just use this competitor's device and it'll kind of give you some of the impression of it. I think they would want developers developing on their hardware,

taking advantage of the things that only Apple can do that are, oh, you couldn't do this on a Quest because we have fancier screens or better processing or whatever that thing is. But they've also done on-site developer things before with other devices, like the Apple Watch. There was no developer kit Apple Watch. They had an early access thing where if you were a developer,

once they were available for public sale, you could buy it from the developer team to get around supply chain issues where it was in short supply. But that's a very different thing. But before that, they had some on-site labs where people could go and try out the Apple Watch ahead ahead of time which i could see working with this but i mean it's just such a complicated device yeah but if there's one platform we don't want to feel like it's the apple watch when it comes to development right like that was a really tough really tough version one.

And apps were not running very well for a long time, right? And it was because of, was it WatchKit? Was that the name of it then? Yes. Right. So like it won. Yeah. So it wasn't even really a full OS, right?

Like things weren't running on the device. I'm trying to get my head around. You know. Yeah, I think that's a very good analogy for some of the questions I have about this device is in WatchKit 1 the app ran on your iPhone and was being projected onto the screen

of the Apple Watch. It was almost like streaming the screen to the watch in real time as you were using it, which meant there was lots of latency issues. There were a lot of complexity issues around that. And that's interesting for a device like this because you have the same kinds of questions like, is this device going to be reliant on another device?

Is it going to be you pair it with your iPhone and then the iPhone is doing some of the heavy lifting for it? Or is it entirely independent? You could use it without any other device. If you don't have an iPhone, you don't have a Mac, you could just get this and put it on and use it. Is this going to be different for developers than for Apple? There may be Apple apps that it could run completely independent and untethered with. But if you want to run developer apps, then you need to have an iPhone that you could load them on.

You know, is the app store on your iPhone or is the app store inside the headset? Like I think these are some of the questions that will be very telling for what the initial developer experience is going to be like and what the initial kind of breadth of what's possible is going to be there. Because, I mean, similarly with watchOS, like the early versions of WatchKit were so limited in what they could do. And it wasn't the technology that Apple was using

to build their own apps. It was this separate kind of second-class citizen version. And it wasn't really until fairly recently in terms of the lifespan of the watch that maybe about four years ago, five years ago, we could really make proper apps using the same technologies that Apple was using. And so all those kinds of questions

really will, I think, define the early days. And we'll have to wait and see too. We're assuming that Apple wants really broad, wide, rich developer experience. And it's like, maybe they do, maybe they don't. Maybe the early days are they're trying to be focused on their own apps, their core experiences that they'll be able to tune to perfection. Because if they're really up against the edge of performance and battery life, that's going

to be much easier for them to optimize for than to build something that third parties are able to do. And then you have to build so many more guardrails and things because you know a developer makes a bad choice and then suddenly it just like destroys the battery life from the device that reflects badly on apple as well as the developer and so you know if that's the experience that people have when they first get it it could be dangerous as well i feel like as you're explaining this to me i'm coming around to the idea there has to be some kind of development kit solution because there just isn't anything that can simulate this because

even if you used headset by another brand it's not going to have all of the features like one of the big features is foveated rendering right which I think at the moment i know the psvr too just has but obviously that's not going to work but this is the idea that like it's using eye tracking to see where you're looking and then just rendering that part now that for certain types of experiences is going to be a very important part of the process right like especially if you're like a game developer and stuff like seeing how that works inside of the environment that you're building will be important it'll be important for testing it's like you won't be able to test for things like that the same for the hand tracking

right so the idea that we've been hearing time and time again is this device is all about hand tracking hand tracking hand tracking hand tracking. Well, no other device that's available on the market right now has very reliable hand tracking. So the entire, this would be like saying, oh, here's the new iPad, but it has no glass on the screen, right? It's just like, and you have, it's like the entire interaction method

relies on the good hand tracking. So unless Apple can get you hardware you can't use anything to reliably simulate that yourself yeah and i think the only thing that i can think of there in terms of like without a developer kit or before general release of the device that I could imagine is like you could imagine them taking iOS components and letting you use them on this device without necessarily developing them on the device. If you imagine like a live activity or a widget or some of those kind of rendering modes they already have in iOS, you could bring those and have them available on a headset without needing a developer kit because they're doing all of the complexity around rendering and placing them places and you know dealing with with user interaction with

finger tracking and all the things that are going on there but if they just make it so that hey if you want to have uh you know a live if you want to have a live activity, you want to have a widget that you pin into your virtual world or whatever, that could be totally reasonable to be developed without a developer kit. They add a new type in the way that with the iPad, when they added widgets there, they added an extra large type. And I didn't need to have an iPad running iOS 15 in order to test that and be pretty confident that I had it right. I could do that in a simulator. It's just two-dimensional and very similar to something we already have.

Those kind of experiences, I could see them very easily developing without having a developer kit. And so it's entirely possible if they're going down that road, where initially there is limited developer support in terms of what's possible, or at least the broad experience they're hoping from developers are existing experiences, like things like widgets or live activities, that could make a lot of sense. It's like if they but if they want a rich experience that everyone's keeping taking advantage of not just the select game developers that they're, you know, bringing to the developer center in Cupertino and working with on-site. Like that's where it gets much more of a complicated story. And I certainly don't envy them trying to find that balance. As a developer yourself, right? So just developer part of David Smith, how are you feeling about the prospect of there being another platform that Apple produces, like for you and for your workload like you you

would you know i know you you don't develop any mac os apps right uh not really no right so you're on ios ipad os watch os we'll call this reality os how does that feel kind of sitting here right now what's interesting about that is that i feel like this there's a deep tension I feel as a developer around new platforms or new system capabilities have been the cornerstone of my ability to be successful as a developer. By being an early adopter, by being out on the cutting edge of whatever Apple's putting out there, that's something that as a small, one-person independent developer, that's what has allowed me to shine in a way that I can move more quickly. I can get out onto these new platforms very quickly because I'm not a big team who has to get approval. I'm not like the Google Apps team who's trying to, I'm sure it's very complicated and difficult to get approval and to be there on day one.

