PodSearch

ATP

697: The Chart Is Terrifying

 

00:00:00   What you can do, John, is you can just sit this one out, because everyone knows that the show is really about Marco and me, and you're just a Klingon. So you can just sit this one out.

00:00:07   From Star Trek?

00:00:08   No, not that kind of Klingon. Anyway, all right, let's get the show on the road. Hey, Marco, I have incredible, incredible news for you and for me. And I know we briefly talked about this privately, but I wanted to bring it up one more time publicly. Marco, on Nugs.net, which we've talked about a fair bit recently on account of me being freaking obsessed with Goose these days,

00:00:30   They're really on a roll, too. It's really not helping you.

00:00:33   It's not helping at all. Anyways, Nugs is finally, one day later, broadcasting or archiving, whatever verb you want to use. They're offering up for listening, for streaming, Dave Matthews Band's concerts, which is incredible news for you, boy, because I have to tell you, the only way I used to be able to get a Dave Matthews Band concert was to either trade for it when I was a child, you know, with tapes or CDs,

00:00:59   or there's some places where you can find torrents and the idea is that that sticks with the spirit of the band's taping policy, which is that you have to share, because with a torrent, you're sharing, but you're not supposed to go, like, stream it from somewhere or something like that.

00:01:12   That's a pretty fine distinction.

00:01:14   We're going to move right past that. We're going to move right past that. But these are all audience recordings. So this is people who bring, it used to be DAT tapes, and now it's like mixed pre's, like I'm talking to you through right now, and hook them up to microphones on towers.

00:01:28   And the quality has gotten way better than it was in 96 when I started listening to Dave Matthews Band, but certainly pretty crappy still in the year, in the year 2026.

00:01:37   Yeah, it's still a microphone in the audience. Like there's only so good that can be compared to the soundboard.

00:01:40   That's exactly it. But these are soundboard recordings from Dave Matthews Band. They have finally joined the ranks of their peer jam bands, and we're just going to cruise right on.

00:01:53   They're peer jam bands, and they are offering their concert recordings the following day. This is monumental news for me, and I'm sure for you.

00:02:00   Yes, obviously. What has been missing from my life is a way to get higher quality Dave Matthews Band recordings of all the things I want to listen to from the Dave Matthews Band.

00:02:10   Exactly right. So our long national nightmare is over. Our long international nightmare is over. And I just wanted to make sure that I called that to your attention, Marco, that you can now stream Dave Matthews Band concerts the following day on NUGS.

00:02:23   Yeah, and to be fair, even though Dave Matthews Band is not my cup of tea, I have been so much enjoying for many, many years now that Phish and then later Goose would put for sale for about $10 each every live show they did, direct masterings from the soundboard with very high quality.

00:02:44   With the same mastering you would get if a band released a live album to iTunes and Spotify, that same kind of mastering level. They're doing that every show, and they release it the morning after the show or later that night for $10.

00:03:00   Phish started doing this years ago, and as Goose has gotten popular, they've been doing it, and it is such a joy for your favorite band to basically release a new set, a new album, every night of a tour, so you get, what, 20 or 30 a year?

00:03:16   It's amazing to have that much music coming from a band that you love, and because these are jam bands, plus Dave Matthews Band, the shows are all different.

00:03:27   So you get a different performance, a noticeably quite different performance, quite different set list every night.

00:03:34   Obviously, there are some songs you hear a lot, but you're getting a different performance and a different work every night they play.

00:03:44   By being a fan of these bands, you end up getting a huge amount of new music every year, and it's just a wonderful thing to be a fan of these bands.

00:03:54   So for Dave Matthews to have finally joined that mechanism, I'm happy for the fans like you who get to enjoy that same thing.

00:04:01   Also exciting, you had a vacation that you just went on, and I would love to hear an after-action or trip report, or I guess I should say, in the vernacular of our dear friend John Siracusa, what was your vacation results?

00:04:14   Overall, pretty good.

00:04:15   Spent about five days in Italy, spent four days in Florence, and it was wonderful.

00:04:22   I can strongly recommend Florence as an American tourist.

00:04:26   They make it very easy on us.

00:04:27   It is full of other American tourists, and the Italians are very kind and very accommodating of Americans.

00:04:35   And I can especially call out the Gucci restaurant in Florence is surprisingly good.

00:04:42   It has one Michelin star.

00:04:45   I think it deserves two.

00:04:46   And so that was wonderful.

00:04:47   Nice restorative trip.

00:04:48   You know, a lot of good history to see there and sights to see.

00:04:53   And so that was wonderful.

00:04:54   The art is great.

00:04:55   The shopping is great.

00:04:56   The food's pretty good.

00:04:58   Although, honestly, I kind of prefer New York-style Italian food.

00:05:00   Oh, please email Marco.

00:05:02   Do not email us.

00:05:03   Please.

00:05:03   I mean, look, I also prefer American-style coffee.

00:05:05   Like, you know, I couldn't get, you know, regular drip coffee anywhere in Italy because everything's espresso.

00:05:10   So they had very good espresso.

00:05:12   Anyway, and then, you know, on the way out, we were flying out of Rome.

00:05:17   So we spent one day in Rome right at the end of the trip.

00:05:20   So, you know, we finished up in Florence, took the train to Rome.

00:05:23   The trains are amazing, delightful.

00:05:24   Get to Rome.

00:05:26   I get on the metro to go to the hotel.

00:05:28   It's a crowded metro.

00:05:30   Everyone's kind of cramming into the stuff.

00:05:32   I kind of got, like, pushed over on the way into the car, like, into the subway car.

00:05:36   Now, you know, look, I've read the New York City subway all the time, so I'm used to this.

00:05:40   The doors close, and I realize my wallet's gone.

00:05:43   Oh, no.

00:05:44   I was in Rome for five minutes, and I was pickpocketed.

00:05:49   Gracious.

00:05:49   Well, you know what they say.

00:05:50   What do they say, John?

00:05:52   Apparently, you don't.

00:05:53   No.

00:05:53   I just hope the listeners can finish that thought for me.

00:05:56   What, when in Rome, like that whole thing?

00:05:58   There you go.

00:05:59   It just took a little while.

00:06:00   I was right there with Marco.

00:06:01   I was deeply confused.

00:06:02   Do as the Romans do.

00:06:04   Get your pocket picked.

00:06:05   The Romans did to me what the Romans do.

00:06:07   Did you have it in your front or back pocket?

00:06:10   I had it in my front pocket, but it was super hot.

00:06:14   I was wearing shorts, and so it's a little looser, you know.

00:06:16   And when I'm traveling, I always have, in my front pocket, I always have my wallet and my passport.

00:06:23   My passport never leaves my person when I'm traveling, like, when I'm traveling internationally.

00:06:28   And so the passport was between the wallet and my leg, and so I think the combination of shorts plus that made me not feel it instantly.

00:06:36   But I noticed, like, obviously part of the, like, I did a bunch of research afterwards.

00:06:41   Obviously part of the scam was when I was shoved into the subway car and I stumbled slightly.

00:06:47   That was obviously a physical distraction to get it.

00:06:49   So I didn't know any of this.

00:06:50   This is, like, oftentimes a multi-person operation.

00:06:54   Locketing in apparently specifically Rome and Paris is, like, a really professionalized thing.

00:07:00   Like, it's very, very well done.

00:07:04   There are many YouTube videos on this if you haven't seen them already.

00:07:07   I haven't, but, you know, as I, at first I was, I mean, again, like, I realized instantly, like, oh, that's it.

00:07:13   It's gone.

00:07:13   Like, I knew, but, you know, I didn't want to, like, get into, like, you know, an altercation with anybody in the train in case it was any of the people behind me.

00:07:19   So I'm like, I don't want to, like, well, now I know, okay, now let's go start canceling cards, you know.

00:07:25   But fortunately, I didn't have that much in there.

00:07:27   And even more fortunately, I still had my phone and my passport, which are the two things that would be harder to replace in that context.

00:07:34   And I really didn't need cash or cards for anything on that trip.

00:07:41   You know, the only real loss was, you know, a couple hundred dollars worth of cash that I probably shouldn't have even been carrying.

00:07:45   And the annoyance of having to replace a handful of cards at my driver's license.

00:07:49   But ultimately, it could have been a lot worse.

00:07:51   It was more of an offense and more of a, like, I literally just, like, I was in Rome for five minutes.

00:07:57   And Rome said, F you, right back to me.

00:07:59   And I'm like, okay, you know what, Rome?

00:08:01   F you.

00:08:01   Like, I can't imagine wanting to go back to Rome now.

00:08:05   Like, it kind of ruined it for me, whereas Florence was delightful.

00:08:08   But Rome was like, you know what?

00:08:09   You know what Rome really made me appreciate?

00:08:12   New York.

00:08:13   Because New York City is safer, nicer, like, in terms of, like, the people.

00:08:19   The people are way nicer in New York City.

00:08:20   The people in Florence were nicer, though.

00:08:21   I got to give them credit for that.

00:08:22   And I like the Italian food better in New York than I do in Rome.

00:08:25   So, you know what, Rome?

00:08:26   Please, my apologies to Federico Pettici.

00:08:28   He's delightful.

00:08:28   But my first impression of Rome was really bad.

00:08:32   So, I'm very happy to be a New Yorker right now.

00:08:35   I don't blame you.

00:08:36   And when we were in Rome, this was pre-Declan, we stayed near some, like, plaza where apparently

00:08:43   the pickpockets were legendarily really, really bad.

00:08:45   And I remember I had my hands deep in my pockets every time I walked through there because I had

00:08:51   the fear of God placed in me, including but not limited to when I was walking to the Vatican.

00:08:55   But anyway.

00:08:56   And, like, I now know, like, what I should have done is had, like, you know, a zipper

00:09:01   pocket or one of those front-carrying bags that I'm, like, you know, clutching.

00:09:05   But, like, I ride the New York City subway usually at least once a week, and I have never had any

00:09:12   problem.

00:09:12   This is a Rome problem.

00:09:13   And, like, it sounds like the Italians really don't intend to put any kind of effort behind

00:09:18   reducing this.

00:09:19   So, like, this is their problem to own.

00:09:21   Again, New York is not a perfect place by any means.

00:09:25   America is not a perfect place by any means.

00:09:26   But it did make me appreciate how nice New York is in this one particular way.

00:09:33   Yep.

00:09:34   I totally hear you.

00:09:34   And I am sorry to hear that.

00:09:35   But now you get to go down the path of potentially, if you so choose, buying 84 new wallets to figure

00:09:41   out which one you like the most.

00:09:42   Nope.

00:09:42   Instead, I reordered exactly the same one I had before.

00:09:45   That was the other option.

00:09:46   Fair enough.

00:09:47   All right.

00:09:49   Let's do some follow-outs.

00:09:50   John, you guested on upgrade number 624, which was really a Designed in California segment.

00:09:58   Can you tell me about that?

00:09:59   And if you'd like, or I can handle it, tell me, what is Designed in California?

00:10:02   Yeah, so they're releasing bits of Designed in California in the upgrade feed, both separately

00:10:08   and integrated into upgrade episodes.

00:10:10   But that's going to end once Designed in California gets its own feed.

00:10:12   Designed in California is a spinoff, I guess, podcast series by Jason Snell and Mike Hurley

00:10:18   that is in the style of...

00:10:21   The rest is history.

00:10:22   The rest is history.

00:10:23   It's a history podcast about 50 years of Apple's history.

00:10:26   And they've done a series so far about the Apple II, and they did a Kickstarter, which

00:10:30   we will link in the show until by the time you listen to this, the Kickstarter may be over

00:10:33   because it ends in five days.

00:10:34   If you just go to designed.fm, that's fast, past tense, designed.fm, that currently redirects

00:10:41   to the Kickstarter, and I'm assuming eventually it will redirect or be an actual website.

00:10:44   They're going to do something like 50 episodes, at least 50 episodes of the history of Apple.

00:10:49   And they were kind enough to ask me to guest on some of those episodes.

00:10:53   I'm not sure if they've released everything that we recorded so far, but anyway, whenever

00:10:58   they need me, I'm available.

00:10:59   The bit in Upgrade 624, which we will also link, was about Apple's need for a new operating

00:11:07   system.

00:11:07   So they're having me come on to talk about Mac OS X, surprise, surprise.

00:11:10   So if you want to hear that, check it out.

00:11:12   And there will be more.

00:11:14   There will be much, much more from Designed in California in the coming year or two.

00:11:18   Indeed.

00:11:19   Yeah, I've already backed this.

00:11:20   I backed it immediately.

00:11:21   I enjoy the couple of rest of the rest of history episodes that I've listened to, but I don't

00:11:27   know.

00:11:27   It didn't really rev my engine like it seems to rev everyone else's.

00:11:30   But these done by good friends of mine about stuff I really, really, really care about.

00:11:35   That's Chef's Kiss.

00:11:35   So I definitely encourage you to check it out.

00:11:38   Designed.fm and back the Kickstarter.

00:11:41   You should.

00:11:42   They also made a like pin or something.

00:11:44   I forget exactly what it was now.

00:11:45   But they made this California bear cow thing that I am still grumpy about after having

00:11:52   seen it several days ago because it's so freaking perfect.

00:11:54   So I think it's pictured somewhere, I'm sure.

00:11:57   I can't find it on their Kickstarter right this second.

00:11:59   But they made a, I think it's a pin that you can get if you back the physical stuff.

00:12:03   Like the, one of the tiers is getting physical things from them.

00:12:05   And I think that this, this California cow bear thing is just freaking perfect.

00:12:10   And I'm still grumpy about it, but that's all right.

00:12:12   Anyway, go to design.fm and check it out.

00:12:14   You should, you very much will like it.

00:12:16   Apple intelligence.

00:12:17   When booting Mac OS from an external drive, Doug Winefield writes, there may be a workaround

00:12:22   for doing exactly that.

00:12:23   And there's a link to itecheverything.com and Om Shashad writes, this process involves

00:12:29   modifying a file on the system so that we can trick it into thinking that your external

00:12:32   drive is actually an internal drive.

00:12:34   This cannot be done unless we disable system integrity protection.

00:12:37   Fortunately, the good news is that you can turn it back on after you're done.

00:12:40   I didn't try this, but I'm glad to see that someone figured out how to make it work.

00:12:45   I would be scared to try this, not because of the system integrity protection, but just

00:12:48   because essentially lying to the system and telling them that your external drive is an internal

00:12:51   drive just smells like a formula for disaster, you know, that is just lurking out in the

00:12:57   future.

00:12:57   Are you afraid Eddie's going to show up at your house?

00:12:59   So anyway, I partitioned my, not partitioned, whatever.

00:13:03   I made another volume on my internal drive, so that's how I'm going forward.

00:13:05   But if you are determined to install Golden Gate on an external drive, you can follow these

00:13:11   instructions.

00:13:12   They're long and complicated.

00:13:13   Good luck.

00:13:13   I can say I still have Golden Gate on my portable laptop, and it is so far pretty enjoyable

00:13:21   with the exception that it does keep filling up its hard drive.

00:13:23   Cool.

00:13:24   Oh, that's a bummer.

00:13:25   Mine hasn't had that problem so far, although I have to say that the first beta of Xcode

00:13:29   27 running on Golden Gate could not launch my app in debug mode at all.

00:13:35   So that's bad.

00:13:36   Like, I was trying to figure out, like, is it launching?

00:13:41   What's happening?

00:13:41   It was launching the process, but like the debugger wasn't attaching to it.

00:13:44   So I had to uncheck the debug checkbox.

00:13:45   So then I could run it from Xcode, but not debug it, which really puts a damper on debugging.

00:13:50   Yeah, I'd say so.

00:13:52   All right.

00:13:53   And let's talk about indexing for Spotlight in the 27 OSes.

00:13:56   Steve Riggins writes, I'm on day six and a half of indexing my iPhone with the iOS 27 beta

00:14:01   one.

00:14:02   Alex writes, same here.

00:14:04   I've even had it plugged in for multiple hours each day.

00:14:07   Yikes.

00:14:07   That could be what's filling Marco's drive or something we could be doing haywire.

00:14:11   But if you're wondering how long will it take to do the indexing?

00:14:14   Apparently a long time.

00:14:15   And like we said last episode, the rumor slash word on the street is that Apple is going to

00:14:21   make these indexes for everybody who upgrades to 26.6.

00:14:26   Whenever 26.6 comes out, they will spend, I don't know, five, 10 days or whatever indexing,

00:14:32   probably when our phones are plugged in at night or charging at night so that we don't

00:14:35   have to wait a week after 27 comes out to have the indexes built.

00:14:39   Righto.

00:14:40   Then with regard to Apple's AI tech talk, this was the thing that happened right after the

00:14:44   keynote, what, a couple of weeks ago now.

00:14:46   Yoken Marshall writes, what I didn't get from the tech talk discussion is whether I will be

00:14:51   able to casually tell Siri AI something like, I put the screws for the bed in the closet in

00:14:55   the guest room and have it still remember that two years later, or will I still have to

00:14:59   keep sending emails to myself?

00:15:00   I would guess that that's not something Siri will remember.

00:15:02   Yeah, that doesn't sound like an AI thing.

00:15:05   That sounds like a, like a data storing app thing.

00:15:07   So like that's the kind of thing I would put in Apple notes or something, you know, like

00:15:10   this is, this is going to happen when people like people who haven't been playing with the

00:15:15   LLM chatbots get this on their phone because everyone or, you know, Mac users or Apple users

00:15:20   or whatever, iPhone users have had Siri on their phone for ages.

00:15:23   So either they use it or they don't, but it's not new to them.

00:15:26   And as far as they're concerned, they change Siri again in some way.

00:15:30   Like they don't see it as a different thing.

