00:00:00 ◼ ► From Relay, this is Upgrade, episode 622. This show is brought to you by Sentry, Decagon,
00:03:23 ◼ ► She did note to me that maybe it's just like the NBA is sucking up all the discourse on the internet,
00:03:54 ◼ ► I think this time, logistically, it's maybe a bit more complicated than your typical World Cup, right?
00:04:42 ◼ ► This has changed over time, and there's just much more news in the summer now than there used to be.
00:04:48 ◼ ► So often in the summer, we have to come up with topics and special kind of evergreen one-off episodes
00:04:55 ◼ ► to kind of help us continue to produce the show every week and for it to be the best that it can be.
00:05:04 ◼ ► Summer of Fun runs from the week after WWDC to the week before the iPhone draft, which seems, when you say it like that,
00:05:21 ◼ ► But also, the other thing that happens is scheduling issues in the summer because people are traveling.
00:05:28 ◼ ► So like you and I had a, while I was driving up here, actually, we had a back and forth about some travel that you're doing.
00:05:37 ◼ ► We had to, like, it's like we would swap this over here and then do this here and pre-record this.
00:05:42 ◼ ► But one of the things that happens for us to do those episodes, sometimes there are guests,
00:05:49 ◼ ► But also, it is a time where we sometimes do a pre-record, where we'll record maybe a week early,
00:06:02 ◼ ► And it has to be something fun and evergreen so that we can hang on to it and run it late because one of us is traveling.
00:06:14 ◼ ► Because I'm going back, there was no other time to record it because I was traveling up here and then I'm traveling back.
00:06:21 ◼ ► And so it pretty much had to be today because my son's graduation is literally at the time of upgrade.
00:06:36 ◼ ► Joseph wrote in and said, Jason mentioned on Six Colors that he took notes in a traditional reporter's notebook at WWDC this year.
00:06:44 ◼ ► I'm curious why he chose that over a laptop, how well it worked in practice, and whether he'd do it again.
00:07:06 ◼ ► I thought, so there's a thought process here, which is like sitting and typing in notes.
00:07:12 ◼ ► I feel like I can do it and I end up with a document that can be turned into what I referenced during the podcast we do about the event.
00:07:36 ◼ ► I had to write fewer things because there were just – there's no way I was going to get more down on paper.
00:07:41 ◼ ► And so I had to kind of pick my spots and also do the thing where like something happens and you do a circle around it or you underline it or like put an arrow next to it and say, this is actually really interesting.
00:07:52 ◼ ► And I've done this throughout my career, but, you know, it has become more common to just sit there with a laptop.
00:07:58 ◼ ► And I just – I thought maybe this time I would try that out to not over note it and see if I could do the pen and notebook thing.
00:08:11 ◼ ► I, you know, again, it was more of an experiment than anything else, although I'm pretty happy with it.
00:08:18 ◼ ► It's different because I don't have that notes document that I can refer back to directly.
00:08:24 ◼ ► Instead, like I was writing my story about this, which I called emptying the notebook because I got to literally empty the notebook.
00:08:31 ◼ ► And, you know, I was writing it and I had to go get the notebook and bring it out because like iCloud doesn't sync the pages of the notebook.
00:09:04 ◼ ► I think you were either using the original one or they ended up producing it as a regular product.
00:09:58 ◼ ► When I'm at Apple Park, I want to kind of soak in the feeling of being there a little bit more
00:10:34 ◼ ► But I then left my notebook in my bag and put my bag behind my chair while we were recording the last episode.
00:10:41 ◼ ► And I didn't want to get up because I was going to have to move all the microphone and everything.
00:10:55 ◼ ► And I do agree with that process that sitting and making a note about something actually kind of enforces it in your brain a little bit more.
00:11:05 ◼ ► There's, you know, as with everything in our world, there's this big narrative about like, oh, you should never take notes on computers.
00:11:21 ◼ ► I'll shout out also, I mean, Grouper is a great example because he, I've always said publicly that I am fascinated by his choice because it says, I'm here, I'm not here to break news.
00:11:39 ◼ ► And I've always admired that, but there was a long stretch in my career where I was there to cover the breaking news and feed things out to people and all of that.
00:11:58 ◼ ► So it's a really nice pen with a clip that's small that can clip to the little rings inside the reporter's notebook and then pull it out and magnetically attach it.
00:12:16 ◼ ► So we mentioned in the keynote that Apple showed some tools for letting you dynamically resize your apps to different window sizes.
00:12:22 ◼ ► Like it's got a particular shout out obvious what that's going to be for, which is for the folding iPhone.
00:12:29 ◼ ► So they want you to kind of start getting ready to have your, uh, iPhone apps being able to be shown in different sizes.
00:12:35 ◼ ► Um, it turns out that one of the things coming to Mac OS Golden Gate is the ability to also resize iPhone mirroring windows.
00:12:48 ◼ ► Um, like some of apps, I think the weather app is one of them where you can just stretch it out.
00:12:53 ◼ ► So that is actually a nice thing you can do to test this, I think, uh, in, if you want to test it, but I've already seen some developers kind of playing around with these new tools in Xcode and kind of like stretching their app UIs out a little bit.
00:13:06 ◼ ► Um, so I think it's good that they're providing some of this stuff now rather than providing it in September, which is definitely a thing they could do, but, you know, would just be like, Hey, and now it's also for developers to get ready for this phone.
00:13:18 ◼ ► You've got three weeks, uh, at least this way people can start to get the basics of what their UI might want to be.
00:13:30 ◼ ► And they have multiple ways that they can expose this without saying, Hey, everybody, we're doing a folding phone.
00:13:53 ◼ ► We're going to talk about betas in a minute, but, um, but yeah, it's, it's, this is the classic Apple thing, right?
00:13:58 ◼ ► This is where you look and you say, ah, Apple is prepping without actually saying that they're prepping, but we all know that they're prepping for the folding phone.
00:14:13 ◼ ► I'm not sure, but, uh, they also are adding support for higher resolution and better, uh, capabilities for, uh, for, for webcams that use the HomeKit secure video.
00:14:29 ◼ ► And somebody pointed out to me, that would be a good feature to have if Apple's going to do a webcam, which is again, rumored.
00:14:43 ◼ ► And it's just like, yeah, I don't think Apple put that in there just for a car, a sake, right?
00:14:51 ◼ ► No, I know that this is the, the year of like, you know, performance updates or whatever, but that feels like something a little bit different because it's like, this is also for up for third parties to use.
00:15:10 ◼ ► One is what features are laying groundwork for products that we know that are coming that Apple is not going to announce.
00:15:16 ◼ ► And the other game is what longstanding features got improved because Apple needed them to be improved for other features that they're doing, which there was a lot of this year because they needed to update Spotlight and shortcuts and all sorts of stuff so that their features could use those things better, even though we'd been asking for those things for a long time.
00:15:35 ◼ ► Number one customer of Apple is Apple, other parts of Apple, then sometimes they can get them, not all the time, sometimes they can't, but sometimes they can't.
00:15:46 ◼ ► And the App Store is like, sorry, no, but Spotlight, we need you to redo the Spotlight.