I've had a lot of success with doing that. Like that's how Pedometer++, that's how Widgetsmith, how those two, you know, like those apps are successful because they were there on day one. And so I'm always excited when Apple adds a new platform, when they add something into this that I want to be there. But it's complicated for this one where I feel like it seems like such a narrow use case and such a narrow audience for this platform that it's hard to be excited about it because I think it's going to be very complicated to sort of have it be something that is kind of worth the time and worth the investment in doing this because it's like if you only have a plan you know it's i don't think they're going to sell nearly as many of these as they sold apple watches and the apple watch is a difficult platform to kind of get a good return on as a developer and so if you have

a tiny fraction of that if you're talking about you know a user base in that should be measured in maybe hundreds of thousands to start with yeah like you have to sell to a lot of them to make any income back. And you're going to need to be selling at a pretty high price probably too in order to kind of recoup anything there. And so it's like, I'm excited about it in that regard, but I'm also kind of skeptical that it's going to be kind of like worthwhile. And I feel like it may, I think it's potentially more likely that it ends up being like a tvOS or something where there's a very particular type of app where it makes sense to develop for.

And then for the rest of the apps, it doesn't really kind of actually fit in. That in practice, there's not a good market there. There's not a lot of reason to show up unless you happen to be whatever platform that is. And it's like, if widgets are a big thing on a headset, then I'll totally be there and that'll be super exciting and it's a new platform. But if it's all about media consumption and communication, say, like it's, you know,

if it's for video apps and like, you know, Skype, Zoom, those kinds of apps, then it just wouldn't be for me. And it would be, I'd be happy with that in some ways, but it's, that's kind of the tension that I'm feeling right now. But this is like a change in your attitude, right? Where I feel like maybe David Smith of 2016, 2017, you would have been like scrambling to try and think of like, what is an app that works in this?

Where now you really do have a couple of very successful apps and you need to just work out how you can apply those to different things. As you say, right, like if they're like, hey, if you make widgets, you can bring them over and, as you mentioned a minute ago, like pin them to your virtual space. Like you're there, right? Like day one, you're in the app store, right? Because that is 100% your thing.

And as you mentioned, if they did do that, it probably wouldn't be very hard, right? Because that's the kind of stuff where they're like, oh, it's all SwiftUI. We're using SwiftUI here too. If you have this kind of application, very easy to compile for, and then you just do some testing and bug squashing,

and you're done, right? I'm simplifying, of course. But it's a different aspect to if you were to come up with a walking simulation app, right? Yeah. And you did that, right? Like that is a whole different kettle of fish completely. Yeah.

And I think I've been burned in the past a bit and part of the, both in terms of, just in terms of like, there's a, I could imagine coming up with like wild ideas and spending a lot of time, you know, say this summer, that's like, that's a, I could imagine coming up with like wild ideas and spending a lot of time, you know, say this summer, that's like, that's all I'm working on. And I think it's difficult to imagine there being a big enough audience to, for that to make a lot of sense is as I think the maturity that I've come to as a developer, that it's fun to do as a, like as a hobby, as something

that's interesting to play around with. But if you're viewing it as a business coming to, as a hobby, as something that's interesting to play around with. But if you're viewing it as a business, as a developer, coming to this thing as a platform that you want to support, if it's a couple of thousand dollars as a device and it's potentially very limited in scope and is a completely new paradigm for the vast majority of people, but they're not going to consider themselves as computing is sitting down in a chair or standing up with a headset on

and closing yourself off in that way. That's just a very, it's a totally new thing. And so that limiting of the space is I think something that makes me tentative going into it in a way that obviously we'll have to see. I could be totally wrong about this and they come back and they say, we found a way to make this device and it's super cheap and it's super amazing like and it's going to have a wide audience and that you know there's millions

of these selling everywhere and it's super it's like the hit thing and so finding anything you can do to be on that platform will be important and will be kind of get a good return like that certainly possible like apple has done that before of all the companies who could pull that off i'd you know certainly put apple at the top of the list but it's the ipad right you know we you know we're going to mention this before and we're going to mention a million more times leading up to this product's launch but these were the rumors for the ipad it was going to be really expensive like over a thousand dollars and it came in at 500 and it blew up and everyone was very excited do you have interest in this device personally i mean not particularly like it's i think the only use case for a device

like this that i see myself potentially interested in is the use cases i've heard of around using it as a virtual display, like if you're traveling or something like that. So when I go to WWDC, I bring a whole bunch of stuff to set up my work environment to be as productive as I can be away from my actual office. Right now I'm sitting in front of a Pro Display XDR with my laptop on a stand and I've got a keyboard and a mouse and all the things that I use to ergonomically be productive and effective. And if there was a way to have a virtual Pro Display XDR with me anywhere in the

world that I, you know, is relatively small and packable in a suitcase, like that sounds super exciting. I'm skeptical that it would be an a similar experience. But if that's the case, like that would be awesome. Like that's a use case I could I could imagine. But it's, I think, tricky to imagine myself and that's even in that situation of wanting to have something like this on my face for eight hours at a time, or, you know, sort of for long stretches, without that being something that is very fatiguing and difficult ergonomically and like challenging on your neck and all those kinds of things. And, you know, sort of playing games on a device like this isn't a use case that particular, like I'm not, I don't play many games as it is.