00:15:31   But this question, you know, points towards like, but it is kind of a different thing because

00:15:36   in the past, if you were to do something like this with Siri, your assumption would be, I

00:15:40   I have to tell Siri to put this in reminders or notes or like, I have to instruct, I have

00:15:46   to like, or say something like remind me and then it will use reminders.

00:15:49   But if you are actually experienced with these type of chatbot things, you're like, well, can

00:15:53   I just ask it to remember stuff?

00:15:55   Because a lot of the chatbot things have what they call a memory.

00:15:59   And you can say things like, remember that I like to use spaces and not tabs in my code,

00:16:03   stuff like that.

00:16:04   And it will quote unquote, remember it.

00:16:07   And the memory feature, as far as I've been able to tell, because it's very difficult

00:16:11   to tell how all these things are implemented, but it's essentially like, it's just going

00:16:14   to write words in a text file and put that into and add that to all the other crap that

00:16:19   goes before the thing that you enter.

00:16:21   So it's going to be the system prompt, all your memory stuff, any skills you have, your

00:16:25   whole previous conversation, and then the last thing that you wrote.

00:16:28   And that all goes into the giant pachinko machine and comes out the bottom.

00:16:32   And so in a sense, it is remembering it by writing it to a text file off to the side.

00:16:37   But the memory is limited and the context window is limited.

00:16:41   And I wouldn't trust that little text file that the thing writes off to the side behind

00:16:46   the scenes to survive for a long term, not just with Siri, but with any of these things.

00:16:50   I don't know what the policy is for compacting or truncating or keeping long term the things

00:16:57   I tell these chatbots to remember.

00:16:59   So my advice is do not assume that even if this works, which it might work for various

00:17:04   chatbots, do not assume that, oh, that's safe forever now.

00:17:07   Tell it to use the app of your choice to store this information so then you can open the app

00:17:13   and look at it and say, yeah, there is now an Apple note that says, here's where I put

00:17:17   the things in the closet for the guest room or whatever.

00:17:20   Have a note that already does that and have the agent write to that note and then look at

00:17:24   the note and see that it's actually there.

00:17:26   And even then, I would be a little bit careful because someday it might wander by and say,

00:17:30   oh, look at this note.

00:17:31   I can fix this for you by summarizing it or some crap like that and you lose all your information.

00:17:34   So be careful out there.

00:17:36   Yeah.

00:17:36   In general, like don't count on the current world of AI to for like long term stability of

00:17:45   any sort, because it's just it's everything is moving so quickly and changing so quickly.

00:17:49   People are, you know, like the big companies are changing their products left and right.

00:17:53   Things change every two weeks.

00:17:54   You know, it's everything is like quicksand right now.

00:17:58   And so for something like like long term memory, long term note or data or fact storage, use

00:18:05   an app that has a proven track record of doing that, like Apple Notes or, you know, bear, you

00:18:10   know, like one of those any kind of like personal shoebox data app like that.

00:18:14   And maybe that is your email.

00:18:15   If you want to email it to yourself, that's another option, too.

00:18:17   But, you know, put the data in something like that and then use AI to index and search that

00:18:23   data in whatever form that can take.

00:18:25   You know, if you're using something like, you know, Gmail, maybe Gemini can search it in a

00:18:29   really good way or something like that.

00:18:31   If you're using Siri, you know, use use Apple Notes and, you know, Siri will be able to read

00:18:35   that.

00:18:36   I think that's the that's the much better way to structure your data like that.

00:18:39   And you can ask the agents to add it to a note, to add it to reminders.

00:18:43   You could ask it the email.

00:18:44   I'm not sure which of those things with Siri will do successfully.

00:18:46   But then you're just using it as the same way you would do, say, remind me when I get

00:18:50   home to blah, blah, blah.

00:18:51   It's it's Siri using the features of the reminders app.

00:18:55   So that's perfectly fine.

00:18:56   But the storage medium needs to be an app that you trust.

00:19:00   You introduced, John, some interesting thought technology to me, I think in Rectifs, and I

00:19:05   think it was forever ago, but the idea of a squirrel list, which I have a pinned note in

00:19:09   Apple Notes called Squirrel List.

00:19:11   And that's where I write where things are that I know I will forget where I put them in the

00:19:15   future.

00:19:15   And that has saved my bacon many times.

00:19:18   And then it's also every time I go to my squirrel list these days, it's like, why did you

00:19:22   not I use it as a verb now, like Googling something?

00:19:24   Why did you not squirrel list this?

00:19:25   Usually I come out ahead, like I had to replace a part of my kitchen faucet.

00:19:30   And I'm like, I've done I've replaced this part before.

00:19:32   And when I replaced it, I know I bought extras because I could tell this is a new newish faucet

00:19:37   and like a year or two into it.

00:19:38   This thing is already having a problem.

00:19:39   I'm probably gonna need more than one of the weeds.

00:19:40   So I bought spares.

00:19:41   But where in my house are the spares go right to the squirrel list.

00:19:44   Is it there?

00:19:44   No, of course it's not.

00:19:46   Of course.

00:19:46   I did find it, though.

00:19:47   And you know what I did after I found it?

00:19:49   I put it on the damn squirrel list.

00:19:50   So set a time or three years from now when that stupid magnetic ring.

00:19:54   And the Delta faucet rusts out again.

00:19:55   I know where to find the spares.

00:19:57   Perfect.

00:19:59   There is a bootleg video, if you will, of the aforementioned Siri AI talk where Craig

00:20:06   is talking, at least in the beginning.

00:20:07   This is via Sayrer, S-A-Y-R-E-R.

00:20:09   This guy that Sayrer links to has three such videos, but says, Sayrer, I saw one from straight

00:20:17   on that has the opening quote that we talked about.

00:20:19   I can't find that one, unfortunately.

00:20:21   And then we have a link to a YouTube short where one of these videos can be seen.

00:20:25   Yeah, see, they should go on Nugs, man.

00:20:26   Like, imagine if there could be an official video instead of audience recordings.

00:20:30   That's exactly right.

00:20:31   It's a whole new world for me, John, I tell you.

00:20:34   All right.

00:20:35   And then with regard to some of the new tools and photos, Kevin Buderbaugh writes,

00:20:41   is the new spatial reframing feature in the 27 releases geographically aware.

00:20:45   For example, my wife and I recently had our picture taken together by someone else, as

00:20:48   in not a selfie, while we were standing on the pier in somewhere unpronounceable in California.

00:20:52   The person taking the picture was facing south, so Morro Bay and Morro Rock were in the background.

00:20:56   The person who took the picture managed to frame it such that my head hides Morro Rock.

00:21:00   If I were to use spatial reframing on it, would it know to fill in Morro Rock, or would

00:21:04   it just generically fill in matching coastline?

00:21:06   I have to assume it would just fill in coastline.

00:21:08   Yeah, I don't know for a fact, but I would put very good money on, no, it has no idea where

00:21:13   you are, and it knows your location, but it's not going to use any kind of awareness of photos

00:21:18   taken by other people in that location, which, again, would be a privacy nightmare that Apple

00:21:21   wouldn't really delve into.

00:21:23   It's just going to use the training data of all the photos that its image generator has

00:21:26   seen and say, here's what plausibly could have been behind you.

00:21:29   Now, it could be that the training data includes lots of images of Morro Rock.

00:21:33   You're like, wow, it is location aware.

00:21:35   When I move my head, it shows me exactly the right shape of Morro Rock.

00:21:38   That just means Morro Rock was in the training data.

00:21:40   I don't think it is basing it on your GPS.

00:21:42   Now, again, I don't know for a fact, because I didn't make this feature, so if someone at

00:21:46   Apple knows otherwise, but I would put good, good money that that's not the way these things

00:21:50   work currently.

00:21:51   It's just going to use the model, and whatever the model puts is what the model puts.

00:21:55   Yeah, concur.

00:21:56   All right, let's talk about Tesla and CarPlay, or in other words, let's have Casey pop off.

00:22:01   Alec Hertel points us to an article on NotATeslaApp.com, which is entitled, Apple announces Maps feature

00:22:09   that could finally bring CarPlay to Tesla.

00:22:10   And Nahal Malek, from the aforementioned website, writes, reports emerged last fall of Tesla actively

00:22:17   pursuing native CarPlay integration.

00:22:18   Follow-up reports from earlier this year indicated that CarPlay integration was still in the works,

00:22:22   with Tesla reportedly working directly with Apple to bring the interface to its vehicles.

00:22:26   At the time, the main holdup was said to be how the navigation would be handled between

00:22:30   Tesla's, whatever they think, their full self-driving, except it isn't really full self-driving system,

00:22:36   and CarPlay's own maps.

00:22:38   Because supervised self-driving relies heavily on the car's native navigation, Tesla and Apple

00:22:44   were working together to ensure that turn-by-turn directions stay in sync across Tesla's and

00:22:47   Apple's mapping platforms.

00:22:49   If this supervised automatic driving doesn't know where the CarPlay map is going, features

00:22:55   like automatic lane changes, which I don't understand why that's a problem, but anyway,

00:22:58   and supervised self-navigation simply can't function.

00:23:01   During a WWDC 26 session covering the latest updates to CarPlay, Apple announced a new feature

00:23:07   called route sharing.

00:23:08   And we will put a timestamp link to the YouTube video in the show notes.

00:23:11   Route sharing allows navigation to pass, navigation app, excuse me, to pass a trip to the vehicle

00:23:16   as an array of route segments, which are geographic coordinates that are sent to the vehicle whenever

00:23:20   the trip changes.

00:23:21   Apple notes that, quote, some vehicles with driver assistance systems work best when the

00:23:25   intended route is known.

00:23:26   For example, vehicles may support automatic lane changes or adjust their guidance systems

00:23:29   to more closely match the route shown in your app, quote.

00:23:32   So, I really, really want this to happen, not because I want Teslas to be more appealing

00:23:38   to anyone, if that's even really possible, but because if Tesla finally caves, because they

00:23:45   were, you know, the kings of, no, we're too good for CarPlay, then maybe Rivian will cave

00:23:50   and maybe GM will cave.

00:23:52   And yes, I am living in a fantasy world where things go my way.

00:23:56   And no, that's not the reality world that I'm also living in.

00:23:58   But man, I can dream, can't I?

00:24:00   Well, maybe.

00:24:02   I like how they didn't name Tesla in WWDC session, you know, because they're, but it's like the

00:24:07   facts match up exactly that apparently this was what Tesla was waiting for.

00:24:10   Now, does this mean Tesla will actually do it?

00:24:11   We'll see, but this rumor has been cooking for a while, like, and the holdup being we

00:24:16   can't do this because we don't, you know, there's not enough sharing around information

00:24:19   to support our other features makes technical sense.

00:24:22   But I still wonder about how committed they are to doing this, but we'll see.

00:24:28   Yeah, I just, I do think it's interesting because this does not strike me as, if you're, if you're

00:24:34   coming at this as a Tesla fan, which is a fantasy world that I find hard to inhabit, but if you're

00:24:39   coming at this as a Tesla fan, this does not strike me as something you would do from a

00:24:43   position of strength because you've banged this drum for so long that, oh, we're too good

00:24:47   for CarPlay.

00:24:48   Our stuff is better than CarPlay.

00:24:49   We have the best software in the whole car industry because we are amazing.

00:24:53   And yet...

00:24:55   And they do it in that voice too.

00:24:56   And they do it in that voice.

00:24:57   And yet here they are potentially caving and saying, sure, we'll support CarPlay.

00:25:02   I am struggling to find any reason to do that other than, man, it would be nice if we sold

00:25:07   more cars these days.

00:25:08   Maybe that's just me.

00:25:10   I have incredibly, I have, you know, Marco, this is your episode.

00:25:14   It just occurred to me because not only do you have access to what will eventually be just

00:25:19   hours and hours and hours of Dave Matthews band on goose.net or excuse me, nugs.net, but

00:25:24   the U S has approved a new sunscreen ingredient reading from scientific American.

00:25:29   The U S is finally getting a new, better sunscreen ingredient.

00:25:32   The Food and Drug Administration added bemotrisinol, there, we'll go with that.

00:25:38   And I tried this earlier and I already forgot what I tried, but bemotrisinol, an effective

00:25:43   chemical filter that's used in sunscreens made in Asia and Europe for decades to the list

00:25:47   of permitted active ingredients and over-the-counter sunscreens.

00:25:49   The list hasn't seen a new entry in more than 20 years.

00:25:52   Yeah.

00:25:53   So this is, you know, as I was going through my sunscreen journey, I believe it was last

00:25:57   year or the year before I had realized that, um, avobenzone, which is the chemical filter

00:26:02   used in almost every U S sunscreen, um, except for the mineral ones, but all the chemical sunscreens

00:26:07   or the hybrids, they all had avobenzone.

00:26:09   And it turns out avobenzone is both not incredibly effective relative to modern, uh, sunscreen filter

00:26:15   chemicals.

00:26:16   And also the problem that I was having was it's incredibly irritating to most people's eyes.

00:26:22   My eyes would just burn and get painful, uh, you know, redness and everything all day.

00:26:27   If I, if I had sunscreen on, because even if like a little tiny bit traveled, oh, and avobenzone

00:26:33   travels through, through the skin through short distances too, um, it migrates.

00:26:37   And so if you had sunscreen anywhere on your face, odds of it getting in your eye, uh, were

00:26:42   pretty high.

00:26:43   So at that point I switched over, uh, I tried European sunscreens and Japanese sunscreens.

00:26:48   Um, and I found all these other chemicals that are more modern that are much, not only

00:26:53   more effective in most cases, but also allow much nicer formulations of the sunscreen.

00:26:58   Like, you know, like how does it absorb?

00:27:00   Does it have good textures, you know, stuff like that.

00:27:01   And then it doesn't migrate through your skin and it's not as irritating to eyes, like by a

00:27:05   huge, huge margin.

00:27:06   So the one that I use the most and love is the Roto Super Moisture UV Gel.

00:27:13   It's a Japanese brand.

00:27:14   I'll link in the show notes.

00:27:16   Um, sometimes you can find people on Amazon selling it.

00:27:18   It's more reliably found on eBay, or you can just order it from Amazon Japan if you're

00:27:21   willing to pay a lot of shipping.

00:27:22   Um, but one of the filters it uses is this one.

00:27:26   Uh, this is also called, uh, Tinosorb S.

00:27:29   Uh, it's available in a lot of European and Japanese sunscreens.

00:27:33   There's also, there's a whole community of Korean sunscreen enthusiasts.

00:27:36   I didn't get that far because I, I basically got to Japan and was happy and stopped.

00:27:40   Um, so, uh, but there's this whole world of sunscreens that now the U.S. has approved one

00:27:46   of the chemicals that they use.

00:27:48   And so therefore we are likely to see better U.S. sunscreens that are not irritating to people's

00:27:54   eyes or not as irritating to people's eyes, uh, and work better against, you know, different,

00:27:58   different forms of UV.

00:27:59   So this is great news for the sunscreen enthusiast world.

00:28:02   I'm going to still keep buying my, uh, my Roto Super Moisture UV gel just because I like

00:28:07   it a lot.

00:28:07   Like the formula is good.

00:28:08   The bottles are convenient.

00:28:10   It, it has, it has a scent, but it's a very, very weak scent.

00:28:13   Um, so it's fairly neutral and, uh, and it worked great for me.

00:28:17   So I'm going to keep buying that, but, uh, hopefully I'll have more options soon.

00:28:20   You have a lot of faith in the U.S. sunscreen industry.

00:28:23   I just assumed looking at this story that, uh, yeah, they'll, they'll put this ingredient

00:28:27   in it and they'll also include Avobenzone.

00:28:30   Well, to be fair, like, I don't think, I mean, I don't, I haven't looked at it exactly, but

00:28:34   I don't think it would make sense to have both just because they probably cover similar UVA

00:28:38   and UVB bands.

00:28:39   So like, I don't, it probably wouldn't make sense for the same product to include both.

00:28:43   I'm not saying it makes sense.

00:28:44   It just seems like a thing that the manufacturers would do.

00:28:46   Yeah.

00:28:46   Like they would put a trace amount of this so they could put it on the label and raise

00:28:49   the price, but just keep it the same as it was.

00:28:51   Yeah, probably.

00:28:52   I mean, you wouldn't believe like how many sunscreens, how many like U.S.

00:28:55   sunscreens I tried first that claimed to be like sensitive or, you know, all these

00:28:59   different marketing words.

00:29:01   And at the end of the day, they all had the exactly the same problem because they were all

00:29:04   using the only, uh, U.S.

00:29:07   chemical filter that was rated for UVA and UVB.

00:29:09   All right.

00:29:10   We are going to dig up some freaking ancient follow-ups that we're finally going to, uh,

00:29:17   exhume and, and get through.

00:29:19   These are not that remarkable, but it's hilarious to me how long these have lived in our internal

00:29:22   show notes.

00:29:23   So like 18 years ago, we were talking about tracking Apple's environmental progress.

00:29:28   And Andrew Leahy writes, your mention of diffing the environmental impact report on the latest

00:29:32   Apple products gave me an idea.

00:29:34   I fed historical environmental impact reports into TOS tracker, which is a terms of service

00:29:39   tracker, if I'm not mistaken, and we'll keep track moving forward.

00:29:41   Mind the bugs.

00:29:42   Some report PDFs are still being ingested, although this was like three months ago.

00:29:46   Uh, and you can go to the TOS tracker website and we will put a link in the show notes where

00:29:50   you can see this and Andrew continues.

00:29:52   If you go to an individual product, you can compare directly across whatever metrics Apple

00:29:55   provides.

00:29:56   And so there will be another link for, for example, the M1 MacBook Air versus the M5 MacBook

00:30:01   Air.

00:30:01   There you go.

