PodSearch

ATP

647: You Get One Exclamation Point

 

00:00:00   John, something occurred to me, I was listening to directives earlier today, and it was very funny to me that you two talked about Das Boot, because there is a forthcoming and comparable episode that I was on where we talked about it. I also had never seen Das Boot before.

00:00:15   But it occurred to me as you were talking about this and talking about your graduation results, that you two are going to be empty nesters real freaking soon, which I think I knew, but I never really thought about it. You are almost there. And how are we feeling about that?

00:00:30   Does it count as empty nester if your kids are still living at home, they're just at school? I thought it was like when they move out.

00:00:36   And I think strictly you are correct. I think it's only when they move out permanently that you are officially an empty nester, but you're an effective empty nester in a couple of months, and that's wild. How are we feeling?

00:00:45   I don't know. I feel like I just haven't looked that far yet, still trying to always get over the next hurdle, and it's just like getting the kids off to school, getting my daughter settled in her first college dorm, getting my son off for his senior year, figuring out what he's going to do.

00:01:01   Like, it just feels like there's still a lot of stuff to overcome. But once they're settled in, and I've stopped doing the 15 trips back and forth to ferry all the things they inevitably forgot, I don't know, maybe ask me then.

00:01:12   And see what it's like. But mostly, I think it'll be good for everybody involved, and it'll be a fun change. But we'll see how it goes. Maybe I'll just be bored.

00:01:21   Let's do some follow-up. Tahoe is apparently dropping Firewire support, and we should pour one out. Did we talk about this already?

00:01:30   We did. We sure did. Wow, I jogged your memory. I was waiting for the little gears to kick in.

00:01:34   It was like three days ago.

00:01:35   It was so long ago when we recorded that episode.

00:01:37   This is a thing. This is a thing we've done. But we have more information about it. That's why I brought it up. See, I'm a professional, I swear.

00:01:43   That might be in follow-up, you know, because it's a thing we talked about on the previous show.

00:01:48   Well, with you, you never know. Everything that's ever been thought about in the last 30 years is somehow follow-up.

00:01:54   No, no, no. I mean, this was literally two days ago, so I'm going to say this one is pretty valid as follow-up.

00:01:59   It is. It is. I'm trying to blame you, even though this is 100% on me.

00:02:03   All right. So anyways, Timo Hetzel writes,

00:02:06   I was shopping for audio interfaces about 10 years ago.

00:02:08   I found a box in the store with a sticker on it that read,

00:02:10   Now with Thunderbolt!

00:02:12   Turns out they just added a free Firewire to Thunderbolt adapter to the existing Firewire-based model.

00:02:18   The World of Audio is one that came up when we were discussing Firewire support going away.

00:02:23   They're very conservative over there, to the point where if they have a device that works,

00:02:27   you want to make a version of it with Thunderbolt?

00:02:30   Just put an adapter in the box.

00:02:31   And it made me wonder, like, if they do that, though,

00:02:33   does that get around the lack of Firewire support?

00:02:36   Because as far as the computer is concerned, it's not Firewire, it's Thunderbolt.

00:02:39   I think it does, right?

00:02:40   I would assume, but you would be the one to know that.

00:02:42   The way Thunderbolt works is basically, it's like PCI Express over a cable.

00:02:49   It's like the simplified version of how it works.

00:02:51   And so it's almost like you're plugging in a PCI Express card with a Firewire controller on it, I think.

00:02:57   Like, you know, logically for the computer.

00:02:59   So if the computer drops support for Firewire as a host protocol,

00:03:04   I think any way you connect it through Thunderbolt would also not work,

00:03:08   because all Thunderbolt is doing is exposing the Firewire controller to the computer,

00:03:12   saying, here, here's a Firewire controller.

00:03:13   And so if the computer has no idea what to do with that...

00:03:16   I was thinking that it would just expose the device as if it's connected directly to the PCI bus.

00:03:20   You know what I mean?

00:03:20   Yeah, but does Tahoe drop support?

00:03:23   Like, if you have a Mac Pro with card slots,

00:03:25   and if you put a Firewire card in one of those PCI slots,

00:03:28   I bet that'll also stop working.

00:03:29   Well, I don't know.