And so having even more sort of having a much more involved experience that requires a lot of commitment for a game is going to be even more difficult. And so it's like, I find it really difficult to be excited about this personally in that way. And so, and maybe that's also part of, you know, why, you know, I have a little bit of sort of skepticism about it in some ways in general, but I think I look forward to be, I would love to be proven wrong in that way. Like I'd be fantastic to, you know, have this device unveiled, have them show me what it does and sort of have like the aha moment of like, oh, wow, no, that's amazing. Like I really, you know, sort of like I'm blown away by what it can do. But as it is now, I'm interested in it mostly from a professional perspective as a platform that I want to support

rather than, you know, as a personal matter, being a device that I'm like super excited to to get and to use and you know would look forward to and camp out to day one for at a personal level and as a developer of apps on Apple's other platforms how do you feel about the prospect of WWDC completely being overshadowed by this? Like if they're going to, in a WWDC keynote, they're showing off a new platform. I can't imagine there's much time for anything else or it's going to be a two, two and a half hour WWDC keynote. Like how does this make you feel of like,

you know, I have all this other stuff that I do. I have, you know, ways i want to push my apps forward but apple will be spending their time talking about this instead yeah i mean i think there are two versions of that scenario and they would play very differently to how i would feel i think in one scenario it is it overshadows like the main keynote because there's the most to explain there but the other platforms still have large meaningful robust announcements to be had like there's lots of things happening still on watch os on ios um and it's just they didn't talk about them in the keynote and there's you know

at after the keynote at the state of the union and at the videos that are being released throughout the rest of the week there's still tons of stuff and it's super exciting. I'd be fine with that. I don't care if the keynote is slim on what's on watchOS, which is probably the platform closest to my heart. That's fine to me, as long as I'm getting a lot of stuff and what's new in watchOS. If it's the other version of that,

where WWDC is overshadowed by this because that's all there is, or there's very minor incremental updates to the other platforms, I think I would feel kind of more sad by that. That if it felt like the other platforms were being put on hold so that they could launch and focus on this new platform, I think I would have a bit more frustration in terms of it would feel like some of the other platforms that I'm perhaps more directly and immediately excited about and whose changes and improvements will have a bigger impact to my users and to me, that would feel a bit more crummy.

So my hope is it's the former version of that, that even if it does have this big, it's the talk of the show in terms of Apple's messaging and their PR strategy, like that's fine as long as the actual story behind the scenes that there's a lot of cool stuff that I can do and that I can look forward to launching in September.

It would surprise me if they weren't able to still do other things on other operating systems because of this. I would understand it maybe more with the hardware, just because it's just complex and they maybe needed to pull people in from all over. But the software stuff, it's not like, I mean, I just imagine,

it's not like everybody who's working on iOS and watchOS had to be pulled to develop the software for this thing. But I don't know. Yeah, it's certainly possible that that would be the case but it seems unlikely that there are certainly like there's a i'm sure there's dedicated people working on watch os on widget kit on health kit on all the other you know on swift ui on all the other platforms and if there's a there might be some drain to that but it doesn't seem like it should be you know that they couldn't do both at the same time would seem like a very strange staffing place for apple to find themselves that

oh we just can't have enough you know we only have 20 engineers and so we have to focus them all on the news thing rather than being able to have different teams you can you know take care of things and you know sort of juggle multiple platforms uh being updated at the same time. And I will just say personally, as a person who produces weekly podcasts about Apple, I'm really bummed out that they're not showing this off next month. Because I was like, my next few months are going to be easy. You know, March to WWDC, we could just talk about the headset. But now I've got to go all the way to June.

Come on. Yes, it is definitely frustrating in that perspective that it is not. I mean, I think anytime you have these big, like unanswered questions, I get, you know, it's just, you want the answers as soon as you can, both from your perspective, from having content to talk about. And from my perspective, from having, you know, having certainty about the future just feels nice in a way that it's difficult to have this like, is it, what's it gonna be? Who knows? And having to wait longer and longer is not a great place to be.

This episode is brought to you by Squarespace, the all-in-one platform for building your brand and growing your business online. You can stand out with a beautiful website, engage with your audience and sell your products, services, or even the content that you create because Squarespace has got you covered. It's so easy to get started with Squarespace. You just sign up and you choose from one of their beautiful, best-in-class, award-winning website templates.

You can just browse the categories that they have to find the site type that you have or maybe the business that you have to find the perfect start in place. Everything is customizable with just a few clicks. You can choose branding, you can choose fonts. It's really easy to then move and drag things around on the page, all in your web browser or on their app, to really customize it to just look and feel like you want it to. And it's so simple to then add in additional functionality. If you want to sell products, physical, digital, it doesn't matter.

They have all of the tools that you need to start selling this stuff online and they have powerful blogging tools to share photos videos updates and more you can categorize share and schedule your posts to make your content work for you i have been using squarespace personally for over a decade at this point i mean it's got to be getting close to 15 years which is just like a wild thing for me to think about now. But it is because if I want to put something online, I don't want to have to deal with any of the like complicated, hard stuff. I want to make the content. The content is what I care about. And Squarespace gives me all of the tools to be able to make that content sing. And so I can just focus on what it is I care about and not all of the stuff that I don't. So go to squarespace.com slash upgrade and you can sign up for a free trial today with

no credit card required. Then when you're ready to launch, just use the offer code upgrade. You'll save 10% of your first purchase of a website or domain. That is squarespace.com slash upgrade. Then when you sign up, use the offer code upgrade to get 10% of your first purchase and show your support for the show. Our thanks to Squarespace for their support of this show and all of RelayFM.

Everyone's talking about AI these days, right? AI all over the place. I got access to the Bing, the Bing AI search thing today. I was playing around with that. I'll actually say I'm trying to plan a vacation, like me and Nadine are looking to take a vacation. And I got some good suggestions from Bing. But one thing that they've done,

you probably saw the wild things it was doing last week where people were breaking through. Ben Thompson's Stratechery article is not just one of the best things I've read from him it's like one of the best things I've read in years so it's called From Bing to Sydney I'll put it in the show notes you should go and spend some time reading it

it's unbelievable also Kevin Roos at the New York Times had the column where Sydney told him that it loved him which was like a whole other thing because of this the column where Sidney told him that it loved him. Yeah. Which was like a whole other thing. Because of this, the way that Microsoft, I think, is solving this for now is you can only ask five questions

and then it makes you restart. Which was really annoying for me today because I was trying to refine some things and I found that quite frustrating. Nevertheless, you have been dabbling with another of OpenAI's projects, which is, I think, currently one of the less sexy ones. It's called Whisper.