00:30:02   The magic of technology.

00:30:03   Yeah.

00:30:03   These, these long lasting, uh, follow up, like lots of things have died out here, but these

00:30:07   are the ones that survived.

00:30:08   It's just because I thought like we talk about this and diffing like this is a pain and who

00:30:11   has time for that?

00:30:12   Well, if there's a website for it, you just bookmark the website.

00:30:14   And if you're wondering what kind of progress is Apple making on the environment stuff, are

00:30:18   they just doing the same stuff year after year and just retouting it over and over again?

00:30:21   And then finally, uh, in episode 680, we were talking about, or we had questions about how

00:30:27   Apple's AI servers work.

00:30:29   Uh, this was keyed off of a wall street journal video that we were talking about in again, episode

00:30:34   680 back in February and friend of the show, Guy Rambeau writes, no need to rely on rumors

00:30:38   or leaks for what Apple's AI servers are running.

00:30:40   The software images they distribute to security researchers include details about the hardware

00:30:44   with regard to how each ultra chip is addressed.

00:30:46   We know that as well.

00:30:48   Each one of those chips is a single PCC node or private cloud compute node running as a standalone

00:30:52   computer with cloud OS, which is a variant of iOS.

00:30:55   And there's an orchestration layer that handles distributing work within that ensemble, which

00:31:00   is what Apple calls it.

00:31:01   You can also see there's a cloud or was at the time anyway, a cloud OS job listing, which

00:31:06   includes the cloud OS team is responsible for all facets of delivering OS and system

00:31:10   services on Apple Silicon servers, including driving hardware and software initiatives to

00:31:15   enable new Apple Silicon based systems in data centers.

00:31:18   Finally, Zoe Knox contributes.

00:31:20   Here's Apple's documentation on how to run the PCC virtual environment.

00:31:23   And we will put a link to that in the show notes in there.

00:31:26   It says the private cloud compute virtual research environment or PCC VRE is a set of tools and

00:31:32   images that can boot a version of PCC software and simulate a PCC node on a Mac with Apple

00:31:37   Silicon.

00:31:37   You got to get this item in right before I would

00:31:40   imagine potentially this whole tech stack becomes irrelevant.

00:31:44   It's not, we don't know that this is going to happen, but now that they've moved to using

00:31:48   NVIDIA servers and other people's data centers, I do wonder how long they're going to continue

00:31:51   the effort of building their own servers.

00:31:54   That Wall Street Journal video like showed us inside the factory where they have those big

00:31:57   giant rack mount servers that had a bunch of, you know, M2 Ultras or whatever inside

00:32:00   them building their hardware, racking the hardware, supporting it, writing an OS called

00:32:06   cloud OS that goes on there doing the security.

00:32:09   Like this is a hell of an effort for what I have to imagine at this point is just not

00:32:14   competitive at all in any measure like an M2 Ultra.

00:32:17   It was a great chip back in the day, but boy, we've moved on compared to today's NVIDIA chips

00:32:22   that, you know, Google's got in its data centers.

00:32:25   Is Apple going really going to keep up with the state of the art or are they just going

00:32:30   to say, you know, what Google is doing?

00:32:32   We're also calling that PCC.

00:32:34   We'll make sure it has all of the same properties as the PCC that we did where, you know, they

00:32:38   already in WWC is that we're not distinguishing between one or the other.

00:32:41   It's all PCC to us.

00:32:42   Right.

00:32:42   And they promised that by the time the 27 OS is shipped, Google's PCC will be just

00:32:47   as good as Apple's PCC.

00:32:48   And I just feel like that's the clearly the path forward.

00:32:51   But for now or in the past, they have their own servers that they built in their own factories

00:32:57   that they ran cloud OS on and they were had job listings for it.

00:33:00   So I don't know how it's going to go.

00:33:01   Tune in next year to see if this is still a thing.

00:33:04   I can't imagine it would be like there's there is not a lot of worlds where this makes sense

00:33:10   for Apple to do long term while, you know, Google and AWS and everybody and, you know,

00:33:15   Microsoft, like all these other companies are running giant data centers that are specialized

00:33:19   for exactly this purpose with the newest cutting edge, everything from everybody.

00:33:23   Like it just doesn't make sense for Apple to do this themselves.

00:33:26   Yeah.

00:33:27   And I kind of see how they got there, like in the old regime, the old regime that was wiped

00:33:32   away and replaced with this new regime that gave us, you know, WWC 2026.

00:33:37   In the pre-WWC 2024 regime, the idea of PCC is like, what if we could do this thing where

00:33:42   we can do stuff in the cloud, but in a secure way and where we have great silicon and like

00:33:47   it's very efficient and we could make our own servers and we could put our own OS on it.

00:33:52   And like then we would have it's basically like, well, you could run on your device and

00:33:55   you can also run in this thing, which is basically as secure.

00:33:58   Can we make something that's not on your device as secure as being on your device?

00:34:01   That's the whole idea between PCC.

00:34:02   Like just like the stuff that stays on your devices, you know, we can't see it.

00:34:05   Nobody can see it as encrypted, whatever.

00:34:06   Can we do that?

00:34:07   But on the server, like it makes so much sense as a thing to do.

00:34:10   And it made sense to like the WWDC 2024 timeframe to announce it.

00:34:15   If only all of that actually ever shipped and it didn't.

00:34:18   And the world moved on and it's like, okay, well now the servers are still M2 ultras and

00:34:23   the world, like the world has moved on.

00:34:25   This is, you know, I'm not going to bang this horn drum too much, but like if Apple wants

00:34:30   to compete in the high end of hardware anywhere, including on the server, you have to make new

00:34:34   new high end chips on a fairly regular basis because the state of the art in things that

00:34:39   run in data centers and do AI stuff is moving rapidly.

00:34:43   The M2 ultra is positively ancient.

00:34:45   Even if they've replaced them with M3 ultras, also pretty ancient.

00:34:48   And where's the M5 ultra sometime in 2026.

00:34:52   Meanwhile, the world is racing forward.

00:34:54   So I just don't think Apple has it in them to compete hardware wise.

00:34:58   And if you can't compete hardware wise, the whole effort is pointless, even if your software

00:35:01   stack is really cool.

00:35:02   So yeah, just have Google do it with NVIDIA GPS.

00:35:06   We are brought to you this episode by Squarespace, the all-in-one website platform designed to

00:35:12   help you stand out and succeed online.

00:35:13   Whether you're just starting out or scaling your business, Squarespace gives you everything

00:35:17   you need to claim your domain, showcase your offerings to the professional website, grow

00:35:22   your brand and get paid on one place.

00:35:24   This is quite simply the best place to build a website, especially for your business.

00:35:28   So Squarespace has cutting edge design that's super easy to use.

00:35:32   So you can make a beautiful professional website with all of their built-in tools.

00:35:36   Everything is visual.

00:35:37   You're not like coding anything.

00:35:38   You're not dealing with software updates or anything like that.

00:35:40   And it's super easy for anyone.

00:35:42   So whether it's you or someone who like comes to you as the nerd and was like, hey, help

00:35:45   me make a website, you can say go to Squarespace and they can figure it out themselves, which

00:35:49   is super empowering and leaves your time free, honestly.

00:35:52   So it's great for businesses too.

00:35:53   They have all these amazing business features.

00:35:55   You can, of course, sell products on Squarespace, physical or digital goods.

00:35:59   They have all sorts of things for physical goods like shipping integrations, tax integrations,

00:36:03   things like that.

00:36:04   And then for digital goods, it goes even further, of course.

00:36:06   Because it's way more flexible.

00:36:07   You can have things like member gated areas.

00:36:10   You can have member content.

00:36:11   You can have digital e-books and stuff you're selling.

00:36:14   You can have private podcasts, newsletters, like all sorts of stuff.

00:36:18   Even if you're a consultant or something, you can build a website that allows you to book

00:36:21   time slots for your customers.

00:36:23   It's amazing.

00:36:24   All this on Squarespace is backed by things like powerful analytics and SEO tools that,

00:36:28   of course, most businesses want to have.

00:36:30   Email campaigns.

00:36:32   All sorts of amazing features on Squarespace.

00:36:34   And again, it's all super easy to figure out for you or anyone who's non-technical.

00:36:39   So it's an amazing service.

00:36:41   So you can focus on growing your business.

00:36:43   Go to squarespace.com slash ATP.

00:36:47   And you can save 10% off your first purchase of a website or domain by using that link.

00:36:51   And you can try it all in trial mode first.

00:36:53   So you can build it, see how it works for you.

00:36:55   Check it out.

00:36:55   Squarespace.com slash ATP.

00:36:57   Thank you so much to Squarespace for sponsoring our show.

00:37:05   On June 17th, which is we record this on the evening of the 25th, it was a week and a day

00:37:10   ago, Tim Cook took one on the chin for John Ternus and did an interview with Rolf Winkler

00:37:17   of the Wall Street Journal.

00:37:18   Rolf writes, Apple plans to raise prices on its products to offset the surging cost of memory

00:37:23   and storage chips, Chief Executive Tim Cook said in an exclusive interview with the Wall

00:37:28   Street Journal.

00:37:28   Quote, unfortunately, price increases are unavoidable, Cook said.

00:37:32   We're doing our best to mitigate the huge increases that are being passed to us.

00:37:35   And we've been trying to shield our customers from the increases, but the situation has become

00:37:40   unsustainable.

00:37:41   Cook declined to offer details on timing or scale of the planned price increases, nor which

00:37:47   products would be affected.

00:37:48   Just you wait.

00:37:49   But they included a chart in the Wall Street Journal, price changes from the first quarter

00:37:53   of 2023, and the baseline, or it starts at, like I said, 2023.

00:37:58   And there's, you know, little blips and blurbs for DRAM memory and NAND storage, how expensive

00:38:03   they are.

00:38:04   And it got a little bit more in 25, and then toward the end of 25, it just hockey sticks.

00:38:09   And the top of this chart, which is what DRAM is estimated to be at the end of 2027, 900%

00:38:18   above what it was in 2023.

00:38:21   Now, again, that's an estimate, but holy jamolies, not great.

00:38:25   This is one of those charts where you're like, please do look at the y-axis.

00:38:28   It is rooted at zero.

00:38:29   And you're like, yeah, but these graphs always exaggerate stuff.

00:38:32   Yeah, it looks like they go up a lot, but I bet it's like 0, and then the top of the

00:38:35   thing is 0.01.

00:38:37   No, the top is 900.

00:38:38   It's a percent.

00:38:40   It's a terrible, terrible, scary graph.

00:38:44   And by the way, Casey, it pains me when you skip my terrible puns.

00:38:47   You need to put them in.

00:38:48   Which one?

00:38:49   Maybe it's my pun.

00:38:52   This is on me, because my pun is so bad that Casey didn't even recognize it as a pun, which

00:38:55   I'm going to take.

00:38:56   That's on me.

00:38:56   That's not on Casey.

00:38:57   The way I described this was Tim Cook takes one for the Ternus.

00:39:01   Takes one for the T.

00:39:02   He's taking one for the team.

00:39:03   I know team and Ternus, they both begin with T.

00:39:05   It's a stretch.

00:39:06   It's on me.

00:39:07   It's not on you.

00:39:08   No, I'll take some of that blame.

00:39:10   That was a group effort because my reading comprehension, as always, has failed me.

00:39:13   I didn't see for the Ternus.

00:39:16   I just read that as takes one for Ternus.

00:39:18   Yeah, I should have done your thing and had team and strikethrough and then had Ternus.

00:39:22   Then you would have figured it out.

00:39:23   Anyway.

00:39:23   That's true.

00:39:23   Again, this was June 17th.

00:39:25   Obviously, we know where this goes because today is June 25th, and stay tuned.

00:39:27   Not there yet.

00:39:29   Yeah, but when this came out on June 17th, everyone was like, oh, that's nice of Tim.

00:39:35   Because we all know this bad news is coming.

00:39:38   When we talked about the RAM crisis and everything and what Apple's doing, and we talked about

00:39:45   how Apple has long-term contracts for its parts.

00:39:48   They don't buy them the day they make the things.

00:39:50   They lock in a price for a certain amount, and they do that potentially a year or two in

00:39:54   advance.

00:39:55   But that timer runs out eventually.

00:39:57   Like, eventually, all those deals that you made one or two years ago, those end.

00:40:00   And then you have to buy things at market prices.

00:40:02   And then Apple is not magical and immune from market forces.

00:40:05   So even though we saw that Apple was holding the line on prices, it's like, well, if the

00:40:10   commodity prices don't turn around and Apple runs out of all its deals, eventually, they're

00:40:15   going to have to make some hard choices.

00:40:18   Tim Cook could have said, yeah, well, you know, we'll just leave this and John Ternus

00:40:22   can announce the increases.

00:40:23   Now, we don't know whether, like, they had to be announced now, and so it was always going

00:40:29   to be Tim Cook because Ternus doesn't take over until September, or whether he was doing

00:40:33   what people thought he was doing on the 17th.

00:40:35   Because remember, on the 17th, we didn't know when the price increases were coming or what

00:40:37   they would be.

00:40:38   So on the 17th, people could have been saying, well, it'll probably be the iPhones in September,

00:40:42   and, you know, it'd be bad if Ternus' first keynote, he's going to announce all the

00:40:45   iPhones are, like, hundreds of dollars more expensive, so let Tim Cook on its way out the

00:40:49   door, soften everybody up, and then when Ternus announces it, it's like, well, we've already

00:40:53   known this was going to happen, right?

00:40:54   Now, you know, that turned out not to be the case, as we see in a second, but this is, you

00:40:59   know, this is one of, like, the, instead of a strategy tax, it's like a strategy credit

00:41:04   or whatever, not quite the same thing, but, like, when you have something like this happen,

00:41:07   you have a transition, there's lots of downsides to a transition, change is scary, and there's

00:41:13   lots of instability, and you're not sure how things are going to go.

00:41:15   But one of the advantages of a transition is you can do stuff like this, have the guy

00:41:20   on his way out the door do stuff that is unpalatable.

00:41:24   No, I'm not talking about the Trump stuff, he did that while he was not out the door.

00:41:27   Like, announce, be the bad guy, announce the price increases, don't make your new CEO do

00:41:34   it, and that is totally a Tim Cook move.

00:41:36   And as we'll see, it wasn't really taking one fraternist, it seems like it might have been

00:41:40   just like something that needed to be announced now, and Tim Cook is currently the CEO, so

00:41:43   it is what it is.

00:41:44   But yeah, and so that chart, the chart is terrifying, please, we'll put a link to the image, but

00:41:50   you can just go to the article, I think you can see it, even if it's paywalled.

00:41:52   But if not, look at the image in our LinkedIn, our show notes.

00:41:55   And then there were some tangents here to the story to say just this chart here, again,

00:41:59   that goes up to 900%, how, this has ripple effects throughout the entire world, and so

00:42:04   this Wall Street Journal story continued with some more fun details on this.

00:42:07   Righto, so from the Wall Street Journal, three companies dominate the market for DRAM memory,

00:42:12   Samsung and SK Hynix in South Korea, and Micron in the US.

00:42:16   Makers of NAND storage include those three companies, as well as Keoxia and SanDisk.

00:42:21   Their stock prices, along with their profits, have exploded over the past 12 months.

00:42:25   You don't say.

00:42:26   Micron and SK Hynix shares have risen more than 800%, while Keoxia and SanDisk have risen

00:42:31   4,600%.

00:42:36   Anybody buy any SanDisk stock in 2024?

00:42:40   Good.

00:42:41   Not me.

00:42:41   Gravy.

00:42:42   Good gravy.

00:42:43   So this has led to some interesting corollaries, like John was saying.

00:42:46   So reading from CNBC, few workers can say that their bonuses have been so large that the

00:42:52   company's central bank takes notice.

00:42:54   But in South Korea, the phenomenon is playing out as workers from tech industries receive

00:42:57   bonuses worth millions of won, prompting the Bank of Korea to warn of the upward pressure

00:43:02   of inflation.

00:43:03   According to an unidentified union source cited by Reuters, a memory chip worker with a base

00:43:08   salary of 80 million won, or about 52 grand US, is expected to receive a total bonus of around

00:43:13   626 million won or 410,000 US dollars this year.

00:43:19   Holy crap.

00:43:20   Imagine you're working in the memory chip factory for 50k a year and your bonus is going to be

00:43:25   400k.

00:43:25   That's a pretty good year.

00:43:26   Yeah.

00:43:27   Good for them.

00:43:28   I tried to find more details.

00:43:30   I think like some of those, their union labor and then unions had negotiated a, like a profit

00:43:35   sharing deal before all this happened, which was like, you know, we get some piddling percentage

00:43:40   of the company's profits as a bonus.

00:43:41   Uh, little did they know that the profits are about to go off like a rocket ship.

00:43:47   And so what seemed like a reasonable deal of profit sharing with the union now becomes,

00:43:51   uh, guess what?

00:43:52   You get a 400k bonus this year.

00:43:55   Diamond hands all the way.

00:43:56   Uh, and then, uh, there was a Reddit post that I presume John stumbled upon, which is quite

00:44:01   funny.

00:44:01   Someone took a screenshot of a chat GPT conversation.

00:44:04   And I guess what they're doing is they're building a custom PC or something like that.

00:44:08   And again, this is a screenshot from chat GPT.

00:44:11   Chat GPT says to this individual, that is a very solid build, but I spotted one thing

00:44:17   immediately and then big warning, rotating light Ram prices wrong.

00:44:22   Your screenshot shows, you know, blah, blah, blah, blah, blah, 32 gigs of Ram, $420.

00:44:27   That has to be a pricing error or the company picked some weird seller.

00:44:31   The 32 gigs of Ram should be around 80 or $120, not 420.