00:15:52 ◼ ► One of the stories leading up to WWDC was wondering what Apple's developer story for AI would be.
00:16:00 ◼ ► And, you know, one of the things is that the foundation models now are significantly better than the foundation models that they're replacing.
00:16:07 ◼ ► But there was also private cloud compute, like what was going to happen to private cloud compute.
00:16:17 ◼ ► However, if you as a developer have any app in your account that has amassed more than 2 million downloads, you cannot use private cloud compute at all.
00:16:39 ◼ ► It feels to me like they just haven't got their ducks in a row with exactly how to charge yet.
00:16:52 ◼ ► It's either that they don't have their ducks in a row for charging or that they are deathly afraid of how much compute they're going to need.
00:17:06 ◼ ► And that until they get a really better gauge on what that looks like, they're reluctant to open the gates wide to popular apps, basically.
00:17:18 ◼ ► The most logical thing here is to treat this like weather kit and say, you pay per use.
00:17:54 ◼ ► It would be kind of nice if developers of Apple only apps especially had access to that in a way that it's easier,
00:18:07 ◼ ► If you're a developer, you can kind of feel comfortable from the privacy perspective, right?
00:18:28 ◼ ► So, Jason, we are halfway through the Kickstarter campaign for our new podcast designed in California that will launch later this year.
00:18:37 ◼ ► So I wanted to just touch on a couple of bits and actually answer some frequently asked questions that we've been getting.
00:19:09 ◼ ► We're looking forward to showing off the artwork that we're working on, but that's still very much under work right now.
00:19:15 ◼ ► And then we mentioned this a little bit already, but if we hit $200,000, we're going to commit to two live events, one in San Francisco, one in London.
00:19:28 ◼ ► We're not exactly sure what they will be yet, but we will do something in both cities, which we're excited about.
00:19:34 ◼ ► I wanted to just reference the bonus episodes that we're putting out on the show so far.
00:19:40 ◼ ► People seem to really be enjoying them, so thank you so much to everybody that is listening to them.
00:19:46 ◼ ► The last episode, so the one that came out last week, was about Jobs and Wozniak traveling to the computer show PC76 and the beginnings of the breakdown of their relationship, which unbelievably happened before the Apple II even came to be.
00:20:00 ◼ ► And then in our next episode, the one that's coming out this Thursday, we're going to dig into exactly what the Apple II is, like functionally, how Woz put it together, and then how all of the other components that were needed to bring that computer to the market.
00:20:23 ◼ ► The episode three is the one where, and this is, I had somebody ask me, like, well, you know, you're doing all the writing.
00:20:39 ◼ ► I mean, you helped me get to a really interesting point at the end of episode three about how the Apple II is the first glimpse, I think, of what we've come to think of as Steve Jobs.
00:20:50 ◼ ► Like, oh, that's Steve Jobs doing Steve Jobs stuff right there with Apple's second product.
00:21:03 ◼ ► I see that you've been doing them so far, but I wasn't sure if it's just because they're episodes of Upgrade.
00:21:09 ◼ ► Yeah, we actually invisibly passed our video stretch goal, which was at 1.50, and we couldn't post it in time that we went over 1.50.
00:21:19 ◼ ► What you've been seeing in the Upgrade feed on YouTube, those are being produced in the same method that we produce Upgrade,
00:21:36 ◼ ► I would say it's not the finished Designed in California product, but as a video version on YouTube, as a sample, we're both okay with it.
00:22:10 ◼ ► The beauty of this is, because it's a Kickstarter, we can say we're going to do it, and then we've got until the show launches to figure out exactly how we're going to do it.
00:22:21 ◼ ► I have a proof of concept of episode two that we looked at this week that matches the audio edit perfectly, which is, that was the big question, is could we do that?
00:22:38 ◼ ► So now we're constructing a plan, and yes, I hope that part of that will actually be that it will be more so that the video version will have, like if we're referencing products or events and there are photos that will, which is not something we have time for with Upgrade, because Upgrade is all about being timely.
00:22:54 ◼ ► And I had somebody sort of snidely comment that if we really wanted our YouTube version of Upgrade to be better, it would have a bunch of annotations and photos and stuff.
00:23:06 ◼ ► Design in California, we're going to have a multi-week production process, and part of that process is going to be making the video.
00:23:19 ◼ ► But again, I will also emphasize we are treating it as an audio podcast, first and foremost, and that's how the edit starts.
00:23:32 ◼ ► I actually think that will enhance the audio version as well, because if we're going to do all that image research and license images and do whatever we need to do, we'll probably pop those in to the show notes or even to the chapter art for the audio podcast as well.
00:23:46 ◼ ► So that's all because we passed 150, a stretch goal we didn't even get time to announce.
00:23:55 ◼ ► This has been one of the great things, I'd say one of the hidden great things for us of doing this Kickstarter campaign is that, you know, really we did this because we wanted to see if there was an audience for the show.
00:24:13 ◼ ► And so as we've been raising more money, we're able to commit to more complicated things because we can.
00:24:22 ◼ ► So when the show launches, it's going to be so much better than if we would have launched it on its own.
00:24:32 ◼ ► It was like, this will allow us to do the show, but I would say at its most basic level.
00:24:44 ◼ ► And, you know, if it's a Getty Images image or something, you know, license the image so we're not ripping off images or, you know, like, I don't want to make those choices.
00:24:52 ◼ ► We can have the ability to say we're going to do this and we're going to do it right in a bunch of different ways that are, again, it's a little bit, it's almost invisibly making the show better, which we always would have done if we had been capable of it.
00:25:09 ◼ ► Rob asked, will it be possible to buy membership to Designed in California after the Kickstarter has ended?
00:25:47 ◼ ► But if you want to save, the Kickstarter is the way to save because you can get it for $60 for a year.
00:25:51 ◼ ► And Sriram asked, it would be awesome if you were to include the Origin of Apple episode of Upgrade in the Designed in California feed for continuity when it launches.
00:26:06 ◼ ► The plan from the beginning has been, we are going to launch a Designed in California feed, not Designed by California.
00:26:21 ◼ ► The plan always has been, we actually had Jim Metzendorf, our editor, hold on to the Applet 50 edits.
00:26:30 ◼ ► And we're going to generate a Designed in California, now I'm going to say Designed by California.
00:26:38 ◼ ► Jim's going to generate a Designed in California version of Applet 50 that will be Episode 0, basically.
00:26:44 ◼ ► And then the ones that we've been putting in the feed this month will be Episodes 1, 2, 3, 4.
00:26:51 ◼ ► So yeah, we're going to preload, before the episodes that you haven't heard launch, we will preload that feed with Applet 50 and the four parts of the Rise of the Apple 2.
00:27:02 ◼ ► So when the feed goes out, it'll already have that extra long episode and the four episodes that we've been putting in the Upgrade feed.
00:27:11 ◼ ► So there'll be a complete Designed in California RSS feed for members and for the general public with the free version.
00:27:23 ◼ ► We'll have that out there before we start dropping brand new episodes, which we, somebody asked me about that.