I feel like not a lot of people are talking about this. It's not Dali. It's not ChatGPT. It's the same company, OpenAI. Whisper is a speech-to-text engine. And you have been feeding a selection of podcasts into Whisper. And I kind of wanted to get, one, why are you doing this? And two, what has been your experience of this so far?

Yeah. And so I think that AI is definitely sort of like the technology of the moment in so many different ways, whether it's the image generation, the text generation, or in this case being able to convert spoken audio into text. And I feel like it's, there's the reason it's so kind of like of the moment and feel like it's, it's transformative in a way that it wasn't, you know, many of these technologies have existed for a long times. They're just, there's something different about it now. And that difference

seems to have been this crossing this point where suddenly these things that previously were just too computationally intensive to reasonably and pragmatically do kind of at scale. Now suddenly we've been able to kind of cross that point where, as is so often with the case with technology, it goes from something that is niche and limited and requires very specialist tools and very specialist hardware to something that can be run much more broadly and just sort of at a scale that is different,

that you can have hundreds of whatever, tens of millions of people talking to ChatGPT and it doesn't completely implode on itself and say anything with mind journey, a mid journey or any of the stable diffusion models and with in whisper specifically's case it's this difference of you know speech to text is something that has existed for a very long time i mean like nuance and drag and dictate um have been doing this you know for a very long time you could be able to do automation on the mac even with your voice and you you know, be able to speak to rather than type, you know, there's some great accessibility things that are

coming out of that for a while. But what made Whisper different and is interesting as a technology is that it is making this thing that previously was either very proprietary, or very sort of expensive and complicated, something that anyone can run on almost any device in a way that is pretty performant and certainly way more performant than it was before. That, you know, so what I do with my podcast searching system is I've done this before. Previously, what I actually used to do is I tried this before and YouTube was the only place that would do kind of wide scale automatic speech to text. And so I would actually take podcasts, turn them into videos, upload them as private videos

on YouTube, wait for it to do the automatic like subtitling, download the subtitles, and turn that into the transcript, which is terrible and awful in lots of ways. And the actual quality wasn't very good. And then Whisper comes out. And instead, I can write a script that just downloads the podcast episode runs it on my, you know, runs it right on my MacBook Pro. And I get the result in a few minutes. And, you know, it doesn't cost me anything. It's there's no infrastructure overhead involved in this, you know, you can run the same model on an iPhone even and have it work with some degrees of accuracy.

And so it's this very different place to be. It just wasn't possible before. How does this work? Like, how do you generate these transcripts? Yeah, so I mean, Whisper is just a technology that you feed it an MP3 file, basically, and it transforms it in a few ways, and then it gives you a transcript, as best as it can come up with, of what that file is,

what the people in that are saying. And in my case, my goal was to make a keyword search quality podcast retrieval tool. And so I make it so that this is fine for us. It's not a true transcript in the way that if you wanted to read the content of this show, but if you wanted to find out when we talked about RealityOS

and you search for RealityOS, it'll be able to transcribe it and give you the place in the episode where we talked about it and so it just isn't you know it's this i don't know it's a magic black box that you feed mp3s into and get the textual versions of that um coming back to you on the other side i did a search for reality os and it currently doesn't find anything because i guess it's just i don't even know because this it, right? Because they're not human.

No. There are things that are missed. Yeah. Right? So because a human would hear that, and if they knew what we were talking about, they would put it in there. And I know we've spoken about,

oh, I found it here, reality space OS. Right? By looking through the transcript like search results but like I know and you know that's one word, right? But like that's what you mean about this. This is not a it cannot currently create a comfortable

readable transcript but it can almost I think of it almost like an index. It's doing like an index of the words to the best of its ability and if you think of it almost like an index. It's doing like an index of the words to the best of its ability. And if you think about it like smartly in those ways, you can kind of find what you're looking for, right?

Where like if I was doing a search on something that I thought was perfectly transcribed, type reality OS and it said no results, be like, okay, there's nothing here. But because I know what the limitations of the machine are, I can just search for the word reality and I will find what I'm looking for.

Right. Yeah. And I think what's so interesting with so all of these systems is based on the input you give it is so determinant on what the output is. So RealityOS currently doesn't get transcribed correctly because it's not a public thing that exists on the internet. In like, say Apple announces it at WWDC, a year from now when they've updated their language model, and that's a like a proper pronoun that exists in the English language, it'll

probably transcribe it 100% correctly. That there's so many situations where now it gets brand names, it gets trademarks correct, because presumably it's indexing, it's gone out and looked at all of the content on the internet and it's worked out, well, what is in the English language? What are these words? And then how can we build a model to predict when someone was saying one of these words? And so I feel like it's the same thing with all of these, you know, with the chat systems or the image generation, it's so much a trend, it's just transforming the input into some kind of useful output in an automated fashion. And so it's really complicated. Like if you're trying to transcribe something that is a, you know, a very standard

use of the English language or with, you know, nouns and or with nouns and proper nouns that are known, it'll do a great job. And it'll be kind of amazing in that way. But otherwise, it's just sort of guessing and doing its best. And that's just like the fundamental reality of all these tools, I feel like, is they're incredibly powerful, but they're not intelligent. We call it artificial intelligence, which is not really what's happening. It's just really powerful, careful automation

that can enable people to do cool things more quickly. But it's not actually intelligent in that way. It's just giving you more handles or more inputs into information that you previously wouldn't have been able to do that. The reason I got started with going down this road is like, I love listening to podcasts. And very often, I think to myself, you know, Oh, what episode did they talk about that? Like, Oh, you know, I remember this thing. And there's this episode of connected where Federico talks about how he got, you know, this very particular device or something like that. Ooh, when was that? And like, that's really, that's a really complicated question to answer.

