PodSearch

ATP

706: No One Is Getting a Good Deal Tonight

 

00:00:00   So I don't want to spoil the uh main chunk of this episode, but

00:00:03   John, you're welcome.

00:00:04   Four seconds is up.

00:00:06   I said I had a four-second pre-show uh before we started recording.

00:00:11   Brutal.

00:00:11   I knew I knew better than to say that to John Syracuse, but here

00:00:14   we are.

00:00:14   Um you're welcome, uh not only for taking longer than four seconds, but also

00:00:19   uh yesterday I wore my Mac Pro Believe shirt, which I have had since

00:00:23   it was new, but have only worn a handful of times.

00:00:26   You didn't wear it enough.

00:00:27   Apparently not enough.

00:00:29   Your fault.

00:00:30   But see, here's the thing.

00:00:31   Here it is.

00:00:31   I think I'm giving you the gift of new computers, and in reality, I'm

00:00:34   I'm getting flack because I didn't give you the gift of a new Mac

00:00:37   Pro.

00:00:37   I see how it is.

00:00:38   Well, t-shirts can last a long time if they're lightly worn, so you know

00:00:41   it could rise again in a decade or something.

00:00:44   Marco, what do you have going on?

00:00:46   I have um a much more victorious situation, and and nobody will care about

00:00:51   this unless it on the off chance you out there, somebody out there, might

00:00:56   care about this.

00:00:57   Wait, wasn't that the uh John Syracuse biography?

00:01:00   Nobody cares about this but me, but I care about it a lot or

00:01:02   something like that.

00:01:03   Kind of sort of close.

00:01:04   I forget anyway.

00:01:05   I mean what can top installing a ceiling fan though?

00:01:07   So let's see.

00:01:08   Same room, same category of general home improvements and and life improvements.

00:01:14   Okay.

00:01:15   I have in my office here, I have casement windows.

00:01:18   Generally, I cannot say I enjoy casement windows.

00:01:22   Now remind me, I I've heard this term, but what does that mean?

00:01:24   It's the kind that rather than being able to open and close up and

00:01:27   down, they have a little crank and they swing out.

00:01:29   Ah, yes.

00:01:30   They open like a book.

00:01:30   Yep, yep, yep.

00:01:31   Exactly.

00:01:32   Thank you.

00:01:32   Um, and there's I know people like these, I don't like these.

00:01:36   Um, but some of the many reasons that casement windows are, you know, a

00:01:40   weird trade-off is like, how do you attach a screen to a casement window?

00:01:45   And they all have these weird little like clips and levers and crap, and

00:01:49   I've never found one that was attached in a good way.

00:01:52   The ones here have these little metal clips that hold them on.

00:01:57   We have to, as part of the warranty for the windows, we have to

00:02:00   get them cleaned every so often, and we have to get them serviced every

00:01:59   so often.

00:02:04   And these screens are taken off multiple times per year for w one of

00:02:09   those two tasks.

00:02:10   The problem is, these little tiny stupid metal clips, first of all, good luck

00:02:16   ever putting one, ever putting a screen back yourself.

00:02:18   And even the window service people don't really do a very good job of

00:02:22   it.

00:02:23   Um secondly, these little clips get lost all the time because they fall out

00:02:28   whenever you try to pop the screen, you know, back in.

00:02:31   Like if you like jostle the clip a little bit, it'll fall off the

00:02:33   window and fall to its doom somewhere and get buried in sand.

00:02:36   And so you never find them.

00:02:37   And every single time I get I have the window servers, I'm like, please,

00:02:39   can you bring like 50 of these?

00:02:41   I'll buy them.

00:02:42   Just please.

00:02:43   We there's never enough clips to hold the screens on, and as a result,

00:02:46   the screens in my office are usually like crooked, or there's like a small

00:02:50   gap somewhere in like between like the screen ridge and the window, such that

00:02:55   mosquitoes can come in.

00:02:56   So I'm just like, okay, this please, for the love of God, I just

00:02:59   need these to be good and to be effective and to be able to

00:03:02   hold my screens against my windows.

00:03:05   I decided to stop waiting.

00:03:07   Stop waiting for the window guys to bring more.

00:03:09   They never do, or they'll bring one, even though I'll say, like, please bring

00:03:12   50.

00:03:13   I'll pay for them.

00:03:14   Like, they're just little metal clips.

00:03:15   These are the most scarce items in my house.

00:03:20   I can now 3D print them.

00:03:23   I took it upon myself a few days ago to replicate the shape of

00:03:27   one of these clips using some uh some semi-rigid TPU, the the bamboo TPU

00:03:33   for AMS, so it is very easy to print.

00:03:35   It is a little bit rigid, but it's a little bit flexible and very

00:03:37   strong.

00:03:38   And so that means I can now have as many of these as I

00:03:42   want.

00:03:42   I printed 30 of them earlier tonight, and my screens are being held up

00:03:44   perfectly, and I have tons of them, and they're a little bit flexible, so

00:03:50   that when I'm attaching the screens, I don't cut my fingers, and they're actually

00:03:54   it's actually like a reasonably pleasant experience.

00:03:57   And instead of having three of them on one window, because I have to

00:04:00   conserve them.

00:04:00   I have eight of them on this one.

00:04:02   It's amazing.

00:04:04   Again, people wonder like, how do I what do I use my 3D printer

00:04:07   for?

00:04:07   It's little crap like this.

00:04:09   Like we have uh we have a gate, and sometimes the wind will blow

00:04:13   the sliding piece so that it kind of blows it half open, and then

00:04:16   deer can get can get in and eat all of our plants.

00:04:19   Um and so I made a little clip out of again a 3D little

00:04:24   custom 3D printed thing, and I just made a little clip that holds the

00:04:28   two pieces of the gate together.

00:04:30   So the wind does not blow them open.

00:04:32   That's the kind of thing I love having little solutions to little problems like

00:04:37   this.

00:04:37   It is so incredibly satisfying.

00:04:39   So that that's my victory lap.

00:04:40   I know John has a bit of a different victory lap, maybe coming up

00:04:43   later, but well, if you can call it victory.

00:04:45   Yeah, right.

00:04:46   Maybe it might be a complex emotion, but um, but at least this for

00:04:52   almost no money, I solved a very big problem.

00:04:55   That's incredible.

00:04:56   That's what that's super cool.

00:04:57   Uh I'm a little surprised that plastic does the trick.

00:05:01   Like for some reason I feel like this should be metal, but I'm not

00:05:03   trying to argue with you.

00:05:04   If it works, it works.

00:05:05   So that's that's really great.

00:05:07   So it's and there's there's obviously like different um forgive me, I'm gonna get

00:05:10   briefly very nerdy about 3D printing, but I swear it's gonna be brief.

00:05:13   Is it gonna be four seconds?

00:05:15   No.

00:05:15   Long enough.

00:05:16   There's a lot of different properties that you can optimize for with both the

00:05:18   material selection and the design of a print.

00:05:21   So like even just things like what direction on the plate that you print,

00:05:25   like these prints standing up because the layer adhesion between layers is usually the

00:05:29   weakest bond in a 3D printed part.

00:05:32   And and also I'm by using a TPU material, I also have very strong

00:05:37   layer adhesion.

00:05:38   And by it being flexible, it makes it actually pretty difficult to, you know,

00:05:42   destroy or crack or you know otherwise break it.

00:05:45   I will say for those of you out there who who were 3D printing

00:05:48   people, don't sleep on the uh the bamboo TPU for AMS filament.

00:05:52   If you go into it expecting it to be very, very flexible, like TPU,

00:05:56   you're gonna be disappointed, but it's incredibly easy to work with.

00:05:59   It is, as the name suggests, compatible with the AMS, so it's very easy.

00:06:02   And it is surprisingly useful.

00:06:05   I it is one of my most used filaments because if something can be

00:06:08   a little bit flexible, you can make it almost indestructible if you make it

00:06:12   out of this stuff.

00:06:13   And to answer your question, Casey, like whether it can be metal or not,

00:06:15   the only the advantage of the metal clips is that they can be very,

00:06:18   very thin.

00:06:20   And there's part of the clip that like like the kind of the part

00:06:23   that goes like vertically up and down between like in the middle of it,

00:06:27   that has to be very, very thin because the thicker that is, I'd be

00:06:31   creating a gap between the screen and the window frame.

00:06:33   Oh yeah.

00:06:34   So with metal, that can be very, very thin.

00:06:36   I c this is as thin as I can make it with it with

00:06:38   a printer like like the one I have, but uh that is the trade-off.

00:06:41   But it so so far, mosquitoes cannot get through that gap, so I'm very

00:06:44   happy still.

00:06:46   You know what else is exciting is a new HP member special, which we

00:06:50   have.

00:06:50   We have your August member special.

00:06:52   This is ATP movie club Nausica in the of the Valley of the Wind.

00:06:57   This is a Miyazaki film.

00:06:58   Uh John, why don't you tell us a little more about this, please?

00:07:01   This is the I think the third one we've done in uh our series

00:07:05   on Miyazaki introducing both of you to it and seeing what your reactions are.

00:07:09   Uh uh tune in to find out how it went and what the future

00:07:13   holds for Studio Ghibli slash Miyazaki Club on ATP Movie Club.

00:07:17   We haven't the last one we did was a while ago, so um we

00:07:20   don't do them very often, but uh yeah, we might come back to this

00:07:24   well uh in the future.

00:07:25   So and if you like these kind of movies and you want to hear

00:07:27   uh two newbies watch it for the first time along with uh an old

00:07:31   grizzled veteran, uh that's me.

00:07:33   Uh tune in and definitely watch the movie.

00:07:36   Like all these things, obviously we talk about the whole movie and everything that

00:07:38   happens in it.

00:07:39   So the idea is that you should watch the movie before listening to the

00:07:42   episode.

00:07:42   And you may think I'm never gonna listen to what uh watch this movie,

00:07:45   but hopefully we do some kind of job of making you at least interested

00:07:48   in the movie.

00:07:49   So watch it ahead of time.

00:07:50   Uh it's a good one, I think.

00:07:52   Yep, you should check it out.

00:07:54   Uh Marco, if one is not a member, but this sounds deeply appealing.

00:07:57   What should one do in order to become a member and get access to

00:07:59   this and all of our other member specials?

00:08:01   Just go to atp.fm slash join today, right now it's atp.fm slash join.

00:08:06   ATP.fm slash join.

00:08:08   Sunday, Sunday, Sunday.

00:08:10   So many of our sponsors want us to repeat the call to action three

00:08:13   times.

00:08:13   And normally one of them is spelled out in this case, well, I guess

00:08:16   slash J-O-I-N.

00:08:17   I can say if I could it's already it's already spelled out otherwise.

00:08:20   Yeah, and uh the and yeah, if you want to hear the back catalog,

00:08:23   like obviously we have tons of member specials and we've done a bunch of

00:08:26   movies, and you can listen to the other movies as well if you want

00:08:28   to get caught up on our uh you know Miyazaki Studio Ghibli run through.

00:08:33   Yeah, we don't have a count of all the member specials on slash specials,

00:08:36   do we?

00:08:37   We have the list of all of them.

00:08:38   No, we don't.

00:08:39   I think it's like 35, 40 now.

00:08:41   I don't know, something like that.

00:08:42   Yeah, I'll have to uh file a ticket with the uh website administrator, the

00:08:45   webmaster, if you will.

00:08:46   Maybe we can burn a bunch of AI tokens somewhere trying to count up

00:08:49   rows on a page.

00:08:52   All right, let's do some follow-up.

00:08:53   And we have some more macOS 27 UI tweaks.

00:08:57   And starting starting with Mario Guzman, who writes Golden Gate got redesigned window control

00:09:02   buttons in beta six.

00:09:03   They sort of grow and shrink when you click.

00:09:05   And uh Mario included a video.

00:09:07   John, tell me more about this, please.

00:09:09   So, first of all, uh Golden Gate's coming out soon, as we'll soon find

00:09:14   out in this episode.

00:09:14   But like, you know, fall is approaching.

00:09:16   This is really late to be changing the window widgets, don't you think?

00:09:20   Like this is the type of thing that like when they first introduced the

00:09:24   OS, like these are the window widgets, and maybe they tweak them slightly, but

00:09:27   usually not at all.

00:09:29   Really late for UI change.

00:09:30   Not that I'm complaining, like it's cool and fun or whatever.

00:09:32   But anyway, the window widgets look more glassy uh than they did.

00:09:37   They used to be kind of just the their regular, you know, uh red,

00:09:41   yellow, and green colors, and now they look like little red, yellow, and green

00:09:44   gems.

00:09:45   Um there was also a video posted of someone like presumably I can't see

00:09:50   I don't think you can see the cursor in it, but clicking on them

00:09:53   and then dragging and they stretch out a little bit like they're flexible, like

00:09:56   they become ovals instead of circles, and then they spring back.

00:09:59   I haven't been able to do that.

00:10:01   Uh well, we'll link to the video in the notes and you can see

00:10:04   what it's supposed to look like.

00:10:04   But I even tried it on touch, even went to macOS 27, beta six.

00:10:08   I connected my iPad through sidecar because that's all like the app get touch

00:10:12   controls and everything are exposed if you use a uh that's that's how they

00:10:15   presented at WWDC.

00:10:16   Hey, support sidecar, not touch max.

00:10:19   Anyway, um, I tried it with my finger, it didn't do it.

00:10:21   So I'm not sure where the dealers of that video, but you can watch

00:10:24   it and see.

00:10:25   Maybe in a future beta, they'll start getting all squishy and uh it'll be

00:10:29   the equivalent of what is it, Casey?

00:10:31   The uh the fried egg headlights on the uh what which the model is

00:10:35   that the 996 that had the fried egg headlights?

00:10:37   Oh god, it's broken.

00:10:39   Broken egg headlights for all the car people are yelling at us.

00:10:41   I'm sorry.

00:10:42   I don't know.

00:10:42   I'm so triggered by this because the group chat that I've brought up many

00:10:44   times on the show, I and I think I talked about this when I

00:10:47   was in my car search.

00:10:44   All they do is send 911s back and forth to each other.

00:10:52   And one of the two people in the group, or one of the two

00:10:54   other people in group chat hates the fried egg headlights.

00:10:58   I understand what you're saying.

00:10:59   I don't know what generation it was, and I don't care.

00:11:02   I'm not a 9-11 headline.

00:11:04   I think I got it right.

00:11:04   996.

00:11:05   Yeah.

00:11:05   There you go.

00:11:06   Anyway, the if you want window widgets that used to be circular but become

00:11:10   not circular when you poke at them with something, maybe that'll come in a

00:11:14   future beta.

00:11:15   But yeah, real real late for changes in Mac OS look and feel, but

00:11:18   they're tweaking it down to the last minute.

00:11:20   And honestly, I think this is I mean, so far it seems like they're

00:11:24   not really breaking it by doing that.

00:11:26   Uh no, I mean they're just cosmetic changes, really.

00:11:28   Yeah, and I think like this to me is this bodes very well, I

00:11:33   think, for the future of Apple's design, their software design, because you know, I

00:11:37   I think obviously, you know, Steve LeMay's uh ascent to power here, uh and

00:11:42   whether it's related to John Turnus's own ascent to power, maybe who knows.

00:11:46   Um, this is all like relatively recent, and so some of these changes I

00:11:51   think are kind of coming in hot, but they're good.

00:11:53   Like so far, all the changes they've made in Golden Gate, uh to Liquid

00:11:56   Glass and to the system design, they've all been for the better.

00:11:59   I can't think of a single change they've made that's been that's been worse.

00:12:01   Um, the little Siri blob that's next to everything or the text all the

00:12:05   time.

00:12:06   Although I just saw someone post how to get rid of that, so maybe

00:12:08   that'll be next week's episode if I try it and it works.

00:12:10   Yeah, those little like Siri like flaps that are next to all the text

00:12:13   areas, that has to go.

00:12:15   Like those are awful.

00:12:16   And to have no option to turn them off.

00:12:19   No, there is.

00:12:20   That's what I was saying.

00:12:21   I think there's uh an option to turn them off.

00:12:22   I just uh system-wide?

00:12:24   Like Yeah, yeah.

00:12:25   Uh someone just posted about it.

00:12:26   Um I haven't had a chance to test it yet because it was literally

00:12:29   like minutes before we sat down.

00:12:30   But apparently it's screen time content and privacy settings and turn off Siri writing

00:12:34   assistance.

00:12:34   And that removes them, supposedly.

00:12:29   I haven't tried it yet.

00:12:36   I'll confirm for next week.

00:12:39   Yeah.

00:12:39   Anyway, so uh I I think it's I think it bodes well because these

00:12:43   changes are all like not only are they good, but I think they show

00:12:49   more more of a return to what we love about Apple and the design

00:12:53   of what used to be called OS 10.

00:12:55   Like it is getting a little bit more aqua y, but like a modern

00:12:59   version of Aqua.

00:13:00   You know, last year when when 26 series OS came out and Liquid Glass

00:13:04   was debuted and it was very controversial because it was some ways good and

00:13:07   some ways bad.

00:13:08   There was a lot of it that we liked.

00:13:10   There were and a lot of it was like, oh wow, this is nice

00:13:12   to have a modern look.

00:13:13   We just didn't like a lot of the implementation details of it.

00:13:16   And granted, some of the foundations of it, I think needed to be rethought

00:13:21   or thrown away, but there was good stuff there.

00:13:23   And I think what we're seeing in the 27 generation now is the design

00:13:28   team not getting rid of all the problems of 26's design, but of actually

00:13:34   giving us like here's the best of Liquid Glass.

00:13:37   Here's here's like the promise of it being better realized.

00:13:41   It still is modern, it still is glassy, it still has some of some

00:13:45   of the good things that Liquid Glass brought, but they're sending off a lot

00:13:49   of the rough edges, they're getting rid of a lot of the worst of

00:13:51   it and evolving it in a direction that not only is more Mac like

00:13:57   in a way, but is also just better overall and has a lot fewer

00:14:01   of those major downsides.

00:14:03   Moving right along with more UI stuff, macOS 27 also has the new preview

00:14:07   icon in beta six.

00:14:08   Yeah, and this preview icon, I'm not a big fan of it, but it

00:14:11   is better than the one that came before it.

00:14:13   So yeah, moving in the right direction.

00:14:14   All right.

00:14:16   Then uh last episode, I think it was, we were talking about how uh

00:14:20   LLMs are not deterministic, and you had made it a kind of off-handed comment,

00:14:25   if I recall correctly, that even a temperature zero, they're not entirely deterministic.

00:14:28   Well, no, I I gave a temperature zero example of them being deterministic and

00:14:31   how that's not pleasing.

00:14:32   We talked we're talking about the the uh AI watermarking thing, like anthropics AI

00:14:36   watermarkings thing.

00:14:37   That was the topic.

00:14:38   And uh yeah, we have discussed uh the role of temperature and LLMs in

00:14:42   past episodes, including temperature zero, but we have some more feedback on that now,

00:14:46   which is just uh basically reiterating something we discussed before, but we didn't mention

00:14:49   it last week.

00:14:50   Right.

00:14:51   So uh when it comes to setting the temperature to zero, you know, things

00:14:54   are still non-deterministic.

00:14:56   Uh, you need to consider that floating point is all over these things, and

00:14:59   floating point is weird.

00:15:01   Uh and there's also a really good uh blog post that John pointed us

00:15:05   to, uh, John Syracuse pointed us to by Horace He, hey, uh who uh

00:15:09   at thinking machines.ai, who writes one common hypothesis is that some combination of floating

00:15:14   point non-associativity and concurrent execution leads to non-determinism based on which concurrent core finishes

00:15:21   first.

00:15:21   While this hypothesis is not exactly wrong, it doesn't reveal the full picture, and

00:15:25   there's a very long blog post that follows that.

00:15:27   That's very good.

00:15:28   Uh additionally, you can check out the IEEE standard for floating point arithmetic, which

00:15:32   is IEEE 754, which we'll link in the show notes.

00:15:34   That talks about why floating point is weird.

00:15:37   Yeah, we got this feedback after maybe the first time we talked about LLMs,

00:15:41   and someone said, hey, even at temperature zero, they're not deterministic because of weird

00:15:44   floating point errors in the hardware.

00:15:45   And you're like, oh, yeah, that totally makes sense because I mean a not

00:15:48   a floating point arithmetic being non-associative when the mathematical operation is theoretically associative is

00:15:54   a problem.

00:15:54   Floating point in general is really weird.

00:15:56   The more you know about how floating point works on computers, the more you

00:15:59   think how does anything ever work?

00:16:01   It's just you know, there are finite resources.

00:15:59   We have to come up with a way to represent floating point numbers that

00:16:06   can span a huge range, and there's just it's kind of like hashing, like

00:16:10   that you can't get blood from a stone.

00:16:12   Like you want to uh hash all the world's data into this number of

00:16:16   buckets, well, you're gonna have hash collisions, right?

00:16:18   And with floating point, it's like you want to represent how many numbers in

00:16:22   what range, and how many bits do you have to do it?

00:16:24   Well, guess what?

00:16:24   There's going to be holes, and not only are there gonna be holes, there's

00:16:27   gonna be weird because there are holes, there's gonna be weird rounding, and there's

00:16:30   gonna be mathematical operations that aren't associative.

00:16:32   I mean, you occasionally see this when someone naively programs a calculator and they

00:16:36   use floating point, and you can do certain additions and subtractions of simple decimal

00:16:39   numbers that an elementary school student could do, but your calculator will get it

00:16:42   wrong because of the magic of floating point.

00:16:44   Well, uh, all LMs are doing is just tons of math all the time,

00:16:48   and it makes sense that these floating port weirdnesses would accumulate.

00:16:52   Uh, so yeah, temperature zero theoretically means completely deterministic according to the algorithm that

00:16:57   LLMs are running, but practically speaking, because of the realities of the hardware and

00:17:01   floating point representations, and possibly also due to concurrency, it's not actually the case.

00:17:07   It's almost deterministic, but not quite.

00:17:10   All right, so with regard to the new photos shared album where you can

00:17:14   get like full resolution photos in that in uh in all the OS's 27.

00:17:17   Tim Hardwick at Mac Rumors writes for years now, Apple has allowed you to

00:17:21   store up to 5,000 photos and videos in shared albums without impacting your iCloud

00:17:25   storage limit.

00:17:25   That's no longer the case in iOS 27.

00:17:27   Perhaps more significantly, if you invite someone to a shared album and they add

00:17:30   their own photos and videos to it, as the owner, this too will count

00:17:32   against your iCloud storage limit.

00:17:35   On the plus side, uh the OSS 27 and introduce a free alternative, you

00:17:38   can create temporary shared albums that allow users to collaborate on short-term projects without

00:17:43   eating into your iCloud storage.

00:17:44   But after 30 days, any photos and videos not saved in your library are

00:17:47   deleted.

00:17:48   Apple has made it easier to invite people to shared albums.

00:17:50   The company says friends and family can now join and collaborate on the web

00:17:53   even without an Apple account or Apple device.

00:17:55   Existing shared albums are unaffected by the changes, but the Photos app does include

00:17:59   an upgrade shared album option for those who want to improve the quality of

00:18:02   its contents at the cost of iCloud storage.

00:18:04   Friend of the show, Greg Pierce also writes it's worth noting that if you

00:18:08   create a shared album on iOS 27 or any of the OSS 27, it's

00:18:11   not shareable with devices on older OS versions.

00:18:14   Yeah.

00:18:14   I mean, I I think all these trade-offs are a reasonable and acceptable, although

00:18:18   it has always annoyed me that the way they count storage is I think

00:18:21   well it doesn't say it here, but I'm assuming they're continuing the practice of

00:18:24   counting it towards everybody's storage.

00:18:26   You know what I mean?

00:18:26   Like, oh, it counts towards your storage, but I think it also counts towards

00:18:29   their storage, and even though we know Apple's only storing one copy of this.

00:18:32   Anyway, um, yeah, I had totally forgotten that those 5,000 photo limits don't count

00:18:37   against your iCloud uh limit.

00:18:38   So they were lower resolution and they didn't count accounts against your limit.

00:18:42   So Apple didn't have to, you know, give you too much free storage.

00:18:45   But now that you get full res, uh it's back on you, and uh

00:18:48   it's because it's encouraging more people to buy more iCloud storage.

00:18:52   You know, one of these days we'll get more than five gigs, right?

00:18:55   Yeah.

00:18:55   Right.

00:18:56   Not this year, I don't think.

00:18:58   Probably not this decade.

00:18:59   I mean, well, honestly, like when you have a new CEO, this is on

00:19:04   his list.

00:19:05   Like, you know, Tim Cook had like a list of things that he changed

00:19:08   immediately, like the like the charitable giving, you know, stuff like that.

00:19:10   Like maybe that's on Turnus' list.

00:19:13   I doubt it.

00:19:14   I can't imagine.

00:19:15   I really can't imagine, but you never know.

00:19:17   Say say they change it.

00:19:18   Say they make it, they multiply it by five.

00:19:21   So you get twenty-five.

00:19:22   Doesn't that really make it?

00:19:23   I mean, it's embarrassing they haven't changed it in a long time, but basically

00:19:26   the way it exists is you get a non-zero amount for free, but that

00:19:29   non-zero amount is basically useless if you really start using it for any of

00:19:29   its intended purposes, in particular using it for backups.

00:19:36   So it's like it's like they they they want you to have storage, but

00:19:39   not enough to do anything particularly useful with like if you if no matter

00:19:43   how stingy you are with storage, if you use Apple's anything, uh like if

00:19:46   you take photos on your phone, like you'll you'll fill it and then they'll

00:19:49   want to get money from you.

00:19:49   So I don't think they're ever gonna give you so much storage that it

00:19:52   changes how many people need to buy the storage.

00:19:54   It will just change the timer on how long it takes before their phones

00:19:57   fill up and they get frustrated.