00:44:37   If I'm still quoting now, if you actually paid $420 for 32 gigs Ram, I'd call the police

00:44:43   laughing with tears in my eyes.

00:44:45   Oh, chat GPT.

00:44:47   Incredible.

00:44:47   It's, it's, it's the hot dog, uh, the, the meme with the guy in the hot dog suit saying

00:44:51   we're trying to find the guy who did this.

00:44:52   I'm sure I've seen that, but I don't know it offhand, but no matter what, how do you not

00:44:56   see these memes?

00:44:57   You have children.

00:44:58   It's going to help you old people see memes.

00:45:01   Well, all I know about is six, seven, John.

00:45:03   That's the only thing I know these days that bro is used as a comma anyway.

00:45:06   Uh, yeah, this is incredible.

00:45:08   I mean, it's sad and terrible if you're not an employee of those memory companies, but it

00:45:13   is.

00:45:13   Yeah.

00:45:14   They're going to have to use all that money to buy their next PC though.

00:45:17   Yeah.

00:45:18   It all comes back around.

00:45:19   Right.

00:45:19   So then that was all June 17.

00:45:22   Again, that was, uh, Wednesday, June 17.

00:45:24   It is now as I'm sitting here today, Thursday, June 25.

00:45:28   And guess what?

00:45:29   Apple's raised prices.

00:45:31   So there's another post at the wall street journal.

00:45:33   We'll put in both for the prior one we were discussing in this one, we'll put in an Apple

00:45:36   news plus link if you happen to be a subscriber.

00:45:38   But anyways, one way or another, Apple has said the consumer electronics industry is facing

00:45:43   an unprecedented challenge.

00:45:44   The rapid expansion of AI data centers has created an extraordinary, extraordinary surge

00:45:48   in demand for memory and storage.

00:45:50   We have never seen a component price increase this much, this quickly.

00:45:53   We have shielded our customers from these increases so far, but we have now reached a point where

00:45:58   we need to begin raising prices on a number of products, including today's increases for

00:46:01   iPad and for Mac.

00:46:03   We know this is not welcome news and we are working tirelessly to find solutions.

00:46:07   We will put some links in the show notes.

00:46:08   Uh, our dear friend, Stephen Hackett had a good rundown of this.

00:46:12   There's a good post on Mac rumors.

00:46:13   And additionally, Gruber had a post that we'll end up discussing here in a moment.

00:46:17   But the, the 50,000 foot view is that basically everything that isn't an iPhone, if I'm not

00:46:24   mistaken, um, has gone up at least 10 to 15% and in some cases as much as 60%.

00:46:33   And I'm just going to jump to the, the, what I was going to say later, but I can't resist.

00:46:36   I'm going to say it right now.

00:46:37   The Apple TV that is ancient has gone from $129 for the base model to $200, which is a

00:46:47   $70 increase or 54%.

00:46:51   The nice Apple TV, the nice Apple TV that has thread and ethernet is $250 to put things in

00:47:00   perspective.

00:47:00   And I know this doesn't mean as much to everyone else as it does to me, but I like

00:47:03   to think we've all joined in my familial journey over the years to put things in perspective.

00:47:08   My daughter, Michaela was not in elementary school when the last Apple TV was released.

00:47:13   It was, I forget how many days ago now.

00:47:16   I don't have the buyer's guide in front of me out, but it was thousands of days, like 1300

00:47:21   days ago or something like that.

00:47:22   And, uh, she was not even a kindergartner yet when that was in, when that was released.

00:47:28   Let's see.

00:47:28   What, what is it?

00:47:29   As I stole for time, 1,346 days ago when the Apple TV was most recently released October of

00:47:34   2022.

00:47:34   My daughter, who was not yet in kindergarten, just grad, or I shouldn't say graduated, just

00:47:40   finished second grade, has gone through three years of school in the time since we have had

00:47:45   a new Apple TV.

00:47:46   What the fuck is going on?

00:47:48   And why the fuck do they think this thing is worth $250?

00:47:50   Are you out of your minds?

00:47:52   What is happening?

00:47:53   I think the more expensive one went up by 67%.

00:47:56   Oh God, it's even worse.

00:47:58   But I have to say this, I find this one of the least objectionable price increases, actually.

00:48:02   Obviously, absolute value.

00:48:04   It's not a lot because yeah, it's 70 bucks.

00:48:06   Percentage wise is big, but it is a cheap product.

00:48:08   But the thing about the Apple TV is like, I'm honestly, I don't quite know why Apple hasn't

00:48:13   been gouging us even more on it.

00:48:14   It's because as time passes, it's competition in the realm of thing you use to watch streaming

00:48:20   video on your TV has just gotten like worse.

00:48:23   And that is fair.

00:48:24   And so it's just like, what would you pay for an Apple TV?

00:48:27   And the answer is I'd pay more than 250 to not have to use like smart TV software or like

00:48:32   a streaming stick or some other weird thing.

00:48:34   I'm not encouraging Apple to increase prices.

00:48:36   This is all terrible.

00:48:37   But that particular one doesn't bother me that much, although it is, I believe, the highest

00:48:41   percentage increase.

00:48:42   And what we're going to mostly be talking about here are base prices, because obviously we're

00:48:46   getting to upgrade prices in a second.

00:48:47   But the base prices have gone up.

00:48:49   And this is for like, they say Macs and iPads, but Vision Pro is in there too, which is great.

00:48:53   The average price increase is $258 and the average percentage increase is 21.5%, which is

00:49:01   bad.

00:49:01   This is bad going up.

00:49:02   But like, depending on what you're looking at, as the base price goes up, the percentage

00:49:07   increase, the absolute value goes up as well.

00:49:09   So Apple TV, I think is the worst with the base price going up 54%.

00:49:13   Some of them only went up like 18, 17, 15, the big expensive Macs with the big CPUs in

00:49:19   them, like the M3 Ultra Mac Studio, 32.5% increase, which isn't that bad.

00:49:25   But that's an additional $1,300 to the base price, which is rough.

00:49:30   Vision Pro got out easy with only a 5.7% increase.

00:49:34   So they just added $200 to the Vision Pro.

00:49:36   And you can look at this chart and kind of see like products with RAM and SSD, the more

00:49:41   RAM and the more SSD, the more the increases go up.

00:49:44   So it's brutal.

00:49:46   And the percentages are weird and uneven.

00:49:49   But if you look at the price increase, it's like $200, $200, $100, $200, $300, $500, like

00:49:54   they tried to make all their increases round numbers.

00:49:56   There's a $150 thrown in there as well.

00:49:57   And Apple TV is $70 and HomePod mini is $30, right?

00:50:00   But they tried to make the round numbers.

00:50:01   And this makes me start thinking about the timing of this, like the whole, you know, we're

00:50:06   going to have to increase prices announcement on the 17th.

00:50:08   It seems like something came to a head at this point where like we can't wait until September.

00:50:15   Like we can't just wait until September and just have like the new phones be more expensive.

00:50:18   We have to raise the price on other products too.

00:50:20   And there's no good time to do that.

00:50:21   So why don't we rip off the Band-Aid and raise the prices on all of them?

00:50:24   Now, does that mean every single one of these products just ran out of its one or two year

00:50:28   contracts for SSDs and RAM?

00:50:29   No, but it's better than doing it piecemeal.

00:50:32   So it could be that some of these products have been coasting on not having Apple style margins

00:50:37   for like a month or two.

00:50:38   And other ones could have another year of runway with based on the RAM and SSD chips they already

00:50:43   have or they already built for the thing.

00:50:44   But they're not going to do it based on like exactly when the things like they're not going

00:50:49   to wait till the last minute.

00:50:49   Like if any if any one of these products needs to be increased so we don't lose our precious

00:50:53   margins, we're going to do all of them now and we're going to do all of them by round numbers.

00:50:57   Are these round numbers exactly how much more Apple has to pay to build them?

00:51:02   Absolutely not.

00:51:03   Like there's no way it would magically come out.

00:51:05   So it just so happens the low end products get like 100 added and the other ones get no,

00:51:08   it doesn't make any sense.

00:51:09   That's but they just pick the round number.

00:51:11   So their announcement uses careful words as always.

00:51:15   It says Apple's statement says we have shielded our customer from these increases so far.

00:51:19   I think shielded is the correct word because Apple's deals have shielded the customers from

00:51:25   these increases in the same way that they've shielded Apple from the increases.

00:51:28   Because, hey, if we have a contract that says you're going to give us this many RAM chips

00:51:32   at this price and we haven't used all those RAM chips yet or the contract doesn't run out,

00:51:37   we're shielded from the increase and so are you as the consumer.

00:51:40   But the second that contract is up and we have to buy at market prices, we're not going to

00:51:45   eat that cost.

00:51:46   Right.

00:51:46   And so they were shielding and then they're no longer shielding, but they're not no longer

00:51:51   shielding themselves either.

00:51:52   What they didn't say is we have been eating that cost for you.

00:51:55   That's not what the statement says.

00:51:56   Now, they may have been eating some of that cost.

00:51:58   Like I can think maybe like on the Neo or some other products that like, you know, got

00:52:02   caught by surprise by this or a super low price that maybe they actually have had reduced margins

00:52:06   on some of these products.

00:52:07   But then I look at the price increase list and I say, well, whatever products you had been

00:52:11   suffering decreased margins on, I think you're in for another few months of what I imagine

00:52:17   will be potentially better than normal margins, especially on the products that still have locked

00:52:22   in contracts for components, but their prices still went up by 20%.

00:52:26   That's going to be some good margins on those and those will offset the other ones.

00:52:29   So anyway, we'll see in their financial calls.

00:52:31   I know we don't cover the minutiae like that, but a, and across the board price increase like

00:52:35   this on basically every Mac and every iPad plus division pro in round number amounts is

00:52:41   totally an Apple thing to do, which is like, let's do this.

00:52:43   Let's do it once.

00:52:44   Let's get it over with and let's make the numbers big enough that we don't have to do

00:52:47   it again.

00:52:48   And I would assume, and we'll see, but I would assume that when they come out with new Macs

00:52:53   and new iPads in the future, they will simply adopt these prices as the new normal prices.

00:52:58   Like even if and when commodity prices go back down, I don't expect Apple to have an announcement

00:53:02   and say, Hey, remember in 2026 when we had to increase all our prices because commodity prices

00:53:07   went up?

00:53:07   Well, guess what?

00:53:08   the AA bubble popped and now commodity prices are back down because there's a glut of RAM

00:53:11   manufacturing capacity because it's 2030 and all these factories have been built.

00:53:15   They're not going to lower the prices back down.

00:53:18   It's just going to keep them where they were.

00:53:19   So, uh, this is the new normal for us probably, uh, for multiple reasons.

00:53:24   One, uh, more capacity for SSDs and RAM is coming online in a few years.

00:53:30   So don't expect like this will be over tomorrow.

00:53:33   So the only thing they can save us is the factories get built in several years.

00:53:39   Or the AI bubble burst before that and AI companies have to start actually selling services at

00:53:44   the cost that makes me, you know, like the whole giant influx of free money to, with which

00:53:49   to buy all the RAM chips in the world stops flowing for whatever reason.

00:53:53   So what do you think will happen first that the new factories will get built in three or

00:53:57   four years or that the air bubble will pop?

00:53:58   I don't know.

00:53:59   I'm not predicting, uh, this is not financial advice, but I would expect these prices to be

00:54:03   with us for a while, which is terrible news for me.

00:54:07   Yeah.

00:54:09   Yeah.

00:54:09   Not a good time to need to buy a very high spec Mac.

00:54:13   I mean, like certainly everyone's focusing a lot on the base prices, but where you see

00:54:17   even more of a hit is on those upgraded RAM and SSD sizes.

00:54:21   Like I would, I looked today like the, like a maxed out, uh, MacBook pro right now is now

00:54:27   a little over $10,000, uh, which, you know, before I think they maxed out around the 7,000

00:54:32   range.

00:54:32   Um, so it's, it's, it's, it's especially like I was looking the, the eight terabyte SSD on

00:54:37   a MacBook pro is now a $3,000 BTO option.

00:54:41   Good God.

00:54:42   So it's, it, these are, these are pretty big numbers.

00:54:46   Uh, most of those are up, you know, 50 to 60%.

00:54:49   Um, now that being said, you know, Casey, you brought up the Apple TV example and the Apple

00:54:55   TV was last updated as far as I could tell quickly in 2022.

00:54:57   Um, and in that time there's been about 15% inflation, like just inflation alone covers a

00:55:04   lot of these for the, for like the old, the older stuff.

00:55:06   And like Ben Thompson has been making this point a lot that basically in inflation adjusted

00:55:10   terms, Apple products have actually been getting a lot cheaper over the last, you know, five,

00:55:14   six years.

00:55:14   Well, the phones in particular, some, some of the max questionable, but definitely the phones.

00:55:18   Yeah.

00:55:18   Because you know, the, the reality is like, this is, we're talking about one reason why,

00:55:24   why a particular type of thing is now more expensive, but we've been going through a pretty

00:55:31   significant period of inflation, certainly in the U S.

00:55:34   and I think a lot of the world has too, um, that, you know, look around our lives, nothing

00:55:40   costs the same now as it did even three years ago, because everything is more expensive.

00:55:45   You know, we have not to get, you know, political much about this, but like we have, in addition

00:55:51   to, you know, general fiscal dynamics and, you know, interest rates, things like that.

00:55:55   We also have ridiculous tariff chaos from our current president that throws the entire supply

00:56:01   change into chaos in a given moment.

00:56:02   Then he started a war and, you know, ruined oil prices for a while.

00:56:06   And it's like, all these things are going on in the world around us.

00:56:09   Of course, everything's going to get more expensive.

00:56:11   There is nothing in my life that costs today what it did three years ago, except I guess

00:56:16   iPhone based prices, but that's probably not long for this world now either.

00:56:19   Right.

00:56:19   So, like, I do think it is, you know, I think Apple has been, um, basically eating inflation

00:56:26   rises for us also for all this time.

00:56:29   I don't think this is going to be that temporary.

00:56:31   Uh, I, I think this is just what these things are going to cost for the foreseeable future.

00:56:36   Um, you know, when more capacity comes online in two years, maybe supply and email get in balance.

00:56:42   Maybe not.

00:56:43   Maybe the capacity that is being built today isn't enough.

00:56:48   We don't actually know that.

00:56:50   You know, it's also, it's equally possible it's way too much and there'll be a big glutton.

00:56:53   These companies will go out of business.

00:56:54   It's like there's, there's certainly a lot of, a lot of, you know, unknowns about this

00:56:58   dynamic, but we are in the middle of a very inflationary period of all costs of everything

00:57:04   going up.

00:57:05   And then also we're adding specifically problems with common computer components.

00:57:11   Um, this is just going to be our world for a while.

00:57:14   I think Apple has done a great job so far of not letting it show, uh, and, and not having

00:57:20   it affect them or us particularly much, but that time is over.

00:57:24   And this is just what things cost now.

00:57:26   So I think this is a good opportunity to, you know, extend the lives of devices you have.

00:57:32   Um, you know, look for deals on used things maybe if you need, you know, if you need to

00:57:36   get that, get a discount to, to make, to put something back into your price range, um, you

00:57:42   know, do what you can to be more efficient and, and pressure app developers and, and other

00:57:47   people who were, who were making very memory hungry, uh, apps and stuff like, you know, maybe

00:57:52   pressure them to, uh, to start being a little more.

00:57:54   Concern with that, but this is a really good time to already have a computer that you like.

00:58:00   This is not a good time to need one.

00:58:02   And this is also going to affect Apple kind of broadly by whenever they release new products

00:58:08   in the coming years, demand is going to be a little bit softened by the fact that they're

00:58:13   going to cost substantially more than they used to.

00:58:16   like, you know, this, this fall, we're heading into a pretty significant iPhone season.

00:58:20   Like they're, they're going to launch this foldable.

00:58:22   Foldables are expensive.

00:58:24   Even before this, you think like now imagine like, you know, so what, you know, the iPhone

00:58:30   pro, you know, you go, you go in now for about 1200 bucks.

00:58:33   It was a thousand, 1200.

00:58:34   What's the base price on the iPhone pro?

00:58:36   Something like that.

00:58:37   Yeah.

00:58:37   You know, that's, I figured that's probably going to be at least 1400 now.

00:58:41   And then what's the foldable going to be?

00:58:43   Like people were expecting before these price hikes that it might be $2,000.

00:58:47   Um, yeah, I'm thinking that might be conservative.

00:58:52   Like maybe it's $2,500.

00:58:54   Well, it depends on what they do with margins because like speaking of the upgrade price

00:58:58   table, the Gruber had in his thing, the upgrade prices for this stuff are a concentration of

00:59:04   this problem to like, because like the whole, you know, your whole Mac, your whole iPhone

00:59:08   has lots of different components that not all those components are getting more expensive,

00:59:11   right?

00:59:11   But specific ones are.

00:59:13   If you're upgrading one of those specific components, like, oh, I want to get a Mac, but

00:59:18   I want more RAM in it, those are not 20%, 15% increases.

00:59:22   They can look at the Gruber chart.

00:59:24   Some of them are a hundred percent increases because that's where the problem is in the,

00:59:29   in the supply chain right now.

00:59:30   If you want a bigger SSD, that's a 50, 60%, 70% increase in price.

00:59:35   And it's not just because Apple is gouging you because that is the source.

00:59:38   That's the source of the problem.

00:59:40   Um, and speaking of the, the foldable phone and stuff, the foldable phone is more expensive

00:59:44   because it's got more screen.

00:59:47   It's got more battery, but screens and batteries aren't the things that are going up in price.

00:59:50   Things are going up in price of RAM and SSD.

00:59:51   So just the foldable and have more RAM and SSD than the pro phone.