00:27:30 ◼ ► And I'm like, you know, I think we would like to start doing that maybe beginning of September, but we'll see.
00:27:39 ◼ ► Yeah, because we want to, obviously, we have to start the production process for the next set.
00:27:49 ◼ ► We want it to, we want it, you know, essentially we have to launch the show again as well.
00:27:53 ◼ ► So there is a benefit in us stopping talking about it for a little while because we have, because we then need to launch the show to everyone for free.
00:28:06 ◼ ► But at the moment, we're kind of estimating a like 10 week, probably production process for each episode.
00:28:26 ◼ ► But yeah, this is a, it's the opposite of upgrade, which is so we're trying to be as timely as possible.
00:28:33 ◼ ► And it means we have time to build a whole production process, which will be interesting.
00:28:37 ◼ ► It reminds me of magazines back in my magazine days, because really you're going to, we're moving cards across a Kanban board and they're going to be episodes in different stages of production.
00:28:48 ◼ ► And that's, that's, you know, we have, we have the luxury, which we generally, you and I never have of having a, building a production process that includes me writing a script and you giving notes and other people giving notes.
00:29:02 ◼ ► And then me finishing the script and then we do a record and then there's like the edit pass and then there's another edit pass.
00:29:12 ◼ ► Also keeping in mind that when we release, if we do a four episode block about something, the release date is not episode one and then a week later, episode two, and then a week later, episode three, because all of those will drop for members right at once.
00:29:33 ◼ ► It's actually, I'm really excited about it because this is a different kind of structure.
00:29:40 ◼ ► I mean, Cortex has a pretty chunky production timeline, but it's nowhere near what this will be.
00:29:50 ◼ ► But also like, you know, you're saying like a magazine, also that there will be a lot of movement too.
00:29:56 ◼ ► Like I had the thought the other day that, you know, maybe we would work on a one shot of Snow Leopard, right?
00:30:02 ◼ ► That we would want to put in, like, because Snow Leopard's on everybody's lips right now.
00:30:49 ◼ ► Building with AI is exciting until something breaks and then you have no idea why that happened.
00:31:34 ◼ ► I think that this is very clever from Sentry because I know so many developers are using these tools right now and it can provide them with added availability and ability to be able to get their work done.
00:31:47 ◼ ► But when you're letting something else into your code, it can introduce more issues, which is exactly not what you want.
00:31:55 ◼ ► So I think it's amazing that Sentry will allow you to pull these tools in and then also observe them alongside all of the other stuff that Sentry is giving you about bugs, about issues, about crashes, things that are slowing your apps down, which are making it not great for your users.
00:32:09 ◼ ► And everybody knows, like I know, you want your app to run well, smoothly, and first time for everybody that uses it.
00:32:18 ◼ ► Millions of developers use Sentry, including your favorite companies, AI and otherwise.
00:32:27 ◼ ► They have a free dev plan and listeners of this show can use the code UPGRADE26 to get $100 in Sentry credits if you're a new user.
00:32:51 ◼ ► We've both read and heard many more thoughts now and had conversations while in California.
00:33:00 ◼ ► So I wanted to kind of see where are you, what are you feeling right now about what we saw last week?
00:33:12 ◼ ► Because it is, I keep coming back to like, it's like sand running through your fingers.
00:33:17 ◼ ► That there's like the big picture issue, which is so big, in fact, sometimes that I think that because of all the drama about Apple intelligence, we've missed some of the big picture.
00:33:27 ◼ ► And then there's all the little things, which are a challenge that is the sand running through your fingers.
00:33:36 ◼ ► Adam Angst at Tidbits wrote a piece where he actually like tried to very intelligently catalog all the things that are in that big slide.
00:33:46 ◼ ► And it's like, and I really, when the public data comes out, I'll get a PDF from Apple PR of like, here's what the new features are.
00:33:54 ◼ ► And I'm wondering how, I do not envy the person who has to put that together because there's so many things that are little and they're all important or potentially very important.
00:34:07 ◼ ► Because it's just like, this is better and this is better and this is better and trying to find trends and all that.
00:34:11 ◼ ► The big picture that I keep coming back to, that I just want to remind people, like, we don't know.
00:34:20 ◼ ► But it's not just Apple makes good on their promise of two years ago, which I think is part of this.
00:34:39 ◼ ► It's just that they got a model that actually works and they replaced all of Siri with, you know, Mike Rockwell said in that post-kenote thing.
00:34:47 ◼ ► To me, that's the quote of the week was, we tore Siri down to the ground and rebuilt it a year ago because it just didn't do it good enough.
00:35:01 ◼ ► Which is, according to Mark Gurman, that, like, Rockwell thought that Siri was going to be a cornerstone of Vision OS and it couldn't be because it just was bad.
00:35:09 ◼ ► And that's the bigger picture that I wanted to say is, just step back for a second and say, how long has it been that we've all been talking about how Apple needs to do a new Siri because Siri is bad?
00:35:24 ◼ ► This isn't just about kind of like making good on the Apple intelligence promises from two years ago.
00:35:29 ◼ ► So it is literally, if they did it right, like they did it, everybody, they did a new Siri that is good, or at least certainly the bar is so low, certainly is viable.
00:35:45 ◼ ► Unto itself, that part of it is amazing if it's true, because we've been waiting for that a lot longer than two years, right?
00:35:51 ◼ ► That's been, you know, 10 years, eight years that we've been all sort of like looking at our watches and waiting for a new Siri to appear.
00:35:57 ◼ ► So I think that that's one of the things that stuck with me is, like, it really is not just Apple intelligence.
00:36:09 ◼ ► Yeah, I was thinking it was on the Vergecast, even David O'Neill said, if Apple two years ago would have said, this is our plan, we're going to rebuild Siri.
00:36:21 ◼ ► It's going to dig into your personal contacts, your messages, your emails, but it's going to take us two years to do it.
00:36:37 ◼ ► And it's like, it's going to, they had an idea, they're now executing on that idea, but it took two years of work to get to this point.
00:36:45 ◼ ► And talking about Mike Rockwell saying that somewhat gleefully, I mean, you could really make the argument that it was Apple's complete failure in 2024 that gave him the latitude to say what they've never, I think, been able to say before, which is always like, let's patch Siri, let's fix Siri, let's change this part of Siri.
00:37:05 ◼ ► Because they so abjectly failed two years ago, it let Rockwell say, tear it all down, you know, tear it all down.
00:37:32 ◼ ► And, you know, it, I'm reminded of like, when they redid the Bay Bridge, like they said, well, we can fix the Bay Bridge in six months if we shut down the Bay Bridge for six months, but we need the bridge to remain open.
00:37:45 ◼ ► Like, construction projects are way more complicated when you can't just knock down the old thing and build a new thing.
00:37:51 ◼ ► You've got to keep the new, the old thing there while you build the new thing next to it and then figure out a way to move them, switch from one to the other.
00:37:59 ◼ ► Well, Rockwell had the, the, the latitude because it got so bad to just knock down Siri.
00:38:05 ◼ ► I mean, I, I know it's still running, but it's, this is the metaphor is he didn't, he didn't have to sort of like make piecemeal improvements to the existing Siri.