00:19:59   Yeah.

00:19:59   So what's gonna happen first?

00:20:00   Uh Apple is carbon neutral by 2030, or we get more than five gigs

00:20:04   of iCloud storage.

00:20:05   Carbon neutral.

00:20:06   Easy.

00:20:07   Yeah, honestly, I think you might I mean one is a lot easier than

00:20:11   the other.

00:20:11   Um, but they're not gonna do the easy one because it'll be too expensive.

00:20:15   Yeah, like the because they haven't done it in so long.

00:20:17   Obviously, nobody cares.

00:20:17   I mean, again, maybe if it's on Ternus's day one list, then that's the

00:20:20   only way it'll happen.

00:20:21   But if it's not on his day one list, I don't see that bubbling

00:20:23   up to his attention in the first five years or so that he is

00:20:27   that he's CEO.

00:20:28   Hey, you know what?

00:20:29   This is our last episode with Tim Cook as CEO.

00:20:32   It is.

00:20:32   Oh, that's true.

00:20:33   I didn't even think about that.

00:20:34   Uh huh.

00:20:34   So we don't even know the day one list could literally be in time

00:20:37   for our next episode.

00:20:39   Like some stuff could start happening.

00:20:41   I mean, it's unlikely that we would hear about it or that it would

00:20:43   be done that fast, but yeah, you wouldn't think there would be anything happening

00:20:45   until like after the iPhone event, which we'll get to, but uh yeah.

00:20:48   Some people did ask if we wanted to do like an exit interview for

00:20:51   Tim Cook.

00:20:52   We already did.

00:20:53   Yeah, I kind of think like we've covered that well.

00:20:56   Um I I think mine would kind of be more like a uh, you

00:20:59   know, don't let the door hit you kinda kind of thing.

00:21:01   But uh Yeah, we should find maybe for next week we'll try to find

00:21:04   the episodes.

00:21:04   But we did we basically did an exit of your interview of Tim Coke

00:21:07   over multiple episodes.

00:21:08   We really covered this topic thoroughly once we knew he was out the door.

00:21:11   And since then he hasn't done anything additional worthy of discussion, I feel like.

00:21:15   All right, continuing with follow-up, uh, with regard to last week's overtime, where we

00:21:19   spoke about Stephen Robles's video uh showing how his son uses his first iPhone.

00:21:24   Uh a lot of people wrote in to say, oh, well, clearly that was

00:21:27   sped up because in the the implication being nobody could use an iPhone that

00:21:30   quickly.

00:21:31   Uh Steven wrote to us and said, I didn't speed up anything.

00:21:34   I blurred some things for privacy, but everything was 1x, uh, which was incredible

00:21:39   to me.

00:21:39   And then additionally, uh, a lot of people thought it was very interesting, his

00:21:43   son's preferences with regard to lowercase i versus capital I, especially in combination with

00:21:47   the fact that uh autocorrect was turned off.

00:21:49   And Steven said also on lowercase i versus uppercase i, that's not a text

00:21:53   replacement.

00:21:54   He actually presses the shift key to capitalize it manually every time.

00:21:58   But I told him about text replacement, and he seemed intrigued.

00:22:02   We are sponsored this week by Tail Scale.

00:22:05   I know you've heard me talk about Tail Scale a lot, and I do

00:22:07   that even times not like right now, because I really and truly love it

00:22:12   that much.

00:22:12   Tailscale has compensated us for me to talk to you about them, but truly,

00:22:16   there is no amount of money that I will accept for me to say

00:22:19   this is possibly my favorite software of 2026.

00:22:22   And I write software.

00:22:23   Tail scale is incredible.

00:22:24   And what Tail Scale brings to the table is imagine that all of your

00:22:27   devices, your servers, your phones, your tablets, your PCs, God help you, your Macs.

00:22:33   Imagine they could all talk to each other always, irrespective of what continent they're

00:22:38   on, what network they're on, what firewalls stand in between the two of them.

00:22:43   Tail scale makes all of that go away.

00:22:45   So all of the devices on your tail net, all of the devices on

00:22:49   your network can always talk to each other, pretty much no matter what.

00:22:53   It is like freaking magic.

00:22:55   I really mean it.

00:22:56   Uh again, they've they've compensated us to talk to you, but truly, it is

00:23:00   like magic.

00:22:59   They they can't pay me to say that.

00:23:02   It is absolutely incredible what Tailscale can do.

00:22:59   Now, you might think, oh, I don't need another VPN.

00:23:06   I already run WireGuard or I already run OpenVPN or I already run and

00:23:10   have Ubiquity Teleport or whatever.

00:23:12   Well, Tailscale is WireGuard, yes.

00:23:14   And so strictly speaking, it's kind of sort of a VPN, but it's a

00:23:17   mesh VPN, which means instead of having a hub and spoke model, all your

00:23:20   devices just talk to each other directly, which means throughput is very quick.

00:23:24   Tailscale is truly incredible.

00:23:26   It's absolutely worth trying, and it is free for personal use.

00:23:30   Tailscale's personal plan will be free forever for up to six users and unlimited

00:23:35   devices.

00:23:36   No credit card required.

00:23:37   They will not ask you for one.

00:23:38   Go to Tailscale.com slash accidental tech to get started, and it's free.

00:23:42   Why wouldn't you try?

00:23:43   If you'd like to bring Tailscale to work, you can use the code accidental

00:23:46   tech for three free months of any paid plan.

00:23:49   Thousands of companies like Microsoft NVIDIA and the New York Times have already made

00:23:52   the switch.

00:23:52   So please, again, Tailscale.com slash accidental tech, T A I L S E A

00:23:57   L E dot com slash accidental tech to get started.

00:24:00   Thank you so much to Tailscale for sponsoring the show.

00:24:06   All right, we got some news earlier today as we record this.

00:24:08   The Apple event surprise and shine has been announced for Wednesday, September 9th at

00:24:14   uh 1 p.m.

00:24:15   one true time zone.

00:24:16   Uh your three intrepid hosts will be watching at home as per usual.

00:24:20   Uh and I don't know, I don't have too much more to say about

00:24:23   that.

00:24:23   I don't know what you guys have.

00:24:25   Herbert pointed out that the reason it's on Wednesday instead of Tuesday is because

00:24:27   of Labor Day.

00:24:28   So when Labor Day falls on the Monday, they shift it.

00:24:31   Uh, so that's uh why it's one day off.

00:24:33   I it's a surprise and shine thing.

00:24:34   Like, I mean, this event, I've as usual with September events these days, like

00:24:38   everyone knows every single thing that's gonna be announced except for like minor stuff

00:24:41   that nobody cares about.

00:24:42   Um, you know, right down to the the specific size measurement appearance, potentially the

00:24:47   colors, everything about the foldable phone.

00:24:44   If it's not name ultra, that'll be a surprise.

00:24:51   But anyway, here's the invitation here.

00:24:54   What's the surprise and shine?

00:24:56   Like, I get it's a rise and shine pun, but like, is there anything

00:24:58   particularly shiny or involving light that is supposed to be announced?

00:25:02   Is there anything even surprised?

00:25:03   I mean, this is the least surprising Apple event of the entire year, usually.

00:25:06   Like, yeah, we almost we can almost always predict, like it's almost to the

00:25:10   day, if not exactly to the day when it will be.

00:25:12   We know pretty much everything that's going to be announced in it.

00:25:16   Like, it's like because the iPhones leak more than anything else leaks.

00:25:19   And even the leaks of like a the AirPods with the cameras gonna be

00:25:21   there.

00:25:21   German says no, they're delayed until next year.

00:25:24   So anyway.

00:25:24   Uh, but yeah, I don't I don't know what shine.

00:25:26   I don't know what the shine is about.

00:25:27   I don't I don't quite get are the screens really bright or something.

00:25:30   I don't know.

00:25:30   We'll see.

00:25:30   I mean, in the sun, they can they can be reflective, which is actually

00:25:34   not a great thing usually.

00:25:35   But yeah, I don't know.

00:25:36   I mean, I mean, I guess the main uh there's Garmin had a whole

00:25:39   big story about uh people who had supposedly used the foldable phone, but like

00:25:43   you know, but uh we're not ever gonna cover it because by the time

00:25:45   we do another episode, they'll they'll have been announced.

00:25:48   Uh but one of the open questions, it's not that much of a question,

00:25:52   but anyway, like uh the foldable phone, the crease, what is the screen feel

00:25:56   like?

00:25:56   I mean, the expectation is that it will be like all the other phones

00:25:59   that use all the same components that Apple's using, and that they don't have

00:26:02   anything special that anybody else doesn't have.

00:26:03   By this point, Samsung has already come out with their own, exactly like the

00:26:07   Apple foldable phone phone, you know, like the same size and shape, you know,

00:26:12   that same form factor, presumably using the same screen technology they're selling to Apple,

00:26:16   unless Apple has some kind of exclusive deal.

00:26:18   So that'll be interesting.

00:26:19   Like, does Apple even make uh an announcement that, like, oh, our foldable phoney

00:26:23   is better than the other ones because of X, Y, and Z?

00:26:25   Because at this point, like, not only is it leak so much, their competitors

00:26:28   have had time to already release their iPhone fold clones.

00:26:32   And Apple hasn't even released theirs yet.

00:26:29   So when theirs comes out, it's like, oh, that's exactly like the Samsung, whatever.

00:26:37   It's like, yeah, yeah, it is.

00:26:39   Then we have some big and semi-breaking news.

00:26:42   This all happened yesterday.

00:26:43   Thank you, Apple, for considering ATP's recording schedule.

00:26:46   New Mac minis, new Mac Studios.

00:26:49   So we'll put links to the newsroom posts in the show notes.

00:26:52   Uh, but John, I figured you were going to want to drive this.

00:26:55   I will just try to quarterback as best I can.

00:26:57   And you wanted to start with the chips.

00:27:00   So there are a bunch of new chips announced, including the M6.

00:27:03   Into interrupt me when you're ready.

00:27:05   Uh, this is two nanometer.

00:27:07   It has dual 16 core neural engine, up to two times the peak compute

00:27:13   over previous generations.

00:27:15   This is one of the most interesting things about the M6.

00:27:17   Again, this is the plain M6, and remember, there's not going to be, according

00:27:19   to the rumors, there's not going to be an M6 Pro and M6 Max

00:27:22   or M6 Ultra.

00:27:22   This is it.

00:27:23   This is the only M6.

00:27:24   And there are a bunch of interesting things, interesting things about this.

00:27:27   But I think Apple basically led with this, is like they doubled the neural

00:27:30   engines.

00:27:31   And I think this is the only one with double, other than obviously the

00:27:34   ultras, which are two Macs stuck together, so then you get double neural engines.

00:27:37   But uh, yeah, the littlest no suffix plain M6 has double the amount of

00:27:42   neural engines in it.

00:27:43   I mean, it makes sense.

00:27:44   Like they use them a lot.

00:27:45   AI is popular.

00:27:46   What can we do to beef up the plain M6 on AI tasks without

00:27:51   putting on a honk and GPU that would make it too big, too hot,

00:27:54   too expensive?

00:27:55   The neural engine is not that big.

00:27:56   Put two of them in, and so they did.

00:27:58   And again, this is the very first two nanometer chip from Apple.

00:28:01   So maybe that gives them a little bit of room and leeway to put

00:28:03   in a second neural engine.

00:28:05   Red Other have 12 core CPUs, which is two more cores than the plain

00:28:10   M5.

00:28:11   There are two super cores.

00:28:13   If you recall, we talked about this a few months ago.

00:28:14   Uh, that's two less than the M5's four super cores.

00:28:19   But they the M6 has four performance cores, which is the same as the

00:28:24   M5.

00:28:25   Or excuse me, I'm sorry, with the M5 has none.

00:28:27   And then six efficiency cores, which is the same as the M5.

00:28:31   Yeah, I remember the confusion about this.

00:28:33   Like I wish I I got myself confused and had to regoogle for all

00:28:36   this again.

00:28:37   When we say performance cores in 2026, we mean the M cores, the medium

00:28:43   ones.

00:28:43   They're in between.

00:28:44   The super cores are the big ones, the efficiency cores are the little ones,

00:28:47   the performance cores are the medium ones.

00:28:49   And so this has fewer big cores, a bunch of medium ones that the

00:28:54   M5 didn't have, and then same number of the small ones.

00:28:57   I still hate this naming of these cores terrible.

00:28:59   So much we should we should just say big, medium, and small.

00:29:03   Yeah, I mean, so this the M6 has really thrown a monkey wrench into

00:29:06   what I hope would happen, which is like, oh, there's this transition period where

00:29:09   they renamed them and some of them have the old ones and new ones,

00:29:11   but like eventually they'll get all on the same page and we'll be back

00:29:14   to two sizes, the big and the little.

00:29:15   But nope, nope, they're keeping three sizes.

00:29:17   They're keeping three sizes with the same stupid names.

00:29:19   We're never going to escape it.

00:29:21   It's just it's big, medium, and small.

00:29:23   If they had just called the medium one something with medium in it, like

00:29:26   they're called M cores under the cover.

00:29:27   So it's clear someone was thinking medium, but alas, that is not the case.

00:29:31   So we have this these stupid names are going to be with us for

00:29:34   a while because they're pumping out chips with three different tiny cores, and it

00:29:37   seems like they're not going to stop.

00:29:38   Yeah.

00:29:39   And and it just it still bothers me that like the P cores used

00:29:43   to be like like the big used to be the big ones, and they

00:29:46   renamed them.

00:29:47   The M five used to have the M5 was announced with P cores, but

00:29:50   those same cores changed names to super cores once Apple decided the big ones

00:29:53   have to be called super cores.

00:29:54   Oh my god.

00:29:55   Yeah.

00:29:55   But anyway, I I I think I don't mind there being three different kinds

00:29:59   of cores.

00:29:59   Just, you know, there's the same thing.

00:30:02   Yeah, the temporary uh and by temporary, I might mean like years long weird

00:30:06   transition period of us getting used to what these names now mean because they

00:30:09   now mean something different.

00:30:10   But actually having three different cores, I don't think we I don't think we

00:30:15   really know yet how much this matters.

00:30:17   Like is obviously like, you know, as you say, the the M6 has actually

00:30:21   reduced the number of its largest cores uh from four to two.

00:30:25   But then it has four of the mediums and it had no mediums before,

00:30:28   and the mediums are pretty capable.

00:30:30   So like how does this I mean in this arrangement, like Apple presumably knows

00:30:34   what kind of workloads do these chips experience and what's the best combination of

00:30:38   cores to handle those workloads with the minimum power.

00:30:40   So I trust them.

00:30:41   This doesn't look like a ridiculous combination for a plain no suffix M chip.

00:30:45   Uh I think this will be good, especially with the two neural engines, but

00:30:48   um yeah, we'll we'll see when people start getting these.

00:30:50   Yeah, because I as far as we know, there's no benchmarks out anywhere yet,

00:30:53   right?

00:30:53   I haven't I wasn't able to find any.

00:30:54   I I haven't seen them leave.

00:30:55   Maybe they leaked while we're recording this, but I haven't seen any yet.

00:30:58   Yeah, because so we we really have no clue how this actually performs besides

00:31:01   Apple's kind of you know lofty claims, which app Apple's claims for performance usually

00:31:07   are correct, but maybe narrow, and like it's hard to get a really good

00:31:11   idea of how it performs broadly until you actually see it in in lots

00:31:15   of different applications.

00:31:16   All right, so uh there uh is a quote from Apple the world's fastest

00:31:21   single threaded performance and up to one point two times faster multi threaded performance

00:31:25   as compared to M5.

00:31:26   And that makes it sound like the supercores are exactly the same speed as

00:31:30   the M5 supercores, because the only thing they're touting in terms of like a

00:31:34   an increase is 1.2 times faster multi-threaded.

00:31:37   So you know, it's the world's fastest, just like the M five was, which

00:31:40   is fine.

00:31:41   Single core was good in the M5, it's good in the M6, and a

00:31:43   little bit faster in multi-threaded.

00:31:44   And a little bit faster in multi-threaded is like it does have two more

00:31:47   CPU cores.

00:31:48   Granted, it has two fewer big ones, but it has four mediums, so it

00:31:44   works out to be better.

00:31:52   Again, that's I think that's a reasonable claim.

00:31:54   Oh, yeah, because well, it says the world's fastest single thread of performance, but

00:31:58   that could technically be true, even if it's tied with the M5.

00:32:01   Yeah.

00:32:02   I mean, it's probably a little bit faster, but uh who I mean, it

00:32:05   depends.

00:32:06   Like, you know, they they have a process shrink here.

00:32:08   How did they, you know, how did they quote spend, you know, the additional

00:32:12   uh capacity of a process shrink?

00:32:14   They probably, you know, they added cores, it looks like, you know, so they

00:32:18   they probably spent some of the some of the gain just putting more transistors

00:32:22   down and having more cores.

00:32:23   Um, but you know, did they also ramp up speeds in any different ways,

00:32:27   or is there more cache somewhere on the chip?

00:32:29   You know, that I think that all remains to be seen.

00:32:31   High yield on Blue Sky, right?

00:32:34   What's out in Apple S core or an all new P Core?

00:32:37   While size isn't everything, silicon area per CPU core can tell you a lot.

00:32:41   That's why I made this handy chart, and we'll put a link in the

00:32:44   show notes.

00:32:45   Uh the E core is roughly square, maybe a little bit less than square

00:32:50   than and not that big by comparison.

00:32:52   The P Core is another, like, I don't know, third bigger.

00:32:54   And then the S Core is, I don't know, another 40% bigger than the

00:32:58   P core, maybe more.

00:32:59   These are quite different in size, uh, especially the difference between the E core

00:33:03   and the S core.

00:33:04   Quite a big difference.

00:33:05   There is a 12 core GPU, which is two more cores than the M5,

00:33:09   which is a roughly 30% increase in peak GPU compute for AI as compared

00:33:13   to M5.

00:33:14   50% increased geometry rates for complex graphics.

00:33:18   It also includes updates to the shader core architecture, dynamic caching, and hardware accelerated

00:33:22   ray tracing.

00:33:24   You can get up to 32 gigs of RAM with 170 gigabytes per second

00:33:29   memory bandwidth, which is up 10% from the M5.

00:33:34   Additionally, SME replaces AMX in the M6.

00:33:38   John, can you make heads or tails of this for me, please?

00:33:40   Yeah, this was from a bunch of other people.

00:33:42   I hadn't really thought about this, but there was a post uh from uh

00:33:45   Taras Sakaro saying the M4 was the first Apple chip to support ARM scalable

00:33:50   matrix extensions.

00:33:51   That's SME scalable matrix extension.

00:33:53   Uh although Apple has included matrix accelerator in its devices since 2019, it used

00:33:57   a proprietary instruction set called AMX.

00:33:59   The developers had to use Apple provided numerical frameworks to take advantage of the

00:34:03   hardware.

00:34:04   The M4 chip with SME changed this by allowing programmers to write low-level code

00:34:07   that directly targets the matrix hardware.

00:34:09   So AMX has been around forever.

00:34:11   And that has been a frustration of everybody.

00:34:12   It's like, oh, AMX is a secret instructions.

00:34:14   You can't know about them.

00:34:15   It's Apple matrix instructions, but you're not allowed to like write to them.

00:34:18   Like, yeah, the CPU supports it.

00:34:20   But I think they might reject you from the App Store if you directly

00:34:22   use them.

00:34:23   Obviously, you're outside the App Store, you can do whatever you want, but they're

00:34:25   like, don't use these instructions.

00:34:27   Like they're private.

00:34:28   We reserve the right to change them.

00:34:30   What you're supposed to use is Apple has frameworks that under the cover use

00:34:33   the AMX instructions, and they're like Apple's just use these frameworks.

00:34:36   But obviously that's somewhat limiting, uh, especially if you're doing really low-level stuff.

00:34:40   SME is ARM, like the people who make the architecture that the Apple chips

00:34:44   are based on.

00:34:45   The ARM company, they have their own scalable matrix extensions.

00:34:49   And those are not specific to Apple.

00:34:51   They're on any ARM chip, they're available to anybody who wants to make them.

00:34:54   And Apple has incorporated them, and now they're basically replacing AMX.

00:34:58   So Longhorn on uh Mastodon said, Goodbye, Apple AMX.

00:35:01   It was certainly interesting, but SME takes things further from here.

00:35:04   M4 and M5 both had both uh SME and AMX present hardware.

00:35:09   But AMX and uh SME run on the same compute units, but support for

00:35:13   the AMX uh instruction set has gone in the M6.

00:35:16   So they have this transition period where the chip supported both of them, but

00:35:19   the M6 is say goodbye to the AMX instruction set.

00:35:22   It's all SME from here out.

00:35:24   And I think the reason Apple did AMX is like either the SME wasn't

00:35:27   available or wasn't satisfactory to them, so they made their own thing.

00:35:30   And they said and part of the reason they were trying to keep it

00:35:32   private is like we would prefer to maybe switch to the standard the ARM

00:35:36   standard matrix uh instructions when they're available.

00:35:39   And so they kept AMX on and they kept it private.

00:35:41   They said, please you go through this framework, and presumably now if you use

00:35:44   the framework on a chip that doesn't support AMX, it will just use SME

00:35:47   under the covers.

00:35:48   So uh a pretty good technical transition for Apple, but it's worth noting uh

00:35:52   that M6 is the first one that no longer supports AMX at all.

00:35:55   So if you did write directly to the AMX instruction set and uh broke

00:35:58   Apple's rules, your thing is not gonna run on M6.

00:36:02   So a very quick side quest introduction here.

00:36:05   We are asked, or people imply to us on a regular basis, you know,

00:36:08   you know that all the people at Apple listen to this show, and we

00:36:12   know that there are some people at Apple that listen to this show, and

00:36:14   my understanding is most of them are you know engineers that are just trying

00:36:18   to get their job done.

00:36:19   I don't think we have any like executives that listen to us or anything

00:36:21   like that.

00:36:22   Oh, but then I don't know.

00:36:23   Well, I don't know.

00:36:24   We have I mean we've had executives on the show, but yeah, I don't

00:36:27   know if we still do, but I know we have.

00:36:30   All right, well, either way, occasionally something will happen in the world of Apple,

00:36:35   and uh occasionally it makes me think, huh?

00:36:39   That's weird.

00:36:40   So Daniel Luce points out uh with a very, very specific chart from Apple.

00:36:47   M6 compared to previous Mac mini models.

00:36:50   Faster long form audio transcription, Mac Mini with M6, 10.3x, Mac Mini with M4,

00:36:57   4.8x, Mac Mini with M1, not very good.

00:37:00   I guess that's 1X.

00:37:01   This, I genuinely have no idea if this is pointed at Marco slash us

00:36:59   or not.

00:37:06   But man, that is conspicuous.

00:37:08   No, it's not.

00:37:09   I mean, the reality is that long form audio transcription done at large scale

00:37:16   is a problem that not just Overcast has.

00:37:19   Although I will say I for earlier today I was I was um looking

00:37:23   at some of my numbers, and uh Overcast transcribes about 30,000 hours of audio

00:37:29   per day.

00:37:29   Gracious.

00:37:30   Um so it's it's kind of um a lot that I do, but um

00:37:35   ultimately like there are lots of problems out there that that need long form

00:37:40   audio transcription as part of their problems.

00:37:43   Think about like think about how many business meetings there are and how many

00:37:47   products there are out there to try to take your business meeting notes and

00:37:52   summarize them.

00:37:53   And every, you know, Zoom and all those different platforms that host video conferencing

00:37:57   and everything, they all have this.

00:37:58   Like, this is a very common need.

00:38:01   Um, what about don't forget the entire, you know, the intelligence community.

00:38:05   Um if they they might have uses for this.

00:38:08   Yeah, there's lots of uses that are that are less than above board, uh,

00:38:12   but there's a lot of need for long form audio transcription.

00:38:16   And it's a it's a pretty exciting, you know, recent thing that computers recently

00:38:20   have gotten a lot better at than they used to be.

00:38:22   Um plus chances are it probably shows off the capabilities of the new, you

00:38:29   know, the new neural engine architecture or the new GPUs, like you know, whatever

00:38:33   it is.

00:38:33   Like they that that probably happens to be a very, very strong benchmark for

00:38:37   these chips, and that's probably why they chose it.

00:38:39   Not because, you know, and Lal Shampy listens to this show and decided to

00:38:44   do this for this part of the thing.

00:38:44   Like, probably not.

00:38:46   Yeah, and this is the biggest benchmark for this because like uh M6, uh,

00:38:50   well, I mean, they're comparing it to M4, I guess, but like it's double

00:38:53   from M4.

00:38:54   They don't compare it to M5, but it shows that they are getting dividends

00:38:57   from that second neural engine, uh, presumably, because I don't think this this could

00:39:01   come from the CPU and GPU changes alone, but who knows?

00:39:03   Maybe it does.

00:39:04   Like that's the thing about these benchmarks.

00:39:07   It's like, does uh audio transcription use the neural engine at all?

00:39:11   Does it depend on what engine you're using?

00:39:13   We just see the bar charts.

00:39:14   We don't have the real benchmarks from people yet, so we'll wait and see.

00:39:17   Yeah.

00:39:17   By the way, it doesn't take 48 Mac minis to transcribe 30,000 hours of

00:39:21   audio a day.

00:39:22   Um, it turns out I have dramatically overprovisioned how many Mac minis I need.

00:39:27   Uh well, because remember I mentioned like during the story that at some point

00:39:30   I fixed a bug that was causing repeat work to be in queued?

00:39:33   Well, when I did that, the usage dropped by like two thirds.

00:39:37   Well, luckily those Mac minis are just going up in value.

00:39:39   So basically it's like getting a bunch of solid gold bars.

00:39:42   Well, the reason I was doing these numbers earlier, I was trying to figure

00:39:44   out like, you know, what how can I best use this excess capacity that

00:39:48   I have?