Because you can't just search Google for it in a way that if, you know, Federico had written an article on Mac stories, well, I could just Google and like, say, site macstories.net, and they're good, you know, boom, it's gonna find it. But, you know But audio was just not indexable in that way, but now it is. And that transformation is just the amazing part of this. How much energy or power does it take on your Mac to produce one of these transcripts, say, for an episode of this show? Sure. So for an episode of Upgrade, typically maybe maybe like an hour and a half

long something along those lines um i can generate a transcript on my m2 max macbook pro in maybe about 20 or so minutes 20 25 minutes um which is reasonably fast as a you know as these things go. But during that time, it is completely using every core of the device. And the fans come on and it is, if you ever wanted to cook an egg on a MacBook Pro, this is how you do it. And I can say going back and re-indexing the entire back catalog of Upgrade, of Connected, of the accidental tech podcast,

of the talk show there was a lot of heat being generated on macbook pros in my office for quite some time just like slowly wearing those claws down yeah i mean i i it definitely made me a little nervous that i was like huh like should i should i you know should i be worried that i'm gonna like you know overcook my macbooks but i'm like i'm, no, this is like, it says pro in the name, right? This is if I was a, you know, a researcher doing some kind of those people, they always show in Apple keynotes of, like these, you know, scientific people who use Mac pros to do protein folding or something like they would be doing this, this is well within the spec, but it definitely made me a little nervous to be,

you know, seeing my my device be running at that, you know, that hot for that long, you know, cause I think it was, you know, the course of maybe about two weeks, 24 hours a day. So, you know, it's definitely not something that is like lightweight in that way where, you know, you can do it. And if you wanted to, you can run this on an iPhone and it works, you know, but if that's one of those, you definitely wouldn't want to do it without be you know without being plugged in because it would absolutely destroy your battery and like the back of the phone you know you can feel exactly where the chip is on the iphone's back because if you run your finger down it'll be like scorching hot like it's you know because especially there because it's entirely passive so all of the heat is just trying to doing its

best to radiate out otherwise. So it's definitely, the amazing thing is that it's possible, that I don't need some dedicated, it's not like, oh, you need to get a Mac Pro with an afterburner card and then you can do this. Or like how Google and Amazon, et cetera, build these specific chips for right? For machine learning. Like they were put in servers.

So, you know, like it is, I guess it's taking advantage, I guess it's taking advantage of the neural chips inside of these machines. I don't know that. Is it? So Whisper has not yet been sort of ported or adapted to take advantage of that in a way that like apple themselves with

stable diffusion they like apple came had you know created a version of cable of stable diffusion that could work with the you know the all of the core ml stuff that apple has been adding to their chips for years but it's right now i don't believe there's a version of whisper that can do that okay i believe i've heard that somewhere but i could be making this up now so don't attribute it to me that apple are looking at that like in the same way they have other things of like how can they make sure that this stuff can run yeah i mean if they do it'll be amazing because that's so like purpose built it's almost like you know this very purpose-built bit of hardware that would both speed up the process

and probably, as a result, do it so much more efficiently than having to do it on the CPU or the GPU on a device now. And especially if they wanted to make this something that worked well on iOS devices, it would be even more of a huge win there. Because I feel like every iPhone for years has been shipped with these neural engines that do nothing, like 99% of their life. There's so few use cases where it makes sense to. But with things like stable diffusion, with Whisper, with these kind of new AI things,

finally, it feels like the payoff is coming for all this, you know, all the silicon that's just been sitting dormant in iPhones for so many years. And like, that's exciting and interesting for me looking forward to like the next year of development is are there going to be like there was a moment where all the profile like avatar generators were super hot on the App Store. And most of those had to be run on the server side. But if Apple did some work in iOS 17 to make those kind of things possible to run on an iPhone in super real time, very lightweight, very optimized for the neural engine, how

amazing would that be? And how kind of cool and interesting would that be as an unlock that it's not this like we're talking about with the headset it's not the situation where only the new people who are going to have it it's like every person who has an iphone since you know whatever it is the iphone 10 or something has has a neural engine of some variety you know variety in their phone that they could use to take advantage of this and so that would be just amazing that would be an interesting thing for apple to do for 17 right to to show off some kind of machine learning capability that they that people would be interested in now and maybe a way they wouldn't have been before yeah right because like hey you remember that thing that

is that you you know you know that thing that everybody's love to do right now well you can do it all on device and how much would that fit in with apple's kind of mentality right the on device-ness of it all yeah and like i could see i could you could imagine a world where like i have no idea how whisper compares to their current dictation system in ios but it seems very good it seems very impressive and makes me wonder about if apple would benefit potentially from you know making their own version using the same technology and concepts behind this and you know building it into ios and then if they do that for themselves, then they can make that available to developers through some kind of very lightweight,

wrapped, hyper-performant system. And then it starts to get really interesting for what that opens up for, where you have a podcast app that can show you live subtitles for what you're listening to. Or when you hit share, the shareable clip includes all of the live index transcript

of what's happening. Like these things, it's so different when you can suddenly do them in a lightweight, performant way on device, do it right where the user is. You can do it offline. All of those sort of advantages start accruing. That is where it starts to get very exciting for me. So there were, Mark Gurman reported in his power on newsletter apple had an event like an internal event um a few weeks ago which is their like ai summit it's kind of internally dubbed as like nicknamed as wwdc for

ai but it's just for apple employees sure Sure. And I've got a quote here. I'll just read from Mark Gurman. In a brochure for the event, Apple's AI chief told employees that machine learning is moving faster than ever and the talent we have here is truly at the forefront. It's an interesting, they're at least focusing on it.