00:38:21 ◼ ► Obviously they should have done it before, but it does seem like this was the final straw where he finally got the ability to say, let's just rip it apart and good for him.
00:38:29 ◼ ► One of the things that I thought was really interesting from the article that you wrote kind of when talking about Apple's AI ambitions is saying that you would like to see them have more.
00:38:45 ◼ ► I was, I was talking to Ina Fried from Axios as we were walking up to the Steve Jobs theater and she was saying how she's not like, both of us were agreeing that like out, the question is like, is Apple cut out for the AI pace?
00:39:02 ◼ ► And I, I just kept thinking about that, that I think their challenges that culturally Apple has for a long time been on this annual cycle.
00:39:18 ◼ ► Just like they got into a groove where like the iPhone comes out every fall, the OS versions need to come out for the iPhone because the new iPhone runs the new OS version because it's got new features in it that the new iPhone needs.
00:39:31 ◼ ► So they get on this annual pace because it used to be like Mac OS came out kind of randomly, right?
00:39:45 ◼ ► And, and, and even if you say, well, we're going to, we, we, we've got a year's worth of work that we're going to work on over the next year, but six months in, there's a trend in AI and you need to adjust, right?
00:39:56 ◼ ► And since Apple, since 2024 happened, has been, I think more focused on announcing features at DubDub that ship in the fall, ship by the end of the year, because they don't want to make that mistake again of over promising things that are too far out.
00:40:13 ◼ ► I think it actually gives them an opportunity to think about what's coming in the spring and adjust based on what they see in the AI world.
00:40:28 ◼ ► Because the judgments they made, I don't know, in January about what was going to be in 27 may not be judgments they would make in September about what was going to be in 27.3 or 27.4.
00:40:39 ◼ ► And so that's, that's why like, and you know, the 26 cycle added some stuff in the spring.
00:40:52 ◼ ► There are lots of things pointing toward the idea that Apple can spread their cycle out a lot more and have it be okay.
00:40:58 ◼ ► And I think that's good because I think Apple needs to have the flexibility to sense the trends in AI and adjust their strategy for the second half of the 27 cycle instead of setting it and then not touching it for a year, which is, I think, what Apple was very comfortable doing back in the 24 and before era.
00:41:27 ◼ ► I don't think Apple is going to work at the speed of AI companies, nor should they really, but there should be more than just once a year because they should be developing their technology more quickly than that.
00:41:45 ◼ ► Yeah. And maybe it's stuff that they're working on and they decide to prioritize something that otherwise was going to wait for the next cycle.
00:41:56 ◼ ► The idea, yeah, they're not going to be like super nimble and they don't need to be, but they need to have the ability.
00:42:03 ◼ ► Like I said, if there's a trend in AI and there's a project they're working on and they have a moment where they're like, actually, this is kind of becoming big.
00:42:18 ◼ ► But if I were them, I would think, okay, we can't get a lot more agentic stuff or more help for people who are running things like OpenClaw in 27.0.
00:42:32 ◼ ► But maybe there's a decision that's made in the spring sometime of this year that says that actually needs to be a higher priority.
00:42:46 ◼ ► I'm not talking about drop everything and we're going to ship a new feature in a month.
00:42:50 ◼ ► More like keep your eyes on the horizon, see what's going on, and don't make it a cycle where you don't have the ability to steer the ship a little bit mid-cycle in a way that maybe you didn't plan.
00:43:05 ◼ ► They don't need to be at the pace of AI, but they need to get used to making those decisions and phasing in features that were not, you know, there should be something in 27.4 that they didn't think was going to be in the 27 cycle, but that they moved it up because it felt like they needed to do that.
00:43:24 ◼ ► Because also they should be continually improving their models, and those should bring with them improvements on new features, capabilities that are now available.
00:43:35 ◼ ► And even something like making a big deal out of improving the Apple intelligence model in May, because now it can enable you to produce better images.
00:43:46 ◼ ► You can do more with your photos, you can do more with your photo edits, you can, you know, Genmoji is better.
00:43:51 ◼ ► Like, these things should be doable, and they can talk about those even if they want to hold big features for WWDC.
00:44:16 ◼ ► So they should continue to keep making it better all the time, which is one of the reasons why I thought that they would have an app in the App Store,
00:44:28 ◼ ► But it seems like, you were pointing this out, you're clearly right, it's embedded so deeply into the operating system.
00:45:24 ◼ ► Yeah, yeah, this is, and Snow Leopard, I actually got, I found Stephen Hackett in 2009 linked
00:45:38 ◼ ► So I actually asked, and it was in the CMS, and it just didn't, so I asked my editor at
00:45:58 ◼ ► But it was famously a, very much this release, in the sense that people, over time, started
00:46:21 ◼ ► I think I was thinking about this from the standpoint of how our reaction to the 27 announcements
00:46:36 ◼ ► And actually, I thought Joanna Stern, who was on the talk show on Tuesday night, the latest
00:46:57 ◼ ► And that, I think, that shows you what the difficulty is with doing a bug fixes and improvements update.
00:47:05 ◼ ► Is that it's hard to describe what they are because there's a bunch of little things and they're
00:47:13 ◼ ► And a barrage of, like, little things is super boring because they're all just these little
00:47:24 ◼ ► It's because, like, it was actually incredibly meaningful and will drive huge amounts of
00:47:36 ◼ ► And here, this is why I'm saying we are our own worst enemies because we look at the thing,
00:47:43 ◼ ► we get the thing we've been asking for, which is please slow down and just make things run
00:48:31 ◼ ► surprised about all of the, this is faster, this is faster, this is faster, so many things
00:48:36 ◼ ► And so it's like, well, did something happen at Apple that allowed things to go faster?
00:48:48 ◼ ► And the sense that I got was, even though they have things like the CPU scheduler that helps
00:48:52 ◼ ► a lot of things in a lot of places, the impression I got was that so many of these games were
00:49:06 ◼ ► And that was, you were, they were allowed to do it rather than, hey, you need to create
00:49:31 ◼ ► And this is, I do think having a few tentpole features allows everybody else time to look
00:49:52 ◼ ► And then also this is a year where we're going to, you know, make everything faster and better
00:50:04 ◼ ► is why companies are driven to do tentpole features and not prioritize things like performance
00:50:17 ◼ ► And, you know, it's a little bit of us taking our medicine, but like, it's going to be great,
00:50:24 ◼ ► It's going to, it's already the, the, the reports about developer beta one are very positive.
00:50:49 ◼ ► So stability, you should be, even in your least stable state, you should still be stable.
00:50:56 ◼ ► I had like two big picture thoughts that I wanted to share one, like when thinking about
00:51:08 ◼ ► Like, so, you know, in that the keynote was slim, they focused on what they wanted to focus on.
00:51:14 ◼ ► They could have spent more time talking about the platform improvements if they wanted to,
00:51:23 ◼ ► They could have done some demos of some stuff and like, you know, if they wanted to kind of, uh, make the keynote longer.
00:51:28 ◼ ► But I think they really wanted to just do what they wanted to do and keep it as, as like slim as possible.