00:39:49   So should I do things like you know, summarization and auto chapterization?

00:39:54   You know, what does that look like?

00:39:55   How how can I do that on a bunch of 16 gig M4 machines?

00:39:59   Like, what are my options for that?

00:40:00   Should I outsource it?

00:40:01   You know, do and that's that's why I was doing that kind of that

00:40:04   that math.

00:40:05   All right, let's talk M5 Ultra.

00:40:08   There is now an Ultra for the M5 generation, which uses Ultra Fusion to

00:40:13   connect two dual die M5 Max chips to form the quad die architecture.

00:40:19   Now, is that the same as the M3 Ultra or is that different?

00:40:24   Well, the the M5 is the first one that uses the chiplets, the M5

00:40:27   Max and M5 Pro.

00:40:28   They separate the CPU and GPU dials die.

00:40:31   So this is this is really just a sort of cruel twist of fate

00:40:33   of like, oh, you finally got the quad.

00:40:29   No, not the way we wanted.

00:40:36   It's still just two maxes stuck together.

00:40:38   It just so happens the max itself split into two pieces into a CPU

00:40:41   and GPU die.

00:40:43   So they do tout it as a quad die architecture, which is accurate, but

00:40:46   it's still two maxes.

00:40:48   They just can't stop sticking those maxes together.

00:40:50   They love it.

00:40:51   Yeah.

00:40:51   I was briefly very excited, and then I'm like, oh wait, no, that's that's

00:40:55   not what I want it to be.

00:40:57   All right.

00:40:57   This has a 36 core CPU.

00:41:00   The M3 Ultra has 32, 12 super cores, 24 performance cores, uh, no efficiency

00:41:09   cores, which is just like the M5 Max's that it's comprised of.

00:41:13   The M3 Ultra has 24 of the old performance cores and eight efficiency cores.

00:41:18   Yeah, I don't know why there were so many efficiency cores in the Ultra

00:41:21   back in the day, but it's it's it's such a weird comparison.

00:41:23   I mean, setting aside what Apple compares to, they love to compare things to

00:41:25   the M1 to show big numbers or whatever, but they also compared to the

00:41:28   M3 Ultra because that's the previous Ultra.

00:41:31   There never was an M4 one.

00:41:32   So you get some weird comparisons like like the core names in the comparison

00:41:36   don't even match anymore.

00:41:37   So uh yeah, and and uh you know, I it's pretty easy for the

00:41:41   M5 Ultra to beat up on the M3 Ultra because you know it's a

00:41:43   two-generation later, those maxes that are stuck together, but it sure does.

00:41:47   Compared to the M3 Ultra, up to 1.25 times higher single threaded performance, up

00:41:54   to 1.3 times higher multi-threaded performance, which is not bad.

00:42:00   Uh, in terms of GPU, it is the same as the M3 Ultra, 80

00:42:05   cores, 80 cores, which is bananas, up to four and a half times the

00:42:10   peak GPU compute for AI versus the M3 Ultra.

00:42:13   So that's such a like you look at the other numbers, 1.3 times 1.25

00:42:17   times for the CPU.

00:42:17   It's like, all right, the CPU is better, but like you know, not that

00:42:14   much better.

00:42:20   But then you get to the GPU thing, same number of cores, up to

00:42:24   four times.

00:42:24   That's because all these the cores have like the neural accelerators in them, and

00:42:28   it's a totally different GPU architecture.

00:42:30   This is the big leap in the M5 Ultra is that the GPU, even

00:42:33   though it looks even though it has quote unquote the same number of cores,

00:42:35   those are not the same cores at all.

00:42:37   And there's a big difference in that, you know, the in the Apple's press

00:42:40   release, like, oh, everything's great for AI and local AI and blah blah blah.

00:42:43   And yeah, the M5 Ultra is going to be a lot better at at

00:42:45   that stuff with its GPU than the M3 was.

00:42:49   There is a quote unquote more capable media engine.

00:42:53   It has dedicated hardware enabled H264 and HEVC encode to code, four ProRes encode

00:42:58   to code engines, hardware accelerated AV1D code only.

00:43:02   But John, what's up with that?

00:43:03   Yeah, I read that in the press release and I was trying to figure

00:43:05   it out.

00:43:06   And then I went to Apple's compare page and it was like, this is

00:43:08   exactly the same as the M3 Ultra.

00:43:10   Like they do say, like, oh, they're better or updated or improved, but like

00:43:14   on Apple's own compare page, they use literally the same text for the M3

00:43:18   Ultra's media engine and the uh M5 Ultra's media engine.

00:43:21   Maybe the media engines are the same, maybe they just have the same number

00:43:24   of components, but those components are clocked tire or are more capable.

00:43:27   Again, we'll find out with the benchmarks.

00:43:29   Then there is up to a half terabyte of RAM with 1.2 terabytes per

00:43:34   second memory bandwidth.

00:43:36   So that means you can dump all of your RAM in a second.

00:43:39   Is am I reading that right?

00:43:41   Uh I mean, sure, yeah.

00:43:43   But dump it where into the CPU and out of it again.

00:43:46   Yeah, right.

00:43:47   I mean, it's just the point is that you can move it so much,

00:43:49   you can move faster than all of the than reading.

00:43:53   Well, one second is enough because if it would only goes up to five

00:43:55   twelve gigs and it would does double that per second of memory bandwidth, yeah.

00:43:58   That's bananas.

00:43:59   Uh anyway, Apple says that it's 50% higher than before.

00:44:03   Uh the M3 Ultra, John, you noted in our show notes the M3 Ultra

00:44:06   was 819 gigs per second.

00:44:08   So this is roughly 32% faster by John's math.

00:44:11   Yeah, I don't know where they're getting the 50% higher than before.

00:44:13   Maybe I'm doing the math wrong, maybe they are, but again, we'll see when

00:44:16   the benchmarks come out.

00:44:17   I think we are, though, before we leave this this little segment, I I

00:44:20   think we are kind of seeing, though, maybe why we're not getting the quad

00:44:25   chip that we wanted, which I guess now would be the Octo chip uh

00:44:28   with this new construction method.

00:44:29   Um, because when you look at how the CPU performance is scaling between the

00:44:35   max and the ultra, the M3 max to ultra, instead, you know, it didn't

00:44:40   get 100% faster by doubling the number of cores with that whole second chip.

00:44:43   It got about 58% faster on geekbench.

00:44:47   So if that's whatever that's representative of, you know.

00:44:49   So, you know, there is some losses as as you scale up.

00:44:52   It doesn't scale linearly with CPUs.

00:44:54   GPUs scale a lot better.

00:44:55   CPUs really don't.

00:44:57   Um by that math, though, with Apple's claim here that it's up to 1.3x

00:45:03   higher multi-threader performance, when you compare the performance of the M5 uh Max, which

00:45:08   we do have benchmarks for, that's actually scaling worse than the M3.

00:45:13   Well, how can you yeah, I think you have to do more math to

00:45:15   figure that out because this is this is compared to the M3 Ultra.

00:45:17   It's 1.2 times faster than the M3 Ultra and 1.3 times faster in multi-threaded

00:45:21   than from the M3 Ultra.

00:45:22   And then you have to go, all right.

00:45:23   So the M3 Ultra had 32 cores and this is 36.

00:45:26   I haven't done all this math to figure out.

00:45:28   Well, that's what I was doing when you were talking about memory bandwidth.

00:45:30   So instead of listening to that, I did this.

00:45:32   Um, so it 1.3 times the the M3 uh ultra multi-threaded performance would be

00:45:39   about 50,000 on Geekbench, but the same ratio scaling up from the M5 Max

00:45:44   would be about 55,000.

00:45:46   So it's actually scaling, I think, kind of significantly worse in that way than

00:45:51   the M3 Max did if this math is all right and if these claims

00:45:53   line up to what what they should.

00:45:55   But we'll know more with the benchmarks.

00:45:56   But it does seem like the ultra similar to how the ultra has always

00:46:01   been, the ultra really scales CPU performance fairly poorly.

00:46:06   And so if what you need is maximum CPU, that's it's not a very

00:46:10   efficient way to get that.

00:46:12   It's much more efficient to cover larger resources and larger GPU tasks.

00:46:17   But like what Apple does with the Ultra is they always say this, and

00:46:19   it's true.

00:46:20   Where else are you gonna go?

00:46:21   It's our fastest chip.

00:46:22   Like, yeah, it's not as efficient as it as you could imagine.

00:46:24   It doesn't scale linearly, but it's still is the fastest.

00:46:27   So if you want the best performance, this is what you have to buy.

00:46:30   And it's a shame, and I don't agree with it.

00:46:32   That's why I was hoping so much that this would be the year that

00:46:34   the ultra would not be two maxes stuck together, but it totally is.

00:46:37   And with all the trade-offs and compromises that brings, and Apple gets away with

00:46:41   it because it still is their fastest chip.

00:46:44   It is faster than their previous fastest chip.

00:46:46   And if you want the most performance, this is the thing that gives it

00:46:48   to you.

00:46:49   So tough luck.

00:46:50   High yield has another diagram uh this time with a visual explanation of how

00:46:55   the M5 Ultra is designed.

00:46:56   Uh, we will put a link to that in the show notes for you

00:46:59   to check out.

00:46:59   I don't know if this is based on knowledge, but it shows basically the

00:47:01   GPU sides touching each other.

00:47:03   So you got a CPU die and a GPU die on your max, and

00:47:06   you take the two maxes and you stick them together, and it seems like

00:47:08   the GPU side is the side that's touching with the silicon interposer.

00:47:12   Presumably this is known because you could look at the interposer on the maxes

00:47:14   or something, but anyway, that's this is the only diagram I saw of this,

00:47:17   so I included it.

00:47:18   All right, the Macs.

00:47:19   Uh both will ship with macOS 27 Golden Gate, as you would expect.

00:47:23   As you would expect, maybe not, because that OS is not done yet.

00:47:25   So now that puts a timeline on all this stuff.

00:47:27   It's like they're gonna ship with a with a beta OS.

00:47:30   No, they're gonna ship with the release version of the OS, which means Apple

00:47:33   has to finalize the release version of the OS in time to stick it

00:47:36   on these computers in time to stick them in boxes in time for them

00:47:38   to be sale on their for sale date, which we'll get to in a

00:47:40   little bit.

00:47:41   So this puts a line in the sand and says, hey, Mac OS Golden

00:47:44   Gate, it's shipping soon.

00:47:46   All right, so they will have or the Mac Mini will have the M6

00:47:50   or the M5 Pro.

00:47:52   In the M6, you can get uh with regards to RAM 16, 24, 32

00:47:56   gigs, or with the M5 Pro, you can get 2448 or 64 gigs.

00:48:00   Storage for the M6, 256, 512, 1TB, 2TB for the M5 Pro 512, 124,

00:48:07   and 8 terabytes.

00:48:09   I mean, as you go through these, you can see the strangeness that Apple

00:48:12   has brought on themselves, which is the one with the higher number after the

00:48:15   M is the low end Mac Mini.

00:48:17   It's the M6 and the M5 Pro.

00:48:20   But the M5 Pro is the more capable, more powerful, more expensive, has more

00:48:24   RAM, has more uh storage.

00:48:26   The M5 Pro is the quote unquote good Mac Mini, but it doesn't have

00:48:30   the double neural engines and the other, you know, differences we just talked about.

00:48:35   So there may be some benchmarks I can imagine that might actually run faster

00:48:38   on an M6 than on M5 Pro if they take advantage of the neural

00:48:41   engines.

00:48:41   That'll be interesting to see the benchmarks as well.

00:48:43   But yeah, the M5 Pro is the fancy, good, expensive, more capable Mac Mini,

00:48:49   even though it's got a five after the M instead of a six.

00:48:52   With regard to connectivity, both have the Apple N1, which is Wi Fi 7

00:48:55   Bluetooth 6.

00:48:56   They both have two and a half gig Ethernet, but 10 gig is available.

00:48:59   Two USB 3 front ports, uh, then the M6 has three Thunderbolt 4 back

00:49:06   ports.

00:49:06   The M5 Pro is three Thunderbolt 5 back ports.

00:49:09   This is usually the case with the NoSuffix M chips, they don't get as

00:49:12   much like PCI E lanes or whatever, so they have you know lesser Thunderbolt

00:49:16   capabilities, and so you get an M6 with Thunderbolt 4, but an M5 Pro

00:49:21   with Thunderbolt 5.

00:49:22   Uh that's the way it shakes out.

00:49:25   Yep, which means that my if I get more base model Mac minis with

00:49:28   the M6, I still can't do that thing where they cluster together to run

00:49:31   big LLMs.

00:49:32   Yeah, they want you to get a studio for that.

00:49:34   Or an M5 Pro version, which I which can also work, yeah.

00:49:38   Uh with regard to video, Tim Hardwick had a really good summary over at

00:49:41   Mac Rumors.

00:49:42   The M6 Mini can run two displays at up to six K60 or 4K

00:49:46   165 Hertz, up from 144 Hertz in the M4, plus a third display at

00:49:51   up to 5K 60 Hertz over Thunderbolt or 4K 60 Hertz over HDMI.

00:49:56   The M5 Pro can run all three displays at up to 6K60 or 4K

00:50:00   165, which is up from 144 again.

00:50:03   In a two-monitor setup, the M5 Pro model's second screen now supports 5K120 or

00:50:08   4K200, whereas the M4 Pro supports 6K60 or 4K 144.

00:50:14   The M5 Pro Mini supports all three displays through one Thunderbolt port, while the

00:50:18   M6 supports two per port.

00:50:21   Uh single display support is unchanged.

00:50:23   Both M6 and M5 Pro support AK60, 5K120, or 4K240.

00:50:27   I like that Apple continues to push this.

00:50:29   Like, oh, we did one uh 44 Hertz last time.

00:50:31   Now we're up to 165.

00:50:32   Yeah, keep going.

00:50:33   But Apple sells nothing to help them.

00:50:35   Like they they are not Apple is not selling 240 Hertz displays, 200 Hertz

00:50:41   of sprays, 165 Hertz displays, one they don't sell any of that at any

00:50:44   price.

00:50:45   Other people do, and you can connect them to your Macs, but it's fascinating

00:50:48   that like the people who work on the video on Macs on the video

00:50:50   drivers and the video hardware are just like pushing as much as they can,

00:50:54   increasing the specs, you know.

00:50:55   And these are again these are these aren't minis.

00:50:56   This is not a video powerhouse, but like they keep increasing the numbers, meanwhile,

00:51:00   the team that makes the displays is like 60 hertz is probably fine for

00:50:59   everybody.

00:51:04   All right, if you have 3,000, you can get 120.

00:51:08   3500.

00:51:09   I forget how much that one costs.

00:51:10   It's so much.

00:51:11   Uh luckily the rest of the monitor market exists.

00:51:14   With regard to pricing, the M6 starts at 900 bucks, which is 200 more

00:51:18   than the M4 Mac Mini.

00:51:20   The M5 Pro starts at $1,700, which is $200 more than the M4 Mac

00:51:25   Mini.

00:51:26   And uh there's some people have done some science and mathematics with regard to

00:51:30   the quote unquote best config uh available for pre-order.

00:51:33   Uh the M5 Pro, 64 gigs RAM, eight terabyte SSD, $6,800.

00:51:37   There's also a chart uh with pricing.

00:51:40   Just pause that for a second.

00:51:41   Yeah.

00:51:42   A six thousand eight hundred dollar Mac Mini.

00:51:44   What's what are we doing here?

00:51:46   Like I mean, I could figure out some things to do with it.

00:51:48   The Ramageddon is really just like distorting products beyond all reason.

00:51:54   Because like in the old days when you know when RAM and storage were

00:51:57   not inflated by uh multiple times their normal price, the Mac mini kind of

00:52:03   stayed within a Mac mini-ish range, no matter how you spec'd it, because like

00:52:07   they were just they would confine the options, right?

00:52:09   And you couldn't make a Mac mini that expensive, but now it's like I'm

00:52:12   sorry, if just you know, as you add RAM and uh storage, it very

00:52:17   quickly just dwarfs everything else that has to do with this product, and the

00:52:20   prices become ridiculous.

00:52:22   Yeah, and and as usual with the Mac mini, uh the base model has

00:52:26   always been by far the best value.

00:52:30   And as you try to scale things up, like you very quickly explode the

00:52:35   price of that computer into something way larger than the than the initial price

00:52:39   uh suggests that it's like in the class of.

00:52:42   Um, but I mean that this is why I own a very large number

00:52:45   of base model Mac minis, and almost none.

00:52:44   I I think oh, I have one.

00:52:49   My one that I use as like my archive uh server in my house

00:52:53   has like the 512 SSD.

00:52:56   And that's it.

00:52:56   Like I have only base models because the base model is a great value

00:53:01   compared to everything else.

00:53:02   And you spec it up at all, and that starts quickly becoming less of

00:53:05   a good value.

00:53:06   Uh 95 Mac also has a really great uh chart of pricing history, which

00:53:11   is worth checking out.

00:53:12   And then John, you took the liberty of doing inflation adjusted versions of that

00:53:15   same chart.

00:53:16   Uh again, both of these will be in the show notes.

00:53:18   Yeah, so the like the pricing on the Mac Mini, it's started at 499,

00:53:22   which is the lowest that it's ever been, but that was back in 2014,

00:53:25   right?

00:53:25   And the the chart just showing the number shows it goes up and then

00:53:28   down and then it goes back up again, and you're like, wow, it's the

00:53:30   highest that has ever been, or the base model being 899.

00:53:33   And it makes sense, you know, the RAM crisis, storage, blah, blah, blah.

00:53:36   The highest it's ever been.

00:53:37   And that's why I was interested in doing an inflation adjusted one.

00:53:39   Inflation adjusted shows that no, the 2018 model was the highest that it's ever

00:53:43   been.

00:53:43   Inflation adjusted, the 2018 base model is 1,076.

00:53:48   And uh, yeah, that's that's higher than the current 899.

00:53:51   But yeah, it has gone up and down.

00:53:53   The the the lowest price ever base model is 712 in today's dollar.

00:53:57   So the actual lowest one is a 2024 model and inflation adjusted at 650.

00:54:02   Um, so I mean, that's change that's kind of the changing face of what

00:54:05   the Mac Mini is and how much it costs to field any kind of

00:54:08   computer with any kind of chip, especially Intel chips in it.

00:54:11   But um, yeah, the base model is still trying to be a good deal,

00:54:15   but just everything is every every uh reasonable uh concept of pricing is thrown

00:54:21   out the window because RAM and storage are the only thing that matters anymore

00:54:24   in terms of pricing.

00:54:25   Then uh this Mac Mini is not available for lease uh through Apple Upgrade.

00:54:30   Is that right?

00:54:31   Yeah, I checked.

00:54:29   I saw people saying that I went to the Apple upgrade page, and there's

00:54:29   no Mac minis there.

00:54:36   So I don't know why they're not again, it can go up to like

00:54:38   $7,000, but nope, can't lease it.

00:54:40   That could be a yet thing.

00:54:43   Maybe.

00:54:43   I mean, I checked a couple times today.

00:54:44   Maybe by the time you hear this episode, it'll be on there, but it

00:54:47   wasn't when I checked.

00:54:48   You can pre-order right now as we record this.

00:54:50   Uh uh you could pre-order that same day when it was announced.

00:54:54   Uh they arrive starting on September 22.

00:54:56   So there you go.

00:54:57   Uh MagOS 27 Golden Gate.

00:55:00   I'm assuming it will be uh released to the general public by September 22nd.

00:55:05   If not, people who get uh Apple hardware uh with a golden gate on

00:55:09   it are going to have their own private copy of Golden Gate that nobody

00:55:13   else has and nobody else can get.

00:55:15   Uh so I would imagine, yeah, that's that's their deadline.

00:55:18   You know, assuming people get their hardware on day one, which is September 22nd,

00:55:21   uh Golden Gate should be released to the public by then.

00:55:24   Yeah, I'm guessing we'll learn to the public probably on September 18th or somewhere

00:55:27   around there.

00:55:28   Because you know, we have, as we mentioned, like the iPhone event is September

00:55:31   9th, and usually they give us yeah, maybe about a week between the iPhone

00:55:35   event and the OS releases.

00:55:36   So I'm guessing, yeah, sometime around September 18th, probably for Golden Gate and all

00:55:40   the 27 releases.

00:55:42   All right, with regard to pricing, uh the CPU for the new Mac minis.

00:55:47   This is over if you want to get the M5 Pro instead of the

00:55:50   M6, that's adding $800 for the binned one or a thousand dollars for the

00:55:54   unbinned one.

00:55:55   That's a difference of 20 bucks over the M4 Pro.

00:55:59   RAM, 24 gigs is plus $200.

00:56:02   Uh 400, excuse me, 32 gigs is plus $400, $48 gigs is plus six

00:56:07   hundred dollars.

00:56:08   Uh that one is sixty dollars more than it was for the M4 Pro.

00:56:11   And so those other ones you're reading, like those offsets.

00:56:13   Apple did, I think, a good job of cranking the prices on the previous

00:56:17   model so that now we get the new ones, and it's like, well, the

00:56:19   upgrades are ridiculous, but they're they're almost the same as they were in the

00:56:23   previous one.

00:56:24   Like that you'll see this in a lot of the pricing comparisons.

00:56:27   It's $20 more expensive than the upgrade was to the Pro chip before.

00:56:32   $20?

00:56:33   What?

00:56:33   And this is $60.

00:56:34   You can see Apple, like Apple likes round numbers, but they're they're also trying

00:56:39   to like hold the line on these increases.

00:56:41   So again, it's not increasing by you know $47 or something, but $60 and

00:56:46   $20 are weird increases.

00:56:48   And it's only the $48 gig upgrade that is $60 more expensive than it

00:56:52   used to be.

00:56:53   Um but yeah, it's it's like I I see that and I see like

00:56:57   Apple's pricing wizards struggling to try to like because I think their plan was

00:57:02   hey, we'll we'll increase the prices on everything.

00:57:04   And then when the new models come out, we don't have to increase them

00:57:06   again because we already did it.

00:57:07   That's not true of the phones, by the way.

00:57:09   But but for the Macs, they did they did increase all the prices.

00:57:11   But it turns out they couldn't actually hold the line.

00:57:13   And for some of like the larger RAM configuration, they're like, no, actually, we

00:57:17   undershot a little bit.

00:57:18   RAM prices still go up.

00:57:19   We have to be $60 more expensive than that upgrade used to be.

00:57:22   Uh with regard to storage for the M6, uh, over the base 256 gigs,

00:57:28   it's $200 for a half gigabyte, excuse me, half terabyte, $500 for a terabyte,

00:57:33   which is $30 more than it was for the M4, and a thousand dollars

00:57:37   for two terabytes, which is eighty dollars more over the M4.

00:57:40   For the M5 Pro, this is over the base five, 12 gigs, a terabyte

00:57:45   is 300, 2 terabytes is 800, 4 terabytes is 1800 and eight terabytes.

00:57:49   If you're a fool and did that last time when you bought a computer,

00:57:53   that is now three thousand eight hundred dollars, which is three hundred and eighty

00:57:57   dollars more than the M4.

00:57:58   Yeah, and look at these deltas.

00:58:00   $30, $80, $180, and $380.

00:58:02   This is some weird numbers.

00:57:59   Some weird increases.

00:58:05   But like that's that's what the that's what the market will bear at this

00:58:08   point.

00:58:08   They would have liked to have kept these increases the same as they were

00:58:11   before, but they couldn't.

00:58:12   And you know, again, eight terabytes, it's gonna be $380 more than it used

00:58:15   to be for a nice round $3,800 added to the price of your Mac

00:58:19   Mini.

00:58:20   And to be fair, eight terabyte SSDs from Apple have always been absurdly expensive.

00:58:25   This is more expensive than I think they have been, but yes, it is.

00:58:30   But not by like, I mean, what's the I mean it's fine.

00:58:32   I think it used to be $2,400, wasn't eight terabytes.

00:58:34   I forget what it was, but here's the thing.

00:58:35   Or maybe maybe $3,000.

00:58:37   I mean, I think I think when they first were introduced, they were probably

00:58:40   actually near this price, but uh they've always been like multiple thousands of dollars.

00:58:44   So this is more than they were, but it's not we're not talking 10

00:58:49   times more.

00:58:49   We're not, you know, we're not talking that level of a difference here.

00:58:52   Well, well, when this was first happening, everybody's RAM prices and storage prices were

00:58:55   going up, but Apples weren't yet because they hadn't done their their price increase.

00:58:58   It was like, hey, Apple's prices are starting to seem reasonable, but Apple fixed

00:59:02   that.

00:59:03   All right.

00:59:04   With regards to Ethernet, you can get 10 gigabit for 100 bucks, which is

00:59:07   uh 10 bucks more than it used to be.

00:59:09   Uh with with regard to Mac mini pricing, Gruber writes if you configure an

00:59:13   M6 Mac mini with two terabytes of storage, the SSD upgrade alone, which is

00:59:16   a thousand bucks, costs more than the entire base model computer, which is 900

00:59:21   bucks.

00:59:22   So too, with the four terabyte SSD upgrade for the M5 Pro Mini, which

00:59:26   is an eighteen hundred dollar upgrade for $1,700 computer.

00:59:29   Cool.

00:59:30   Just because you want more storage, you're like, oh, the price will go up

00:59:32   a little bit if I add storage.

00:59:33   And now it's like if you s touch that storage tier thing in the

00:59:36   configurator, you double the price of your computer as soon as you touch it.

00:59:39   Don't do it.

00:59:40   Yeah.

00:59:41   Uh there's also a new Mac mini ad, which is like the stop motion

00:59:44   thing, which is very cute and very cool.

00:59:45   Um, there's also a behind the scenes uh Instagram reel reel, which you should

00:59:49   check out.