00:59:54   No, but it does have more screen and battery.

00:59:57   So maybe that will actually take less of a hit because the things that it has more of are

01:00:01   not appreciably hit by the particulars of the RAM and SSD crisis.

01:00:06   Obviously just there's inflation or whatever, but like, and it's a new product and it's complicated

01:00:10   to build and it's folding.

01:00:11   So it will definitely be expensive.

01:00:13   But, uh, yeah, the, the, the, when I look at these, these tables of the upgrade stuff,

01:00:17   all I can think about is like Apple strategy of on-device AI seems so smart a couple of years

01:00:23   ago.

01:00:23   It's like, well, they have great Silicon and everyone else has to have these big expensive

01:00:26   data centers.

01:00:27   And Apple's gonna, you know, if Apple tried to run data centers for its billions of iPhone

01:00:31   users, it'd be really expensive.

01:00:32   But if people can do it on device so that, you know, we got the Apple intelligence RAM

01:00:37   windfall, uh, uh, a year or two ago, or Apple put more RAM in all its devices so they could

01:00:43   run Apple intelligence.

01:00:44   Well, now we're in a world where Apple needs even, even more RAM.

01:00:48   Like they're increasing their RAM even more above and beyond what they did for like the

01:00:51   old Apple intelligence for the new one, because they want to run stuff on device.

01:00:54   Well, guess what's expensive now specifically because of AI data centers.

01:00:58   It's RAM because they're stealing it all and Apple needs more of it than ever on its

01:01:03   devices to run its local models.

01:01:05   So Apple strategy of we're going to run a local models because we have the best silicon is

01:01:09   great until RAM is prices are destroyed.

01:01:11   So I don't know if they're going to reconsider that policy.

01:01:15   Like the rumors of all the upcoming devices are they will in fact have more RAM than the

01:01:19   old ones specifically so they can run the, whatever, you know, the, the Apple foundation

01:01:24   model core advanced, blah, blah, blah.

01:01:27   But like that Apple is putting more RAM and all this devices above and beyond what that

01:01:31   used to be the baseline for Apple intelligence so they can run this new stuff.

01:01:34   That's going to hurt prices way more than you would think because yeah, it seems like those

01:01:40   had to buy, they had to buy that RAM at prices that were inflated.

01:01:43   We'll see.

01:01:43   We'll see what happens with the phone stuff.

01:01:44   But these Mac things, I feel like they're running it, probably running it in inventory and

01:01:48   some of these are going to be replaced and their deals are expiring or they didn't have

01:01:51   long-term contracts for these things.

01:01:53   But with the phone, I still have some faith that maybe some of the phones introduced this

01:01:58   September will benefit from locked-in rates for certain components.

01:02:01   But we'll see.

01:02:03   As with the big price increases we just saw here, it probably behooves Apple that even if

01:02:08   they do have good locked-in rates for their components for the, for the iPhones in September,

01:02:12   just increase all the prices by $200 anyway.

01:02:14   Because guess what?

01:02:15   In a year, this will probably still be going on or potentially the, like the rumors are

01:02:19   the rumors that predictions are this is going to get worse next year, not better, but like

01:02:23   it's going to get twice as bad as it is now in 2027.

01:02:26   So buckle up for that.

01:02:28   And that means Apple should probably increase the prices on all of its phones by a big amount

01:02:32   so they can keep that price the same in 2027 and maintain their incredibly high margins.

01:02:39   To go back several steps, I priced what is a modern facsimile for my current laptop.

01:02:45   So I have an M3 Max, 64 gigs, 8 terabytes because like Marco, I'm a fool.

01:02:51   I want to say I paid roundabouts of $5,000 for this computer when it was new, when the M3 Max

01:02:57   was brand new.

01:02:58   This was late, like two or three years ago, late in the year, two or three years ago.

01:03:01   So it was $5,000 then today, although I did option the nanotexture display because I want

01:03:07   it, but leaving that aside, that's the only real difference.

01:03:09   $8,250.

01:03:11   Yep.

01:03:12   Mine went from $7,200 to an even $10,000.

01:03:16   Good freaking God.

01:03:18   So I'm really, right now I'm really happy with my M3 Max right now.

01:03:21   It's like, oh, do I want to spend $10,000 to have the modern version?

01:03:26   Not particularly.

01:03:28   For the past month, I've been pricing out used and refurbished Apple Silicon stuff, including

01:03:34   a used M2 Max Studio with a third-party 8 terabyte SSD in it and everything.

01:03:41   And I wonder if those prices are going to go up too, because I didn't pull the trigger on

01:03:45   anything because I'm like, oh, it's still a huge amount of money.

01:03:48   And do I want to pay a huge amount of money for an ancient computer?

01:03:51   Maybe I'll just keep holding out because it's what I'm good at.

01:03:53   But speaking of that, Apple has, by the way, also raised the prices on their refurbished

01:03:59   Macs and iPads.

01:04:00   And you would think, wait a second, the refurbished ones, those are already built and assembled

01:04:05   and probably pre-owned.

01:04:06   Why are they going up?

01:04:07   Well, market prices, supply and demand.

01:04:10   People want them because they don't want to buy the new ones because they're too expensive.

01:04:14   So the demand for the refurbished ones goes up, which raises their prices too.

01:04:18   Although I did see someone complain about this and I did not confirm it, but a question

01:04:21   is, have they also increased the trade-in prices?

01:04:24   And the person was complaining that Apple hasn't increased the trade-in prices, but I did not

01:04:28   confirm whether that's true or not.

01:04:29   But that's something to watch when you go to buy a new whatever.

01:04:32   Has your trade-in gained more value just because the price of all these things has gone up?

01:04:36   We'll see.

01:04:36   Narrator, no.

01:04:37   Right.

01:04:38   I'm just saying I didn't check it.

01:04:39   I'm just saying I didn't check it.

01:04:40   I mean, I'm sure it goes up.

01:04:42   I'm sure the value is up like on eBay or something, but not in Apple's trade-in world.

01:04:45   No, I seriously doubt that will go up.

01:04:48   It certainly hasn't yet.

01:04:48   And then we should note that this all happened, like I said, today as we record this, Apple

01:04:54   stock is down to 6.15% today on account of all this.

01:04:58   So not a great day for Apple either.

01:05:00   I mean, yeah.

01:05:01   It's also worth noting, we mentioned it before, but like these are Mac and iPad increases.

01:05:05   There were no price increases for iPhones, watches, or AirPods.

01:05:09   iPhones, you just assume they're waiting for September to increase.

01:05:13   You know, the new ones will be more expensive.

01:05:14   Watches, probably the same deal.

01:05:16   And AirPods don't have SSDs, have the tiniest piddling amount of RAM.

01:05:23   And so maybe aren't actually affected by this, but watch for that.

01:05:28   Watch for the AirPods price increase and try to explain that.

01:05:31   I mean, again, like inflation adjustment.

01:05:33   It's not as if Apple has never tracked inflation.

01:05:35   They do tend to increase the prices of their products over time.

01:05:38   They've held the line on the phones mostly because the margins are so massive and they

01:05:41   find ways to, you know, decontent them to keep their margins the same while keeping the price

01:05:46   the same.

01:05:46   But the phone is such an important product.

01:05:48   They just like anything having to do with the phone is like multiply by a billion.

01:05:52   And it's like, it's scary to mess with that.

01:05:54   But the other products they've, you know, adjusted over time.

01:05:57   I will be watching the AirPods to see.

01:05:59   I mean, obviously they're going to come out with the AirPods with the whatever camera, IR camera,

01:06:03   whatever.

01:06:03   And those will be more expensive simply because, you know, they have more stuff in them and they're

01:06:06   the fanciest and the newest, but the plain old AirPods, the ones that are like the ones

01:06:11   we have now, which I assume they will continue to sell in some form.

01:06:13   I wonder if those will actually stay more or less the same simply because they don't use

01:06:17   any of the components that are affected by this crisis.

01:06:19   But the phones and watches certainly do.

01:06:21   So watch out for that.

01:06:22   And yeah, Wall Street says we don't like this, Apple.

01:06:24   We we we think this is not good for your business, because if you raise the prices on all

01:06:30   of your products, probably going to make the demand for them go down.

01:06:33   And that doesn't bode well.

01:06:34   But we'll see.

01:06:35   We'll see how Apple does.

01:06:36   It may be a buying opportunity for people who want to get in and on the stock.

01:06:39   And, you know, because in Apple's next earning call, they're going to be like, we sold more

01:06:42   phones than we've ever sold it ever, you know, and then it goes back up again.

01:06:45   But anyway, this is, again, not financial advice.

01:06:47   Don't play the stock market.

01:06:48   I would also point out, though, like, you know, when you're thinking like the AirPods are

01:06:51   not going to be as affected.

01:06:52   First of all, you know, they do they do have some RAM.

01:06:55   I mean, almost every electronic device has RAM, NAND storage or both.

01:07:00   And I do think, though, you know, you have to consider that, OK, well, right now these component

01:07:06   prices are going up.

01:07:07   So that means that now Apple's products are going up.

01:07:10   Well, the companies that, you know, like all the different companies that make all the components

01:07:15   that go into AirPods, all of their costs are going up around them.

01:07:20   Like all of their components are going up, their own, you know, R&D costs are going to

01:07:24   go up every as everyone in the world is facing inflation on all of their expenses.

01:07:30   They have to demand higher salaries so they can afford regular stuff.

01:07:34   And so everyone's costs go up broadly and diffusely across everything.

01:07:41   So even things that are not directly affected, they get the ripple effect.

01:07:46   So there I don't think there's any major product category, you know, in the same way

01:07:50   that like fuel costs affect everything.

01:07:52   I don't think there's any major product category that's going to escape some effect of this.

01:07:58   Everyone is going to have to raise their prices on almost everything, almost everywhere.

01:08:01   Yeah, it's it's really not great.

01:08:05   And I mean, again, I'm I'm very grumbly at Apple for just cranking up the cost of this

01:08:12   freaking ancient Apple TV.

01:08:13   But other than that, I mean, I don't really blame them for most of this, but there is

01:08:17   some good news at the end of the story.

01:08:19   We had three people either reach out or we had seen three people talking about how they

01:08:24   had great timing.

01:08:25   So Team McGeary writes, I owe Marco a beer.

01:08:27   I bought a MacBook Pro a few weeks ago based on his warning about Apple prices increases

01:08:32   on ATP.

01:08:33   Thanks.

01:08:33   Aaron writes, four or five episodes ago, Marco recommended that if you were planning on buying

01:08:38   a MacBook, now is the time I'm writing to thank you.

01:08:40   I bought a new MacBook Air for my wife in late May and that same machine is four hundred

01:08:44   dollars more today.

01:08:45   And then a friend of the show, Greg Pierce wrote, guess I pulled the trigger on a new M5 Pro

01:08:50   MacBook Pro at the right time.

01:08:52   The specs I bought two weeks ago for thirty four hundred dollars is now eight hundred dollars

01:08:56   more expensive.

01:08:57   Wow.

01:08:58   Yep.

01:08:58   That's that's what happens.

01:09:00   I don't think the end is in sight.

01:09:02   I also don't think this is the last price increase to happen in this wave.

01:09:07   Again, we're going to see more across everything and we're going to see significant ripple effects

01:09:14   of of those price races.

01:09:16   So, you know, not only everyone's cost going up and everything like that, but also we're

01:09:20   going to think like how will this change the market for certain types of devices at all?

01:09:24   You know, as we know, when computing resources get very cheap, that enables stuff like Raspberry

01:09:30   Pies and like all sorts of like fun and new new types of gadgets, new types of new uses

01:09:37   for hardware that used to be expensive or that used to be expensive and then got cheaper.

01:09:41   Now we see the opposite of that.

01:09:44   what categories of hardware are now just not going to be worth it or are going to be,

01:09:51   you know, such low volume customers of these component vendors that they won't even be able

01:09:56   to get allocation of stock.

01:09:58   So they won't even be able to exist.

01:10:00   So, you know, this this could affect things like, you know, obviously, like on the medium

01:10:04   sized market, this could affect things like game consoles significantly.

01:10:08   Yeah, as Xbox just announced the price increases and we'll talk about the steam machine next

01:10:13   episode probably right like, you know, what we're seeing in that world is like consoles

01:10:16   are being, you know, price hiked or delayed.

01:10:19   You know, people are kind of just kind of try to get more of what more more out of the current

01:10:24   generations because it's just not a good environment to try to launch a new electronic

01:10:28   that's not super critical.

01:10:29   This could affect other stuff like, you know, my my beloved e-ink tablet world or e-readers.

01:10:35   Like there's all sorts of electronic categories that like now that hardware is becoming more

01:10:41   expensive and harder to even come by supply wise.

01:10:45   Some of these categories aren't going to make it or they're going to kind of just be put

01:10:50   on ice for a few years until things stabilize.

01:10:52   I think this is going to be in some ways a pretty dark time for the electronics business,

01:10:59   while in other ways like also AI is very exciting and opening up new possibilities for lots of

01:11:05   people in lots of contexts. So it's a really turbulent time and some of it's positive and

01:11:11   some of it's not. If your business or hobby life relies on certain electronics being available to

01:11:18   you, you know, maybe move some of those purchases up quickly.

01:11:21   It might be too late. Yeah, no, on that topic and a call back to the episode we did about why

01:11:25   everyone hates AI. Yeah, it's like, oh, that's all well and good. You talk about that episode,

01:11:29   what other people think, but now they come for your Apple hardware and all you Apple fans like,

01:11:32   well, now suddenly I care about AI. It was fine when they were increasing prices, stuff that I don't

01:11:37   care about, but now they're increasing Apple product prices. This is this can't stand. But

01:11:40   this is basically it's all the same thing, which is like, obviously, there's a massive investment

01:11:45   investment across the entire industry in AI. That massive investment is allowing those companies

01:11:50   to buy up all the RAM in the world, like literally all the RAM in the world, right? They're not doing

01:11:55   that because they're so rich from the profitable, from the profit they're making from selling their

01:11:59   products. They're doing that with the money they have invested because people think these companies

01:12:04   are the future. And so they're piling all this investment in and they take that investment money and

01:12:08   they shove it to Nvidia and all the RAM companies and everything. That is, that's the bubble. It's

01:12:12   distorting the market with all this money flowing into it. And if and when this market distortion

01:12:19   comes for the thing that you care about, our case being Apple hardware, but it could be anything,

01:12:23   as Marco said, it could be, you know, garage door openers. They have chips and RAM and everything,

01:12:27   like it's anything, right? And, you know, as Marco said, that just the general price increase

01:12:31   everywhere, as we saw with a COVID-19 lockdown, people will take any opportunity to price gouge,

01:12:36   even if their costs haven't actually increased. If everyone else is doing it, they're just going to do it

01:12:39   anyway because it's a thing to do. And you've got cover. It's like, oh yeah,

01:12:42   AI RAM crisis. That's why, you know, when that happens to you, people are naturally going to

01:12:48   think, as I am right now, is this trade-off worth it? Oh, AI is cool and exciting and everything,

01:12:54   but there are things I hate about it. And also, would I be happier with a slower growth in the

01:13:00   AI industry in exchange for sane prices in Apple hardware or garage door openers or Raspberry Pis or

01:13:07   e-ink tablets or whatever you have? And I think a lot of people would say,

01:13:11   hell yeah, like, AI is exciting and everything, but do we need to destroy the entire world economy to allow

01:13:17   that industry to grow as fast as possible? Because people would look, it's like, what am I getting in exchange for that?

01:13:21   Okay, say you love ChatGPT. It's the best thing since life spread, but then you look at the things,

01:13:25   the costs on the other side of it, and you're like, even me as the biggest fan of this, I'm not sure this

01:13:29   balances out. Say you hate AI and everything about it, then you're looking at this and saying, oh,

01:13:33   so the thing I hate is making all the things I love worse? This makes me hate it even more. So,

01:13:39   yeah, this, like, I feel like this is, you know, it's, it's not imminent that something is going to

01:13:44   happen, but things are going to have to come to a head. Like I said, it's either the case where we

01:13:49   wait out three terrible years, which, like I said, will probably be worse than this year. Like, next year is

01:13:53   predicted to be worse than this year, not better. We were at all these years and they build capacity

01:13:58   or whatever. And then eventually the bubble burst or there's consolidation or whatever. And then it

01:14:02   comes back to sanity or maybe the bubble burst sooner. Maybe chickens come home to roost. Maybe

01:14:08   these companies go public and it turns out when you're public, you actually have to turn a profit

01:14:11   in a more reasonable time. Although look at Amazon, they spent a long time not turning a profit and they

01:14:15   did fine. So again, not financial advice, but like it seemed the current situation seems both terrible

01:14:20   and unsustainable in obvious ways that everyone is looking at. And that just like the financial

01:14:25   people are like, well, this is just the horse race. This is just, yeah, we know this is all going to

01:14:28   end in some kind of consolidation and it's not going to go on like this forever. And there'll be winners

01:14:33   and losers and blah, blah, blah. But this is what we're doing now. And screw the people who get

01:14:38   sideswiped by the damage we're making. And then the rest of us are just out here, like hitched to this

01:14:42   wagon. Like, I mean, nothing highlights this more than what was it? The stupid SpaceX IPO, which you may

01:14:47   think, what does rockets have to do with AI? Uh, Elon Musk has combined all this

01:14:50   companies into one giant cluster of evil. Um, anyway, when companies are that big, uh, these

01:14:57   funds that buy like, uh, you know, index funds for your, like your retirement accounts end up owning

01:15:02   them just because they're so big and they buy like the top 500 companies or whatever, as part of their

01:15:06   index thing. And now SpaceX is part of that instantly because it's so big by market cap, because it's a

01:15:11   combination of Twitter and XAI and blah, blah, blah. And so all of a sudden people's retirement

01:15:16   accounts, people who do not buy individual stocks, this may be a more US centric thing, but anyway,

01:15:19   we have retirement accounts here because we don't have any kind of social safety net to speak of.