00:51:35 ◼ ► The operating system quality is at a level that they really want to shout about, right?
00:51:51 ◼ ► And also the response to Siri AI has been very positive and they want people in the media to use it.
00:52:04 ◼ ► And like, in a way that I'm not sure that really gets across as much, but they seem very keen, very keen for people to try it.
00:52:14 ◼ ► This feels like a very different company to the last two WWDCs where 24, they were kind of on the back foot because it seemed like they had to get some things together quickly.
00:52:35 ◼ ► Like, look, I don't know if I said this last week, but like the, the thing I talk about, about how when you're riding high, there's no reason to change.
00:52:47 ◼ ► Like, if you're, if you're succeeding at everything and you think there's, and somebody comes to you and says, well, there's this one area where we're not doing as well as we could.
00:53:01 ◼ ► And once you have a moment where you hit a bump, where, you know, something happens that's bad, you finally get an opportunity to revisit and question and have open the conversation about what we could do differently.
00:53:14 ◼ ► And Apple has been on such a role as a business for so long that it feels like it really took them, you know, a punch in the gut, like 24 failing.
00:53:47 ◼ ► So the screen time stuff and the parental control stuff has apparently been broken for a long time.
00:53:57 ◼ ► Basically, since it shipped, you know, I would often get like 24 hour screen time days because a web page on a Mac, for some reason, was counting as 24 hours in complete use.
00:54:08 ◼ ► And then also, you know, I said this to you, but somebody wrote in like a week before the keynote and was like, if you speak to anyone at Apple this next week, please ask them to fix parental controls because they're so broken.
00:54:20 ◼ ► And it seems like, you know, people that are using this stuff, they've been having issues with it.
00:54:26 ◼ ► So in their year of wanting to improve OS stuff, I would assume they realized, similarly to Siri, the only way to fix the bugs in these features was to start over again with the underlying architecture, which turned into a big job.
00:54:47 ◼ ► So then they just kind of, at that point, was like, you know what, let's go all the way, right?
00:55:01 ◼ ► To combine also the government issues around this and realizing, I wonder if they had a moment where they realized governments were going to ask things and screen time was totally broken.
00:55:19 ◼ ► Yeah, you could argue that this portion of the keynote was very much just like the Apple intelligence portion, where it was, we actually have to make good on a promise that we kind of failed to deliver on.
00:55:31 ◼ ► But Apple's never going to say, we're sorry, we know screen time has been busted and we fixed it.
00:55:36 ◼ ► They're going to say, hey, we have a new focus and some great new features in screen time that we're even calling, not even mentioning the word screen time because it's so embarrassing, which I thought was also kind of funny.
00:55:47 ◼ ► Oh, this is a little tangential, but I wanted to mention, I talked to somebody who used to work with John Gianandrea.
00:55:54 ◼ ► And I just wanted to say that I got confirmation from them that our theory that he was always a more academic, big thinking kind of guy who was about research and development.
00:56:11 ◼ ► And that he was the wrong person to be in a suddenly having to rapidly iterate and ship products to customers role.
00:56:30 ◼ ► That he was fundamentally miscast in the role of shipping products to customers at Apple, which Apple suddenly found itself in.
00:56:44 ◼ ► But I talked to somebody who used to work with JG and he was like, oh, yeah, that's that.
00:57:02 ◼ ► And build some features that are not the features that you like or are interested in building.
00:57:11 ◼ ► Anyway, there's a new head of AI and ML now who also came from Google and worked on Gemini.
00:57:15 ◼ ► And I imagine that that guy is more, I forget his name, but he was on stage at the Tech Talk.
00:57:43 ◼ ► Speaking of which, actually, I wanted to just point people to something Federico wrote just because it's so beautiful.
00:57:52 ◼ ► Me and you both called it an essay in the links that we – because it was much more of an essay on Mac Stories called Between Seasons.
00:58:09 ◼ ► Yeah, he was reflecting on his 10 years covering WWDC, which I thought was, like, that's very impressive.
00:58:29 ◼ ► There's also – I want to talk about – I want to get into the details and talk about some of our experiences.
00:58:35 ◼ ► But I do want to note there is some drama brewing about the Apple, the EU, and the DMA all over again.
00:58:41 ◼ ► I want to try and get into that maybe next week because just in reading this morning, it's like, oh, there's like 17 different points of view on this one.
00:58:50 ◼ ► And so I'm hoping that maybe if we give it a few more days, maybe it will become a bit clearer.
00:58:59 ◼ ► So hopefully next week, some point in the next couple of weeks, we'll get into this a bit more.
00:59:12 ◼ ► The best customer experiences happen when customers feel like they're the only one, almost like they have their own concierge.
00:59:19 ◼ ► But once more and more requests start to come in, a lot of companies turn to AI, but then they lose that personal touch.
00:59:30 ◼ ► Decagon helps companies create personalized concierge-style experiences with AI agents across chat, email, voice, and SMS.
00:59:39 ◼ ► They're available 24-7, they feel natural to talk to, and can resolve customer requests on their own
00:59:54 ◼ ► And because workflows can be updated using natural language, teams can make changes themselves about long engineering cycles.
01:00:01 ◼ ► This means less time spent managing workflows and more time focused on customers and ROI in weeks.
01:00:09 ◼ ► Once live, agents can pull up-to-date information, take actions on requests, and understand the context behind any conversation.
01:00:16 ◼ ► Plus, Decagon gives your team full visibility into why agents make decisions and what's happening across every conversation
01:00:27 ◼ ► Decagon helps power millions of conversations every single day for brands you know and love,
01:00:39 ◼ ► Go to decagon.ai slash upgrade to get a personalized demo and see what Decagon can do for your team.
01:01:11 ◼ ► Oh, I am currently running, I have a, I have a loaner MacBook Pro that is running Golden Gate.
01:01:35 ◼ ► However, hearing all the positive response, you know, I'm starting to already get the, you know, get the itch to use this with real hardware that I actually rely on.
01:01:47 ◼ ► I had the realization the other day that I am currently not, I've been planning on doing something with it, but I haven't put those plans into effect yet.
01:01:57 ◼ ► I still have my M1 Mac Studio just sitting unplugged, and I thought, that may give me the strength to update to Golden Gate on my Mac.
01:02:09 ◼ ► Because if I need to record podcasts, and something is bad in a beta, I can always, I can use that for the podcast side of it while living with the beta.
01:02:24 ◼ ► I spent time doing some Siri stuff on that iPhone, because the iPhone has all my data on it.
01:02:30 ◼ ► It's a migrated version of my other phone from last fall, so it is not up to date like my main phone, but it's got most of my data on it, so it was usable.
01:03:12 ◼ ► I usually wait to about 3, but I was hearing people over the few days, you know, like, talking about the beta, feeling pretty strong, and I realized I really wanted to try out Siri AI, and realized, you know, if I'm going to do this, it has to be on a device that I'm using every day that's got all my stuff on it.
01:03:35 ◼ ► And I was thinking, well, I could put it on my iPad, but I don't have my iPad with me all the time, and I want to get into the habit of thinking of and asking questions of it.