00:59:44   Uh good stuff.

00:59:51   I really enjoyed those.

00:59:52   All right, Mac Studio.

00:59:53   Maybe this will be John's future computer.

00:59:55   The SOC is an M5 Max or M5 Ultra that with regard to RAM,

00:59:59   the M5 Max comes in 364864 or 128 gigs with different CPUs uh for

01:00:05   the 4864128 that requires the big boy CPU.

01:00:09   M5 Ultra has 96 gigs, 256.5, 12 gigs, uh if you have the M5

01:00:14   Ultra with the 80 core GPU.

01:00:17   Additionally, a five twelve the five 12 gig memory option for the M5 Ultra

01:00:21   is coming in late October for reasons.

01:00:24   Well, yeah, it's not available in the configurator now.

01:00:26   You can't even see how much it's gonna cost.

01:00:28   The answer is more than you want to pay.

01:00:30   Exactly, right.

01:00:31   If you have to ask, it's too much.

01:00:32   Yeah, and they and so this is when I was doing this price comparisons,

01:00:35   like obviously the like the Mac Studio, the configurations that you could get the

01:00:39   Mac Studio in have just been disappearing over the past month or two, right?

01:00:43   And when I went to do the comparison, it was like there was nothing

01:00:46   left of the M4 uh based Mac Studio.

01:00:48   It's like you get like one RAM configuration per CPU and that's it.

01:00:51   And the M5 model has uh increased your options again, but you're still constrained.

01:00:56   So if you want the binned M5 Max, which is like the one with

01:00:59   uh 32 core GPU, you get one RAM choice and that's it.

01:01:03   The second you want any more RAM, you just got to go with the

01:01:06   unbind one, which increases the price a lot.

01:01:07   So a lot of people have been tripped up by this.

01:01:09   Gruber had a poster where he's like, oh, I my math on the the

01:01:11   pricing was wrong because it says, oh, add $300 for this CPU choice.

01:01:14   Oh, I should do that.

01:01:15   That that's a good deal.

01:01:15   But as soon as you touch that option, it's like, well, a bunch of

01:01:18   other stuff is changing too.

01:01:19   Uh you can't have the binned one uh thing anymore.

01:01:21   And yeah, anyway.

01:01:22   Um, but there are more, all I'm saying is there are more options than

01:01:25   there used to be, but there's still not as many options as there was

01:01:27   back in the day when RAM and storage wasn't messed up.

01:01:30   But it is a relief from the the end of the M4 days, which

01:01:29   is like you got choices of like two computers, and that was it.

01:01:36   With storage, M5 Max gets five 12 gigs, one, two, four, or eight terabytes.

01:01:41   The M5 Ultra gets one, two, four, eight, or even sixteen terabytes.

01:01:45   I it I shudder to think about how much 16 terabytes will cost.

01:01:48   We'll find out in a little bit, but yes, the answer is too much.

01:01:51   All right, connectivity, Thunderbolt 5 in RDMA, which is remote direct memory access.

01:01:55   That's the thing where you can chain them together, right?

01:01:57   To make like uh put a big stack of Mac Studios and connect them

01:02:01   together so they can run a model bigger than any single one of them

01:02:03   could run.

01:02:03   It works on Thunderbolt 5.

01:02:05   That's what you can do with your Mac Studios if you have a lot

01:02:07   of money.

01:02:09   The next generation SSD architecture is built on PCI Egen 6, according to Apple.

01:02:14   Apple's N1, as we said, Wi-Fi 7 Bluetooth 6.

01:02:17   These are the first time this and the mini are the first time these

01:02:19   have come into Mac.

01:02:20   Uh, up to eight displays or up to four studio display XDRs at who

01:02:25   who even cares about those anymore.

01:02:26   Anyway, at full 5k resolution and 120 hertz.

01:02:30   Actually, by the way, on that topic, oh no.

01:02:32   On on one of my XDRs, I actually don't even know which one it

01:02:35   is, whether it's the the land the uh mainland John one or the Beach

01:02:39   one, but um I'm still paying for the annual Apple care renewal because I

01:02:43   have it like on the annual you should.

01:02:45   I wish I had not let mine lapse.

01:02:47   It's like 180 bucks a year.

01:02:49   I'm like, and I'm wondering like what happens if it breaks.

01:02:52   Like, do they give me a new one somehow?

01:02:54   I bet they have parts.

01:02:56   Maybe like I I'm just I'm gonna keep paying it as long as I

01:03:00   can.

01:03:00   Like, how long will they let it go?

01:03:02   I I have no idea.

01:03:03   Yeah, I I really wish I was still paying for that, but I'm I

01:03:06   let it lapse, totally forgot about it, and now I'm just living in fear.

01:03:10   Great.

01:03:11   Uh, let's see.

01:03:12   This also supports Genlock over USB C for precise synchronization between a display and

01:03:16   a professional camera capture like the iPhone 17 Pro.

01:03:14   I think the Mini does too.

01:03:20   I think Genlock might have done beyond both models, but yeah, this is not

01:03:23   a feature that I'm interested in, but they're touting it on their new Macs.

01:03:26   Quote from Apple Some it's simultaneously, play up to 33 streams of 8K Pro

01:03:31   Res 422 at 30 frames per second on M5 Ultra, which is cool.

01:03:36   Pricing the M5 Max starts at 2500 bucks, which is 200 more than the

01:03:41   M4 Max.

01:03:42   The M5 Ultra starts at $5,500, which is $600 more than the M3 Ultra

01:03:48   Max Studio.

01:03:49   Uh lease pricing, M5 Max starts at $49 a month.

01:03:53   M5 Ultra starts at $110 a month.

01:03:56   The quote unquote best config available for pre-order.

01:03:59   Everyone better sit down.

01:04:06   The just eminently affordable $18,300.

01:04:12   Holy balls.

01:04:13   That is an expensive computer.

01:04:15   I do wonder, I mean, we're not aware of these prices because we don't

01:04:18   shop for these things.

01:04:18   And two of us don't.

01:04:20   No, no, not these prices.

01:04:21   The prices about to say like the the prices that uh people pay for

01:04:25   like the big giant cost as much as the car NVIDIA things that run

01:04:30   LM training and inference, you know, those things, those have always been tens of

01:04:33   thousands of dollars.

01:04:35   But I'm assuming those have also increased.

01:04:37   I mean, I don't I don't know how it works, but like the whole

01:04:38   reason the RAM crisis is there is because they bought up all the RAM

01:04:41   for those, so maybe those prices are steady.

01:04:42   But anyway, when I see like an $18,020 M5 Ultra config and it's only

01:04:47   got $256 gigs of RAM, I'm like, well, if I have $20,000 and I

01:04:50   want to do LLM stuff, why don't I buy like one of those rack

01:04:53   mount LVD NVIDIA things that has way more RAM and is way more powerful,

01:04:57   and then that's like $30,000.

01:04:58   Maybe you're getting more bang like if that's what you care about.

01:05:01   If running obviously, like they don't run Mac OS, like you can't use them

01:04:59   as your desktop computer, so I get it.

01:05:06   But if these prices, once you start saying $18,000 for $246 gigs of RAM

01:05:12   and 16 terabyte SSD, which is probably you know the main thing is inflating

01:05:15   this price, it's like that's not that great a deal if you just care

01:05:18   about AI stuff.

01:05:19   You're and you know, to like to Marco's point earlier, you're probably better off

01:05:22   getting some lesser uh M5 something Mac Studios and connecting them together with that

01:05:26   RDMA thing than buying the one big Honkin one.

01:05:28   But if you want to, Apple will sell it to you for almost 20,000.

01:05:31   Your pre-order now uh arrives September 22nd.

01:05:35   Grober has some comments here as well.

01:05:37   It kind of stinks that there are no RAM options for the studio between

01:05:40   the 96 gig base and the $4,000 256 gig upgrade.

01:05:45   Jesus.

01:05:46   I mean, 96 gigs is basically enough for anybody who's using it as a

01:05:49   desktop computer, but if you want to use it for AI stuff, that extra

01:05:52   RAM really pays, and yeah, you jump from 96.

01:05:55   Next choice in the little configurator is add $4,000 to go up to 256.

01:06:00   Which again, 256 gigs of RAM used to not be that obscene.

01:06:04   Like the computer I'm sitting next to in theory supports 1.5 terabytes of RAM.

01:06:08   Obviously, uh a lot slower than the RAM that they're talking about here, but

01:06:12   still, if just you know, we're I hate going backwards.

01:06:15   I hate having things that are the numbers are lower than they used to

01:06:19   be, even if again it's not the same kind of RAM.

01:06:21   And the prices are way higher, so that's not great.

01:06:23   Yeah, I recently was looking uh like I was looking earlier, like, I wonder,

01:06:27   could I with my next purchase, could I fit comfortably into 96 without noticing

01:06:31   a degradation of performance?

01:06:33   Because I currently have 128 because when I bought this, you know, M3 Max

01:06:37   MacBook Pro, whatever it was three years ago, like 128 at that time was

01:06:42   a lot, but it wasn't like obscene.

01:06:44   It was like just, oh, that's a that's like a high-end desktop.

01:06:47   Um, and I currently use about 92 gigs most of the time.

01:06:52   So it's like, well, I can I can fit into 96.

01:06:54   Like I it would it would still feel like annoyingly like a downgrade, though.

01:06:59   So I I just uh what a what a shame to have to buy

01:07:02   a computer in this day and age, John.

01:07:04   Yep, it must really suck.

01:07:06   Uh then before we move on, there's a new Mac Studio ad, which I

01:07:10   actually kind of liked.

01:07:11   Um, and they do a lot of like play with the the words on

01:07:15   the screen.

01:07:16   It's a brief ad, and I think you should check it out.

01:07:18   Not as fun as the Mac mini one, but still not bad.

01:07:20   Yeah, I think that Mac mini mat is really fun and interesting.

01:07:23   I mean, I don't know how how well it sells the computer or even

01:07:26   relates to the computer, but it's a cool looking short piece of animation.

01:07:32   Uh do we want to save your trip down the configurator for overtime?

01:07:36   No.

01:07:36   I'm kidding.

01:07:37   I'm kidding.

01:07:38   I'm kidding.

01:07:38   Yeah, most of our listeners don't care about this, right?

01:07:41   Right.

01:07:41   I'm so tempted to legitimately do that, but I'm not that evil.

01:07:44   Uh so John, all right.

01:07:46   Uh you can handle this however you choose.

01:07:48   I know I speak for me and probably Marco to say I am dying

01:07:50   to know if you've pre-ordered anything.

01:07:52   Um I I did ask you to walk us through your thought process with

01:07:57   regard to configurations.

01:07:58   If you want to jump to the end and tell us what you've done,

01:08:00   or if you want to walk through Wait, wait, wait, before he does, Casey,

01:08:03   do you want to take bets on what he did, if anything, and what

01:08:06   and you know the specifics of that?

01:08:07   Oh I I do, I do, but I'm not prepared to answer this question,

01:08:11   so I'm gonna have to walk through the configurator real, real fast.

01:08:14   Hold on.

01:08:14   Let's do it together.

01:08:15   Okay.

01:08:15   So first of all, I I'm gonna bet that he did place an order

01:08:20   today.

01:08:20   Or at least, hmm.

01:08:22   He probably okay so indecisive.

01:08:24   He's going to buy this Mac Studio, but he probably wants to wait until

01:08:29   somebody gives him some kind of like insider discount.

01:08:31   And that might not be available yet this second.

01:08:29   That might that might take like a week or two or something.

01:08:36   I'm guessing though, he has decided he will buy this, and he has decided

01:08:39   what configuration he will buy.

01:08:42   So let's step through this the configurator now together.

01:08:45   Um I think M5 Max, not M5 Ultra.

01:08:50   Oh, I disagree.

01:08:51   He'll absolutely go for the M5 Ultra.

01:08:52   Really?

01:08:53   Well, let's let's see.

01:08:54   Before we get there, let's see, because like maybe some of his storage options

01:08:56   might require it.

01:08:57   Okay, so I'm gonna say memory.

01:09:00   I'm gonna say he probably won't wait on John.

01:09:03   How much memory is in your Mac Pro?

01:09:04   Are you clicking?

01:09:05   You should both be clicking on the options because as Marco said, the stupid

01:09:07   configurator changes stuff all the time.

01:09:09   It throws up oh, you want to change that?

01:09:11   Well, you gotta change these.

01:09:12   It's exactly like buying a car.

01:09:13   Yeah, you need the premium package now.

01:09:15   Um my Mac Pro has 96 gigs of RAM.

01:09:18   Yeah, so I'm I'm gonna say he went either 96 or 128.

01:09:22   Let me see.

01:09:23   Well, there is no 128.

01:09:24   There is 128 on the max chip, not on the ultra.

01:09:27   Okay.

01:09:28   Uh so I'm gonna say he went 128 max.

01:09:31   I disagree.

01:09:32   I think he's going 256 uh ultra.

01:09:35   What are you clicking on the you clicked max, but then you have to

01:09:38   pick.

01:09:38   I hate this.

01:09:39   You have to pick options before the other one's like ungray.

01:09:41   Yeah, yeah.

01:09:41   Which did you click?

01:09:42   Did you click on the bind or unbind max?

01:09:44   I I click I think I just clicked unbid because I think unbind is

01:09:46   required to hit 96 or to hit 128.

01:09:49   See, if he goes, the problem is if he goes, let's see, if he

01:09:53   wants 96, he has to get the ultra.

01:09:56   Oh my god.

01:09:57   I'm telling you, he's going ultra and it's unjust.

01:10:00   Just like cars, man.

01:10:01   You need the you need the leather package and the all-weather package if you

01:10:03   want the heated mirrors, but if you want the heated steering wheel, you can't

01:10:05   get the sunroof.

01:10:06   No, you're exactly right.

01:10:07   All right, so Casey, you're so you're you're betting ultra unbend with 256 gigs.

01:10:13   Okay.

01:10:14   And eight terabytes.

01:10:15   That's my vote.

01:10:14   Yeah, I think the eight terabytes is probably a given.

01:10:14   Um, either way.

01:10:20   So you're gonna go, you're going ultra unbend.

01:10:24   I already lost track.

01:10:26   256.

01:10:27   8 terabyte.

01:10:28   What's your total on that?

01:10:29   Wait, wait, wait.

01:10:30   What about trade in?

01:10:31   Are you clicking the trade-ins thing?

01:10:32   Uh we're we're gonna skip that for now.

01:10:34   Well, you can't skip that.

01:10:35   That's an essential part of the thing.

01:10:36   Well, I don't have your like serial number and all that jazz.

01:10:38   I'm gonna say no.

01:10:39   You don't you don't trade anything in, you keep it in the attic.

01:10:41   Oh, that's true.

01:10:41   Yeah, either way, you're not gonna trade in.

01:10:43   Yeah, uh so mine says buy for $14,300.

01:10:51   And it's so funny to me because John is the cheapest.

01:10:55   Sorry, one of the more frugal people I know, but when it comes to

01:10:59   this and only this, John knows no limits.

01:11:03   And so I think we are looking at at least $14,300 for his computer.

01:11:09   It would not surprise me if he actually spends more than that.

01:11:13   All right, I'm gonna go with M5 Max, Unbind, so the 18 core and

01:11:19   40 core.

01:11:20   It's like the the best M5 Max, because I don't think John can accept

01:11:23   a bin chip.

01:11:24   I think he wants the best version of whatever chip he has.

01:11:27   So, M5 Max, full cores, 128 gigs, eight terabytes for $8,059.

01:11:37   That is a much better deal bang for the well, it's both of these

01:11:40   are hilariously overpriced.

01:11:42   No one's getting a good deal here tonight.

01:11:44   No, no, no.

01:11:44   But given that you are at a little bit less uh or a little

01:11:47   bit more than half of what I've got, that's a much better choice.

01:11:51   But I know, I feel like I know John, and I feel like as

01:11:54   frugal as John is, there is no turning back when it comes to his

01:11:58   computer.

01:11:58   And this is how he's going to get over the fact that there is

01:12:01   no Mac Pro anymore, is by buying a truly asinine Mac Studio.

01:12:06   So, all right, we will I will leave the configurator up because I knowing

01:12:09   me, I will forget what I've selected in the next five minutes.

01:12:12   John, walk us through what your thought process was.

01:12:14   If you are ready, or if you'd like to take this a different direction,

01:12:17   that's fine too.

01:12:17   All right.

01:12:18   Well, so the first thing I did was I looked at the configurator as

01:12:21   you were doing and and got frustrated with the limitations and the options.

01:12:24   But anyway, I immediately went to the storage section to see like what are

01:12:26   the storage options.

01:12:28   Uh and the storage options on both of the models of uh Mac Studio

01:12:33   are unchanged in price.

01:12:34   Like if you want a four terabyte, it's at 1800 under 1800 bucks.

01:12:37   If you want an eight terabyte, it's at 3800 bucks if you get the

01:12:40   max, and the ultra is add fifty at 1500 and at 3500 because of

01:12:44   the vague reset.

01:12:44   But anyway, these are exactly the same as they were for the M3 Ultra

01:12:48   and the M4 Max.

01:12:50   So storage prices for the Mac Studio are the same as they have ever

01:12:54   been.

01:12:54   They're awful, but when Apple increased the storage prices on the M4 Mac M4

01:13:00   slash M3 Ultra Mac Studio before, they were able to hold the line on

01:13:04   those.

01:13:05   And so they're not worse than they were, but they're not good, obviously.

01:13:09   Um for the um for for the the CPU and stuff, mostly unchanged, like

01:13:16   is there are things tied together.

01:13:18   So if you get like an uh an unbin 64 gig with four terabytes,

01:13:22   um, it's like 5300 bucks or eight terabytes is 7300 bucks.

01:13:26   If you bin the chip, you get minus 300 for the bin CPU, which

01:13:30   is not that big of a savings, but then you have to go down

01:13:31   to 36 gigs of RAM.

01:13:33   Like you can't, you know, the RAM is tied to the it's just it's

01:13:36   super annoying.

01:13:37   Um for the M5 Ultra.

01:13:39   Uh, uh if you go with the binned M5 Ultra, it's seven grand, which

01:13:43   is more than the M3 Ultra was.

01:13:45   If you go bin 96 gig, it's seven grand.

01:13:47   These are configs that I was trying.

01:13:43   Um, it's 200 more than the M3 Ultra config was.

01:13:52   And if you go eight terabytes, it's nine grand, which is again 200 more

01:13:55   from the M5 Ultra.

01:13:56   So these combos of like bin 96 gig, if you want an ultra, but

01:14:00   you're not gonna get the fancy one, but you want 96 gigs of RAM,

01:14:03   like it that's 200 more.

01:14:05   And then unbined 96 gig, the prices are the same as they were on

01:14:09   the M3 Ultra, which is nice because like M3 Ultra to M5 Ultra is

01:14:12   a big upgrade in the CPU.

01:14:13   Um, you know, four terabytes is 8300 and eight terabytes is uh 10,300.

01:14:18   So these are these are like the configs I first looked at m5 max,

01:14:22   binned uh unbined 64 and bin 36 because I have to like I'm never

01:14:26   gonna get a bin 36, but just FYI, but they make you do that.

01:14:29   Um, and then M5 Ultra, bin 96 and unbinned 96, and then I looked

01:14:34   at all those numbers and then I looked at everything over.

01:14:36   Um like I I don't know.

01:14:39   I mean, I thought about this a lot before these computers came out, but

01:14:41   it really depended on what the prices did, and having the prices not move

01:14:44   too much with the exception of a couple of 200 more expensive options.

01:14:48   Um, and again, those 200 more expensive options are for a big upgrade from

01:14:52   an M3 Ultra to an M5 Ultra.

01:14:54   Like, I'm you know, I didn't buy the the M4 Mac Studio because I

01:14:59   waited too long.

01:14:59   By the time I realized there was gonna be nothing else by the time

01:15:02   the Mac Pro was killed, so on and so forth.

01:15:04   Uh by the time I realized nothing new was coming anytime soon, it was

01:15:07   like, well, now it's too late to get an M4 Mac Studio.

01:15:10   Although I tell you, I did think about it a lot because for reasons

01:15:13   I'll get into later in the show.

01:15:14   Like I'm doing a bunch of development stuff.

01:15:17   Mac OS 27, I have to test my things on.

01:15:20   I can't do it on my Mac because Mac OS 27 doesn't run on

01:15:22   my Macs.

01:15:23   And I had 27 on the MacBook Air, and then as we discussed in

01:15:25   the earlier episode, now I've got 27 my wife's computer, so I'm using her

01:15:28   computer and she doesn't like that.

01:15:29   I love like you could solve this problem by buying a MacBook Neo, and

01:15:32   you won't do that.

01:15:34   Well, I thought, I mean, that's what I'm saying.

01:15:35   I thought about that.

01:15:36   I was looking at Mac minis.

01:15:37   What's the cheapest ARM Mac Mini?

01:15:39   I was looking at eBay, like M2 Mac minis, you know.

01:15:42   Like, I was looking for like literally anything.

01:15:43   They're not that cheap.

01:15:44   You'd be surprised.

01:15:45   I know that's what I'm getting at.

01:15:47   So, and even and even the Neo, like the Neo, yeah, it's it's like,

01:15:50   is that the cheapest R Mac?

01:15:51   Even that one increased in price, and I don't want that with a laptop

01:15:53   and they come with not a lot of RAM.

01:15:54   And if I'm gonna be doing stuff, I mean, you know, I just the

01:15:57   Neo was kind of out, and the minis like I again, I'm looking for

01:16:01   like used Mac minis.

01:16:02   Like I don't care.

01:16:03   I just needed something ARM on my desk that I could just plug into

01:16:06   my monitor and have a thing.

01:16:07   But they were just not that they're super expensive, but it's like, well, if

01:16:10   you're thinking about getting a new computer sometime soon, don't spend 800 bucks on

01:16:14   a computer you're gonna use for two months, you know what I mean?

01:16:16   Yeah.

01:16:16   But I'm kind of desperate for something ARM.

01:16:19   I need something ARM in my life.

01:16:21   Um, and the price is not changing, is like, oh, this actually kind of

01:16:24   makes it more difficult because, like, I don't know.

01:16:27   Like, I mean, a couple things make this more difficult for my decision making.

01:16:30   One, that the M5 Ultra is just two Macs stuck together, so it's not

01:16:34   interesting or novel.

01:16:35   The scaling is not better than it used to be.

01:16:38   It's you know, it's not exciting.

01:16:39   It's the kind of chip that's probably not going to get uh chip design

01:16:43   on the back of a future t shirt.

01:16:44   Just saying.

01:16:45   If they if they had done something exciting, I might have drawn another chip

01:16:49   on the back, but now it's not happening.

01:16:50   There'll still be a shirt though, don't worry.

01:16:52   Thinking I won this bet.

01:16:53   I think you did win this bet.

01:16:54   I'm shocked.

01:16:55   I'm not so uh and then I'm looking at these configs and I'm like,

01:17:01   you know, this is this goes into my well, two things.

01:17:04   One, as discussed on past episodes, I screwed up.

01:17:07   I waited too long.

01:17:08   I made poor decisions.

01:17:10   I think Marco and I have told you that for the last three years

01:17:13   running.

01:17:14   Right, sure.

01:17:14   But like it just like no, I mean, and then the thing is like

01:17:14   Oh my god, Casey, he bought a used M4 version.

01:17:20   The excuse is you know, by the time we saw the AI RAM crisis

01:17:25   coming, it was already too late.

01:17:26   You know what I mean?

01:17:27   Like by that point, is it wasn't too late to buy, like I could

01:17:29   have bought you know a Mac at that time, but it was too late

01:17:32   for me in the cycle.

01:17:32   It was like I said, it was too late to buy an M4 Mac

01:17:34   Studio because the new ones were coming out.

01:17:36   But it was also like, hey, if you wanted to save money, you should

01:17:39   have bought them before they increased all the prices, right?

01:17:41   So I was I was in a fix, and it was it's like in

01:17:44   a way that I remember everyone on the internet and YouTube are all making

01:17:47   fun of me.

01:17:47   Like, oh, you bought a Mac Pro and then they changed chip architectures.

01:17:50   Haha.

01:17:50   This thing has lasted like a champ, like longer than most people's computers last.

01:17:54   Ironically, Apple transitioning chip architectures was less of a disruption to my normal buying

01:18:00   patterns than the AI induced RAM and storage crisis.

01:18:03   Because that has screwed things up, screwed me over way more than the chip

01:18:06   architecture change ever did.

01:18:07   Do you buy enough Macs to qualify as a buying pattern?

01:18:11   Yeah, I do.

01:18:12   I mean, if you go over from 1984 till now, it you know, it

01:18:15   definitely counts as a pattern.

01:18:16   Um the pattern is I keep them for a long time.

01:18:19   That's the pattern.

01:18:21   Um yeah, I mean, this is this Mac Pro is gonna end up being

01:18:24   like six, seven years, seven years old or so by the time I get

01:18:26   something new.

01:18:27   Um, so all this is to say, kind of like with my constant uh

01:18:31   monitoring of the TV world and the uh the automotive world, this is the

01:18:36   not not the right time to buy a new computer.

01:18:38   It's the from in my from my perspective, you know, from with my rule

01:18:41   set the way I buy computers, I'm always looking.

01:18:43   Is this the right time to buy?

01:18:44   Is this the right, you know, with my TV?

01:18:45   Is this the right one to buy?

01:18:46   Should I wait till the next one, or should I do it?

01:18:48   Like, when is the right time to buy?

01:18:49   For me, this is not the right time to buy a new computer on

01:18:53   multiple fronts.

01:18:54   Even it even before you get into the RAM and storage, you know, crisis

01:18:59   with pricing, it's still not the right time to buy.

01:19:02   Apple kind of agrees that it's not the right time to buy a new

01:18:59   computer.