01:15:23   Um, if you have an individual retirement account or a 401k or whatever, and you have a bunch of index

01:15:28   funds, this is a wide ranging paragraph. Yeah. You have an index fund. And now that index includes

01:15:33   SpaceX, which is an AI company because they went public. And when this bubble bursts, oh, your 401k

01:15:37   gets screwed too. So we have that to look forward to as well. Not only enduring all the pain of this,

01:15:43   but when it does quote unquote end and things consolidate and the bubble burst and winnings,

01:15:48   losers are picked and companies have to actually start turning a profit. All our retirement accounts

01:15:52   get screwed. He's like, I never consented to buying any of this crappy AI stock. Well, tough luck. If you're

01:15:56   buying the S&P 500 or some other index fund, SpaceX is going to be in there. So yeah, things are not

01:16:02   looking great financially for us. And like I said, for me in particular, because I do have this Intel Mac

01:16:08   here. And at this point, any kind of arm based Mac is looking like it's going to cost a lot of money. So I may be

01:16:16   cruising on this for a while. I mean, we made fun of my old Mac pro that I had for over 10 years. And it's like,

01:16:21   well, that won't happen again. They did a processor transition right after you bought your new Mac.

01:16:24   Surely you'll get rid of that one pretty soon. Well, I'm on year seven. So tune in in three years

01:16:31   to see if I'm still running this Intel thing. Well, you can't on account of the OS, right?

01:16:35   I can just keep running the OS. I'm not running 26 now. I mean, yeah, I'll do increasing amount of

01:16:40   dev work on my M1 MacBook Air, which is what I'm doing, which is what's booted into Golden Gate right now.

01:16:45   Oh my God. I'm sorry, John. I really am. I mean, the good news is you have precedent for spending

01:16:49   as much as a civic on a computer. So yeah, I don't, I, but I don't like this, but your voice

01:16:54   is haunting me. Casey's remember on the previous episode or a couple episodes ago, you were like,

01:16:58   do you think you're going to spend as much money as your Mac pro and your new Mac? I'm like, no,

01:17:02   that can't because I subtract $6,000 because I'm not buying a new monitor. Knock on wood. Um,

01:17:08   but now maybe you're right. Maybe, maybe the eight terabyte M5 max max studio with a 64 gigs of Ram

01:17:18   is going to actually cost as much as my entire Intel Mac pro setup from 2019.

01:17:22   John, it feels good to be right, but I did not want to be right about this.

01:17:26   I mean, I'm really hoping now, cause if that's a situation, like maybe, like, maybe I'll just get

01:17:31   like the, the, you know, is the max the smallest one? That probably is going to be the smallest one.

01:17:36   The problem is they don't let you get eight terabytes in the Mac mini. I'd be like, I'm going to

01:17:39   get a Mac mini. Yeah. If you want eight terabytes, you're, you got to get a max ship these days.

01:17:42   And then, and the smallest CPU that comes with is the max, right? Uh, yeah, I believe so.

01:17:46   Yeah. Yeah. And you have to get, and you have to get the fancier max because there's at least on

01:17:51   laptops, there's two maxes. And I, when I was pricing it a few minutes ago, I went to eight

01:17:56   terabytes and it was like, Oh no, no, no, no, no, no, no. You need to get the super baller CPU before

01:18:01   you can get eight terabytes. Thank you very much. Maybe I have to become an external disc

01:18:04   person again and just have a bunch of discs and divide up my world into little pieces. I'm not

01:18:08   looking forward to it. If that is a problem for another day, because I mean, this just makes it

01:18:12   so much worse that like the, this, this exact same crisis essentially caused, if the rumors are to be

01:18:17   believed, the Mac studio that would have been announced to WWC to be pushed back because, you know,

01:18:21   because of this whole thing. And now it is pushed back past the price increase horizon. So if it

01:18:27   one Apple does roll out an M5 max based Mac studio, it will benefit from the price increases in both

01:18:33   the base price and the upgrade prices that we just read, because they are going to be this bad or

01:18:37   potentially worse. It's not great, Bob. No. And, and, oh, and, and by the way, just for the record,

01:18:45   in case this is unclear, space data centers make no sense because of heat and radiation and everything.

01:18:52   And so, yeah, that doesn't work just, just for the record, just putting that out there.

01:18:57   That's not a thing. In case, in case that wasn't clearly a stupid idea.

01:19:01   Right. Like in case anyone doesn't know science, just look into what it takes to radiate the amount

01:19:07   of heat that a typical data center rack needs to radiate into space. But Margo, isn't space cold?

01:19:13   Yeah. The problem is how do you transfer the heat from the chip to space? And it turns out there is a

01:19:21   way to do it. It's just really large. So, yeah, that's, that's not, that's not going to be a thing.

01:19:27   You can just build on land. It's a lot easier and cheaper.

01:19:29   We are sponsored on this episode by ZocDoc. Now, raise your hand if you've been putting off a dental

01:19:38   cleaning, an annual checkup, or honestly, any kind of doctor's appointment. Is your hand up? Mine is.

01:19:43   When something, you know, when I, when I know I need something, maybe I'm feeling a little bit off,

01:19:47   like I am this week. Honestly, what I usually do is either nothing or I'll search the web and try to

01:19:52   like, you know, diagnose my, my symptoms myself, or I'll ask an AI bot what I should do. But these are

01:19:58   not amazing. If this sounds familiar to you this year, consider doing things differently. Find doctors

01:20:03   you love and book appointments with ZocDoc. ZocDoc is a free app and website that helps you find and book

01:20:09   high quality in network doctors. So you can find someone you love. There's more than 150,000 providers

01:20:16   available across all 50 us States, whether you're looking for primary care for general purpose stuff

01:20:21   or specialty things like dermatology, eye care, even dentistry, or one of the over 200 specialties

01:20:27   offered on ZocDoc. You can easily search by specialty or your symptoms to build whatever care team is

01:20:32   right for you. They have of course, video visits. If you prefer that and in-person visits, if you prefer

01:20:37   that, or if you want to do some, sometimes some, the other times you can do both, whatever fits you.

01:20:41   They have thousands of verified patient reviews to give you a real sense of each doctor and who

01:20:46   they are and whether they'll be a good fit for you. And when you're ready to book, you can see their

01:20:49   real-time availability and click to book instantly. No phone tag, no waiting around, no forms. And the

01:20:55   appointments made through ZocDoc are usually within 24 to 72 hours of booking, sometimes even same day

01:21:00   appointments. So availability on it is fantastic. I know I personally, I often put these off because

01:21:07   it's such a pain to book them. ZocDoc promises to make that easier. Stop putting off those doctor's

01:21:12   appointments and go to ZocDoc.com slash ATP to find and instantly book a doctor that you'll love today.

01:21:18   That's ZocDoc spelled Z-O-C-D-O-C dot com slash ATP. ZocDoc dot com slash ATP. Thanks to ZocDoc for sponsoring

01:21:28   our show.

01:21:28   All right, let's do at least a touch of Ask ATP. And Brian Webster wrote, I don't know, maybe a month

01:21:37   ago or so, given the soaring cost and unpredictability of RAM prices, and given that Apple already

01:21:44   manufactures its own SoCs and many of its own cell modems, would it ever make sense for Apple to start

01:21:48   manufacturing its own RAM? They ship high enough volumes of RAM that they would likely get an okay

01:21:53   return on investment in facilities for RAM production. It would insulate them from the

01:21:57   supply constraints triggered by the AI boom. And most importantly, since they would only be paying

01:22:02   manufacturing costs without extra profit margins, they could finally lower those ludicrous RAM

01:22:08   upgrade prices. Right? Right?

01:22:10   Do you know the meme he's referencing here, which you refuse to read?

01:22:14   Yes, I do.

01:22:15   All right, you know a meme. Marco, do you know this one?

01:22:17   No.

01:22:18   The Anakin Padme meme? You're lowering RAM prices, right?

01:22:22   Right? Okay, well, we're batting 500. Casey knows the meme and Marco doesn't. Again, I don't know how

01:22:28   both of you exist on the same internet as me.

01:22:29   No, Marco does know it. He just doesn't realize he knows it.

01:22:32   Marco doesn't know it.

01:22:33   He doesn't know who Anakin and Padme are.

01:22:34   No, I know. I'm very familiar with those characters from the terrible Star Wars prequels

01:22:39   that inexplicably, she falls in love with him and then dies of sadness. I mean, has there ever

01:22:45   been a worse written woman in all of cinema?

01:22:47   Well, doctors never believe women, so. She probably just had a blood clot like Serena.

01:22:53   Oh, gracious. Anyway, so how do we want to approach this? I mean, obviously, the clear

01:22:58   answer is you can't just spin up, you know, a new RAM factory. And we were talking about

01:23:05   that earlier. Like, it takes two to three years to spin one of those up. So that...

01:23:08   If you're a company that already does that.

01:23:10   Right, exactly.

01:23:11   So you should just read Joe Lyon's answer because he knows the industry.

01:23:14   All right. So Joe Lyon writes, memory is a commodity. That's why Apple doesn't design

01:23:17   it and also why they will never manufacture it. Eventually, sane pricing will return. But

01:23:22   if Apple made their own memory, they would be locked into their own cost structure and

01:23:25   they wouldn't be able to bargain with vendors and multi-source nor meet surge demand. Also,

01:23:29   the Chinese government has spent hundreds of billions of dollars building and propping

01:23:32   up domestic memory suppliers over the past 12 years. And there's still several generations

01:23:35   behind the leading edge. If Apple broke ground on a fab today, it would be five years before

01:23:39   the first wafers came out, 10 years before high volume yield on a decent process, and 15

01:23:44   years or more, or perhaps even never, before it would be cost competitive with current DRAM

01:23:50   vendors.

01:23:50   That's the problem. Remember that episode we did where we talked about like the silicon fabbing

01:23:54   machines from ASML and everything? Like you say, well, they're not making processors, they're

01:23:58   making RAM. It's the easiest thing to make. It's very regular. And it actually is a little

01:24:02   bit of a different process than used for logic and everything. But the base

01:24:05   facts remain the same. This is not a turnkey business. You don't just buy the kit and like

01:24:10   have a franchise and start a thing. These companies that have been doing this have been doing it

01:24:14   for decades and decades. And you can't catch them by just saying, well, we'll just make a

01:24:19   factory and do the same thing they're doing. It's incredibly difficult and complicated. And

01:24:24   anyone who has ever tried to do it, like the Chinese government example is great. What if

01:24:28   you had basically unlimited money, like the Chinese government is funding you, they want this industry

01:24:34   to come up and it's taking them a long time. And they're not a private company that has to have

01:24:38   investment that the Chinese government is backing it saying we want to be contenders here. And they

01:24:43   spent 12 years doing it. And there's still several generations behind those three big companies. And

01:24:47   Apple kind of sort of needs like the good RAM, like the best the industry has for a purpose in an iPhone.

01:24:54   They can't, you know, you can put crappy RAM on the Apple TV, fine.

01:24:57   But the iPhone's kind of got to have the RAM best suited to the iPhone. They're not bargain shopping for that in terms of like, we'll just get a two year old version of it. No, they want the best. And so yeah, Apple, even if Apple decided to do this today, it wouldn't help them unless this crisis goes on for a decade and a half, which God, I hope it doesn't.

01:25:14   But that's the kind of the timeline. But the earlier point, I think is more important that Joe made that like, it's a commodity business.

01:25:20   If Apple is just making RAM for itself, the only way you can survive in a capital intensive business like silicon fabrication is you spend billions and billions of dollars building the factories, machines and expertise to make chips.

01:25:35   And you need those things to be running and churning out money to make back your investment.

01:25:40   And so you can't have them be like a single customer. This is why Intel was in such trouble. You can't have it just be a single customer thing. It's like we just make what Apple needs.

01:25:48   Well, what about when Apple doesn't need any more chips for the year? Do you just let the factory sit there and be idle?

01:25:52   That's terrible that you lose tons of money doing that. You need to keep running that factory and selling people chips.

01:25:59   And so you end up essentially in the RAM business, selling commodities to the entire industry.

01:26:04   And Apple tends not to want to be in the commodity business like we sell components to the world.

01:26:10   But if you build a RAM fab, the only way it's a viable business is you have to sell RAM to everybody.

01:26:15   And suddenly you're not Apple making these special products just for your customers with big margins.

01:26:20   Now you're competing with the existing RAM vendors to sell the world.

01:26:24   And it looks like a great business now because everyone wants RAM.

01:26:27   But again, you got to build the factories and learn how to do that over the process of 15 years.

01:26:30   And who knows what the RAM situation will look like in 15 years.

01:26:33   But Apple essentially never wants to be in the business of manufacturing things for the entire world to use as parts in their products.

01:26:41   They just don't do that.

01:26:42   They have other people do that.

01:26:44   And as Joe points out, they play them against each other to get better rates and everything.

01:26:47   But it's those people's problem what to do with the excess capacity in their factories.

01:26:52   Now, now that now it seems great.

01:26:53   It's like there is no excess capacity.

01:26:55   TSMC is running flat out.

01:26:57   All the RAM manufacturers are selling every chip they can sell.

01:27:00   It's like a great business.

01:27:00   But that doesn't stay true forever.

01:27:02   So if you spend a decade and billions of dollars building yourself up as a commodity business for RAM,

01:27:07   and then eventually the world changes and we're not in the current crisis,

01:27:11   you will either lose money hand over fist or you have to become a different kind of company.

01:27:16   And Apple doesn't want to do either of those things.

01:27:17   So, I mean, this makes sense when you, as I think Brian lays out the question,

01:27:22   like it makes perfect sense.

01:27:23   Look at all these other things they do.

01:27:24   He mentions the SOCs, but they don't make their own SOCs.

01:27:27   They pay TSMC to do it.

01:27:28   TSMC has that factory that they invest billions and billions of dollars into that they need to constantly turn out chips.

01:27:33   And Apple is not their only customer.

01:27:36   Apple used to be their biggest customer, but not their only customer.

01:27:38   So Apple doesn't want to do that.

01:27:40   They want someone else to do it for them.

01:27:42   And they want to be able to, you know,

01:27:44   have things fabbed in Arizona for last year's or two years old chips or whatever.

01:27:48   So it seems like it might be a good idea.

01:27:51   But if you look at it more closely, it is a thing that Apple is never going to do

01:27:54   and would probably be a terrible idea.

01:27:56   And even if they didn't, wouldn't save us from the current crisis.

01:27:59   Anonymous writes,

01:28:01   I remember a while back there were stories about TSA doing searches of mobile devices as people entered the U.S.

01:28:05   Does the new Siri open up any security or privacy concerns?

01:28:08   Does Apple provide any settings that could mitigate these issues while traveling similar to 1Password's travel mode?

01:28:13   I mean, I guess if you allow Siri when unlocked,

01:28:17   and I haven't used the new Siri yet.

01:28:19   I've only got a couple of devices on the beta OSs and I haven't made it through the wait list yet.

01:28:25   But if you allow Siri when unlocked and if it allows you to go searching your personal context without, you know,

01:28:31   like a face ID challenge or whatever, then yes, I suppose that is possible.

01:28:35   But if it were me, even as an American, I would do the thing where you hold the side,

01:28:41   the buttons on either side for a few seconds to force a non-biometric unlock.

01:28:46   I would absolutely do that as I was approaching like immigration or customs or anything like that just to be safe.

01:28:52   And that's how I would handle this.

01:28:53   And at that point, you can't use Siri no matter what until you type in your password and unlock your phone again.

01:29:00   I think that's very optimistic of a solution.

01:29:03   But I think the reality is, so, you know, so that solution of like locking out the phone so biometric authentication doesn't work,

01:29:11   that's great if you are afraid of being arrested in your country and having, you know,

01:29:17   having law enforcement try to get you to unlock your phone.

01:29:20   But that's a different situation than like border control entry.

01:29:24   For a border interest situation, like you have to get through these people to get where you're going.

01:29:30   And they can just say no.

01:29:33   Like if you won't let, if you won't give them access to what they want, they can just deny you access to the country.

01:29:40   So it's a different situation.

01:29:42   And I think in that case, like I think the, for, especially like for non-U.S. citizens trying to enter the U.S.,

01:29:48   I think we are such incredible jerks in that area.

01:29:53   And again, that's only getting worse under the current administration.

01:29:57   They might not have that option.

01:29:59   Like if they lock out their phone, then the border control can say, okay, well, I'm just not letting you in the country then.

01:30:05   Whatever you need to come to the U.S. for, it's not happening.

01:30:07   So it's a very different power situation.

01:30:10   I think if, in this context, if one of our terrible border agents wants to ask you to show them your, you know, your Instagram account or whatever, I think your only real option is to do it.

01:30:22   Or, again, or like cancel your vacation or cancel the conference you were going to and leave.

01:30:28   And that's not really an option for most people entering the border.

01:30:30   So the reality is they're going to just do it.

01:30:33   And whatever technical measures you're going to try to implement, I think, you know, best case scenario is you can hide apps.

01:30:41   Like that's, that I think is, you know, if you, if you don't want them to see that you have a certain app installed, you can just, you can, you can put it in the hidden folder and make it require face ID.

01:30:51   That's, that's the best way to make sure an app, like, doesn't appear to exist on your phone so that if they, you know, make you search for it or whatever, it won't show up.

01:31:01   So that you can do.

01:31:02   But if they're asking for your social media handles and they're going to look at that, like, to say, I don't have Instagram, it's a pretty easy thing to go for them to go verify, you know?