01:03:59 ◼ ► My phone's been getting hot, but it's also doing that indexing, which apparently I've heard can last, like, weeks, depending on how much stuff you have on your phone, and maybe in your iCloud.
01:04:20 ◼ ► I don't know if I needed to do any of these things, but I did them, like setting my region to America and doing that kind of stuff.
01:04:27 ◼ ► I think it kind of helped kick it all into gear, so don't ask for a tutorial because I don't know how to do it.
01:04:32 ◼ ► But I've been spending a lot of time using Siri AI, and this has been for different types of things.
01:04:40 ◼ ► So I've just been asking some questions just to see what the answers would be like, right?
01:04:55 ◼ ► They weren't the most recent emails, but they were the ones that it deemed most important.
01:05:05 ◼ ► And I said, okay, tell me more about that contract and compare it to the last three deals.
01:05:12 ◼ ► And it went through, found the last three contracts over the last year, put them in a table, and then gave me a summary about how the budgets have changed over that period of time.
01:05:34 ◼ ► Like, earlier today, I asked it to find me a message that I knew had been sent a couple of years ago, but couldn't exactly remember the details of it.
01:05:52 ◼ ► So, like, I'm waiting until that's done before I say, like, you can really, like, say, like, one way or another how good it is at mining that personal context.
01:06:03 ◼ ► Like, another example of a thing, you know, I know that Adena sends me shopping lists sometimes.
01:06:30 ◼ ► So it's like, you go take the day, but just let me know some things that you want me to do.
01:06:42 ◼ ► And so it's like, if you have things you want me to do at certain times in a day, you let me know what they are.
01:06:49 ◼ ► And I just said to Siri AI, grab the three most recent iMessages from Adena, add them to an Apple note for me.
01:08:12 ◼ ► that I didn't have the last time that I looked into the network attached storage world,
01:09:16 ◼ ► One of the key reasons for this is like, at first, I was like, oh, I might just get like
01:09:21 ◼ ► a RAID or just some kind of driving closure, but I need to be able to share files with other
01:10:10 ◼ ► Like I gave it, I've been doing this thing where I've been copying requests that I've given
01:10:51 ◼ ► And one of the reasons that we said like, don't is because they've built brand awareness, right?
01:11:08 ◼ ► And I think that this name kind of gives that off, especially because in my experience so
01:11:26 ◼ ► I am increasingly in, I'm going to put a link in the show notes to Stephen Robles' video.
01:11:56 ◼ ► 26.6 isn't out yet, but word on the street is the 26.6 starts the indexing process in the
01:12:15 ◼ ► slowly do the initial spotlight index for 27 when you don't have 27 so that when you get
01:12:32 ◼ ► And it's a very intense re-indexing and they have to start from scratch because it's a new spotlight
01:12:37 ◼ ► And so it sounds like they're priming the indexer in 26.6, which is super, super clever.
01:12:52 ◼ ► I do this silly podcast with David Lohr and Philip Michaels where we watch old episodes
01:13:07 ◼ ► And what we do is we record two at a time and at the end of our recording, we have to decide
01:13:15 ◼ ► And we have a little group chat and it goes way back and we, you know, throw other silly
01:13:31 ◼ ► I switch over to the Magnum podcast, uh, text chat and scroll back until I can find the episodes
01:13:38 ◼ ► And I said, Siri AI, what are the next two episodes of Magnum podcast that we're doing?
01:13:55 ◼ ► Like it, it knew enough to do the right search, to find the right information, to give me the
01:14:00 ◼ ► answer without me having to go and do a thing I could do, but like go to messages, find that
01:14:06 ◼ ► thread, scroll and look at all the text in the thread until I found the one where I said,
01:14:21 ◼ ► Like is fantastical going to supply its calendars to spotlight or am I going to have to add all
01:14:30 ◼ ► Is MimeStream gonna, is there a way for MimeStream to even offer search stuff in spotlight or is, can it not?
01:14:40 ◼ ► And if it can't, does that mean I also just need to run mail in parallel so that, that, uh, Siri AI knows about my email?
01:14:51 ◼ ► These are things that we're going to have to figure out this summer, but, um, but it, it, right now it looks like this feature is useful enough that if I have to build parallel workflows just for spotlight, to feed spotlight information, I will, if the benefits are like this.
01:15:06 ◼ ► One of the features that is only coming to the new phones, because, so the, the newest phones, the 12 gigabyte of Ram phones, they get access to a, I think a bigger model, right?
01:15:21 ◼ ► Um, I think it's been a bit confusing, but I know one of the features that they're, that the newer phones are getting is a new dictation system.
01:15:53 ◼ ► Like if I, if I tried to dictate a message, I talk very carefully, you know, like it kind of slowed down and enunciate,
01:16:01 ◼ ► but I'm able to speak way more fluidly and it sent, and it's doing a much, much better job.
01:16:22 ◼ ► You know, like I've done a few of the extending things and it just adds in stuff, obviously, that's not there.
01:16:30 ◼ ► So I took a picture of me at Apple Park and I extended it and it put me in shorts and I wasn't wearing shorts.
01:16:39 ◼ ► I took a picture of Adina and it cut her arm off and I extended it and it put her arm in and gave her a round analog watch, which is not a thing that she owns.
01:17:24 ◼ ► What you said, like if you're extending somebody's pants and it turns out they're wearing shorts, even though they weren't.
01:17:38 ◼ ► But you can't rely on that as the memory of what actually happened because it didn't actually happen.
01:18:00 ◼ ► Or I do a lot of stuff where it's I needed a little bit wider to put it on six colors and it's just not quite wide enough.
01:18:08 ◼ ► Or the example Apple gave is when you rotate a photo, you have to crop it because there's a portion.
01:18:14 ◼ ► There's a little wedge that has been revealed in the rotation, top and bottom, that don't have anything in them.
01:18:26 ◼ ► And you can generatively fill those little wedges now and keep your photo the same size.
01:18:43 ◼ ► Like, you know, this is one of those things, just because we give you these tools, you still need to think about how you want to use them and what you get out of it.
01:19:05 ◼ ► I don't know if we should say who it was because he was talking to us privately, but a photographer who pointed out that the challenge with the spatial reframe is when you're using the spatial photos in Vision Pro, you know, there's a base photo that's real and then you're just kind of moving your head.
01:19:23 ◼ ► But with spatial reframe, you're moving your head essentially to do the low resolution sort of like set.
01:19:30 ◼ ► And then you press the button and it goes out to private cloud compute and renders the image for you.
01:19:36 ◼ ► And as a result, what you're getting back is based on the original photo, but it's not quite the original photo.
01:19:49 ◼ ► Like it's okay as a kind of like a cheap parallax effect, but when you're actually doing a full render, it can look uncanny because you're taking a, also you're taking a fairly rudimentary depth map.
01:20:02 ◼ ► That's been created by machine learning on your device, which is how they do the spatial stuff.
01:20:07 ◼ ► And it's very impressive as a trick, but like that depth map is not the contours of that face.
01:20:18 ◼ ► And it's again, it's using the pixels that were there, but it has to generate new ones in order to fill it in.