01:19:04   Because Apple's like, hey, we think the M7 and like the AI stuff we're

01:19:09   doing, according to the rumors, the AI stuff we're doing in the M7 are

01:19:13   so good that we are just, as Casey would say, poop canning the M6

01:19:19   Pro Max and Ultra, just so we can accelerate the timeline on the M7s,

01:19:23   because we think we did such a good job on those that we want

01:19:26   to get those out ASAP and just forget about the M6 Pro Max and

01:19:29   Ultra.

01:19:30   That that's like unprecedented.

01:19:31   And part of that is probably the RAM crisis and chip crisis and blah

01:19:34   blah blah.

01:19:34   But also, I feel like Apple, given the timelines on their chips, like the

01:19:39   whole you know, the uh ascendancy of the importance of doing local inference and

01:19:45   stuff on Macs.

01:19:46   If you extrapolate that timeline backwards for typical silicon design timelines, it makes sense

01:19:50   that the M7 is the first one that's like, oh no, local inference.

01:19:54   Like that's like that's where all we need to spend our transistors.

01:19:57   And the rumor is the M7 is really good at that.

01:20:00   This is exactly the wrong time to buy a new powerful computer because Apple

01:20:04   itself, according to the rumors, is like, yeah, our are good are good super

01:20:09   powerful computers.

01:20:09   This isn't it.

01:20:10   It's great that we upgraded.

01:20:11   It's great that you know we never had the M4 Ultra, but now we

01:20:13   have the M5 Ultra and it's fine and it's good and it's better than

01:20:15   the M3 one, and it is the best, most powerful chip we have.

01:20:18   And if you need one right now, now now this is the best we

01:20:20   have, so tough luck.

01:20:21   But the M7 would be the one that I would want to get.

01:20:24   Are you gonna wait?

01:20:25   Is it not that you're gonna buy a used one, but you're just gonna

01:20:27   keep friggin' waiting?

01:20:28   So I could clearly not choose the wine in front of me, Casey.

01:20:31   He's gonna switch to a desktop laptop and get the uh new ultra touch

01:20:34   screen because you love laptops and touch screens so much.

01:20:37   Yeah, I love them.

01:20:37   I love them so much.

01:20:39   Um so all that said, it it demands a specific strategy from me.

01:20:45   What should I do in this situation?

01:20:44   Again, I screwed up, I'm in a bind, I'm in a tight situation.

01:20:51   This is the wrong time for me personally to buy a new computer, according

01:20:54   to everything that I believe in in terms of new product buying.

01:20:57   Again, cars, TVs, computers, everything.

01:21:00   Uh, and yet somehow I need some new thing.

01:21:02   Should I go back on eBay and find one of those Mac minis and

01:21:04   buy it?

01:21:05   I was scouting, I almost like a lot of these on eBay are scams,

01:21:08   obviously.

01:21:08   So I gotta be careful.

01:21:09   But there was like a and I think it was like an M2 Ultra

01:21:13   Max Studio with a huge amount of RAM and like a like a four

01:21:16   terabyte SSD or something for like a thousand bucks or something.

01:21:19   I'm like, this has to be fake, but I was interested.

01:21:22   Right.

01:21:22   Like, not that I'm, you know, because again, you can't find a mini, like

01:21:26   you feel you feel bad buying a crappy like M2 based mini or something

01:21:29   for like 900 bucks, but then you're like M2 Ultra Max Studio for like

01:21:33   less than two grand, you know, like with a buy it now, but it

01:21:36   seemed like it was a scam.

01:21:37   So, you know, you gotta be careful out there, people.

01:21:39   And if it looks too good to be true, it probably is.

01:21:41   Uh but I I looked into a lot of that, and so here I

01:21:44   am.

01:21:44   Should I go back to eBay and find something different?

01:21:46   And let me tell you, things have not gotten better on eBay in the

01:21:48   time since since I've been making these decisions.

01:21:51   Um here's the strategy I came up with.

01:21:56   I need something that I normally never get, which is a laptop.

01:22:01   No, never, which is a short-term Mac?

01:22:06   A tied me over computer.

01:22:07   Oh my god.

01:22:08   Oh no.

01:22:09   I need something to just get me over this hump.

01:22:12   We're in a difficult period.

01:22:13   Sometimes when the chips are down, you're gotta do whatever it takes to survive.

01:22:18   Like in Goodfellas, when you're in the slammer, you could do whatever it takes

01:22:21   to survive.

01:22:21   But when you come back on the outside, you gotta leave the junk alone.

01:22:23   So Pauly says, right?

01:22:25   You know, understand you gotta do.

01:22:26   And so I'm I'm in the slammer.

01:22:28   I gotta do what it takes to survive until until my M8 or M9

01:22:35   Ultra Mac Studio, whatever arrives.

01:22:38   Because that's the computer I want to buy.

01:22:40   I don't want to buy an M5 anything Mac Studio.

01:22:43   I want to buy something that's after the M7, after that's like they really

01:22:46   figure out what these things are gonna be for.

01:22:48   Because, like, this is basically the turn here.

01:22:50   Uh, you know, unless something dramatic changes, uh, Apple is still not particularly interested

01:22:54   in gaming, but Apple seems to really be uh pointing their high performance computing

01:23:01   at AI related stuff, which makes sense in the world that we're living in.

01:23:05   And it just so happens that Apple's hardware is actually pretty good at AI

01:23:08   related stuff.

01:23:08   And I can see a future for myself where instead of paying whatever dollars

01:23:12   per month to some company for their uh coding agent, I can run something

01:23:17   that is as good as what I'm currently paying them to run on their

01:23:20   servers on my M9 based Mac Studio at some point in the future, right?

01:23:24   Obviously, by then you're not buying gun, you're not gonna buy a computer for

01:23:28   like five years.

01:23:28   Is that what we're doing getting?

01:23:29   We're getting there.

01:23:30   Oh my god.

01:23:31   But this is this is this this is what I'm thinking.

01:23:33   This is my strategy here, which is like obviously by then the ones that

01:23:37   run the servers will be better still, so I'll still probably want to use

01:23:39   them.

01:23:40   But there is a certain baseline utility of like, let's say, like uh uh,

01:23:44   you know, current informal sponsor anthropic fable five level model that you could run

01:23:50   on your Mac locally in some years in the future.

01:23:53   Again, the server ones are are gonna be better by then, right?

01:23:56   But something fable caliber locally, basically, for free, a fable caliber thing can do

01:24:03   a lot of mundane tasks coding wise.

01:24:05   Like, like not maybe your super hard ones, not maybe find the obscure bug

01:24:08   or whatever, but like just run of the mill, like write me this automation

01:24:11   script, do this thing, whatever.

01:24:13   I think that is going to be attainable locally on Macs at some point

01:24:14   in the future, but only for people who have the big honk and Mac.

01:24:19   So it's like, why do you ever buy big honk and max?

01:24:21   My answer used to be so I can run PC games at high frame

01:24:23   rates when I reboot into Windows.

01:24:25   That answer is not the case for so many reasons now.

01:24:29   Um, but my answer in the future, I think will be, and this is

01:24:32   this is a use case that was not a thing before coding agents existed

01:24:36   and were good, right?

01:24:37   Now it's like I would like to be able to run powerful coding agents

01:24:41   locally at some point in my future.

01:24:43   That's the big honkin computer I want to buy.

01:24:45   Again, M M8, M9, M10, like who knows what the future holds.

01:24:49   But I feel like in that timeline, that's the one I want to buy.

01:24:53   But I do have to buy a computer now because I can't keep running

01:24:57   Intel.

01:24:58   I have to.

01:24:59   I have to buy something.

01:25:00   So what do I buy?

01:25:02   I need a tied me over computer.

01:25:03   You bought one of Marco's Mac minis.

01:25:06   No, I he should have probably if I had known it all had all

01:25:08   these spares.

01:25:08   I should have just seen if I get the friend price.

01:25:10   That would have really helped.

01:25:11   Right.

01:25:11   I mean, we could talk about it, but you they're all base models, you

01:25:13   wouldn't want any of them.

01:25:14   Yeah, that's the problem.

01:25:15   That's the problem.

01:25:15   So, and oh, and uh, as we've discussed on past episodes, like I've got

01:25:18   the eight terabyte problem.

01:25:19   Like, I have a four terabyte SSD, and it is absolutely full, which is

01:25:22   why I made a program that saves disk space, and I use my own

01:25:24   program all the time to save me from death constantly.

01:25:28   I'm at the ragged edge of constantly every time we do an episode, it's

01:25:31   another few hundred megabytes.

01:25:32   It's not helping.

01:25:33   Like I'm so close to filling my drive, and I have a four terabyte

01:25:37   drive, and the only choice above four, if you want an internal drive from

01:25:40   Apple is you got to go to eight, and then you're mortgaging your house.

01:25:43   So, how do I deal with this?

01:25:44   Okay, so I'm guessing.

01:25:47   You already had your guess, you don't need to guess now after I explained

01:25:49   everything.

01:25:49   All right, all right.

01:25:51   All right, so here, let me just tell you what I've done.

01:25:54   I have ordered a computer and I ordered it like very early in the

01:25:57   day.

01:25:57   Because I I I had sort of gone through this process in my head

01:26:00   multiple times.

01:25:59   And once I saw the reality, I'm like, it's slotted into one of my

01:26:03   decision trees.

01:26:04   Uh I ordered myself a Mac Studio, because I'm not gonna get a laptop.

01:26:11   We're not gonna get a Mac mini.

01:26:13   I ordered myself a Mac Studio, the M5 Max, because it's a tied me

01:26:19   over computer.

01:26:21   64 gigs of RAM, which is less wait, which M5 Max, the bin or

01:26:25   unbend.

01:26:26   Well, so I got the unbind because that's the one nice thing I did

01:26:30   for myself.

01:26:31   Because you were correct, Marco.

01:26:32   I cannot you cannot abide a partial processor.

01:26:35   I cannot stomach a bin CPU.

01:26:36   If it comes with parts, they should all work.

01:26:38   Yes, exactly.

01:26:39   So and and in the grand scheme of things, going for bin to unbin

01:26:42   with my other specs wasn't that bad.

01:26:44   So yeah, the the unbid one, 64 gigs of RAM, which is less than

01:26:48   my Mac my Mac Pro has.

01:26:50   But the I mean the thing is my wife's computer has 64 gigs of

01:26:53   RAM, okay?

01:26:54   And so I'm using Golden Gate with 64 gigs of RAM.

01:26:56   It constantly has four people logged into it, all families logged into it, because

01:27:00   we're all home right now.

01:27:01   Um, and with everybody logged into it, and everybody doing what they normally do

01:27:05   on the computer, and then me using Xcode, 64 gigs, I'm able to get

01:27:08   by in with four people logged in.

01:27:10   So I think again, 96 what I got on the Mac Pro, that was

01:27:12   part of my like indulgence for myself, and I basically said it means I

01:27:15   never have to think about RAM, and it's true.

01:27:17   What I do on my Mac Pro, RAM becomes not a thing that is

01:27:20   in my mind at all.

01:27:21   I don't even know if I even touch swap.

01:27:23   So I don't need 96, unless I was doing LM stuff, which again I'm

01:27:27   not.

01:27:28   Um 64, and you know, basically, it's your only real option because of the

01:27:32   configurator.

01:27:33   And a four terabyte SSD.

01:27:35   What?

01:27:36   Okay.

01:27:37   Why?

01:27:37   Because that's what he has now.

01:27:39   Yes, but you hate it.

01:27:40   I know.

01:27:40   So here's here's the strategy.

01:27:42   Okay.

01:27:42   So first of all, let me just tell you what that that config costs.

01:27:44   Max Studio M5 Max unbined 64 gigs RAM four terabyte SSD is $5300.

01:27:44   Obviously, I'm gonna get discounts.

01:27:52   Okay.

01:27:52   Um leasing, it's $106 per month for three years.

01:27:58   And if you did that lease, at the end you'd have about $1,500 left

01:28:02   to pay on it.

01:28:03   I didn't really seriously consider the lease, mostly because I don't know how it

01:28:07   works with like Apple discounts, you know what I mean?

01:28:10   And I didn't want to navigate that, like getting this other people being nice

01:28:13   and helping me get discounts.

01:28:14   I'm like, oh, but can I also do the lease?

01:28:16   I didn't like I don't know what the deal with that is, so I

01:28:18   just didn't bother figuring it out.

01:28:19   Also, because I don't know what the delivery timelines are going to be like,

01:28:23   which is why I ordered this computer within like an hour and a half

01:28:26   of it being announced, and I did get September 22nd delivery.

01:28:29   Okay.

01:28:30   Now here's the strategy.

01:28:31   How am I gonna live with this?

01:28:32   Have I just screwed myself?

01:28:33   Yeah, you got a tied me over computer, but you haven't improved anything.

01:28:36   You're getting a computer with less RAM and the exact same amount of storage.

01:28:39   Now you're screwed.

01:28:40   Well, here's what I did because I looked at an eight terabyte option a

01:28:42   lot, but it's just it's just so much, it becomes not a tied me

01:28:45   over computer, it becomes like a you know a big beefy computer, but it's

01:28:48   not a big beefy computer that I want.

01:28:50   So, what I did was I bought a Samsung SSD, Samsung four terabyte SSD

01:28:56   for like 580 bucks.

01:28:59   God, that's so much cheaper than it is from Apple.

01:29:02   Now, that SSD used to be like 300 bucks before all this happened, so

01:29:05   it's doubled in price from what it used to be.

01:29:07   But compared to Apple's prices, compared to getting a four terabyte internal and a

01:29:11   four terabyte external from Samsung, it's like a rounding error compared to getting an

01:29:15   eight terabyte internal.

01:29:17   And my plan is photo library on the external disk.

01:29:21   So right away, because my photo library is more than half of my four

01:29:24   terabyte SSD, it's more than 50% of my drive, it's just the photo library.

01:29:28   Okay.

01:29:28   So immediately I have a less than half filled main drive, so makes a

01:29:35   less uh a roughly half-filled photos drive.

01:29:38   So I got the uh it's a Samsung 990 Evo.

01:29:41   We'll put a link in the the show notes.

01:29:43   I think there's only like one or two left in stock, but it may

01:29:45   be lower price than it used to be.

01:29:47   I just put it in the chat room if one anyone wants to buy

01:29:49   the lone remaining one.

01:29:51   Um it's not a particularly fast, fancy SSD, but you can't beat the price.

01:29:55   It was it was like 40 something percent off on Amazon.

01:29:57   That's why I got that.

01:29:58   Okay.

01:29:59   And I got uh I had to find a case for it, which is

01:30:01   way harder than finding the SSD.

01:30:03   Um I wanted a Thunderbolt 5 case, I didn't care about USB.

01:30:08   The case like you can get a case for these, like a Thunderbolt 5

01:30:12   enclosure for uh for an uh SSD that costs like 200 bucks, so you're

01:30:16   paying 600 bucks for SSD and 200 bucks for the case.

01:30:19   It's rough.

01:30:19   So I found a cheaper one.

01:30:21   I got this for like 150 bucks from OWC.

01:30:23   It's it is Thunderbolt 5, but it doesn't like the I don't either doesn't

01:30:27   have USB at all, or if it does, it's super slow.

01:30:29   I think this is the one that doesn't have USB at all.

01:30:31   So if you don't have a Thunderbolt computer, you can't use this, but I

01:30:33   don't care, it's gonna be permanently attached to the back of my Mac Studio.

01:30:37   So I don't care about it.

01:30:38   So yeah, that's my tied me over computer, and I can tell you with

01:30:42   the discounts that I ordered it.

01:30:43   I got out the door for the for just for the Mac Studio, not

01:30:46   counting like the the external SSD, I got out the door for less than

01:30:49   five grand.

01:30:50   This is my tied me over M5 Max Max Studio with the only indulgence

01:30:55   that I gave myself was an unbinned SOC.

01:30:58   Everything else is cut down to the bare minimum.

01:31:01   I think 64 gigs of RAM will be fine.

01:31:03   Again, based on my experience.

01:31:05   I think the split SSD thing will be annoying to me.

01:31:07   Again, my strategy with that is like uh the reason I always like to

01:31:10   have it all in one is because it's just so much simpler from a

01:31:12   backup perspective, but I'm just gonna have to deal with the complexity because the

01:31:14   savings are significant.

01:31:17   If I had if the only change that I had made from on this

01:31:20   computer has gone from four to eight, it becomes a $7300 retail computer from

01:31:24   a $5300, just $2,000 difference.

01:31:27   And I took you know less than like a quarter of that $2,000 difference

01:31:31   and put it towards the external drive and enclosure and everything like that.

01:31:35   So yeah, not a great time for me to be buying a computer, but

01:31:38   I did have to buy a computer, and I did buy a computer, and

01:31:41   it's not the computer of my dreams, but it is a computer, and oh,

01:31:44   here's my other strategy with this.

01:31:45   This is my wife's next computer.

01:31:47   She's got a Mac Studio now, it's fine for her now.

01:31:50   By the time the M8 or M9 is out, that probably won't be fine

01:31:54   anymore.

01:31:54   Maybe Mac OS will drop it, she's got it.

01:31:55   She's got an M1 Max Max Studio with 64 gigs of RAM.

01:31:58   Performs, does everything she needs perfectly fine.

01:32:00   It's a great computer.

01:32:01   It's faster at X code than my computer, obviously, right?

01:32:04   Like everything about it is great.

01:32:05   She has no problems with it, but someday support for it will drop.

01:32:08   This is her future computer, an M5, an unbin M5 Max Max Studio with

01:32:13   64 gigs of RAM and a four terabyte SSD.

01:32:15   Alright, that that le I was gonna try to talk you out of doing

01:32:18   this, but that last bit with replacing your wife's on that cycle, I think

01:32:21   that actually makes a lot of sense.

01:32:22   And that's also the reason that I'm not too uh too worried about not

01:32:25   having the lease, because at least would make sense.

01:32:27   Oh, it's a tie to me over computer, you should totally lease it and

01:32:29   then give it back.

01:32:29   That's perfectly what you should do.

01:32:30   But it's like, no, this computer has a future purpose in life, and it

01:32:34   will be fine for that purpose for her computer, especially if three years from

01:32:38   now one or both of my children are out of the house.

01:32:40   We'll see how that goes.

01:32:41   Yeah, I I think the the only thing I would caution you against is

01:32:47   what if your assumptions about three years from now or whenever that is, like

01:32:52   what if one of two things happens?

01:32:54   Either it doesn't go in that direction, and you know, and they they're just

01:32:59   incorrementally better with three regular years of progress, or and or what if three

01:33:06   years from now computer resources are even more expensive and the the expected fall

01:33:11   from from you know an increase in supply either hasn't happened or demand has

01:33:16   just continued to increase and so we still haven't caught up with supply.

01:33:20   I mean, I mean you kind of know the answer to that, because I

01:33:22   do feel like an M an unbinned M5 Max with 64 gigs of RAM

01:33:27   and a total of eight terabytes of storage attached to it can last me

01:33:31   a pretty long time.

01:33:32   Again, I'm on seven years of using an Intel Mac Pro that was like

01:33:37   similarly spec'd with like you know half the storage and you know ancient CPU

01:33:41   and everything.

01:33:42   I I I could use this computer, this tied me over computer, I could

01:33:45   use it for seven years easy.

01:33:46   I'm not saying you can't use it.

01:33:48   I I I'm just saying, like, if your plan right now is all right,

01:33:52   I'm gonna I'm gonna sacrifice, I'm gonna compromise.

01:33:56   John, you are not a man of compromise.

01:33:58   I am, I totally.

01:33:59   Not with your computer, you're not with your computer.

01:34:02   Here's the thing waiting, waiting is a compromise.

01:34:05   When the when I'm not buying stuff, I'm not getting fancier, faster hardware.

01:34:09   That is a compromise.

01:34:10   Like I always want the to buy at the right time to buy, but

01:34:13   in the times in between, especially at the end of it, I'm using older,

01:34:16   slower, crappier stuff.

01:34:17   That is a compromise.

01:34:18   Like, there's no there there's there's no free lunch here.

01:34:20   You're always taking trade-offs.

01:34:22   Like, I will be, I mean, if it's not three years, if it's four

01:34:24   years, if it's five years, like I'm fine with that.

01:34:27   Um, because that's what I'm forced to do in this situation.

01:34:30   Like, I I think I put it this way I would be less happy

01:34:32   with like a mini or something.

01:34:34   Because I do want something, first of all, I want it to be you

01:34:37   know, faster than any of the Macs that we have, faster than like because

01:34:41   I do feel like the M5 Max is gonna be faster than you know,

01:34:44   this is the other thing.

01:34:45   People are like, why didn't you just get a laptop at that point?

01:34:48   My my own brother even said to me, You what you end up buying

01:34:51   this computer, you could have just bought a laptop like three months ago.

01:34:53   Yeah, but here's the thing.

01:34:55   If you if I was to get the 14 inch version of this, same

01:34:57   specs, that's $800 more.

01:34:59   And if I was to get to 16 inch, it's uh $1,100 more.

01:35:02   Obviously, because they have screens and keyboard, like it makes sense that they cost

01:35:05   more, but there's no way in hell I'm gonna pay even $800 extra for

01:35:09   a screen, keyboard, and trackpad that I despise.

01:35:12   Not the screen, the screen is good, but like whatever.

01:35:14   I was never gonna get a laptop.

01:35:15   So with me getting this M5 Max, it is the fastest, especially if you

01:35:20   count sustain performance Mac that you can buy that's not the ridiculous Ultra.

01:35:24   And so I think I'll be happy with it in that regard.

01:35:26   I mean, and part of the the reason I I know I'll like this

01:35:29   is using my wife's M1 Max and X code, I'm like, wow, this computer

01:35:32   is fast.

01:35:33   Like because I'm I'm using my Intel computer.

01:35:36   Like certain things in Xcode takes so much longer.

01:35:38   Um again, my Mac Pro is slower than an iPhone, right?

01:35:42   So I I'm I'm sure when I get this, I will be overjoyed and

01:35:46   everything will be fast.

01:35:46   And the store, that's the other reason I didn't want the M4 model, because

01:35:49   I knew the M5 models have faster storage.

01:35:51   The storage is faster.

01:35:52   I don't care about the ultimate speed of the photo library.

01:35:55   That's more about bulk storage.

01:35:56   So I I think this will work out.

01:35:58   Yeah, I guess I I guess my my worry here is that your tied

01:36:04   me over computer is just one of your computers.

01:36:07   Like this is just gonna be your computer for the next X years.

01:36:10   And if X is, you know, more than a few, uh I feel like

01:36:15   you might you might be disappointing your future self by compromising now in ways

01:36:20   that if you were buying it to be your your next six to seven

01:36:23   year computer, you would not necessarily want those same compromises.

01:36:27   Well, so here's the thing.

01:36:30   Even though I only have one child in college, I still do have two

01:36:33   and a half more years of paying for that college.

01:36:35   And her college is more expensive than my son's college.

01:36:39   There are other expenses in my life that make it not the right time

01:36:42   for me to buy a eight to ten thousand dollar computer or anything like

01:36:45   that.

01:36:45   As much as Casey thinks I would love to get that computer, like it's

01:36:49   just it's just not financially responsible for me to be doing that.

01:36:52   I'm paying for a whole bunch of things, especially like in these, you know,

01:36:56   this part of the year, and it's just maybe, you know, it's just it's

01:36:59   not it's not the right time for multiple reasons.

01:37:01   So this compromise is something I have to do.

01:37:04   And if I if I use this computer for five years and then replace

01:37:07   it, I will still consider it a tied me over computer.

01:37:09   If I use it for seven, I'm like, well, tidy me over lasted longer.

01:37:12   But I still think I'll be I won't be as in dire streets as

01:37:15   I am with this Mac Pro now in seven years, because seven years from

01:37:18   now, first of all, the M5 Max is not in any danger of being

01:37:22   obsoleted or no longer supported by Mac OS because it is basically their second

01:37:26   newest chip that they have, or third newest, I guess, because you've got the

01:37:29   ultra and the M6.

01:37:31   Um, and it'll certainly have enough power, and I don't I expect there not

01:37:35   to be another architecture transition in the next five to seven years.

01:37:38   So yeah, this basically gives me the leeway to wait.

01:37:43   And like if the M7 comes out and it's not exactly what I want,

01:37:45   I'll be like, uh, I'll wait to see what the M8 does.

01:37:48   And if the M8 isn't what I want, I'll wait until the M9.

01:37:50   Once the N10 comes out, I'll be like, well, maybe I have to buy

01:37:52   this one.

01:37:53   But it's not ideal, but sometimes you gotta do what you gotta do, and

01:37:56   this is what I have to do right now.

01:37:59   I am quite shocked, to be honest with you.

01:38:02   I'm not saying that you're making the wrong choice.

01:38:04   I I I don't love the way you're handling storage, but I do understand

01:38:09   what you're saying, and it is certainly the more frugal, well, not even frugal,

01:38:12   it's the it's the more intelligent approach, I would argue, is to do what

01:38:16   you're doing.

01:38:14   Uh, I just I I wouldn't want to be bothered.

01:38:20   And if it doesn't work out and like having the split storage and third

01:38:22   party external storage drive ends up being a disaster, it'll be good content for

01:38:26   the show.

01:38:27   That is true.

01:38:28   Uh no, I think this does make a lot of sense.

01:38:30   I am very surprised this is the route you're going, but I think it

01:38:33   makes a lot of sense.

01:38:34   So you think sitting here today, and obviously you don't know what what the

01:38:38   future will bring, but you think this is a two to three year computer

01:38:42   would be your guess?

01:38:43   No, it's at least three years.

01:38:44   I would say three to five is my guess, somewhere between three and five.

01:38:47   Um, and if it goes to seven, then I will have uh uh miscalculated.