And we've seen it, right? Like you said, we've seen it. We've seen them do enhancements to the operating system to support and kind of optimize for some of these models. I think it would make a lot of sense for them to have some kind of

what is now called AI-focused features for the next couple of versions of ios because they could it could really catch on with people yeah and i think too it's an opportunity for them to do ai in a way sort of in the way that they would talk about it in the sense of like in a way that only apple can right it's that sense of like when they did their image indexing system in photos, I remember Craig Federici talking about how they did the work to build their own training set of appropriate images for defining what a mountain is or what a dog is. is or what a dog is and did it in a way that then they didn't just scrape the internet and then use that as their basis or use user photos as the basis for it. They did it in a way that is much more predictable, reliable, like legally sound, all those kinds of things. And like, I could imagine

them going down that road with a lot of these types of features and saying, this exists other places. If you wanted to generate an image, you could do it with Stable Diffusion or DALI. But if you do it there, there's all these questions or difficulties or ambiguities about how that works. But if you do it with our system, we've taken care of all that. We've done it kind of like, quote, the right way in a way that everyone else is doing it the easy way or the cheap way. This is the Apple way. We're going to do it the best way.

And that would be very interesting to see where they could go with that because they have devices that are capable of this and they have kind of a history and a pedigree of going down that road. And the results are great. When they did uh you know image indexing on the iphone initially they were doing it in a way that was like totally different than everyone else because everyone else was doing it in the server and kind of doing it using building models based on user images apple went the other way and they did it completely on device and

both their own kind of internal image sets and training models. And like the result is amazing. And the stuff that iOS 16 can do where if, you know, you ask it to show me pictures of the beach and it'll show you all the pictures where you were at the beach. And you can ask it very sophisticated, nuanced questions and it will be able to answer you in terms of what is the content of my images? Is there text in the image? You can select it and search it. And like that kind of stuff is the content of my images? Is there text in the image? You can select it and search it.

And that kind of stuff is just where they're doing, they've done this in a very different way. And it's been very impressive. And so I think it's very exciting for them, for me to imagine that that's how they see themselves in both and it's something that they care about enough that they're putting in their own version of it, that they're not sort of worried about not being first, they'd rather be best.

And I think that's a great place for Frapple to find themselves. Because the incredible growth in AI systems recently in the public sphere, it makes me very curious and intrigued to see where they're going to be on their internal side coming at it from a potentially more rigorous approach. week. You also become a RelayFM member, so you get access to a ton of bonus content, the RelayFM members Discord, which is a great place to be. You go to getupgradeplus.com, you can sign up for just $5 a month or $50 a year, and you'll be helping support the show. On last week's episode, me and Jason and I played a game of Marvel Snap together. In the show notes

for the Upgrade Plus version, there is a video recording of the game with our audio over the top so you get access to that in this week's episode i have a ton of ask underscore questions that came in we're going to do a bunch in a moment but we have tons more that was submitted by relay fm members that we're going to do in upgrade plus so if you want this and no ads go to get upgrade plus.com Are you ready for some ask underscore questions? Beautiful. Look at that. The lasers, they're still here. First question comes from Joe, who wants to know, what additions or improvements are you hoping for WidgetKit this year? Yeah. So I think the main thing that I feel like I would love to see widget kit get in iOS 17 is a broadening of what lock screen widgets are. I feel like they were really

interesting, really, they're a very cool, compelling, popular feature. But what they can be and how they look is so limited and so often limited in weird ways that it's kind of frustrating for me as someone who spends a lot of time trying to make widgets awesome. Like there's only so far I feel like I can make lock screen widgets awesome because there's no color. Layout options are really limited. The size options are very limited. You have a giant lock screen, but you can only put them in these very specific places. And the one above your lock screen has almost no customization options whatsoever. Like it feels really limited. And so I think I would be very excited if widget kit got more options there that if we could do color widgets on your lock screen, you can, or there's

a new, you know, different sizes, or even you could take your home screen widgets and put them on your lock screen or you know crazy things like that would be wonderful and i think very exciting um to kind of push that forward and especially now that we know kind of what it looks like with the always on and some of the other concerns and things that i could imagine if you know going it came into play with lock screen widgets this year, where they were announced before the always-on display was announced. And so there are certain things in there that are kind of inherent and challenging. And they have potentially evolved what they think of with the always-on lock screen, where

initially it always had the image in it. Now it has the image. Or you can go to a black wallpaper when you're in always on. And I feel like those evolutions and refinements mean that there's a bigger possibility for what lock screen widgets could be. And I think that's where I'd be most excited. I mean, if there were any changes to widgets anytime, I'm always excited. Like I'm super up for taking on whatever Apple wants to give me on widgets. But that's the area that I feel like home screen widgets feel pretty good.

They feel pretty robust. And I don't have nearly as long of a kind of wish list on them than I do with lock screen widgets right now. Do you think we're ever going to get interactivity on home screen widgets? I don't know. I think what's strange about that is I, as someone who has a lot of users who interact with widgets on a very regular basis, I don't hear a lot about requests for interactivity. They're not asking you, right? Like, can't I do this?

Can't I do that kind of thing? Yeah. It doesn't come up in the way that I would have expected it to in a way that I think people are pretty happy with them not being interactive obviously like if they became interactive people would be happy and you'd have new things that are possible um I think the things that I hear more are people wish they could be updated in ways that were more live and real time or like being able to show animations or videos or those kinds of things on the lock screen or on the home screen would be compelling but being able to interact with it like putting

buttons there just hasn't been something that i i sort of hear about as much and it's something that yeah sure it'd be cool if it happened but i'm'm not... I think the way they went about it makes a ton of sense technically, that it's this very predictable thing that you can put on your locks. If you're going to put it on the home screen, it needs to be kind of bulletproof. Because if the home screen, you do something weird as a developer and it crashes your home screen, well, your phone is in trouble in a way that if you do something weird in your app and your app quits, well, you're just back to your home screen, well, that's like your phone is in trouble in a way that you have, if you do something weird in your app and your app quits, well, you're just back to your home screen. Like the home screen has to be bulletproof. And I feel like the way that they've done it by making them

pre-rendered and stable is a way of making that like guaranteed. And so I don't expect that necessarily, like sure, if it'd be cool, but I don't hear a lot of requests for it. And it's not something that I feel like is as compelling as it's like conceptually, it would be compelling. But in practice, even back in the day, when we had interactive widgets with the old today style widgets, like it wasn't, I don't think that the interactivity was the killer feature. I think the putting that information in a place that is accessible and then making it customizable by the user is what's cool rather than necessarily being able... If you want to interact, interact with the app.