01:20:39 ◼ ► And so like, you know, you, you gave some examples of some great utility things, right?
01:20:45 ◼ ► So you, you know, you extend it up and a tree appears that wasn't there or the top part of an airplane.
01:20:50 ◼ ► You know, you're, you're on an airplane and now like the roof of the airplane is there.
01:21:07 ◼ ► But anytime that, that important things, i.e. people, pets, all this kind of stuff are any way involved in this, you've broken the memory of it.
01:21:21 ◼ ► And it's like, so if I do one of these things and then save this effect and then look back in two years time, now I've altered the memory of that situation.
01:21:49 ◼ ► So I sent a message today to someone and it recommended, like, do you want to make a Genmoji of this phrase?
01:22:16 ◼ ► So that is a much, that is a better system, I think, than what they were doing before, where it would only take a couple of words or a couple of emoji or whatever.
01:22:45 ◼ ► Like, I gave it an image of me and you at WWDC and it said, why don't you ask for pixel art?
01:22:59 ◼ ► And in the clothing that we were wearing, but it gave us emoji heads, which is an odd choice, but that's what it did.
01:23:25 ◼ ► I wanted to mention, because we got some feedback about this, our dislike of image playgrounds.
01:23:30 ◼ ► And, you know, somebody called us Luddites for not, like, just for criticizing image playgrounds.
01:23:39 ◼ ► And I just wanted to say, like, somebody sent me an image that they used and they said they'd never used image playgrounds before and they used it on the beta.
01:23:46 ◼ ► And it was, like, line art for, like, an invitation for a home backyard barbecue with, like, a pig and, like, text in it.
01:23:56 ◼ ► A thing that, you know, is just, I need some simple imagery here and it's just kind of text in a shape and all that.
01:24:08 ◼ ► My biggest issue is that Apple is even using as examples the idea that you're taking a friend's image without consent and posing them somewhere.
01:24:21 ◼ ► I don't think to completely glide over it and be like, yeah, here's your friend and you're going to put him in a chef's outfit and put him baking a cake.
01:24:30 ◼ ► Like, if a friend of mine did that and then sent that to me, I'd be like, that's really kind of gross.
01:24:37 ◼ ► And reasonable people can differ, but I don't like the idea that, like, you're just making images of me and then sending them around and it's not me.
01:24:45 ◼ ► There are also lots of other issues involving, like, training and are they knocking off artists' styles so that you can replace the artists and not use the artists and not pay the artists, but we've stolen all of their work and used it to train.
01:25:03 ◼ ► And so I would just say, you know, officially for the record, we don't like it, but there are aspects of it that I think are actually fine and there are aspects of it that I think are kind of gross.
01:25:14 ◼ ► And there's plenty of room for debate, disagreement, but what I would say is certain is that it isn't just a settled issue that this is fine and that anybody who has issues with Image Playground is just fighting against the tide and a Luddite
01:25:36 ◼ ► I don't know what world that person is in, but in the world of people that I'm around, including a lot of creative people, it is most definitely not a settled issue.
01:25:50 ◼ ► It's like using somebody else's image, even if they're your friend, like, it doesn't matter.
01:26:11 ◼ ► I think that these are better, though, where you're creating these, like, things that aren't supposed to look like a person if you're going to do it.
01:26:19 ◼ ► But I still stand by the, like, you shouldn't be picking people from images and creating, like, brand new things.
01:26:28 ◼ ► So, like, one of the things in Genmoji now, the thing that I just did is I gave it an image that I had of you, and it made a version of that image.
01:26:36 ◼ ► I don't like that, but what I really don't like is what you were talking about, which is we're going to take someone's face, and you're going to create them in an environment that they were never in.
01:27:01 ◼ ► And, like, I know why we do it, and we did it two years ago, and I know it's going to continue to happen, but I'm just going to be honest.
01:27:17 ◼ ► But, like, somebody sends me a photorealistic generated image of me, and I am repelled by it, right?
01:27:27 ◼ ► But, like, doesn't that say that maybe this isn't right if, like, the response people are having to the image tools that you're giving them is that they feel uncomfortable by it?
01:27:41 ◼ ► I don't like, I now like it even less because they've decided photorealism is where they want to go, you know?
01:28:10 ◼ ► Look, and I can see the utility of it, and, like, I get it, and people are going to do it, but I don't love it, and I don't love that Apple enables it and makes it very easy and demos it and doesn't seem to even give consideration to the fact that taking your friend's image and putting him in different clothes doing a thing and then mailing it to all of your friends might be problematic in some way.
01:28:36 ◼ ► Like, the example that they gave in the keynote, right, I think, or an example that was given was, like, here, you know, it's Frank's birthday party and Frank likes to bake.
01:29:02 ◼ ► I only ever use image playgrounds to make fun of my friends directly to them, and I just don't think that is a good use of software.
01:29:34 ◼ ► If you copy a screenshot to your clipboard and go to iMessage, a button pops up on the top of the keyboard that says paste from screenshots.
01:29:44 ◼ ► So, because that's what you're usually doing, at least for me, I'm usually taking a screenshot and sending it to someone, and it knows that.
01:29:58 ◼ ► For some reason that I will never understand, every few weeks, my ringtone level gets set to zero, and so my alarm doesn't make any noise.
01:30:14 ◼ ► I don't know what is going on in there, but now you can set these things independently, which I think is awesome.
01:30:18 ◼ ► And in 26, they made the Apple Pay card selector, when paying online, significantly worse.
01:30:28 ◼ ► And they kind of, there used to be a button that you could tap to change cards, and you can still do that, but they moved where the button went, and it stopped making any sense.
01:30:40 ◼ ► It's a big image of the card that you're using, and you have a big button that you can tap.
01:30:48 ◼ ► It shows you a visual of all of the cards that you have in your Apple Pay, and you can select any of them.
01:30:54 ◼ ► And when you first open Apple Pay, it shows you the card that you selected, but then has the other ones kind of stacked behind it, like CoverFlow, and they disappear.
01:31:03 ◼ ► So you can immediately know, I can change my card, and then it's incredibly clear for which card you're going to use.
01:31:11 ◼ ► That was a big annoyance of mine that they fixed, and they've actually made the UI so much better.
01:31:20 ◼ ► Things that didn't feel right, or were buggy, or whatever, have been fixed, and been made better.
01:31:44 ◼ ► The sad truth is that this data can be crawled through, collected, and aggregated by third parties.
01:31:52 ◼ ► One of the easiest ways for data brokers to track you is through your device's unique IP address, which also reveals information about your location.
01:32:03 ◼ ► This makes it much more difficult for you to be monitored, tracked, and monetized by your private online activity from data brokers.
01:32:16 ◼ ► First, it's at its lowest price ever, planned starting at just $3.49 a month, which is just $12 a day.
01:32:23 ◼ ► It encrypts 100% of your network traffic to keep your data safe from hackers when you're on public Wi-Fi.
01:32:39 ◼ ► So I just returned from California, and I was able to use and was using ExpressVPN whenever I was connected to public Wi-Fi,
01:32:50 ◼ ► And then I can also reroute my traffic to continue watching things on the streaming services that I'm using at home in the ways that I want to.