01:31:10   So, like, I think that's, that kind of approach will work a lot better on certain types of apps than others.

01:31:15   If you don't want, if you want them to not see that you have, like, you know, signal installed, that's probably a lot easier of a lie to keep up.

01:31:22   Whereas if you say, I don't use any social media, they could very easily just check and find your profile if it's public, and then you're in trouble for a couple of problems.

01:31:30   So be very careful with this kind of thing.

01:31:33   But I think hiding apps is a much better way to do it than assuming that they won't convince you to unlock your phone.

01:31:39   I don't know if the new Siri makes this worse.

01:31:41   Like, in theory, the new Siri is more capable.

01:31:44   So a technically savvy person, if it was working from the unlock screen, blah, blah, blah, could use it to be more successful, but just because the old Siri was so useless that you couldn't use it or anything.

01:31:52   But this assumes a fairly sophisticated attacker in the form of whoever is doing the screening here.

01:31:59   If you want to social engineer your way to mitigate this, my advice would be to drain the battery on your phone and just tell them it's broken.

01:32:06   Because, you know, are they geniuses or are they going to sit there?

01:32:10   Because you know how long iOS devices take when they're like, the battery's totally drained.

01:32:14   When you plug it in, it just shows the little battery with the thing.

01:32:16   That takes forever, and they're not going to want to wait.

01:32:18   And if you just tell them it's broken and they can't turn it on, like, that's social engineering.

01:32:22   It's rolling the dice, but, like, that may be your best bet in terms of mitigation.

01:32:25   And second best is, you know, the password unlock thing or whatever.

01:32:30   But it depends on what situation.

01:32:31   This is talking about TSA, which I assume is just inside the U.S. or whatever.

01:32:34   So, yeah, I would say the new Siri probably does make it slightly worse if you have a sophisticated attacker who knows that they can do things from the lock screen with Siri because you've configured it that way.

01:32:43   But you can not configure it that way, and you could also hold the power button on your phone and require a passcode, and you can also drain the battery completely before going to the airport if that doesn't freak you out.

01:32:52   I know your boarding pass might be on the phone, too, so you're going to have to print it out.

01:32:55   Like, it's a terrible world we live in.

01:32:57   Don't come to the U.S.

01:32:58   It's not a good place to visit.

01:33:00   Our government is bad.

01:33:01   Yeah, pretty much.

01:33:02   Better than Rome.

01:33:03   I don't know.

01:33:05   Rome has universal health care.

01:33:06   That's true.

01:33:07   And will not arrest you for saying something on social media and won't shoot you dead in the street.

01:33:13   Yeah, I guess they do have a couple of things on us.

01:33:15   They have a lot.

01:33:16   A lot.

01:33:16   And they have one more wallet now than they used to.

01:33:18   All right, thanks to our sponsors this episode, ZocDoc and Squarespace.

01:33:22   And thanks to our members who support us directly.

01:33:24   You can join us at atp.fm slash join.

01:33:27   One of our many perks of ATP membership is ATP Overtime, our weekly bonus topic.

01:33:33   Every week, we do about 20 more minutes of content on some topic that just wouldn't fit in the main show because, as you know, a lot of stuff happens in computers.

01:33:41   So, this time on Overtime, we're going to be talking about Anthropics Fable and John's, quote, LLM bug releases.

01:33:49   We're going to hear what that means in Overtime.

01:33:50   If you want to hear that, join atp.fm slash join.

01:33:53   Thanks, everybody.

01:33:53   And we'll talk to you next week.

01:33:55   And you can find the show notes at atp.fm.

01:34:23   And if you're into Mastodon, you can follow them at C-A-S-E-Y-L-I-S-S.

01:34:33   So, John, you wanted to talk to me about a blog post I wrote recently.

01:35:03   What's going on?

01:35:03   We're going to jam it into the Overtime because it was, like, kind of AI-related things.

01:35:07   And I had this Overtime topic in before you wrote your blog post, which was just a happy coincidence.

01:35:11   But we figured we would separate it out because it's more like after-show material because it's you talking about your feelings and your feelings about AI.

01:35:18   Even though, like, again, the Overtime is going to be about Anthropics Fable and all the drama there and me using Fable.

01:35:23   But you've also been using AI.

01:35:25   And before I read your blog post, I was like, oh, this is great.

01:35:27   Casey has the exact same idea as me.

01:35:29   He's going to talk about exactly the thing that I'm going to talk about.

01:35:32   And then I read your post.

01:35:32   I'm like, nope.

01:35:33   It's totally different.

01:35:34   Summarize what you wrote and why you wrote it.

01:35:38   Yeah, so when I was at the beach a week and a half ago, two weeks ago, whatever it was, during WWDC week, I had somebody write in.

01:35:46   And they were very kind and very polite.

01:35:48   And they said that something was broken.

01:35:50   And the particular something, if I recall correctly, was that in call sheet for the last several months now, you can turn on or turn off different sections of the Discover screen, which is to say the screen that you land on when you open the app cold.

01:36:02   And so maybe you don't care about recent TV shows.

01:36:06   And you can turn that section off.

01:36:07   And maybe you want to see now playing information from like Plex or Jellyfin or something, but you don't want that to be front and center at the top.

01:36:14   You want it to be toward the bottom of the screen so you can rearrange them and so on and so forth.

01:36:17   And I had had reports, sporadic reports, since I released that feature that for some people, their settings would get plowed over every time they restarted the app.

01:36:30   And I had no strong theories about what that was other than, well, maybe they're like force quitting the app before User Defaults has a section to like flush its, before it has a chance to flush its queue, if you will.

01:36:46   But I don't know, man, User Defaults is the mechanism by which most apps save like user preferences onto your device.

01:36:52   And it's got to be up there on the most, you know, battle tested, like almost flawless bits of the iOS SDK.

01:37:03   So it was a very weak theory that I held very loosely.

01:37:06   But this person wrote in and, you know, said, oh, it's not working for me.

01:37:09   They were very kind about it.

01:37:10   They weren't jerks about it.

01:37:11   And I asked a follow-up question here and there and then got a log file from them, which was very kind of them to spend the time with me and give them, you know, or give me that log file.

01:37:20   And there was something in the log file that made me go, huh, that doesn't seem right.

01:37:25   And I'm at the beach.

01:37:27   I don't want to sit down and like properly do work.

01:37:30   Like I don't want to bust out Xcode.

01:37:31   I don't want to like really, really do work.

01:37:33   So I thought to myself, well, let me talk to my buddy Claude and let me see what Claude code can come up with.

01:37:39   And so I asked Claude to take a look at this one particular section of my code.

01:37:44   And the literal verbatim query that I asked of Claude was in usersettings.save, I save the user's preferred arrangement of sections of their main or discover screen on call sheet.

01:37:56   Sometimes users will report that this setting is reverted to the default.

01:37:59   I'm completely at a loss as to how or why that could be happening.

01:38:02   Can you take a look and see if you have any guesses, please?

01:38:05   That was the entire prompt.

01:38:06   It obviously has access to all my code, but that was the entire prompt.

01:38:10   It was a brand new session.

01:38:11   I'd given it no other information.

01:38:13   And this had been happening for months now.

01:38:16   And again, I'd glanced at it.

01:38:18   I'd looked at the code here, looked at the code there.

01:38:20   Couldn't freaking figure it out.

01:38:22   So I asked Claude and it churned for some amount of time.

01:38:26   I want to say it was between two, maybe five minutes tops.

01:38:28   And it said, oh, I think I know what this is.

01:38:32   Not verbatim anymore, but I think I know what this is.

01:38:34   And it spelled out what it thought the problem was.

01:38:38   And it wasn't changing code at this point.

01:38:40   It was just saying, here's what I think the problem is.

01:38:44   Actually, so I'm looking back.

01:38:46   This is the compressed version of the conversation.

01:38:48   You know, I had my prompt and it says, I'm going to do this.

01:38:52   I'm going to do that.

01:38:52   Found it.

01:38:53   Let me confirm by looking at discover section state, which is a piece of my code.

01:38:56   And the constant keys, you know, what read for a little bit.

01:39:00   I found it.

01:39:01   Let me confirm.

01:39:01   Nothing else writes the old toggle keyword.

01:39:03   Now I'm getting into specifics.

01:39:04   But then confirmed.

01:39:05   I have a strong guess.

01:39:06   And it clearly explains why only some users see it.

01:39:09   And it turns out that there was a very brief window of time when I allowed users to turn on or off sections of the discover screen.

01:39:18   But I did not yet have the code written to rearrange sections of the discover screen.

01:39:23   And that was a released version of call sheet, not just test flight, like a honest to goodness released version of call sheet.

01:39:30   And I sent that out into the wild.

01:39:32   And then very briefly thereafter, I released into the wild the version that lets you rearrange things.

01:39:37   And it turns out the specifics aren't particularly important, but it turns out that the way in which I was trying to key off of, well, did they have the version that only did the on and off?

01:39:48   Or do they have the version that does on, off and rearranging?

01:39:52   When I was keying off, when I was looking at that and trying to do like basically a migration, I screwed it up.

01:39:57   It was my fault.

01:39:57   I screwed it up such that if you had run the version of the app that did the on off, which again, only existed for maybe a week at most, and then went and did the migration and went to do the newest version at the time, which was the one that allows the on off and rearranging, I screwed up the way I did that migration.

01:40:18   So it would keep migrating every time.

01:40:20   And every time it migrated, it would clear out what was there and it would reset it anew.

01:40:25   So the reason it was sporadic was because it was only people potentially who messed with this setting or at the very least that ran that one specific version of the app.

01:40:34   I don't remember what version it is offhand.

01:40:36   And Claude found it and it found it in five or 10 minutes.

01:40:39   And that was that.

01:40:40   And I thought, okay, well, there's this one other thing that's weird.

01:40:43   Occasionally, when you override the language settings or the region settings and call sheet, similar stuff happens.

01:40:49   Now, that isn't user defaults.

01:40:51   That's, what is it, cloud ubiquitous key value store or something like that?

01:40:56   It's basically like a...

01:40:56   I love key value store, yeah.

01:40:57   Yeah, thank you.

01:40:58   It's basically similar to user defaults, but it's stored in iCloud, right?

01:41:01   And I don't remember the specifics and it doesn't really matter.

01:41:04   But suffice to say, it looked at it for a few minutes and said, oh, this is your problem.

01:41:07   And again, it was me being an idiot at the end of the day.

01:41:11   That's what it boiled down to.

01:41:12   But it was fascinating that here I am at the beach and I'm obviously, like, paying attention to what's going on.

01:41:20   And when it presents a theory, I don't remember if I had it implement the fix, but at the very least, when it presented the theory, I looked at it and then I did end up opening up Xcode because now I have nerd sniped myself.

01:41:32   But it presents the theory and then I look at Xcode and say, oh, no, nope, this sounds right.

01:41:38   Yep, yep, that's got to be it.

01:41:40   And so then I either implemented it myself or I think what I might have had to do is, like, do a first pass and then I rejiggered a fair bit of it.

01:41:47   And then I asked it, okay, I've rejiggered what you've done.

01:41:49   And, you know, does this still look right?

01:41:51   Does this still look right to you, et cetera, et cetera?

01:41:53   But what was fascinating about this was this is two different bugs that, again, wholly my fault.

01:42:00   I'm not trying to deny that they're my fault, but two different bugs that were very pernicious and very tricky and that I had been trying to figure out what the heck was going on, again, on and off for months.

01:42:13   I would look at it, you know, for a few minutes here and then I'd decide I still don't get it.

01:42:18   And then I'd look at it for a few minutes another time and I still don't get it, both of these bugs.

01:42:21   And Claude was able to look through my code and figure it out.

01:42:26   And what was so striking about this, and I was sitting on the couch in the beach house and I was leaning forward because my laptop, my $5,000, soon to be $8,500 laptop, was sitting on, like, the coffee table.

01:42:40   And I've caught myself leaning back against the, you know, the back of the couch and thinking to myself, holy God, I have a coworker again.

01:42:49   I haven't had a coworker in the better part of eight years.

01:42:52   It's, I think, eight years, just a week or so shy of eight years that I've been independent, which is also bananas.

01:42:57   But anyway, I have a coworker again because what this felt like was so, it felt like so much like me saying to a coworker, hey, I'm banging my head against the wall.

01:43:08   Can you take a look at this code and see what you can figure out and working through the problem together?

01:43:13   And let me tell you, one of the things, if not the thing that I miss most about being, about working with coworkers and particularly doing so in person, although, you know, maybe this would be just as fine, just this would work just as well remotely.

01:43:29   But my memory of it was in person because it was eight years ago.

01:43:32   I miss so desperately being in, like, a conference room, standing in front of a whiteboard and working a problem or maybe standing, you know, maybe sitting in front of a projector and, you know, looking at a, looking at code together and working a problem.

01:43:46   And more than I have at any point in the last eight years, I felt like, at least when it comes to code, obviously, I have coworkers with you guys, with Mike, in other cases as well.

01:43:56   But when it comes to writing code, suddenly I had a coworker again.

01:44:00   And as I said in the top of this blog post, as a pundit and as a person, I still don't know what I think about AI.

01:44:08   And I echo what you were saying, John, like, yes, there's so many amazing things that AI provides for us, like this very moment.

01:44:16   But is it worth my computers being twice as much?

01:44:19   Is it worth, you know, possibly systemic inflation, as we were talking about earlier?

01:44:23   I don't know.

01:44:25   As a pundit, I'm really not sure how I feel about it.

01:44:27   But as a developer, this was an incredible feeling that really felt genuinely good.

01:44:34   And as I hear myself talking, I'm a little uncomfortable with how much I feel like I sound like, oh, I just, I found myself a girlfriend.

01:44:42   You know what I mean?

01:44:43   Like, obviously that's not at all the same, but it almost was like a, not a relationship feeling, but like, oh, this is an interpersonal thing or on the verge of an interpersonal thing that I haven't had in so long.

01:44:54   And I know that I could get either one of you or both of you on a Zoom call.

01:44:59   And if I was really desperate, you would work with me, work through these problems with me.

01:45:03   But it takes so long to bring any human being up to speed on my code base and the way I like to write code that even if they would have taken either of you 10 minutes to spot the bug, it would have taken the three of us working together or the two of us working together 10 hours to get up to speed on the way my code works.

01:45:21   And as a developer, thinking myopically as a developer, this is such an incredible boon to my ability to write code and do work and fix problems.

01:45:33   And I'm so enthusiastic about it and so excited about it.

01:45:36   As a pundit, as a human, as a person, I don't know.

01:45:40   As a developer, holy crap, it's so cool.

01:45:42   Yeah, I have found, I'm leaning more into that opinion as time goes on.

01:45:51   You know, when you look at, even just what you're saying now, like even just the bug fixing and testing abilities of modern AI tools, it dramatically changes the landscape of software development.

01:46:04   Like, even if you're writing most of your code yourself, like the other day, I, look, if you're an Overcast user, there's a couple of things that you probably have with problems with the app.

01:46:14   And some of them are probably obscure bugs.

01:46:17   And one of the most common obscure bugs that I get reported is basically unexpectedly large storage usage.

01:46:23   Like somewhere, something is leaking temp files or something.

01:46:26   And what is going on?

01:46:28   Why are some users getting really high storage usage for Overcast?

01:46:32   And I have thought for a while this is related to the background download daemon because it creates these weird like NSURL session D folders in various places and fills them with crap.

01:46:42   And, you know, maybe at some point the temp files aren't being cleaned up properly.

01:46:45   And so I've had logic in Overcast to try to clean those up over time.

01:46:48   And I actually, I rewrote the downloader a few months back in the fall.

01:46:55   And I actually made it a whole separate component.

01:46:58   I gave it a name.

01:46:59   I might open source it at some point.

01:47:01   It's like a whole separate thing because I was so tired of having to deal with downloader bugs.

01:47:06   And I feel like let me actually write a good one and open source it so that nobody has to deal with downloader bugs.

01:47:10   And maybe other people can help me fix my downloader bugs.

01:47:13   And it turns out, you know, most of the things I've open sourced, they don't get a lot of users because I don't, I don't make the kinds of things people want.

01:47:19   And I don't support them in the kind of way that open source software needs to be supported to get a lot of community around it.

01:47:25   Oh, well, but I said, you know what?

01:47:28   The other day, just last week, I asked Claude, hey, here's, here's this component in Overcast that's the downloader.

01:47:34   People are reporting the soft, this, you know, this storage usage.

01:47:39   I can't figure out why.

01:47:41   Look through this component.

01:47:42   It's probably here and try to figure out why and fix it.

01:47:45   And sure enough, it churned away for a while.

01:47:49   I read a bunch of stuff.

01:47:50   It dumped a bunch of garbage to the console.

01:47:52   And then it found a bug.

01:47:55   There was one case where I had like, you know, I was testing a Boolean condition with a knot on it and I had the logic backwards.

01:48:03   Oh, this is like a weekly occurrence for me and I'm not proud of that.

01:48:08   But oh my gosh, I need one of those signs.

01:48:10   It has been zero days since I've screwed up basic Boolean logic.

01:48:13   Oh my gosh, I'm right there with you.

01:48:14   Well, of course, we're programmers and we're humans.

01:48:15   Like that's like programming involves a lot of Boolean logic and it turns out humans aren't perfect.

01:48:22   So it found this inverted Boolean condition.

01:48:26   I never would have spotted this.

01:48:27   The only way this would have ever gotten fixed is next time I rewrote the entire component.

01:48:31   Maybe I wouldn't have made the same mistake.

01:48:32   Yep.

01:48:33   Or if you had tests.

01:48:34   Well, maybe this was this was this would have been a hard one to come up in a test, honestly.