01:38:50   But uh, you know, like I said, even if it goes to seven, I

01:38:53   think I'll be fine because I won't be in a crisis like I am

01:38:55   now.

01:38:55   You know what I mean?

01:38:56   Like in seven years, like it's not gonna be like, oh my god, I

01:38:59   need a new computer, I can't get anything done in this.

01:39:00   I'm using my wife's computer, it's too slow, it's not like I won't be

01:39:03   like that in seven years.

01:39:04   It'll just be like, oh, I guess this wasn't a tide me over after

01:39:06   our.

01:39:06   But like that's that's kind of my line of between what's a tide me

01:39:09   over and I and honestly, after the you know, the discontinuation of the Mac

01:39:13   Pro, like that dream ended with the Mac Pro, and like any computer I

01:39:17   was gonna get was gonna be like, well, I'm forced to get the lesser

01:39:21   offerings.

01:39:21   Like the same thing with like the M2 Ultra Mac Pro.

01:39:23   Like that would have I would have considered that a weird compromise and a

01:39:26   lesser offering that I never would have bought anyway, because it's the same as

01:39:28   the studio for my process, right?

01:39:30   So I'm kind of out of that phase for a while.

01:39:32   And like we said before, with your believe shirt, hey, in a decade, if

01:39:35   they reintroduce the Mac Pro, maybe I'll be back in business.

01:39:38   But I'm it's I'm in a down period now.

01:39:40   I'm uh like it's like the rebuilding decade for like sports teams.

01:39:43   It's like, yeah, yeah, you're not gonna win any championships anytime soon.

01:39:46   So just hunker down, do what you gotta do, and then wake up in

01:39:48   a decade and see if Apple has changed its mind about high-end computing.

01:39:52   Wow.

01:39:53   Well, I'm shocked.

01:39:54   I am genuinely shocked.

01:39:56   Clearly, Marco gets the win on this one, even though both of us were

01:39:59   wrong.

01:39:59   I feel like Marco was much closer to the right.

01:40:01   He had the right instinct, but he just didn't lean as heavily into how

01:39:59   much this is not the right time for me to buy a computer.

01:40:07   Like how much I feel that, how much this is not it's wrong, it's

01:40:10   the wrong time.

01:40:11   Yeah.

01:40:11   I I don't know if I can do the rest of the episode.

01:40:13   I am truly shocked.

01:40:15   But I will soldier on for you, the listener.

01:40:17   Uh, we do have a few more things that were announced.

01:40:19   There are new keyboards.

01:40:20   Uh Mark Ehrman writes.

01:40:22   Apple also launched new Mac external keyboards, but they're virtually identical to the prior

01:40:26   ones.

01:40:26   Minor labeling changes on the keys to match the recent MacBook Neo or Air

01:40:30   updates.

01:40:31   Uh John, you've done some science and some research.

01:40:34   The old one is the which I th I just bought like a month

01:40:37   ago, maybe two months ago, is uh mxk83LL slash A.

01:40:43   And you can tell that that's the one because the home button, this is

01:40:47   for the um extended keyboard.

01:40:48   The home button has the word home on it, uh page up and page

01:40:52   down, have the words page up and page down, and so on.

01:40:56   The new one is MJM 74 L L slash A, where the home end,

01:41:01   page up, page down, among other things, are all symbols now.

01:41:05   Um, this is the magic keyboard with touch ID and numeric keypad for Mac

01:41:09   models with Apple Slocken, USB C US English black keys.

01:41:13   We'll put a link in the show notes.

01:41:14   I'm I don't uh I'm assuming this is what they did on the Neo.

01:41:17   Like again, they added FN on the FN key, whereas like the older model

01:41:20   doesn't have it.

01:41:21   It's just got the globe.

01:41:22   Uh both of these don't have control in the lower left corner.

01:41:25   They have the stupid globe slash FN key there.

01:41:27   I hate, really hate that they don't say home end and page up and

01:41:31   page down anymore.

01:41:32   Because first of all, the home and end symbols, it's like they don't read

01:41:35   as home and end to me.

01:41:37   But second, page up and page down are up and down arrows.

01:41:39   And I know they don't use arrows on the arrow keys, they use triangles,

01:41:42   but it's like that doesn't say page up or page down.

01:41:45   If someone looked at that keyboard, they'd be free to be forgiven for thinking

01:41:48   that was the up and down arrow, but it's not just below it is

01:41:44   the actual up and down arrow.

01:41:52   I think this is a big regression.

01:41:53   I don't like the symbols.

01:41:54   I I mean maybe this hap uh helps with like internationalization or whatever.

01:41:59   Like clear doesn't say clear anymore.

01:42:01   It has like a a box with an X in it.

01:42:03   It looks like someone typed a like the wrong Unicode character, or it's like

01:42:06   the character encoding is like there's no character in this font.

01:42:08   Like it's just shift doesn't say shift, it's just got the up arrow.

01:42:12   Like it's just seems like a savings for money where like they don't have

01:42:15   to put words on it, and there's I don't like it at all.

01:42:18   But I will reiterate as we discussed in the past episode.

01:42:21   I in anticipation of me getting a new computer, saw this when it came

01:42:25   up recently.

01:42:25   I was like, oh yeah, they screwed up the keyboards, didn't they?

01:42:27   They got rid of the in particular.

01:42:29   I can't stand not having the control key in the lower left.

01:42:31   I mean it's there, but it's it's half size and it's not in the

01:42:33   corner anymore.

01:42:35   So I bought on eBay uh the model that I like, which is MMR3LL-A

01:42:41   or slash A.

01:42:43   Uh it's the A2520 model, but that's the particular uh Apple part number for

01:42:47   the particular color and blah blah blah.

01:42:49   But the point is an extended keyboard, it's got touch ID.

01:42:52   The control key is full size, the full double width thing in the lower

01:42:55   left corner, and it says home end and page up and page down on

01:42:58   it.

01:42:59   That's what I want, and that's what I got.

01:43:00   It's sitting here next to me, sealed in the box when my Mac Studio

01:43:03   comes.

01:43:04   That is what I'm connecting to it.

01:43:05   But that's lightning, isn't it?

01:43:06   Or is that also USB C it's lightning on one end, but it's USB

01:43:10   C on the other, and then the Mac Studio doesn't know or doesn't care

01:43:12   what's on the other end of the wire.

01:43:14   That's gross.

01:43:15   And I can use it through Bluetooth if I wanted, but I don't.

01:43:17   Yeah, it's gross.

01:43:18   I don't love it, but I mean I hear I hear the logic there.

01:43:20   I don't love it.

01:43:21   Especially if you're using it wired, it doesn't really matter.

01:43:23   Yeah, it's a desktop computer.

01:43:24   It's like I don't the wire is not impairing my keyboarding, believe me.

01:43:28   And then we have possibly the biggest news of this entire episode.

01:43:29   There is a new polishing cloth.

01:43:33   Emma Roth at The Verge writes Apple just dropped a new polishing cloth, and

01:43:36   unlike its refer refreshed Mac mini models, it comes with price cut, as spotted

01:43:40   earlier by Mac Rumors.

01:43:41   The polishing cloth is still made with soft, non-abrasive material, but costs $9 instead

01:43:46   of 19.

01:43:46   Doesn't look like anything else has changed as the description matches the archived version

01:43:50   of Apple's old polishing cloth, which it no longer lists on its website.

01:43:54   And we'll put a link to both The Verge and Mac Rumors in the

01:43:58   show notes.

01:44:10   That's my complaint about Apple's polishing cloths.

01:44:12   Like I got on with my Pro Display XDR, and they were just so

01:44:14   stiff.

01:44:15   Like who you how do you even use this?

01:44:16   It's just like it, you know, normally you take a polishing cloth and like

01:44:20   you wrap it around your fingers or hands so like you're you're touching the

01:44:23   screen with a small area so you can get some force on it.

01:44:25   But this thing was so stiff, it's like you have to put the entire

01:44:27   pad against the screen or something.

01:44:28   It's just it's weird.

01:44:29   I'm not a fan of the polishing cloths, but yes, they continue to update

01:44:32   them and sell them, apparently.

01:44:33   Uh I would be remiss if I didn't mention uh our friend Quinn Nelson's

01:44:37   shop, Copeland.supply, not a sponsor, but a friend, uh, where he has the cloth

01:44:42   pro max for $18, which is uh very, very good.

01:44:45   I have one and it's great.

01:44:47   He just got undercut by Apple in price, though.

01:44:48   Yeah, sure did.

01:44:49   Because he was 18 and they were 19.

01:44:50   Look, it's cheaper, but now they're down to nine.

01:44:52   You know, there's one thing getting cheaper on Apple's website, it's the polishing cloth.

01:44:55   Well, it doesn't have any RAM.

01:44:57   That's true.

01:44:58   All right, thank you to our sponsor of this episode, Tail Scale, and thank

01:45:01   you to our members who support us directly.

01:45:03   You can join us at atp.fm slash join.

01:45:06   One of the many perks of membership is ATP Overtime, our weekly bonus topics.

01:45:10   So in addition to all those specials that we talked about earlier, we also

01:45:13   do a weekly bonus topic after every episode, a little bit more that we

01:45:14   just couldn't fit into the main show.

01:45:18   This time we're we're going to be covering um a recent benchmark to suggest

01:45:22   uh is Mac OS slower than Linux when performing file system operations.

01:45:26   Join underlisten, HBZ FM slash join.

01:45:29   Thank you so much, everybody.

01:45:30   We'll talk to you next week.

01:45:35   Now the show is over.

01:45:37   They didn't even mean to begin, cause it was accidental.

01:45:43   Oh, it was accidental, accidental.

01:45:46   John didn't do any research.

01:45:48   Marco and Casey wouldn't let him, cause it was accidental.

01:45:52   Accidental.

01:45:53   It was accidental.

01:45:56   And you can find the show notes at A T P dot F M.

01:46:01   And if you're into Mastodon, you can follow them at C-A-S-E-Y-L I S S.

01:46:10   So that's KC Lis.

01:46:12   M-A-R-C-O A-R-Marco Armin.

01:46:17   S-I-R-A-C USS.

01:46:22   It's accidental.

01:46:24   Accidental.

01:46:29   Accidental.

01:46:30   Check podcast so long.

01:46:35   John, you've been busy apparently, and I genuinely know almost nothing beyond that.

01:46:39   So what have you been up to?

01:46:40   Yeah, I apologize in advance for consulting a casey like wind up for this,

01:46:44   but I feel like it's important for conversation.

01:46:45   Alright.

01:46:46   Calm down.

01:46:47   It's just that's uh that you said point of pride.

01:46:49   It's just uh anyway.

01:46:51   So uh we'll link in the show notes to my blog posts about uh

01:46:54   my three uh existing Mac apps, front and center, switchglass, and hyperspace.

01:46:58   And if you read through those blog posts, like they're like the announcement posts

01:46:59   of like, here's this app and here's why I made it and stuff.

01:46:59   And with the exception of hyperspace, the especially the explanation for front and center

01:47:08   and switchglass, it's basically like um I used to use a bunch of apps

01:47:13   on my Mac, and then uh you know, well, and one particular drag thing

01:47:17   by James Thompson.

01:47:18   It stopped being developed, and I kind of needed the functionality was provided because

01:47:23   I'm so used to using my Mac with that functionality, and it was hard

01:47:26   to find in other apps, and so I'm like, well, I mean, that's what

01:47:28   got me to be doing Mac development in the first place.

01:47:30   That's what finally got me to do it after all these years.

01:47:32   It's like no one is making this weird little app that I want.

01:47:35   I'm gonna make it myself.

01:47:36   I used to use drag thing for this, drag thing use no more.

01:47:39   I used to use a bunch of other apps for it, they're also gone.

01:47:41   I'm gonna have to make this myself.

01:47:43   So front and center was the first one, it's like the simplest little app

01:47:45   you can possibly imagine.

01:47:46   And then switch glass was the second one.

01:47:48   And drag thing, uh, for people who don't know back in the day, both

01:47:51   on classic Mac OS and on uh Mac OS 10.

01:47:55   Um, let you put make like these little palettes on the screen with stuff

01:47:59   in them, and one of the kind of palettes you could make was an

01:48:01   application switcher palette, and also just as a side, inside the drag thing settings,

01:48:06   there was like a checkbox that said, and by the way, make Mac OS

01:48:10   10 look work like classic Mac OS when clicking on Windows.

01:48:13   So the features of front and center used to be in drag thing, and

01:48:16   the features of switch glass used to be in drag thing, and drag thing

01:48:19   did tons more stuff too.

01:48:20   But I'm like, these are the features that I miss the most from drag

01:48:23   thing.

01:48:24   I miss the front and center thing, which is when you click on a

01:48:27   window for another app, what happens is does it bring just that window to

01:48:29   the front or all the windows?

01:48:30   They changed that behavior in Mac OS 10.

01:48:32   I could never get used to it.

01:48:33   Both behaviors are useful, but I wanted to have the option to use one

01:48:36   or the other and switch between them by shift-clicking, which is what drag thing

01:48:40   gave me.

01:48:40   So I made front and center.

01:48:41   Switch glass, I wanted a little application switcher.

01:48:43   Why do you want that?

01:48:44   Don't you just use the doc?

01:48:45   It's just, I just could not live without it.

01:48:47   I just got used to it.

01:48:48   The doc has stuff besides other apps in it.

01:48:50   It's got the trash, it's got folders.

01:48:52   I do use those features.

01:48:53   I just wanted an app switcher.

01:48:54   So I'm running switch glass with my app switcher in the upper right.

01:48:57   And the dock also visible.

01:48:58   It's just the way I want to do it.

01:48:59   These are the weird little apps I made.

01:49:01   But you can see how the first two apps I made were just entirely

01:49:03   motivated by drag thing goes away.

01:49:06   Blame James Thompson because I needed that functionality.

01:49:10   And they were just very focused.

01:49:14   You know, drag thing just had a checkbox in there.

01:49:16   Why not just make front and center a checkbox and switchglass?

01:49:18   I'm like, no, I just want little turdy apps that just do one particular

01:49:21   thing.

01:49:21   They're paid up front, there's no subscription.

01:49:23   You buy them once.

01:49:24   People who bought front and center five years ago for eight bucks or five

01:49:27   bucks or whatever I was charging for it, then they just have this app

01:49:30   forever.

01:49:31   And switch glass, it's just an application switcher.

01:49:34   Uh and it's ridiculously configurable in appearance because this is what I wanted, but

01:49:38   whatever.

01:49:38   People probably just use the defaults and probably nobody uses it but me, but

01:49:41   I like it.

01:49:42   And I I run these apps 24-7.

01:49:43   When I'm using my Mac, front and center and switch glass are running.

01:49:46   And hyperspace, of course, I'm using to try to save my disk space, which

01:49:48   is and hyperspace was like, hey, this is my first real app, it's not

01:49:52   super weird because it's like nobody wants front and center in SwitchGlass except for

01:49:55   me.

01:49:56   Okay.

01:49:56   But hyperspace, that's potentially useful to a lot of people.

01:49:59   A lot of people have want to save disk space without deleting files.

01:50:02   So I'm like, finally, I'm out of this stupid corner of making weird little

01:50:07   apps that are just for me.

01:50:09   Hyperspace, even though it is still, you know, it's a Mac app, first of

01:50:12   all.

01:50:12   So right away I'm eliminating most of the population of the planet because it's

01:50:15   just the way it is.

01:50:16   Um, and also it's a pretty nerdy Mac app, but it's not as you

01:50:21   know, narrowly focused as front and center in Switch Glass.

01:50:23   So I thought I had gotten out of the the dark alleyway of making

01:50:28   tiny little apps that only satisfy some need for me.

01:50:32   But I I when was switch glass out?

01:50:29   I guess wow, six years ago, 2020.

01:50:37   Jeez.

01:50:37   Anyway, switch glass apparently has been out for six years, and during that time,

01:50:42   lots of people would have feature requests for you should add this, you should

01:50:44   add that.

01:50:45   And I was like, no, it just switches apps, it's right in the name.

01:50:47   It just it's a switcher.

01:50:49   Yeah, I could add other stuff, but I'm not going to.

01:50:51   But here's the thing there's one more feature from drag thing that switch glass

01:50:56   doesn't do because it's narrow and focused that I miss.

01:50:59   So I basically just use the app switcher, but I also had like another

01:51:02   drag thing palette that had like some folders in it.

01:51:06   And in the Mac OS 10 age without drag thing, what I do instead

01:51:10   is I put folders on the lower left corner of my desktop.

01:51:12   I generally keep a mostly clean desktop, but I always just have like permanently

01:51:16   like the stupid on-screen climate controls and your touchscreen EV permanently in the bottom

01:51:21   left corner of my screen.

01:51:22   I have a set of folders.

01:51:25   Um, but you know, so unlike a drag thing palette or something else, the

01:51:29   desktop gets covered sometimes, and you can't get to the folders.

01:51:32   And I was like, Yeah, I should just extend switch glass with a feature

01:51:36   that says, Oh, and by the way, you can also make a little palette

01:51:38   with some folders on it and put it somewhere else.

01:51:41   And I I like it basically implemented this, like in a in a branch

01:51:44   of switch glass, like I don't know.

01:51:46   Again, I I'm so bad at guessing.

01:51:47   If you'd asked me how old is switchglass, I would never have guessed six

01:51:50   years, but probably three or four years ago, I implemented in switch glass this

01:51:54   feature.

01:51:55   It's like, oh, and you can also have one other palette for the hell

01:51:57   of it that you can put some folders in and click them and they

01:52:00   open the folders and finder.

01:52:01   Like, that's it, that's the whole freaking feature.

01:52:03   Okay.

01:52:03   But I was like, oh, what it was like switch glass is supposed to

01:52:06   be focused.

01:52:06   I don't wanna add this on.

01:52:09   And like, how confusing is it gonna be?

01:52:11   It's like, so you have an apps you switch glasses and app switcher, and

01:52:14   also a little place where you can put folders.

01:52:16   Like, what the hell?

01:52:17   Like, it doesn't make any and like how do I fit it into the

01:52:19   preferences?

01:52:19   Do I make another dialog box for it?

01:52:21   How do I explain the functionality?

01:52:23   Like, it's incoherent.

01:52:25   And so, even though I implemented it, I'm like, this is not the right

01:52:28   place for switch class.

01:52:29   Also, I don't know if you remember, but switch glass again, 2020, it's Swift

01:52:33   UI, like the view itself is Swift UI.

01:52:35   But Swift UI in 2020 on the Mac was garbage, it didn't do anything,

01:52:40   it's still not great.

01:52:41   But like all of like the drag reordering of like collection views and stuff

01:52:46   that you take for granted in like UI kit and app kit and stuff

01:52:49   did not exist in any way.

01:52:51   So I hand rolled every single part of that.

01:52:53   It's like an entirely custom UI.

01:52:56   Hand rolled at like great pain and expense to look and behave exactly the

01:53:00   way I want it to.

01:53:02   And this is this is like this is ancient Swift UI code.

01:53:05   This predates at observable.

01:53:06   It is like, remember at state object versus at state?

01:53:10   Remember at published?

01:53:11   That's what switch glass is.

01:53:12   Yeah, it is the oldest, creakiest uh Swift UI code you've ever seen.

01:53:17   It supports OSs like back really far.

01:53:20   Okay, it was like obviously it's some ordered the OS that was that was

01:53:23   new in 2020, but it supports even back even farther.

01:53:25   Like, and I don't want to cut off support for that.

01:53:28   It's like it's like you should modernize switchclass, it would be so much easier

01:53:30   to write with modern APIs, but like I don't want to just like abandon

01:53:33   the old OS's at this point, and maybe I should eventually anyway, but like

01:53:36   that still doesn't solve the problem of where to add this functionality.

01:53:38   So I was just stuck.

01:53:39   I was like, I'll just go back to the folders on the desktop.

01:53:43   Like, I've I've been using their little folders in the corner of my desktop

01:53:45   forever, I'll survive without it.

01:53:47   And here's the thing: folders in the corner of your desktop, nobody wants that.

01:53:50   Like, that's even worse than front and center.

01:53:52   It's like you're gonna make that a separate app, and you're not gonna put

01:53:55   it into SwitchClass, you're not gonna modernize switch class.

01:53:57   Switchclass like does what you wanted to do.

01:53:59   Why would you go in there and add new bugs by modernizing and cut

01:53:59   like so?

01:54:02   I was just stuck.

01:54:04   Along comes the magic of vibe coding.

01:54:07   I'm like, this is the solution to my problem.

01:54:09   I'm just gonna vibe code a new modern app to make a little palette

01:54:13   with like five folders in it and stick it in the corner of my

01:54:16   screen.

01:54:17   That'll be my solution.

01:54:18   No one else never needs to use this app, it'll be just for me.

01:54:20   I won't muck up switch glass, it's not gonna screw up like my the

01:54:25   purity of my apps, it's not gonna be hard to explain.

01:54:27   No one else even needs to know it exists, and it will solve my

01:54:29   problem.

01:54:30   And you know, and I know this is totally within the realm of vibe

01:54:32   coding because how hard is this?

01:54:33   It's the easiest thing in the world.

01:54:34   So I did.

01:54:35   I vibe coded two seconds, like, okay, there you go.

01:54:37   It's a little thing, and you put make a little palette and you put

01:54:40   some folders in it, and it's great.

01:54:43   Um then you know, because vibe coding is fun, and you're like, oh, you

01:54:47   know, went off and did it and did some other things, then you argue

01:54:49   with it about stuff and blah blah blah, and then eventually, all right.

01:54:51   So it's fine.

01:54:52   You're like, you know, I've got some other ideas that I always was interested

01:54:55   in trying.

01:54:56   Uh as I was like, I was doing this in Mac OS 27, so

01:54:58   I could use all the fancy new APIs.

01:55:01   And I told the uh the my little uh coding agent, I said, you

01:55:06   know, the materials that you get in Mac OS that I give in SwitchClass

01:55:09   and everything, of like the various like different tints of backgrounds and stuff, like

01:55:13   in the and in Tahoe, they added clear glass and regular grass and tinted

01:55:17   glass to add to that collection.

01:55:19   None of those materials is what the doc looks like.

01:55:22   Like Apple does something custom for the doc.

01:55:24   So I'm like, oh, coding agent, I know you love to decompile.

01:55:27   Figure out how the doc oh my gosh, what it does, and then make

01:55:32   me a material that does what the doc does.

01:55:35   And it went off and did its thing, and it found like there's like

01:55:39   these recipe files and materials files in the doc, and there's also a whole

01:55:42   bunch of private APIs, as you could imagine, that only the doc uses that

01:55:45   are not available, you know.

01:55:46   Like, and it found all this stuff, and it was like, hey, uh, I've

01:55:44   implemented the doc material, and it was close, and I was like, it's not

01:55:53   quite the same.

01:55:54   And I, you know, I worked out this whole thing of like, here you

01:55:56   can tell whether you're exactly the doc.

01:55:57   Like, make it make it just like the doc.

01:55:59   And eventually it did.

01:56:01   Obviously, using private APIs and all sorts of like you know, unholy things that

01:56:04   would never fly in the Mac app store or whatever.

01:56:06   Um, but now I had this modern app uh with modern Swift UI that

01:56:12   gave me the little feature that I wanted that I could also make it

01:56:14   look like the doc, which was fun.

01:56:16   And then I got a little bit mad with power, and I was like,

01:56:19   but that's the doc in macOS 27.

01:56:20   The doc looks different in Mac OS 26, and it looks different again in

01:56:23   Mac OS 15.

01:56:25   So I need to make like a little toolkit to let it uh figure

01:56:28   out what how the doc worked on all those other computers and simulate those,

01:56:31   and so I did.

01:56:32   I said, Hey, do that.

01:56:33   I want to be able to pick doc on the current OS doc on

01:56:36   macOS 27, doc on mac OS 26, doc on mac OS 15.

01:56:40   This is the fun of vibe coding because I'm like, I'm not sitting in

01:56:43   front of the computer while it's doing this, I'm like doing other things, and

01:56:45   it did that.

01:56:46   I was like, um, what am I doing here?

01:56:49   This is supposed to just be like my toy project for me, and I

01:56:52   already have what I want out of it, and all this cool doc stuff.

01:56:55   Like, no one's ever gonna see this.

01:56:56   Like, I can't put this on the Mac app store, obviously, it's all private

01:56:58   APIs, nobody cares.

01:57:00   Like, basically, what like what the L the what these LM coding agents did

01:57:03   is they made me like one ten thousandth of a g Rambo, you know

01:57:07   what I mean?

01:57:07   Like, because this is this this is not a skill that I have.

01:57:10   Like, I can't I can't reverse engineer these things, I don't know how to

01:57:13   go spelunking in their files and figure it out.

01:57:15   He could do it, and obviously the LLM can do it somewhat, but like

01:57:19   this is just like a glimpse of what it's like to have like again

01:57:22   one ten thousandths of the skill of someone who actually knows how to do

01:57:24   this stuff, but it was fun.

01:57:26   It was a thing.

01:57:26   Here's I'll put it this way: it's a thing that I personally could not

01:57:30   have done without a coding agent because I don't have those particular skills at

01:57:29   all.

01:57:34   And yeah, you could learn them, but like I feel like at this point

01:57:37   in life, like I'm never gonna become one 10,000th of a gear Rambo.

01:57:40   I've just I've got the coding agent for that.

01:57:42   But it was fun, and I had this thing, and I just didn't know

01:57:44   what to do with it.

01:57:45   And so like I gone a couple of direct different directions here.

01:57:47   I could say, Well, that was fun, I'll use it myself, and it'll be

01:57:50   fine.

01:57:51   But like at that point, I felt like I had built up a significant

01:57:57   amount of functionality.

01:57:58   In particular, I did like I didn't tell you how I started this project,

01:58:00   but the way I started it was I said, Hey, coding agent, here's the

01:58:04   the source code to switch glass.