01:33:04 ◼ ► That's expressvpn.com slash upgrade to find out how you can get up to four extra months.
01:33:26 ◼ ► Apple has announced that the new, more advanced on-device models for Siri AI for more expressive voices and improved transcription require devices running the most recent silicon with 12 gigabytes of RAM.
01:33:40 ◼ ► It would be disappointing if the base iPhone 18 and maybe iPad mini revision would be locked out of these features.
01:33:47 ◼ ► Do you think this likely means that this year's A20-based processor will also meet the requirements for these features?
01:34:07 ◼ ► And then, and well, and yeah, it's like in these features exist, but they're not as good.
01:34:20 ◼ ► It's, it's that we are entering a world where we're going to have multiple models and you will get a better model if you've got a more powerful processor.
01:34:30 ◼ ► And this will go into the future where they will have new models that are cutting edge that don't come to older devices.
01:34:50 ◼ ► So I think Apple has a motivation to make new iPhones get these features, but I would also say, I don't feel like it's Apple intelligence where they feel like they have to put it in every single product.
01:35:09 ◼ ► But the bottom line is like, I think the chip design precedes the existence of these features, right?
01:35:15 ◼ ► And it's not, it is not like Apple intelligence version one, where it was like, you just get nothing if you don't have the base chip level is there are just certain features that aren't the big ones.
01:35:29 ◼ ► On the cheaper devices, older devices, you just don't get the, the as good to transcription as good of a voice on whatever else features they might use in the future with those models that, that, so I think there's, I think it get, again, gives them some pressure to try and have more devices work with the latest and greatest cool models because it kind of, it raises the bar for new models.
01:35:55 ◼ ► But I don't think it's like a, all hands on deck kind of thing like it was with Apple intelligence.
01:36:01 ◼ ► And I think that there will be new products, there will very likely be new products that come out that just don't support this, right?
01:36:26 ◼ ► I, but I, so I think there's a motivation, but I don't think it's enough for them to say future, you know, every product after we make this announcement must support this newer on device model.
01:36:35 ◼ ► I don't think that's practical, but I do think that there's a little bit of pressure like to, to make it work.
01:36:40 ◼ ► If they can, for as many new products as they can, but if they can't, I think they're going to be okay with it because you're not, not getting that feature.
01:36:51 ◼ ► Simon says, one of the reported challenges of AI-ing Siri was to add the open world LLM functionality without breaking the more prosaic Siri turn off the kitchen lamp component.
01:37:11 ◼ ► You've provided in our, in our document, they, they put this on stage at the tech talk.
01:37:17 ◼ ► Yeah. So they, they showed us essentially a flow chart of what happens when a request is made and like how it can go down the stack.
01:37:25 ◼ ► And at the very top is something they called system experience, which has inside of it at a box called assistant experience.
01:37:34 ◼ ► And what, what I think they're getting at is when you ask for a thing to the system, it is then starting to decide the, where is this being kicked onto?
01:37:44 ◼ ► And it feels like from what I can remember and what I worked out from this image, essentially, if you request something like that, the system is going to say, Oh, I can just handle this and do it because it's a basic on device thing.
01:37:57 ◼ ► And if you can't, then you start going out to what they called later, the system orchestrator, which then starts to work out how complex of a thing is this?
01:38:12 ◼ ► And then do I need to kick it out to the on device models or actually does it need to go to the cloud?
01:38:17 ◼ ► So there's like a bunch of things going on here and it feels like the very, very top level is can I just do this thing?
01:38:34 ◼ ► I think it's, I think it is like volumes and timers and a bunch of stuff that's just very simple local requests that need to just happen immediately.
01:38:50 ◼ ► Paul asked, did Apple make a mistake to include an AI subscription within the existing iCloud subscription?
01:39:11 ◼ ► I think iCloud Plus is a great example of Apple gating things behind, do you give us money already?
01:39:18 ◼ ► So that there's the money, the features you get, the functionality you get from buying our product.
01:39:32 ◼ ► That they need to work their business model so that using an iPhone is something you do, you can do freely.
01:39:48 ◼ ► And then you want more times of doing image generation or whatever, then iCloud Plus is a nice way to just kind of throw it in the box.
01:40:03 ◼ ► It doesn't preclude them from saying, if you want to buy more, although at this point, they don't really want to even make that a product.
01:40:11 ◼ ► I think that this is their system-level way of selling kind of enhanced access to existing customers who are obviously generating more revenue for them.
01:40:21 ◼ ► Which means that it pencils out that the added cost of AI queries will basically be covered.
01:41:01 ◼ ► There should be a specific plan or subscription available for something that is showing to be a core operating system feature.
01:41:17 ◼ ► But I do understand that for features that are above and beyond and have incremental cost, that having it be that way.
01:41:30 ◼ ► If you pay 99 cents a month for a little more storage, I think you're iCloud Plus at that point, right?
01:41:46 ◼ ► I wouldn't want them to go down the path of, here are a bunch of features that are in our operating system, but you don't get them until you pay us.
01:42:11 ◼ ► So if you pay them for 99 cents a month for 50 gigs of storage, you are iCloud Plus or anything above that.
01:42:18 ◼ ► If you've got more iCloud storage, if you're in the bundles, you've got that too, then you're in.
01:42:27 ◼ ► Like, basically, it's do you pay Apple for some subscription that includes iCloud in some way?
01:42:35 ◼ ► And you get stuff like private relay, custom email addresses, hide my email, and more storage and stuff like that.
01:42:51 ◼ ► But they seem to have just gated this as iCloud Plus, which means that, yes, you could pay $1 a month and you would be there.
01:43:08 ◼ ► Ian says, you've mentioned a few times in prior episodes to look for canaries in the coal mine at WWDC for features that foreshadow future hardware launches.
01:43:18 ◼ ► There's been a lot of talk about how the ability to change the simulator widths of precursor to the folding iPhone.
01:43:24 ◼ ► But there has been little talk, as far as I've seen, about pull-to-refresh being a precursor to a touch MacBook Pro.
01:43:34 ◼ ► But there is also support for drawing and free-from on the Mac that they say use the trackpad for.
01:43:43 ◼ ► The real canary in the coal mine here is sidecar is a feature that lets you use an iPad as a Mac monitor and they had like pencil support in it and stuff so that you could do things like draw in Photoshop and it was like you were using a graphics tablet.
01:43:57 ◼ ► Now they are adding a whole set of APIs where you can add touch support for sidecar to apps, to like AppKit apps, to like Mac apps.
01:44:16 ◼ ► That's like really the message that they can't say but that everybody knows is use sidecar as a way to experiment with how touch interactions will work on your Mac when we do a touchscreen MacBook.
01:44:37 ◼ ► And then like I mentioned earlier in this episode, the 4K cameras thing feels very much like they're going to have a security camera or a doorbell camera.
01:44:54 ◼ ► If you would like to send us in a question for a future episode of the show, please go to UpgradeFeedback.com.
01:44:59 ◼ ► You can send in your AskUpgrade, your SnowTalk, your feedback, your follow-up, all by going to UpgradeFeedback.com.