Tap on the app and you can get into that. And I think some of the changes they've made with that about being able to launch into other apps and those kind of features have made it really compelling. So like, sure, if it happened, but that's not something that's really on my sort of to-do list or my radar. Yeah, I would say personally, I have not missed,

or like I don't particularly feel like I need interactivity in my widgets. Like I kind of see my widgets more as like little windows into the app. Like if I want to do something, I'll open the app, do the thing. But other than that, it's just here's some information. I don't feel like it's something that I really miss as much as I thought I would when it seemed like that was the way that was going to change

of how widgets were going to be. Rob asks, how do you balance the many apps in your portfolio? I've been happily using Podometer++ and related apps for years. When WidgetSmith became a breakout here, I was happy for you, but worried that the support for other apps that I use would drop to a minimal. I've been very happy to be very wrong about that, but i don't understand why is it for diversification in your business i mean i would say since things like widget smith becoming you know wildly successful and popular when it first had its moment when it went viral on tiktok like i am less diversified and less uh you know broad in my approach as a result of that.

That I think, I'm glad if people think that the other apps are still getting attention and it feels that way to them because they're getting some attention but not nearly as much. And I think this is just the reality of working on a portfolio of things makes sense. I give products the attention that they can pay for.

If they were a consulting client, like back in the day before I made apps, I was a consultant, and clients would buy a certain amount of my time and pay me for it directly. And in some ways, I view um, in a similar vein, just, you know, with, with a slightly different perspective where, you know, WidgetSmith is by far has the broadest audience that I have and has the most potential for me to do things. And so

I spend the most time on it. Like it sort of, it buys that time for me and, you know, pedometer plus plus is gets the next biggest chunk of that. Um, and then, you know, Pedometer++ gets the next biggest chunk of that. And then, you know, my other apps from there kind of falls off very dramatically. That beyond that, apps that I'm working on, I'm primarily working on because they're exciting to me. Like they're more of a hobby or something that I enjoy working on or I have a personal reason to work on them rather than it being like a business decision.

And I think that's an evolution of the way that I work. Like when I started and certainly I've been doing this for 14 years or something, I've been making apps now. And so in the early days of the app store, like the best way to make, you know, sort of to try and make it was to build lots and lots of different things. And that was because not a lot existed. You know, there were no apps on the App Store. So anytime you made something,

it was new and interesting and pushed something out. Whereas now those opportunities I find are much fewer and far between. And there's a much better return on enhancing and going deeper into things that have already found traction. That trying to get traction is so difficult that it's better to just

sort of double down on things. And so I just kind of balance them by based on their general popularity. And then when I have, you know, periods of, you know, there's inevitably in the Apple developer cycle, there are going to be kind of these quiet periods where, you know, there's not much that Apple is doing right now in terms of new stuff versus from June to September. So June to September, WidgetSmith was my primary focus. And I was doing everything I could to have the most compelling thing day one for that. Some of my other apps, not so much.

If they got an update, that's great. But if they didn't, it was fine. And so I think there's just a maturity that you kind of have to have to be realistic and understand that sometimes it's going to disappoint people who care about one of the less popular products are just going to be kind of, you know, it's like if you're a big fan of the, you know, whatever, the Mac Pro, and you're going to get sad that Apple puts all their attention on the MacBook Air, but that's just the reality. That's just what makes sense. And it's, you know, all those same forces are magnified, you know, a thousandfold when it's just a one-person developer trying to make those kind of decisions about how

to best allocate your time oz asks where do you predict the apple watch could go in the next few years do you think we're getting close to third-party watch faces or is this still a pipe dream oh boy third party watch faces i mean third party watch faces are like my that's my dream feature like that is the thing that i i enjoy making them like i make them now they don't like that doesn't exist as a thing that apple has created there is no like third party api for watch faces but i have found ways for years now of making my own. And like, you know, you, you and I have met up in person and you still look, if you look at my wrist, I'm very rarely running a Apple watch face because I like my own better. I've come up with dozens of these over the years. I think they're super compelling and interesting,

but that's just not been a place that Apple has kind of decided to put effort into. And I don't, at this point, it is certainly a conscious choice. There's nothing technically why they couldn't. I mean, I'm making apps that pretend to be watch faces and running them in kind of a way that doesn't make a lot of sense and isn't necessarily the most battery efficient. Like if they wanted to, they could do this so much better. sense and isn't necessarily the most battery efficient. Like if they wanted to, they could do this so much better. But it seems to be a conscious choice that they've not done. And I think what is complicated there is the future of the Apple Watch is whether that is they're holding it back for a kind of using it, you know, they have this trump card in the back that they're trying to play at the right moment where they can get a big impact or they have a quiet year that they want to, you at the right moment where they can get a big impact or they have a quiet year that they want to, you know, turn into something big. And if that's the case, like this

year would be really cool. It's the 10th anniversary of, you know, watch OS, it'll be watch OS 10, like be a great time to have a big splashy feature. But like, otherwise the, you know, the history, if I was trying to predict the future of the Apple Watch, I would say, you know, history shows us that every year what they'll do is they'll add a couple of new workout modes and a couple of new kind of workout fitness related things. You know, it's every year there's something new that they're doing there. There'll be some new capability on the new hardware that is an improvement, kind of an evolution of what's happening there. And they'll add two or three new watch faces of their own. And it's like, if I was going to place a bet on what watchOS 10 and the Series 9 watch

was going to look like, that's what I would predict. I think that's the most likely scenario. I think the only thing kind of pulling it slightly differently this year that, you know, if I'm going to try and do a little wish casting is this year, you know, last year they dropped support for the Series 3 Apple Watch. But that, you know, that was the first year that they did that. And I'm curious to see what a full year of being unencumbered by the series three looks like for watch os that i think that's interesting to me that they'll still have i don't know at what