01:48:39   But yeah, point taken.

01:48:40   And I've had clawed right tests for me, too.

01:48:43   It's wonderful.

01:48:44   On the topic of Boolean logic, though, I've never even proposed this and everyone who has proposed it would have surely been shouted down with this exact argument.

01:48:51   But I come from a language that had the keyword and less.

01:48:55   So you didn't have to do if exclamation point condition you could do in less condition.

01:49:00   And, you know, anyone who hasn't used that thinks it's going to be the end of the world because they're going to oh, it's going to make me screw up the logic because I can't handle it.

01:49:06   Right.

01:49:06   Guess what Swift test.

01:49:08   Guard condition else.

01:49:11   Yeah.

01:49:12   Oh, yeah.

01:49:12   And guard condition else is, I would say, worse than an unless keyword in terms of making you accidentally put the wrong logic in a condition, especially if you convert an if to a guard once you realize it's kind of a precondition.

01:49:25   Don't forget to reverse all the logic.

01:49:27   So all you people out there saying I'm never going to allow unless and Swift.

01:49:30   You would love unless and guard as much as I like it.

01:49:34   And I understand the function that serves in the language guard else.

01:49:39   I don't particularly like that.

01:49:40   I've come to enjoy the guards, but linguistically, it is kind of weird, but I like the function.

01:49:45   Anyway, you know, Claude found this bug and it could very well be related to some weird edge case downloaded behavior and the downloader never cleaning up those temp files that are abandoned by the system.

01:49:58   So like if if the background download demon abandons a file, like if the process crashes, maybe then that file might never be cleaned up.

01:50:06   This found that bug.

01:50:07   And while it was there, it found a couple other little bugs that might affect background download terminations.

01:50:14   Nice.

01:50:15   Oh, and by the way, and it was it was giving statements to me in in like the feedback chat stuff about like, oh, when users delete episodes, I never told it it was a podcast app.

01:50:25   I never the the component it was looking at was not dealing with episodes.

01:50:29   It figured all that out just from the rest of the code like this is amazing.

01:50:35   And, you know, I come from a little bit different angle in that it's been a much longer time since I've worked with other programmers and I never work with that many of them.

01:50:43   Like my my biggest team I ever worked on programmer wise was like six people.

01:50:47   I'm not coming from the corporate world where I'm used to having a lot more infrastructure around me, a lot more like testing and process and things like that.

01:50:55   Like I don't I don't come from that world.

01:50:56   So I don't know it.

01:50:58   The way I see this so far is like not only is this allowing me to write better software, the kind of bugs it's finding, you know, you know how to do next overcast fans.

01:51:10   I'm having a look at my priority podcast logic because the priority podcast insertion into into playlist with priorities set on them ever since the rewrite.

01:51:20   People have been very upset that it has weird behavior that they don't think is right.

01:51:25   And that's a really hard thing to write tests for.

01:51:28   And of course, I'm not going to do it anyway, but Claude can and Claude also found some logical bugs in that.

01:51:35   And so I'm going to in the next few days, probably ship an overcast beta update that includes that download fix and these priority playlist fixes and people can start telling me how they work for them.

01:51:44   And these are these are like really obscure little logical things like the priority playlist thing was like a greater than less than was flipped backwards.

01:51:54   Little things like that that are just really hard for humans to find.

01:51:59   And it's and it found them like the first time I asked it to.

01:52:03   So it's allowing me to write better software and it's allowing it's allowing me to continue being one person team for longer.

01:52:12   Now, I don't know how long I want to be a one person team with overcast.

01:52:15   To be honest, I've been honestly thinking about maybe trying to hire another programmer to help help me, you know, do more things because I feel a little bit stretched right now.

01:52:23   But Claude gets me some of the way there in terms of like if I have like certain bugs on my hit list, I can have it take a look and it fixes it.

01:52:33   So just disclosure, Claude has been a sponsor of our show and probably still will be.

01:52:36   But honestly, all this also applies to codex and everything.

01:52:40   When you have it, look, it's just like an AI broadly.

01:52:42   Right now, I happen to be using Claude.

01:52:44   But anyway, I also think like from what I've heard, Apple also uses Claude internally a lot over the last year or so.

01:52:53   Well, look at what happened at WBDC this year when they scrolled by those giant walls of text of oftentimes fairly small but nice little fixes.

01:53:03   How do you think Apple did all that in one year?

01:53:06   It's not because they had a change of heart and realized software quality now matters.

01:53:11   No, it's because they are using Claude to fix their own bugs and to help them develop things more quickly.

01:53:20   And that's going on across the industry.

01:53:23   So I think even though this is this is kind of a it's going to be a bumpy ride for lots of reasons, as we've talked about, both from for like a straight up like, you know, AI coding and security perspective, which we'll talk about more in the in the overtime.

01:53:36   Also to just everything we've been talking about, you know, AI itself is very disruptive and turbulent and causing a bunch of turbulence in the world and is not a universal gain for everybody.

01:53:45   But specifically in the area of software and software quality, I think it's a huge boon because everyone can start finding and fixing really weird, obscure bugs that a human never would have found.

01:53:59   And when you see Apple's giant wall of text, WVDC slide, you can tell that's what's going on here.

01:54:05   They are fixing their own stuff and we are all benefiting from it.

01:54:09   So I again, it's as Casey said, like this is not a universal good.

01:54:15   But there are some real benefits to this and specifically this area.

01:54:20   And I am here for that.

01:54:22   Yeah.

01:54:23   And just very briefly, and I also covered this very briefly in the blog post, but I'd like to reiterate here.

01:54:30   In my perfect world, I would have enough money coming in from CallSheet where I could hire somebody either, you know, half-time or full-time to be that person that we can work together to figure these things out.

01:54:44   But CallSheet does not make that kind of money.

01:54:46   It does make enough money to handle a, you know, monthly fee from Claude, which I'm now paying.

01:54:54   I mean, they did sponsor in the past and I was rolling on the freebies that they gave us when they sponsored.

01:54:59   But now I'm paying them and I'm choosing to because it has been that transformative for me.

01:55:05   Now, again, in a perfect world, I'd much rather have a human being to work with on this.

01:55:09   To be honest, I'd still probably ask Claude some of these questions, but I think there's something to be said for having a human and having a human do these things and make these decisions and use what makes humans uniquely wonderful and amazing to figure out some of these problems.

01:55:24   But since I don't have that and I cannot afford that with CallSheet alone, the next greatest thing I could do is ask Claude.

01:55:33   And I think that the thing is, what I'm saying in a roundabout way is, this isn't the question of hiring a human or hiring Claude.

01:55:40   A human was never going to be here.

01:55:43   I just don't, it doesn't, CallSheet does not make that kind of money.

01:55:45   So the question is, do I fix these bugs by using the tools that I have at my disposal or do I not, right?

01:55:53   And that's the thing.

01:55:55   And so because of that, I don't feel too guilty about it.

01:55:59   I feel guilty about AI in the broad strokes like we were talking about, but in this one specific way, I don't feel too guilty about it at all.

01:56:07   And it is so rewarding having what isn't a person, but occasionally kind of feels like a person, to fire questions at and work with and figure out problems.

01:56:16   And again, even a lot of times I don't take the code that Claude writes or I vastly change it to work, to fit my preferences and style and so on and so forth.

01:56:25   But having it identify where a problem is, like Mark, like I said, like Marco said, is incredibly valuable, incredibly valuable.

01:56:33   And it makes my product better, which makes my customers happier.

01:56:37   And I think that's very cool.

01:56:38   Two things on the topics you touched on first, Marco, on the big wall of text and Apple fixing bugs.

01:56:44   I'm sure Apple's using all the same tools everybody is to find their bugs, but you still need the essentially corporate mandate that that's where they're allowed to spend their time.

01:56:51   Because every bug fix requires understanding the bug, finding it, verifying the fix, and the risk inherent in changing any code and blah, blah, blah.

01:56:59   And all of that requires a corporate mandate, an actual decision that, hey, this release, we're going to fix bugs.

01:57:04   Once that decision is made, yes, the tools make bug fixing way more efficient as before.

01:57:09   But I still give credit to management deciding we are going to fix bugs this release.

01:57:14   Maybe the decision was influenced by the increased power of the tools available to do that.

01:57:18   But it's the decision and management that's the most important part.

01:57:21   In other words, if Apple hadn't made this decision and these tools were available, they still would have been forced to do whatever features there have to do for the release.

01:57:29   And they wouldn't have been able to spend time with these tools to fix bugs.

01:57:31   Maybe they would have snuck a few in because they're more powerful, but nothing like what we saw.

01:57:34   So kudos to Apple for deciding to spend time fixing bugs.

01:57:37   And Casey, on your blog post, I do feel like even though you've spent a lot of time talking about the nitty gritty of the bug or whatever, I felt like the thrust of your post was what you were getting at at the end there, which is like the feeling of working with someone else.

01:57:49   Listen, I feel like your post essentially identified or crystallized for you a need that you have.

01:57:58   And I'm sure this is not a surprise to you.

01:58:00   You know you miss working with people or whatever.

01:58:01   But like having that experience with the coding agent was like, you know, I really do miss working with people.

01:58:08   I miss this.

01:58:08   Yeah, this kind of reminds me of that.

01:58:11   And here's a need that I have in my life.

01:58:13   On the flip side of that, I would say that even though this evokes that feeling and reminds you that you want it to, you know, the AI thing is not a person.

01:58:23   It's I know sometimes it fools you into thinking it is, but it's like that's that's a dangerous trap.

01:58:29   I mean, it's a trap a lot of people fall into.

01:58:30   Obviously, it's more dangerous the more these things are able to fool people into thinking that they're a little person that they're talking with.

01:58:38   And so, I mean, in all cases, when someone finds himself in that situation, it is highlighting a need, not fixing a problem.

01:58:46   And the need is I have a need for human context, right?

01:58:49   I miss my coworkers or whatever.

01:58:52   That need is not filled by talking to an LLM agent, unfortunately.

01:58:56   So take if you find yourself using one of these things and finding it's like, I just I find myself just I just love talking to it.

01:59:02   It makes me feel better.

01:59:03   That's highlighting a need that is not fulfilling that need.

01:59:06   So please, everybody, yeah, LMs are not people.

01:59:10   But but the need is real.

01:59:12   And I feel like that experience and like that what you took away from it is not entirely.

01:59:16   Oh, I know I have a tool to make my apps better.

01:59:19   That's part of it.

01:59:19   You just talked about it.

01:59:20   But also, I've identified something that I'm not getting in my life, especially if it's something that you used to have.

01:59:26   I'm like Marco, who is far removed from that and never really had it to the same degree as you.

01:59:30   You could find other ways to get that.

01:59:32   I mean, I mean, working on open source projects or other things where you're collaborating with someone on code doesn't make sense to do it for call sheet for the reasons you stated because you'd have to hire someone to do that because it's not like someone's going to help you with your app for fun.

01:59:43   But that's, you know, that's maybe argues for like, say, say you get super into jellyfin or home assistant or whatever, like dabbling in that environment.

01:59:51   It doesn't become a second job or a third job or a fourth job or whatever.

01:59:54   But like that could give you that experience of working with other people.

01:59:58   And then you also mentioned like you could have had someone else like ask some other programmer, you know, to look at this thing and see if we could figure out the bug or whatever.

02:00:06   Well, guess what?

02:00:07   If you asked another programmer to do that, if you asked me to do that, I would have said, why don't you just ask the agent of your choice to find bugs?

02:00:13   Like that's because that and that highlights the role of these tools from my perspective is they are tools.

02:00:20   There have been tools that will help you find bugs in your program before LLMs.

02:00:24   There's, you know, fuzz testers and Valgrind and all like this is not the first tool that humans have ever made that helps find bugs in programs.

02:00:30   It's an amazing, powerful one that surpasses all others.

02:00:33   Yes, but it's not the first one.

02:00:35   And so if you were asked another human today to come in and help you find bugs in your program, this would be the tool that they would pull out.

02:00:40   They wouldn't spend 10 hours learning your code.

02:00:42   It's like, well, let's do this for five minutes at least.

02:00:45   And if it can't find them, oh, well, then we'll dig in.

02:00:48   But ignoring the tools available to you as a programmer is counterproductive unless you're taking some kind of moral stand against it, which is fine.

02:00:55   But I'm saying most people who you asked would would would grab for the same tool that you were already going to use anyway.

02:01:01   And that's how I view these things.

02:01:02   They are, in fact, tools for programming tools with questionable origins and bad externalities and whatever.

02:01:09   But in the particular case of programming, as I've argued in the past, I personally believe that it is possible, not currently happening, but possible to make a ethical, environmentally responsible, created under ideal circumstances, blah, blah, version of this simply because code like to give an example.

02:01:30   If Apple trained its own coding agent, if it was not terrible at doing that, let's say they were actually good at doing it, they train their own coding agent on all the source code Apple has ever created.

02:01:40   That would be an incredibly powerful agent for writing for Apple platforms just based on the code that Apple owns because it wrote it.

02:01:47   That's what it would be trained on.

02:01:49   And they would do a good job of it.

02:01:50   And they would hire people to like, you know, like just they could make an Apple centric coding agent and then run it on, you know, in data centers run 100% for renewable resources.

02:02:00   You know, like do all the things like it is possible because we see the technology is there and it's like, well, isn't the rest of it just details?

02:02:07   I mean, it's kind of important details.

02:02:08   So if you if you're living in Memphis and you got XAI data centers spewing out smoke next to you, it is not a it is not a theoretical thing.

02:02:14   It is a real terrible thing.

02:02:16   So I get that.

02:02:16   But like my techno optimism says, but I can see how it would be possible to both have this useful tool and not destroy the world and bankrupt us all and blah, blah, blah.

02:02:28   Currently, that's not the path we're going on, which sucks, but it seems possible.

02:02:31   So I do feel like this kind of tool will inevitably become an essential part of every programmer's toolbox eventually, hopefully not in the current form that we're currently currently have it available to us.

02:02:45   And perhaps after a long, dark period of it being so expensive that nobody can use it, because if we were paying the true cost of the things that we're doing right now, none of us would be using it either.

02:02:54   Because in the same way that call sheet can't doesn't make enough money to hire another programmer, it also doesn't make enough money to pay for the tokens you've been using.

02:03:01   If you look at the actual cost of you're getting whatever you're paying for your agent for a monthly thing, that is not the true cost of what you're getting.

02:03:08   The true cost is thousands and thousands and thousands of dollars.

02:03:11   And that's a bubble, folks.

02:03:13   And so someday, maybe by the time that ends, inference will have gotten so much more efficient with dedicated ASICs and everything, and we'll all be running on renewable resource data centers and all that stuff.

02:03:22   But we're not there right now.

02:03:24   So it's a weird time to be in, but I do, I did enjoy your blog post, less for the technical aspects of it, which, by the way, we will get into more in overtime, because it's more in that vein, but more in terms of like, what I felt like your thrust was in the post, which was like, I miss programming with people.

02:03:41   Yeah, which is, which is a very strange conclusion going in to say, oh, I use these cool AI tools to do coding or whatever.

02:03:46   And the lesson is, I miss coding with people.

02:03:49   Not because the tools are bad.

02:03:51   You just go, you explain how the tools are good and helped you, but because it reminded you of what it was like back when you work with people.

02:03:58   And I felt that, too, because I used to work with people as well.

02:04:00   And that's, I mean, if you're ever listening to this program, it's like, well, I want to be a cool indie developer and support myself and do all these things.

02:04:07   There are lots of upsides to that, but there are downsides, too.

02:04:09   And one of them is if you came from a world where you were working with other people, even if you're not particularly social like me, just working with people in their capacity as fellow programmers on a problem or working with people on any kind of problem, programming or otherwise, is fun and, you know, rewarding.

02:04:25   And when you're working on your own, unless you can hire a staff, and even if you can, by the way, because your staff aren't necessarily going to be your friends, especially if you're like your boss or whatever.

02:04:34   It's a downside of being an indie developer is like indie developer loneliness, I guess.

02:04:40   And yeah, that's what I took away from your post, and I thought it was good.

02:04:43   I think the thing I missed the most from working in an office is lunch, like with other people, just being able to hang out.

02:04:50   You can have lunch at home, Marco.

02:04:51   It's okay.

02:04:54   No, I hear you.

02:04:55   I really do.

02:04:55   I mean, I was lucky enough to work in very, very progressive, isn't the word I'm looking for, but places where they did not micromanage every second of my day for the most part.

02:05:05   And I remember like at my most recent jobby job, we, this is when the Switch was brand new or very new.

02:05:13   There were a handful of times that we would go down to the cafeteria and all of us would do like, would, would do eight player Mario Kart races with our Switches for a little while over lunch.

02:05:22   And like that, you can do that to a degree over the internet, but it's just, it's fun and different doing it in person.

02:05:28   And yeah, you're, you're exactly right.

02:05:30   I mean, that was the thing that struck me so much about this experience was it reminded me of having coworkers and I do miss that quite a bit.

02:05:38   I am not, I'm not trying to complain.

02:05:39   I have, I live an incredibly blessed and lucky life.

02:05:43   I am so thankful to everyone listening to my words right now because you, you listening to me makes my life, Marco's life, John's life possible.

02:05:51   Um, I don't, please don't hear what I'm not saying.

02:05:53   I'm not trying to complain, but as John said, there's something to be said for having coworkers, for having people.

02:05:59   Even if they're just work friends.

02:06:01   And once you leave that job, you never speak to them again.

02:06:04   It's still nice to have work friends.

02:06:05   And I, and I miss that in the development capacity.

02:06:08   I'm so lucky to have that with the two of you, with Mike and with others from a podcasting capacity, but I do miss it quite a bit from a development capacity.

02:06:17   Beep, beep, beep.