01:58:06   Do that, but over here and use modern Swift UI.

01:58:10   So I basically I'm pointing it to my existing, all handwritten, you know, ancient,

01:58:14   all handwritten code, and I'm saying, see that do exactly that.

01:58:19   Like, I want it to look like that, I want it to behave like

01:58:22   that, I want it to everything, but don't use you know, at published and

01:58:25   at state object and all this other crap.

01:58:27   And I said, and if you can use the modern like reordering APIs in

01:58:30   Mac OS 27, you should do that.

01:58:32   Um, I had to support back to 15 because I wanted to use it

01:58:35   on my Mac and my Mac is stuck on 15.

01:58:37   Um, so I said, support back to 15.

01:58:39   So you can't use 27 only APIs entirely, but use them if you can.

01:58:43   And anyway, I told it to do that, and that was that was an

01:58:46   interesting, just a brief aside on that.

01:58:47   That was an interesting LLM experience because you're like, I'm pointing it at my

01:58:49   own source code, and I'm saying, basically, copy that.

01:58:52   Like, I've already figured it out, and like it would take a cut at

01:58:54   it, and then I would try it.

01:58:56   And like, I don't know if it's a context window thing or something, but

01:59:00   it was like you missed all of like the performance enhancements.

01:59:04   I spent so much time tweaking my stupid switch glass thing, so it has

01:59:07   decent performance on slow old Intel Max.

01:59:09   And the like, and I'm I'm developing it on M1 Max.

01:59:12   So the first time I brought it over to my Intel Mac and I

01:59:14   like resized the icons, it was like a half a frame per second.

01:59:19   Like, what the hell LM?

01:59:21   Like, I gave you the source code to a thing that performs amazingly on

01:59:24   Intel CPUs, and you missed all of my optimizations and did like the dumbest,

01:59:28   most naive thing.

01:59:29   So I had to go back to the computer and be like, hey, look

01:59:32   at this file in this area.

01:59:33   See where I'm doing this.

01:59:34   I was just like, you're not I did this already.

01:59:36   Just copy it.

01:59:37   Just like it's just anyway.

01:59:39   You can't tell it's like a black box, but it didn't like it got

01:59:42   the gist of what switch glass does.

01:59:44   But I even given the source code, it did not literally copy all of

01:59:47   my optimistics.

01:59:48   So I had to go back in for each thing and say, do this

01:59:50   like this, do this like this, do this like this.

01:59:52   Like, yes, copy all these optimization, a file at a time.

01:59:55   Look in this file, that's where this thing is.

01:59:56   Look in this file, that's where this thing is.

01:59:57   The animations weren't as smooth or is good.

01:59:59   Look in this file.

01:59:59   See when I did this, see this reason I did this over here.

02:00:02   Like, it's all in the comments, and this anyway.

02:00:04   Eventually I got it to work like switch glass, and it just it just

02:00:07   felt like that.

02:00:09   I had something that seemed like it might have some kind of value besides

02:00:12   just this weird thing that works for myself.

02:00:14   But of course, it's all private APIs and crap.

02:00:16   So, what am I supposed to do with this?

02:00:17   And so I said I could have gone two directions, which is one, just

02:00:20   keep it for myself and whatever.

02:00:22   But the other direction I felt like I could go in is what if

02:00:26   I turn this into an app that basically supersedes switch glass, that it does

02:00:32   everything that switch glass does, but with modern APIs, but also makes it have

02:00:38   little folders in the corner or whatever.

02:00:39   Like it basically a general purpose way to make a bunch of palettes around

02:00:43   your screen.

02:00:44   And I know you're thinking, are you just making drag thing?

02:00:47   Is that what you're doing?

02:00:48   There's the whole thing of like you're just gonna reinvent drag thing, you're just

02:00:50   keep making Mac apps till you reinvent drag thing.

02:00:53   I think I'm not doing that.

02:00:56   James Thompson can correct me if I'm wrong.

02:00:58   Yet, I think I'm not doing that.

02:00:59   What do you think, Casey?

02:01:01   Next year.

02:01:01   Yeah, seriously.

02:00:59   But I'm really uh I'm really approaching it.

02:01:04   Because once you can make a bunch of arbitrary palettes all around your screen

02:01:07   that have different things in them, that's basically drag thing.

02:01:11   The only difference is the drag thing could let you put tons of stuff

02:01:13   there and do all sorts of interesting things.

02:01:15   And I'm doing a much smaller set of less interesting things, at least for

02:01:20   now.

02:01:20   So I just decided I have to make this into an app.

02:01:23   And also, by the way, this is related to the computer that I just

02:01:27   bought.

02:01:27   My tradition seems to have been my tradition, like my one-time tradition, which is

02:01:32   buy a really expensive computer and then make a bunch of Mac apps that

02:01:35   eventually pay for that computer over the course of the next several years, which

02:01:37   is what I did with my Mac Pro.

02:01:39   That's my plan here is like make a new app and try to pay

02:01:43   for my computer over the next three to five years.

02:01:45   In terms of because that's the kind of income I things make.

02:01:48   Your map apps make enough to pay for a computer, yeah, over the course

02:01:51   of several years.

02:01:52   Like, that's what we're talking about here, okay.

02:01:55   And so that's my plan.

02:01:56   I'm like, I'm gonna ship this as an app, and it's I it's gonna

02:01:59   be an app that lets you put a little bunch of little palettes all

02:02:01   around your screen.

02:02:02   It could, if you use switch class, this will do exactly the same thing

02:02:04   as switch glass, but like again, better with more modern APIs and stuff with

02:02:08   better code and stuff.

02:02:09   Um, but also it'll let you have a bunch of folders in the corner,

02:02:12   and you can have another little palette over there with other stuff in it,

02:02:15   and you could have like whatever, like it's just you know, I don't it's

02:02:18   it's not a great idea.

02:02:19   Again, it's not drag thing, it's still fairly limited, but it is no longer

02:02:22   like switch class, which is like this is an app switcher, and that's it.

02:02:25   This is instead arbitrarily arbitrary little pality turds all around all your screen.

02:02:32   You really need to work on that messaging, John.

02:02:34   Yeah, I mean, that that's basically what it is.

02:02:37   Like, I don't even know if anybody wants this, but I I once I

02:02:40   had the ability to do this, I'm like, well, I could those folders over

02:02:43   there, that's not the only thing I might want on the screen, especially if

02:02:45   they're auto hidden and they just pop up when they come when I go

02:02:44   to the edge of the screen.

02:02:44   Like I can think of other potential uses for this.

02:02:51   Um and then I just kept adding stuff to it.

02:02:54   Because and by I I mean almost entirely my coding agent.

02:02:58   The only code that I wrote in this is basically all the UI because

02:03:00   I don't know what it is with coding agents, but they cannot lay out

02:03:02   a dialog box to save their life.

02:03:04   And like I'm like, don't even bother.

02:03:05   Just I will do the dialogue boxes.

02:03:07   There, it's it's the dialog boxes are not complicated.

02:03:09   It's like, you know, a label and a toggle control.

02:03:12   Like that's it, but they cannot handle it.

02:03:13   So I did other than that, they're doing everything and I'm just directing them.

02:03:17   Um but what do I do about the doc material?

02:03:20   Like what like I was all excited about that.

02:03:22   Obviously, I can't ship that in the Mac app store, and I'm absolutely not

02:03:25   gonna send this direct to people and make my own payment pro like it's

02:03:27   not that was exactly what I was gonna ask.

02:03:30   Is are you gonna do some sort of other payment processor?

02:03:33   If this was an app that I thought people are actually gonna buy in

02:03:35   any appreciable numbers, I would have considered it, but this is not that app.

02:03:38   So no, I'm not doing that.

02:03:39   It's Mac App Store, and which means I can't do any of the private

02:03:41   APIs and stuff like that.

02:03:42   But I have some precedent in dealing with this, which if you remember in

02:03:45   Switch Glass, um, again, it's just a little app switcher.

02:03:49   Uh, and it does have little right-click menus on all the apps and stuff

02:03:51   that have various actions you can perform.

02:03:53   And the obvious thing that you would want them to do, like when you

02:03:55   right click on uh on an app icon in the dock, you can quit

02:03:58   the app, right?

02:03:58   Right click and quick on an app on the dock.

02:04:00   That's a reasonable thing to want to do.

02:04:02   MAG App Store says no, you can't do that.

02:04:05   Um, so you can do a little daemon thing.

02:04:08   Yeah, there's an entitlement that you need to do that that Apple will not

02:04:11   give you and did not give me.

02:04:13   Uh and so you can't do that.

02:04:15   Um, and so my solution to that was okay.

02:04:18   Well, I am just going to make an unsandboxed, notarized direct download free app

02:04:24   that Switch Glass will look for if it happens to be there, launch it

02:04:28   if it is there, and talk to it until it quit the apps.

02:04:31   And so that app is called QuitchGlass.

02:04:33   It's like quit, but SwitchGlass, QuitchClass.

02:04:35   It is a free download.

02:04:36   If you have Switch Glass and you want to quit apps, you just download

02:04:38   QuitchClass, launch it once.

02:04:40   It is totally invisible.

02:04:42   It doesn't have a doc icon, it doesn't have a menu bar icon.

02:04:44   You never have to launch it manually.

02:04:45   Switchclass will do all the management for you.

02:04:47   The point is if you have QuitchClass installed and have launched it one time

02:04:50   ever, SwitchClass will have a quit in the menu that will work because SwitchGlass

02:04:54   communicates to QuitchClass and says, hey, you do the quitting for me.

02:04:58   And it does.

02:04:58   The magic of not being sandboxed.

02:05:01   So, how can that strategy help me get a doc material on uh on

02:05:09   my app?

02:05:10   Not quite as easy because it's not like you know, I'll send you a

02:05:13   message and you'll do the work.

02:05:14   Like it's the material that's showing behind the little pallets.

02:05:17   And that's I mean, again, if I was Ge Rambo or somebody clever, I

02:05:21   there's probably some way to like have things rendered by another process and put

02:05:24   back into yours and blah blah.

02:05:26   But I I'm never gonna figure that out.

02:05:29   But here's my strategy, and this is probably why I mean if Apple again,

02:05:32   if Apple does listen to this, they're probably going to shut me down and

02:05:35   this will never be a thing.

02:05:36   But until Apple stops me, which may be very soon, here's my strategy.

02:05:41   Ship my app on the Mac App Store with the things that Mac App

02:05:44   Store apps can do.

02:05:45   So no quitting, no doc materials, just tons of other limitations that suck.

02:05:50   And even more new ones that I discovered as I submitted uh builds and

02:05:53   they kept rejecting me.

02:05:54   And I had to remove entitlements, remove entitlements, remove entitlements.

02:05:57   They don't like me.

02:05:57   I've been rejected more on this app so far.

02:05:59   It's not even released just on test flight builds.

02:06:02   Anyway, get something that will go up on the Mac App Store.

02:06:06   Then also make from the exact same Xcode project another version of this same

02:06:11   app with all the the uh forbidden features and forbidden entitlements and all that

02:06:18   good stuff and make that a free download on my website.

02:06:22   But the the trick is that that free download will not run unless you

02:06:27   have purchased my Mac App Store app.

02:06:30   Oh, that sounds unfun, but like a pretty decent solution to the problem.

02:06:34   And I mean, I'm sure Apple hates this, but like as far as they're

02:06:38   concerned, the app on the app store is perfectly legitimate.

02:06:41   It does only the things that it's allowed to do, and that's it.

02:06:44   I'm not gonna link to the like obviously from the app, I can't tell

02:06:47   anybody from my Mac app store app, I can't say, hey, go get the

02:06:50   I can't say mention it, I can't link to it, I can't acknowledge that

02:06:53   it exists, I can't do anything.

02:06:55   And the other thing is my Mac App Store app has no idea that

02:06:58   other app exists, doesn't communicate with it, doesn't look for it, doesn't do anything.

02:07:02   It just has to be there on the disk with its like Mac App

02:07:06   Store receipt, you know, the cryptographic receipt that like happens when you buy an

02:07:10   app or like a paid up front app.

02:07:12   Meanwhile, if you were to wander over to my website and find this other

02:07:18   unsandboxed app that is free for anyone to download, but is entirely useless unless

02:07:24   you have installed this Mac app from the Mac app store.

02:07:27   Just run that app and that app can quit apps, can resolve aliases to

02:07:32   servers, can do doc materials, can do all of the fancy features that you'd

02:07:37   always want to do but couldn't do before.

02:07:39   And so this is my first foray into non sandboxed Mac app development.

02:07:42   And once I knew I could do that, I basically made I'm working on

02:07:46   making this app not a doc replacement, but I'm doing stuff that I didn't

02:07:50   even think was possible.

02:07:51   Like um, I've always said people always say, hey, SwitchClass should show the notification

02:07:55   badges that are in the dock.

02:07:56   There's no API for that.

02:07:57   There's not even like there's nothing for that, but I'm like, well, if I'm

02:08:01   out of the sandbox, I'll do it.

02:08:03   Let's go nuts.

02:08:04   Yeah.

02:08:05   Hey, coding agent, find a way to do that.

02:08:07   And the answer is always as always is accessibility.

02:08:09   Accessibility exposes the badges.

02:08:11   And so my app will look at the accessibility view of the doc.

02:08:14   And when it puts a badge, I'll put a badge.

02:08:16   Same thing for like the quick command is easy.

02:08:19   Yeah, you so you can make a very convincing doc look alike with my

02:08:25   app that a regular person could even forgive them for thinking was the doc.

02:08:29   There are still some features that things the doc does that I can't do

02:08:33   because there's they're not exposed from accessibility.

02:08:35   Like uh basically I'd have to uh inject into the process or something.

02:08:39   But anyway, this is my plan.

02:08:41   I'm going to put this app uh app out when it's it's not even

02:08:44   close to done, but like I'm gonna put this Mac uh out app out

02:08:47   on the Mac app store, and I'm going to have an unsandboxed version of

02:08:51   the app that is free for anybody who has bought the actual app.

02:08:55   And then if Apple doesn't stop me, I'm just gonna continue doing that until

02:08:57   they do stop me.

02:08:58   If they do stop me, I'm not sure what I'll do.

02:09:00   But like honestly, this is not a very mass market app.

02:09:01   But I just I just had so much fun making this and adding functionality

02:09:05   to it.

02:09:05   I'm still having fun.

02:09:07   Um, so yeah, and my plan is as as Casey pioneered on ATP, I'm

02:09:11   going to offer the uh access to the test flight to any ATP member.

02:09:14   I will warn you, ATP members, this app is rough.

02:09:18   It's not done.

02:09:19   The good thing about it is, unlike hyperspace, like it doesn't shouldn't really destroy

02:09:23   your files or anything.

02:09:25   Like the beta's not, it's it look, it's a little palette, a bunch of

02:09:28   pallets on your computer where you put stuff in it.

02:09:30   Like it's not it's not modifying your files.

02:09:33   Like it's it's the the danger, this probably one of the reasons that I'm

02:09:36   very comfortable vibe coding this is like it's not it's not a consequential app.

02:09:41   It is just a UI, uh frivolous little UI thing that you may or

02:09:45   may not like but if you want to try it and find bugs and

02:09:48   report bugs or whatever um at some point maybe by the time you listen

02:09:52   to this but certainly before next episode if you log into at v.fm and

02:09:56   go to your member page it's just atp.fm slash member or click on membership

02:09:59   in the navigation you will see a section appear on your member page as

02:10:03   it has in the past that says hey do you want to join the

02:10:05   Tesla for John's new app click here and then you will join it I'll

02:10:09   also have a link to my Slack for the app if you want to

02:10:12   join and complain about stuff there.

02:10:14   Yeah I'm not sure how I'm gonna describe this app or pitch it or

02:10:18   anything like that but it is basically you can make a bunch of little

02:10:21   dock things all over all your screens and have them contain whatever you want

02:10:25   including the trash can like you can put a trash can in it and

02:10:27   all that stuff like you can you could I'm not suggesting that you make

02:10:30   a doc like I'm not gonna use it for that but I have replaced

02:10:33   switch glass in my daily use with this because it is like switch glass

02:10:36   but better.

02:10:37   Oh the other one I forgot to mention is custom icons man that their

02:10:40   situation I've found so many feedback saying please make an API for this but

02:10:44   like even without being sandbox there's no way to get a lot of these

02:10:48   things I did get the calendar because if you use like a doctile plugin

02:10:51   and if you're Apple I can load Apple's dock tile plugins but I'm not

02:10:54   gonna load doctile plugins from third party apps because I just feel like loading

02:10:58   third party libraries into my process is just a bridge too far.

02:11:00   I could do it but I'm not gonna but anyway calendar will show the

02:11:03   right icon and I might do custom stuff for things like ivory that use

02:11:06   like the catalyst set custom icon but the the the icon api is pro

02:11:11   the private icon apis on macOS just do not expose this stuff.

02:11:15   But still, my switchgrass glass replacement is it's currently showing a badge for Slack

02:11:20   right now.

02:11:21   If I had calendar in there, it would show the correct calendar date.

02:11:23   Uh and it uses all modern APIs.

02:11:25   And my little folders in the side, they're looking beautiful.

02:11:28   They will raise when my cursor goes up them over them, so they have

02:11:29   auto arrays on them, so they'll go in front of other things.

02:11:33   If I slam my cursor against the bottom of the screen, it's all looking

02:11:36   good.

02:11:37   So yeah, that's my uh that's my new app.

02:11:39   Wow.

02:11:40   All right, so I have a question.

02:11:41   What's gonna happen first?

02:11:43   You completely subsume all of drag thing functionality or Apple is carbon neutral.

02:11:49   I'm never gonna do everything that drag thing did.

02:11:51   Like that's the problem with this app.

02:11:52   People are gonna see it and they're gonna be like, I want to do

02:11:55   this, I want to do that.

02:11:56   And now I no longer have an excuse other than I don't want it

02:11:58   to do that.

02:11:58   No, I'm just messing with you.

02:12:00   Yeah.

02:12:01   I honestly I don't think I foresee this drawing attention from Apple negatively because

02:12:08   what you're doing, first of all, it's like, yeah, you're doing a whole bunch

02:12:10   of like private API crap, but like not in a way that would blow

02:12:13   up in their face.

02:12:14   Um, and second of all, with like the Mac App Store thing, like where

02:12:17   you're gonna have this app require the Mac App Store purchase.

02:12:21   That like Apple cares if you try to bring purchases out of their in-app

02:12:26   purchase system, right?

02:12:27   You're trying to bring purchases into it.

02:12:29   I you're trying to you're you're doing a hack for them to make more

02:12:33   money, so I don't think they're going to care about that either.

02:12:36   Here's here's where I think they'll all get tripped up.

02:12:38   And for the first release of this, I'm going to avoid this, but for

02:12:41   subsequent releases, I will not avoid it.

02:12:43   And here's what it is.

02:12:44   When you submit the app review, and you have like you have to have

02:12:47   like URLs in your app for like your support URL and your privacy policy

02:12:50   and crap like that, and also like the help button, like in the help

02:12:54   menu of all my apps, there's a help item.

02:12:56   And if you select that help item, it launches you to a page on

02:12:58   my website.

02:12:58   Oh, yeah.

02:12:59   Then they go poke around to see if you're trying to suck purchases out.

02:13:02   Yes.

02:13:02   And on my website, you can click on the logo at the top and

02:13:06   get the most recent blog post.

02:13:07   And if the most recent blog post is like, hey, look at this unsandbox

02:13:10   app you can download.

02:13:10   They're gonna be like, eh, sorry, you can't link to another app to download.

02:13:13   Like you basically, that is an app store rule.

02:13:15   You cannot in your app say, hey, download this other app from my website.

02:13:19   You can't do that, at least not in the US.

02:13:21   Even if you don't have your own purchase system outside of their system, no,

02:13:24   I know.

02:13:25   But like I what I'm gonna do is I'm going to for the first

02:13:28   release of this.

02:13:28   I am not, I'm gonna make sure there's no links on my like uh

02:13:31   cause the question is how many links do you follow?

02:13:33   Because remember the horror stories of like I link to my website from my

02:13:36   about screen on my app, and 20 clicks from that you can get to

02:13:38   this post that mentions this thing, and Apple's like, sorry, we found our way

02:13:41   to this thing that mentioned another app rejected.

02:13:43   Like that's my fear.

02:13:44   So, one thing one thing that I've done, um, even though I don't think

02:13:48   I am at risk for any problems with with my with the website, but

02:13:51   like for Unforgetful, the privacy policy and terms of service pages, which have to

02:13:57   exist for all subscription apps.

02:13:58   Um, those when you open those in the app, there is no header or

02:14:04   footer or navigation to go to other pages.

02:14:06   It only is those pages.

02:14:08   Um, could you do something like that?

02:14:10   Yeah, I probably I probably will when I put them up.

02:14:12   Like, but the problem is the help page.

02:14:13   I want there to be a header on the help page, like it's part

02:14:15   of my website.

02:14:16   Like, we'll see how it goes.

02:14:17   Like, if I have to, and also it's like, as you know, with the

02:14:19   app review, it depends on what reviewer you get.

02:14:22   Is this the reviewer that clicks three three follows three links deep from your

02:14:25   help page or four?

02:14:26   Or you know, it's just there's no consistency.

02:14:29   So I'm gonna do what I can.

02:14:30   Apple don't listen to this podcast.

02:14:32   As Marco says, see, I'm bringing money into your system.

02:14:34   I don't I don't know what they think of it, but this is my

02:14:36   current strategy.

02:14:37   If it doesn't work out, I'll think of something else.

02:14:38   I mean, the good thing is this is never gonna be a mass market

02:14:41   app anyway, so who cares?

02:14:42   It's just mostly a curiosity and a fun thing to do on the show.

02:14:44   But I'm having fun making it.

02:14:46   So no, this this is very clever.

02:14:48   I did not see this like like a whole sale replacement of Switch Glass

02:14:53   being where this conversation was going, because all we knew was John's new vibe-coded

02:14:57   app.

02:14:58   Uh, I'm surprised this is the where it ended up, but I mean I

02:15:01   think it makes sense.

02:15:01   And and I think if it's working for you, that that's all that matters.

02:14:59   It's a regression for me.

02:15:05   It's a re like I said, it's a regression from me trying to make

02:15:08   useful apps going back to my I'm back on my BS, as they say.

02:15:12   I'm back to making apps that nobody wants that are weird.

02:15:14   They're just bigger than they used to be, and also mostly not coded by

02:15:18   me.

02:15:19   But that's just the place where we are, the place where I am right

02:15:22   now.

02:15:22   So maybe someday I will return to making apps that other people might want

02:15:25   to buy.

02:15:26   But uh, yeah, in the meantime, I'm doing this one.

02:15:28   You haven't asked me about the app's name yet.

02:15:31   Okay, what's the name, John?

02:15:32   It's bad.

02:15:33   Not Switch Glass?

02:15:35   I just assumed you were subsuming switch glass.

02:15:37   No, no, no.

02:15:38   Switchclass will continue on.

02:15:40   Um, as as you know, again, it's it's it's a six-year-old paid upfront one-time

02:15:44   app that if you paid for six years ago, it will keep I'm gonna

02:15:47   keep it running, right?

02:15:48   Um, but I will probably never modernize the code.

02:15:50   So at some point, if like if this version of Swifty Y no longer

02:15:53   compiles, I maybe I'll have the LMs modernize it or something.

02:15:57   But anyway, I I named this app when I thought no one else was

02:16:01   ever gonna see it.

02:16:04   Yeah, yeah, you know how it is.

02:16:07   Um also on that front, I made the icon myself, and it's not good.

02:16:13   Oh, been there too.

02:16:14   Right.

02:16:15   We've all been there.

02:16:16   And here's the thing with this.

02:16:18   I'm probably not gonna change the name because like who cares?

02:16:22   Like, whatever.

02:16:22   It's not it's not it's not as bad.

02:16:24   What was your past one?

02:16:25   Uh fluck up, it's not fluck up bad, but it's not good.

02:16:28   All right, calm down.

02:16:29   The icon, on the other hand, it's not good.

02:16:33   But my plan is if I release this in the Mac App Store and

02:16:37   Apple does not kill it.

02:16:40   Uh once it makes enough money for me to pay someone to make an

02:16:43   icon, that's the like the the first X amount of uh money from this

02:16:44   app is going to pay for its own icon.

02:16:49   So it will not be in the black until it is paid for its

02:16:52   own icon.

02:16:53   Yeah, that's that's where I'm and I honestly making the icon was fun, but

02:16:58   it just like it's not again, it's not in my skill set.

02:17:01   It's like and I and I used AI help, like so the icon.

02:17:05   Well, I did I for the first problem is I have no good ideas

02:17:09   for the icon.

02:17:09   So that's not a good place to start.

02:17:11   If you have no ideas, like if I have an idea, I can fumble

02:17:13   through and like make something, but I have no good ideas, so I went

02:17:16   with bad ideas because you gotta do something, right?

02:17:19   And then I was like, can you AI generate an icon?

02:17:22   No, you can't, they just look terrible.

02:17:24   So I it I did what I've done in a few times.

02:17:27   Like I did this for the beta icon for hyperspace.

02:17:30   I bought stock photos from iStock Photo.com.

02:17:33   Oh you can did you know you can buy stock photos?

02:17:36   You can you pay for them, and they're yours free and clear, and you

02:17:39   can use them in your app icon.

02:17:40   And so I bought stock photos, uh, and then I AI generated one element,

02:17:45   and then I photoshopped it with a bunch of crap that I did in

02:17:48   Photoshop and Affinity Designer plus my stock photos.

02:17:50   Can you see how this isn't making for a not good icon?

02:17:54   So that's how I ended up with what I have.

02:17:56   I will show you.

02:17:57   So what's the name?