point they decided they were going to drop drop the series three like if that was a process something they knew going into all of watch os 9 then you know maybe it doesn't actually make a difference but if it they didn't know that until like january, you know, of last year, and so there were some choices they couldn't make because it would have had to support the Series 3, like that's going to be supported and upgraded regularly or is it just kind of be this thing that happened once and you know will get updated every few years and from a system you know software perspective doesn't get much so it's hard to say but i mean third-party watch faces i'm so there for that like i will be shouting as the top of my lungs and jumping up and down if that ever happens at a wwdc maybe definitely be a new watchsmith there i feel like if that was the case uh i i am intrigued by the idea of there could be something that they are able to do now or a set of things they're able to do now because they can they can be more they can they can rely more on the hardware and what the

hardware is capable of because series 3 to series 4 was a big jump right like yeah the series 4 watches they have remained kind of like better like you know better and over the long term than the series 3 and i guess they also stopped selling the series four and the series right the series three just kept selling it which is dragging everything back yeah and jonathan asks when you want to implement a feature in an app but you don't know how what is the process that you work towards building something into your app do you start with documentation tutorials or example projects and what do you do if there aren't a lot of resources so what is your learning process like how do you understand new things and how do you kind of like test putting them into your applications so i mean i think everyone has a different approach to learning and mine tends to be very experimental and kind of I very much am on the side of rapid prototyping,

get something working, build it as quick as you can. And that's the best way for me to learn. I don't really read a lot of documentation. I think the only bit of Apple documentation that I would actually could say that I've read is probably the human interface guidelines, like the technical, you like documents that sometimes get produced, or like the Swift language guide and things like that. Like I, I just get I just can't do it. Like those hurt my brain. I'm not that kind of developer. And tutorials, I have a similar kind of difficulty, like I struggle just wrapping my head around something unless I can see it and

use it. And I'm in Xcode doing that. And so if there's a new feature I want to build, I just start, I suppose, like I just will think, okay, well, what is like a basic version of this? Or how can I probably first step is, is this a problem that I can solve at once? Or do I need to break it into five smaller problems? And then I break it down into five smaller problems and try and pick, you know, pick the first one and just try and build it. And if I don't know how to do it, I'll tend to just like either experimentally do it or just do a lot of Google searching and, you know, kind of looking around or those kinds of things. Because usually there's most things tend to have, they're not, you're not the first person to solve a problem. Even if it's a new technology, a new platform, there are rules or things that you

can benefit from where you can kind of be taking your past experiences and bringing it forward. Or I try and say like this new thing, how is this similar to something else that I've done before? And in my case, you know, I've been making apps for 14 years now, I've launched something like 60 different apps over that time. Like I got a lot of back, I got a lot of code in the catalog that I can go and look at and pull from and, you know, kind of learn to see if I can take this thing and adapt it into the new thing, which isn't necessarily the best always in terms of I'm not doing these kind of clean room, oh, Apple introduced this totally new technology,

let me just implement it in a totally new way. It's like very often I'll end up saying, can I use this old version and make it better and then refine it and move it forward going from there. But for me, that's what works best. And I feel like for a lot of my best advice, usually to someone who's trying to get into app development or wanting to learn something,

is find a problem that you feel like you are excited and passionate about solving, and then just start trying to solve it. And that is the best way to learn, that trying to learn it theoretically never works for me. There's something powerful about, like the most exciting part of my job is when I hit build and run

and then something that didn't exist before now exists. Like this capability, this feature, whatever that is, like there was a magic in kind of taking this thing that is theoretic in your brain and turning it into reality. And that is like actual in your brain and turning it into reality and that is like actual feels like magic like it is it just you know it's turning something from nothing into something and so i just go through that process and try and do it as many times as i can how can i make this from something that doesn't exist into something that does exist and get started and if you do that

enough you start to get good you kind of learn how to learn um you know through through that process and i feel like if you get too stuck in documentation or feel like you have to understand something before you try it you'll hardly ever start and that's you know definitely worse thank you to everybody who sent in a question for under school if you would like to send in questions for our next episode where you will be asking things of casey liz just go to i don't know why that made me laugh to say that but yeah if you want to make casey answer for his crimes go to upgradefeedback.com and you can submit questions uh for me to ask casey on the next episode uh but most importantly right now i would like to thank you underscore david smith for joining me on this episode.

It has been wonderful and fascinating to talk to you as always. It's been an absolute pleasure. Thank you for having me. If you would like to find David's work in the meantime, here's a few places you can go. If you want to find his blog, David does a great job, by the way,

of you're going through, I know some pedometer plus plus stuff right now, and you've been detailing a lot of the kind of like design diary that you've been going through. You can find out more about that and find out about David's apps at David hyphen, right? Smith.org. That is an interesting URL you've got there.

.org as well. Well, there's David Smith's a hard, it's a very popular name. Can't you get like underscore something? Surely you must have one. I do have underscore davidsmith.com spelled out, but then you have the problem of you can't use underscores in URLs. David-smith.org works. Have to do what you just did and say spelled out,

which is that's no good. Yeah. All right. David hyphen or David dash. That's the word I was looking for. Smith.org so you can find out more there or just if you google david smith what happens do you think i think if you do david smith app i would be the top results but david smith there's a famous sculptor who

was also called david smith so he's usually the number one hit i mean i just got the number one hit is a brit British embassy security guard convicted of spying for Russia who's going to spend 13 years in prison. Not me. Are you sure? That was three days ago. So Google is full of this right now.

So your Google juice is not good right now. Your SEO is bad. Search for Widget Smith. Search for Pedometer++. Those are trademark names that are going to be much more unique there you go and widget smith you know everybody knows it right who's not using it come on you can also listen to david's podcast here under the radar on relay fm and david is underscore david smith you're on mastodon.social right i am yes great you can listen to my shows of course here on relay fm check out my work at I am, yes. Expander for their support of this week's episode. I'll be back next time. Until then,

thank you for listening. Thank you to David Smith. Say goodbye, David. Goodbye.