02:17:57   I will show you the icon first and see if this helps you get

02:18:00   the name.

02:18:00   I will try to show you a small version of the icon so because

02:18:03   it looks worse the bigger it gets.

02:18:07   Uh here you go.

02:18:09   So this uh the the the inspiration of this icon, which I will put

02:18:12   in the uh in the show notes.

02:18:14   Is this air power?

02:18:16   No, no.

02:18:16   So the uh the the menu bar icon, because like switchclass has a menu

02:18:20   icon, and this one has a menu bar icon, the menu bar icon is

02:18:22   essentially like the SF sim, it's the SF symbol for like a computer screen

02:18:26   with dock at the bottom of it.

02:18:28   Like a lower third, yeah, it's like a computer screen.

02:18:31   I forget what the SF symbol is called, but it's like computer screen with

02:18:33   dock, or I don't even know.

02:18:29   But that's that's what it's supposed to be.

02:18:35   It's supposed to be a uh like a round wreck with a line at

02:18:38   the bottom.

02:18:39   And I'm like, that's my app.

02:18:40   It makes docks.

02:18:42   I'm not gonna use docs, I'm using pallets for a reason, like I don't

02:18:44   want it confusion.

02:18:45   Looks like inset filled bottom third square.

02:18:48   Yeah.

02:18:49   Uh but I'm like, but yeah, but like obviously that's so fine.

02:18:52   That's the monochrome uh menu bar icon, but what is the actual app icon?

02:18:56   And so here uh this is what I've got for the app icon, and

02:19:00   it makes sense with the name, kind of sort of.

02:19:02   I tried a lot of iterations in this.

02:19:03   Marco, you have any guesses of the name of this thing, keeping in mind

02:19:06   that it makes a bunch of little uh uh rectangular uh long skinny things

02:19:10   around the edges of your screen.

02:19:12   And and given that you're looking at the icon, which one of you can

02:19:14   feel free to describe the icon.

02:19:15   I won't do it.

02:19:16   So it it's like a it's a wave crashing onto a beach, like like

02:19:19   that like the foaminess of the end of a wave onto a beach with

02:19:22   air power sitting on the beach, yeah, with with a big black sunglasses case

02:19:26   uh sitting or like a like a big like like a like a polished

02:19:29   oval stone.

02:19:31   Um so I'm guessing you probably didn't call it Docker, even though that would

02:19:34   be a great name for it.

02:19:35   That would be I think that's taken.

02:19:37   Yeah, maybe.

02:19:38   Um it's like uh and like the icon, it's like a you know, like

02:19:42   a polished stone from a beach.

02:19:44   So uh is there any uh uh long uh skinny rectangular things related to

02:19:49   beaches that you can think of, Marco?

02:19:50   Docks, yeah.

02:19:51   That's right.

02:19:52   But I'm not using that one, so what's out of it?

02:19:54   Right the answer is right there in front of us, and I'm cheating because

02:19:58   drift fuel.

02:20:00   I know, I know.

02:20:00   Casey, don't tell him.

02:20:02   Oh, the file name.

02:20:03   It's obvious once you hear it though, jetty.

02:20:05   Yep, yeah, jetties.

02:20:07   I grew up.

02:20:08   I grew up going to a beach, uh uh Long Beach on Long Island

02:20:11   on the North Shore that has a jetty.

02:20:13   I played on that jetty all the time.

02:20:15   For those who don't know, a jetty is like a uh uh long skinny

02:20:18   pile of stones or other things that goes out into the water and directs

02:20:22   where the erosion should go so that one people on one side of it

02:20:24   are mad and people on the other side are happy.

02:20:26   Yep.

02:20:26   That's what it that's what a jetty does.

02:20:28   And uh it's a long skinny thing.

02:20:30   And it makes you know, what's a bunch of long skinny things?

02:20:33   That's what I'm I can't use doc.

02:20:35   I can't use anything having to do with doc.

02:20:37   Calling the app Jetty and the uh the unsandboxed version is called Jetty Pro.

02:20:42   This is honestly, I don't see anything wrong with this name.

02:20:45   I mean, I haven't done like you know a trademark search or anything, but

02:20:47   like at first glance, I don't think this is a bad name at all.

02:20:50   I agree.

02:20:51   Uh I was kind of surprised it wasn't taken.

02:20:53   It's but who even knows what a jetty is?

02:20:55   And I the reason this thing is black is because I tried a bunch

02:20:57   of non-black rocks, but the problem is it's small sizes, they just become mud.

02:21:01   And this this at least is a when you shrink this down to tiny

02:21:03   sizes, it looks like the icon in the menu bar.

02:21:06   So you can be like, what is the icon in the menu bar?

02:21:08   Oh, it's this app who's I can see it.

02:21:09   That's why it has the thick black outline around it, like the and it's

02:21:13   it's jet black, like I guess.

02:21:16   Like whatever.

02:21:17   Like, this is a bad idea for an icon.

02:21:19   It's a bad idea for an icon not well executed.

02:21:21   Like, I'm not happy with this icon, but honestly, I don't have a better

02:21:24   idea.

02:21:24   But when I get enough money, so if I get enough money from this

02:21:27   app, I'm going to give it to someone to come up with either a

02:21:30   better idea or a better version of this icon if we can't come up

02:21:33   with a better idea.

02:21:35   Is there a reason you didn't like orient it to be more like a

02:21:38   jetty?

02:21:39   Uh, because this is what the menu bar icon is.

02:21:41   Oh, yeah, it's the whole reason.

02:21:43   Right, right.

02:21:43   The menu bar like I picked the menu bar icon first and I got

02:21:46   used to it.

02:21:47   I'm like, and I couldn't think of any other ideas for the thing.

02:21:49   So I know what you're saying.

02:21:50   Like, it should, but the thing is the the UI in this app doesn't

02:21:54   ever stick out from the screen.

02:21:55   They're always flat against the edge.

02:21:56   And I know that's not what a jetty does, but like again, it's not

02:21:59   I can't I can't I can't uh defend it like it it is what

02:22:03   it is no this icon the this icon is straight trash but the name

02:22:07   I like yeah and and it doesn't even matter like you know like if

02:22:10   you know people who don't live near these things don't know what jetties are

02:22:13   who cares there's lots of names that like people don't know the origin of

02:22:16   the name they just associated it with the app and it's fine.

02:22:18   Yeah you know how many people know what a Google is even if even

02:22:21   if you spell it the other way.

02:22:22   Although I will say that Jetty with no colon suffix was in fact taken

02:22:26   on the app store because everything is that I ever try.

02:22:29   Well well I mean yeah no I think I think this is fun.

02:22:32   Yeah what are you what is your suggestion for the for my colon suffix?

02:22:35   What did you say little turdy palettes everywhere?

02:22:38   I'm not going with little turdy palettes.

02:22:40   See what you what you actually want to do like if you want people

02:22:44   to actually find and download this what you actually want is figure out what

02:22:50   keywords you want that that people would be most searching for for apps like

02:22:54   this and figure out some subtitle that would that would include those keywords.

02:22:59   Like that's that's the actual answer to this but since you are really good

02:23:03   at at not making money with your apps because you you're you're a purist

02:23:08   I I have to I I can only go where my heart leads me

02:23:10   and it leads me to non profitable places.

02:23:22   No I I accept that it's the way it is I had to do

02:23:24   with hyperspace too like it is what it is.

02:23:25   Like I'll do it you know as long as I as long as it's

02:23:28   not on my icon in the app in the applications folder I'm fine.

02:23:31   Yeah, I don't have a clue what I would write.

02:23:33   Oh, here's what I've got, and I'm uh as usual not happy with it,

02:23:36   but uh it will actually probably help with SEO is I've got Jetty colon

02:23:40   custom docs.

02:23:41   Because people are gonna search for doc.

02:23:43   Yeah again I'm not fair I'm not gonna call things doc because in my

02:23:47   UI I have things like material doc, which means make it look like the

02:23:51   doc, right?

02:23:52   Or preset doc, which means make a thing that looks like the doc.

02:23:55   So I do not use doc for the name of my things.

02:23:57   I always use palette, which is I think what drag thing might have called

02:24:00   them.

02:24:00   I don't know.

02:24:00   I'm using palette is what I'm going with.

02:24:02   But you can't put palette on the headline so jetty colon custom docs is

02:24:06   terrible and looks gross and kills my eyes but it will come up in

02:24:09   people's searches maybe I'm will apple let you use the word doc we'll find

02:24:13   out I mean they did let me test flight it and let me tell

02:24:15   you they were rejecting me like crazy in test flights I'm like it's a

02:24:18   test flight like reject me in app review fine but like I'm just trying

02:24:21   things out here and they're like nope we can't even let this test flight

02:24:23   thing up.

02:24:24   You're not getting that entitlement forget it.

02:24:26   Wow so they let they let it up with the subtitle we'll see how

02:24:28   it goes.

02:24:29   Well good luck I I think I think that is a perfectly good subtitle

02:24:32   for this app but uh I would be I honestly I'd be surprised if

02:24:35   Apple used it are you gonna capitalize doc?

02:24:39   Yeah it's title case capital J, capital C, capital D, Jetty colon custom docs.

02:24:44   Other apps in there say doc stuff.

02:24:46   Okay there's docdoor, for example.

02:24:48   I don't know if that's on the Mac app store.

02:24:49   We'll find out if I have to change the subtitle fine but I can't

02:24:51   I would have just used plain jetty obviously if I could have gotten it

02:24:54   but someone else has got it.

02:24:55   This is this is exciting.

02:24:56   I I think this is fun.

02:24:57   Um I think you should I think you should stick with it and see

02:25:01   where it goes.

02:25:02   And you know listeners this is not an official perk of being a member

02:25:06   so uh if this never if the test flight never ever shows up in

02:25:09   membership.

02:25:10   Well all right just make it I'm just trying to leave you an out

02:25:13   trying to leave you an out.

02:25:13   Nope I've I've got the test flight built up.

02:25:15   I've made the group, it's all ready to go.

02:25:17   I just haven't pushed it to the website yet.

02:25:19   Oh, all right.

02:25:19   And like I said, the app is uh filled with bugs, but you know,

02:25:22   the consequences of those bugs is the animations are glitchy in your custom docs.

02:25:25   It's not a big deal.

02:25:26   Real time follow-up, there are actually lots of apps in the Mac app store

02:25:29   that include the word doc, including some like things that seem to be doing

02:25:33   somewhat similar tasks like adding extra docs, basically.

02:25:37   So yeah, there's lots of apps that do this.

02:25:38   And it's like, why don't you just use one of those apps?

02:25:40   Obviously, I need the app that works exactly specifically like I want it.

02:25:44   Yeah, of course.

02:25:44   Well, yeah, I mean, look, I just made a reminders app.

02:25:46   There's lots of reminders apps, including reminders.

02:25:48   Exactly.

02:25:49   But that's the whole that's the whole thing about vibe coding.

02:25:51   Like I never would have bothered doing this whole rigum roll for just this

02:25:54   one little thing, but vibe coding makes it makes it possible to do this

02:25:57   with the amount of effort that uh with an amount of effort that makes

02:26:00   sense for the amount of money this is not going to make.

02:26:02   You know what I mean?

02:26:03   Yeah.

02:26:04   Because like the amount of effort I put into Switch class was tremendous.

02:26:06   Hand rolling all that, learning, learning on the fly, like my very first Swift

02:26:10   UI thing.

02:26:11   Huge amount of work into that.

02:26:12   But that was like a learning opportunity, blah, blah, blah.

02:26:14   And I didn't make I didn't make it back on the app.

02:26:15   Like, but it was, but it was a good learning opportunity.

02:26:17   But now when it came time to do like, oh, revise switch class to

02:26:21   like, you know, generalize it to do this thing.

02:26:23   I implemented it and I realized this is not this is a waste of

02:26:25   my time.

02:26:25   This is not a good use of my time, which is good because like

02:26:27   I didn't do that and I did hyperspace instead.

02:26:29   So but anyway, this is one of the beauties of vibe coding that even

02:26:33   for developers, it opens up the potential to do things that would have been

02:26:36   a waste of your time otherwise, and this definitely qualifies.

02:26:38   Oh, yeah.

02:26:39   I mean, like, like just literally in the last you know, two weeks, I

02:26:43   I not only you know, do I make my regular software, but I've I've

02:26:46   vibe coded two different apps just for myself.

02:26:50   Like one of them was uh I have an app that decrypts audible books,

02:26:54   and there's been like you know, command line utilities to do this.

02:26:56   I want to listen to my audio, my audible books in Overcast.

02:26:58   Okay, of course.

02:26:59   Like I want to use the uploaded them to my uploads and use my

02:27:01   audio engine with my effects and my speed to listen to audiobooks.

02:27:04   So I buy my audible and I use my crappy vibe-coded app that I'm

02:27:07   never gonna release because I don't want the hassle.

02:27:09   Um, you know, but it's just took two seconds, you know, for for this

02:27:13   app to come together.

02:27:14   It's great.

02:27:14   And I don't have to I don't have to remember like all the command

02:27:17   line incantations for how to use the different tools and where do they drop

02:27:20   the little turdy configuration files and how do I keep all the logins straight

02:27:23   and everything.

02:27:24   Nope.

02:27:24   I just had a vibe coded Mac GUI app that stores stuff in Keychain,

02:27:28   and it's great.

02:27:28   And I didn't have to look at any of the code ever.

02:27:31   And that's great.

02:27:32   And then my second one is an app that um you know how like

02:27:36   in the app store, people want you you you basically want to have screenshots

02:27:40   that aren't really screenshots.

02:27:42   They like show the app kind of in a device frame and then have

02:27:44   a bunch of text crap to brag about stuff about the app.

02:27:47   And I years ago, like I tried this with Overcast, I even I think

02:27:50   we didn't even did an under-the-radar episode about it, um, where I did see

02:27:54   like, yeah, they those do actually convert like a little bit better.

02:27:57   It wasn't a huge difference.

02:27:59   I think at the time it was something like seven to ten percent better

02:28:01   conversion rate.

02:28:02   Um, but it it's you know, that matters.

02:28:04   And I just hated having to like make them, and I hate having to

02:28:07   make all the different size screenshots to optimize for different devices, and some some

02:28:12   of that's easier now, but not all of it.

02:28:14   So I in like one giant 15-line prompt, I had Claude make me my

02:28:21   own stupid little admin app to make those screenshots.

02:28:24   It's basically a GUI editor for me to add text and symbols and crap

02:28:29   onto screenshots and have them put into a device frame and then to be

02:28:33   able to export them in this nicely, you know, nicely sized and named list

02:28:37   of ping files that I can drag into App Store Connect.

02:28:41   It and it took that whole thing took a total of I think three

02:28:44   prompts and maybe a half hour.

02:28:47   And now I have this app that again, like, I'm not gonna release it.

02:28:50   It's not, it doesn't make any sense for me to release it, but like

02:28:52   it's vibe coding is really helpful for like making little things to scratch your

02:28:58   own personal itches.

02:28:59   You know, stuff that used to like either only like you could only do

02:29:03   it if you had a lot of free time and a lot of you

02:28:59   know skill in that area.

02:29:07   Um, or if you could find another market for it.

02:29:10   And now you can just make your own stuff.

02:29:12   Like in the same way, like people can make their own spreadsheets, you know,

02:29:15   to solve their own problems for for all these years.

02:29:18   Now you can make your own little apps to solve your own problems.

02:29:20   Like for internal tools, for admin tools, it's very, very valuable.

02:29:24   That screenshot frame thing, probably one of the reasons it probably did such a

02:29:27   good job is there are so many of those apps out there, and some

02:29:29   of them are probably open source, so it has an idea of how to

02:29:31   do it.

02:29:32   Like that was my hope with the um it didn't work out as I

02:29:34   hop uh as well as I hoped it would.

02:29:36   That was my hope with feeding it switch glass as the as the initial

02:29:39   prompt, which is like look at the source code to this app.

02:29:41   Here's the repo, you have all the source code, just do that.

02:29:44   It didn't work as well in my experience as when these things are trained

02:29:48   on open source repos of like a thing that makes uh uh device framed

02:29:51   screenshots with text for your phone.

02:29:53   Like, there's so many of those apps out there, and I get it, some

02:29:55   of them have to be open source.

02:29:56   It makes me think that like training if you know, because obviously switch class

02:30:01   source code is not public, it's not an open source app.

02:30:03   Okay, but but if these models had trained on the switch class source, I

02:30:07   feel like they would have performed better than me saying, here's the entire switch

02:30:11   glass source code in your context window.

02:30:12   Because my experience with doing that is it was like it wasn't even looking

02:30:16   at the source code, or it was like looking at just certain portions of

02:30:18   it, and I had to direct its attention.

02:30:19   Look at this file, look at this function, look at this thing.

02:30:22   It was a little bit frustrating and counterproductive productive, but it was a good

02:30:25   learning experience in how to use these things.

02:30:27   But yeah, for the device frame thing, so many apps are out there that

02:30:30   do that.

02:30:30   And I think the reason everybody makes them.

02:30:32   Again, predating vibe coding people were making these little apps for themselves, people who

02:30:35   did have enough free time to do it, because everyone wants to do it

02:30:38   in their own way.

02:30:39   And again, some of those must be open source, and some of them must

02:30:42   be in the training data.

02:30:42   So when you ask for one, it's like I totally know how to make

02:30:44   that.

02:30:45   I've seen this app before.

02:30:46   Well, I uh I mean I I did actually specify a lot of the

02:30:44   behavior.

02:30:49   Like I specified the data model, I specified the UI and a lot of

02:30:53   that stuff.

02:30:53   But and I I don't know I don't think most of the coding agents

02:30:58   now are like replicating entire apps from their training data.

02:31:02   I I think I think it's more piecemeal and more involved than that.

02:31:05   But uh you know, I I I think it's it's a pretty well-known problem

02:31:10   set.

02:31:10   Like you can just if you know, if they look in their training data

02:31:13   or do web searches for just what these kinds of apps need to be,

02:31:17   what are the app store connect requirements, stuff like that.

02:31:19   Like, that's that's all easy to find.

02:31:22   That's good stuff, Sean.

02:31:23   I am excited to see the test flight and and kick the launch the

02:31:27   app once and never look at it again.

02:31:28   Probably.

02:31:29   But you know what?

02:31:29   I'm excited to see it nevertheless.

02:31:30   Yeah, congratulations.

02:31:31   Like, I uh as I as I have you know recently gone through this

02:31:35   process of launching an app, like launching any app is a significant achievement, and

02:31:41   it's a significant amount of work.

02:31:42   It's way like just you know, my stupid audible decryptor and screenshot annotator, like

02:31:48   that took no time.

02:31:49   If I wanted to release either of those, that's a that's a totally different

02:31:53   ball game.

02:31:54   Yeah, even if you're not writing the code for the app, right the the

02:31:57   pass.

02:31:57   I mean, you could get the LMs to do that too, but I'm I'm

02:32:00   not gonna do that.

02:32:00   But like, yeah, there's so much like overhead.

02:32:02   That's why like you need some you really need some motivation to get something

02:32:06   over the hump of like, am I gonna actually release it to people?

02:32:08   Because it's just such a hassle, and there's so much stuff to deal with,

02:32:10   especially if you're dealing with app store and app review and the website and

02:32:14   support and bug tracking and it's just so much that like keeping stuff private,

02:32:18   like my little trimmer app or these other, you know, I have a bunch

02:32:20   of other little like command line things that I made for myself that like

02:32:23   give me info about files that I use as like debugging tools while developing

02:32:27   hyperspace that have all these vibe-coded tools that like spelunk in the file system.

02:32:29   Like that stuff is great, but once you release anything to the public, it's

02:32:29   like, well, now you've taken on this giant mound of crap into your life,

02:32:40   so it better be worth it for what you want to do.

02:32:42   Yeah, exactly.

02:32:43   And that's that's kind of that's partly why too.

02:32:45   Like, you know, I I see a I see a bunch of um, you

02:32:47   know, people around our circles.

02:32:49   Like, I'll see somebody who we followed in years for Mastodon, and they'll be

02:32:52   like launching apps for the first time.

02:32:53   And these are people who maybe they were power users, maybe they weren't necessarily

02:32:57   programmers, or they weren't this kind of programmer before.

02:33:00   I don't have any problem with people doing this.

02:33:02   Like, you know, it obviously people feel very differently about you know using AI

02:33:07   to code stuff, and and I because I have mentioned that some of the

02:33:11   code I write is AI code, uh I'm I'm now getting one regular one-star

02:33:16   reviews on both of my apps um because it's quote AI slop, even though

02:33:20   I mean I it's like I spent three hours today writing writing code in

02:33:24   unforgetful myself, which I've written most of, and it's like and people, you know,

02:33:29   people will accuse whatever.

02:33:30   Yeah, I'm sure I'll get those reviews in this app too.

02:33:32   But you know, it's like if you want handwritten ones, I've got three other

02:33:35   apps.

02:33:35   Right.

02:33:36   But and the thing is like I I really don't have any problem with

02:33:40   people a hundred percent vibe coding an app and releasing it as long as

02:33:45   it works, and as long as like you know, if there's a bug, they'll

02:33:49   be responsive to it.

02:33:50   You don't have to be handwriting the code to have an app out there

02:33:54   that works and that you're responsive to making sure it continues to work.

02:33:58   So I think like to me, uh again, I know this is controversial, but

02:34:03   I guess I've already lost all these people or or angered them.

02:34:05   This is a very enabling technology.

02:34:08   Lots of people can now make apps who couldn't make apps before, and lots

02:34:12   of people who could make apps before, like us, can now make more apps,

02:34:16   better apps, and apps that just wouldn't have been worth the time to make

02:34:14   before.

02:34:21   So again, not a perfect technology without its downsides, but overall a massively enabling

02:34:27   technology.

02:34:28   And I I'm able to take joy in the fact that it can do

02:34:32   stuff like this.

02:34:32   That like these at like an app like this that you just made that

02:34:35   like that couldn't really exist or that wouldn't have existed before.

02:34:39   Now this exists.

02:34:40   That's a cool thing.

02:34:41   Yeah, I mean, the problem is that you know, people hate the technology, but

02:34:44   like we're talking about this one very narrow corner of the technology where it

02:34:47   excels.

02:34:48   And the technologies are not only used for programming.

02:34:50   And that's part of the big problem.

02:34:52   It's like, okay, that's great for you and your little programming world, but like

02:34:56   people use uh uh AI for other things, and it's extremely problematic in many

02:35:00   of those things.

02:35:00   And that's even before you get down to like, okay, but what about any

02:35:03   you know?

02:35:03   So that whole wad of crap is there, but like we have such a

02:35:06   narrow focus in the show because we're programmers and we're always talking about programming.

02:35:09   And I think sometimes people hear you guys love AI.

02:35:11   It's like I mean, it's we're we're using one very specific thing, which is

02:35:18   a popular thing that people use and people are getting, you know, utility out

02:35:21   of it.

02:35:21   But boy, people are like the majority of people who are using Chat GPT

02:35:25   are not making apps.

02:35:26   No, definitely not.

02:35:27   Let's put it that way.

02:35:27   And and you're right, like there are a lot of I think more problematic

02:35:30   uses for it that that I think uh, you know, w we're not we're

02:35:34   not advocating for everything all of those things, but in this case, uh, you

02:35:38   know, using it for software development.

02:35:39   It's yeah, which is why I like I'm excited about being able to run

02:35:43   like local like open weight models in the future on your Mac without paying

02:35:46   some giant company, without you know, running in data centers and like all that,

02:35:49   you know, like I can see that future being attainable, but uh we're not

02:35:53   there yet.

02:35:53   Yeah.

02:35:54   And and like that's what I was looking earlier.

02:35:56   That's why I was saying about like with you know, trying to figure out

02:35:58   how to use my excess Mac mini capacity.

02:35:59   I was looking at like, okay, well, what if I want if I want

02:36:01   to do like you know summarization and stuff like that with local models running

02:36:05   on Mac minis that I own?

02:36:07   It's like that's that's pretty low impact, really.

02:36:09   Like I mean for the world, like it's not that's not that bad.

02:36:12   But yeah, Apple was touting that by the way, with the new Mac Studios.

02:36:15   They're like, look, in this giant stack of Mac Studios that are connected, it

02:36:18   takes so little power.

02:36:19   Like I think Snell was saying, they said, look, it's just plugged into one

02:36:22   outlet.

02:36:23   Like, I mean, okay, great, but like people usually plug computers into one outlet,

02:36:27   like but for AI, it's like, look, this is this is so energy efficient.

02:36:31   You can you know run this incredibly big model plugged into a single household

02:36:35   outlet.

02:36:35   It's like, I guess this is the bar now.

02:36:37   You can have this computer in your house.

02:36:40   Right.

02:36:41   You don't even need three phase industrial power.

02:36:43   Yeah, instead of whatever, whatever you have in your like uh 240 volt or

02:36:47   whatever dual power strip thing in your data centers.

02:36:50   Right.

02:36:50   And and again, and I do want to just while we're on this topic

02:36:53   briefly, I do want to reiterate everyone saw that big list of features that

02:36:58   Apple did on these OS releases, right?

02:37:00   You know, Apple's using Claude, right?

02:37:02   Like everyone knows that out there.

02:37:04   I I think I think you need to know that.

02:37:06   Like that huge massive list of things that they're adding in this OS release,

02:37:11   that that's not a coincidence.

02:37:12   Like they're they're definitely using, we know that they're using Claude widely.

02:37:16   They have been for a few years, and that's why they're able to get

02:37:19   this much stuff done.

02:37:21   So if you're like taking a super hard line against using anything AI generated,

02:37:25   I I I have I have some bad news for you to use Mac

02:37:29   OS, iOS, probably Windows, and yes, also Linux.

02:37:35   Like, I I think you're gonna have a hard time uh avoiding AI generated

02:37:38   code forever.