00:03:30   I mean, I guess this is one of those things I feel like,

00:03:33   well, someone obviously who knows the technical details could just tell us.

00:03:36   But failing that, when Tahoe does actually ship with no Firewire support,

00:03:43   I'm sure someone will either accidentally or on purpose get upgraded

00:03:48   and realize that Firewire stuff doesn't work,

00:03:50   and then try to find ways to make it work, and then we'll find out.

00:03:52   We shall see.

00:03:54   I'm sure someone will be kind enough to do an experiment for us.

00:03:57   If you wanted to install Tahoe Betas without installing Xcode Beta,

00:04:01   I feel like we've definitely covered this at least thrice at this point,

00:04:04   but that's all right.

00:04:04   Anonymous writes,

00:04:06   there's a specific device support package you can install

00:04:08   that doesn't require installing the entire Xcode Beta.

00:04:11   Device support for macOS 26 Beta 2.

00:04:14   You can install device support for macOS 26 Beta

00:04:16   if installing the macOS seed in a virtual machine fails on a host Mac.

00:04:21   You can get this from developer.apple.com download,

00:04:24   but we will also put in a direct link,

00:04:26   although I do believe you get challenged for login

00:04:28   when you click the direct link.

00:04:30   Yeah, you need to be logged into a developer account,

00:04:32   but it's interesting that, like,

00:04:33   so all these requirements that I didn't know

00:04:34   because I had never done this before,

00:04:35   that you need the Xcode Beta,

00:04:36   but actually you don't need the Xcode Beta

00:04:38   because you just need the device support,

00:04:39   and actually they distribute that separately.

00:04:41   It's one of the things that annoys me

00:04:43   about Apple's developer website.

00:04:44   Finding anything on there in terms of downloads

00:04:47   is just always such a puzzle

00:04:49   because they don't,

00:04:50   they have this weird,

00:04:51   they divide it up into, like,

00:04:52   three different tabs.

00:04:53   It's like OSS applications and other,

00:04:55   or there's, like, a more thing,

00:04:56   and then sometimes there's an additional download section,

00:04:58   and what if you want some older stuff?

00:05:00   Is that still available somewhere?

00:05:01   That's why all these little websites pop up,

00:05:02   basically saving the URLs of where things were

00:05:06   when they were first put up somewhere

00:05:07   on some CDN that Apple controls,

00:05:09   and even when it goes off their website,

00:05:10   as in it's no longer linked from a webpage,

00:05:12   the download is still there,

00:05:13   and you can get it if you can,

00:05:15   and if it's not one of those things

00:05:16   that Apple needs to sign or something.

00:05:17   Anyway, yeah, all things that could have benefited me,

00:05:20   but many different ways to skin this cat,

00:05:23   and apparently device support for macOS 26 beta 2

00:05:27   is one of them.

00:05:28   I wonder if that's going to be updated

00:05:29   with device support for macOS 26 beta 3,

00:05:32   or if it's not beta specific.

00:05:33   We'll find out.

00:05:34   We shall see.

00:05:35   A friend of the show, Guy Rambeau,

00:05:36   with regard to putting older OSs on newer Macs,

00:05:40   so if you buy a Mac that comes with Tahoe

00:05:42   in a couple of months,

00:05:43   and you wanted to put Sequoia on it,

00:05:45   so Guy writes,

00:05:46   on the subject of running an older OS

00:05:48   than the one that a given Apple Silicon Mac

00:05:50   was originally shipped with,

00:05:51   the short answer is that it is effectively impossible.

00:05:54   Even if you patch the restore image

00:05:56   by copying the missing components

00:05:57   from a newer release,

00:05:58   you'd end up with iBoot components

00:06:00   that don't match the kernel cache,

00:06:01   or a kernel cache that doesn't match

00:06:03   the DYLD cache in the root file system.

00:06:05   It would be a mess.

00:06:06   Regardless of the above issues,

00:06:08   you wouldn't be able to get such a restore image

00:06:10   to be flashed onto a DFU Mac,

00:06:12   because Apple's online signing service

00:06:14   would refuse to sign the firmware components

00:06:16   for a board that's not allow-listed

00:06:19   for that OS release,

00:06:20   and that signing is mandatory

00:06:22   when restoring an Apple Silicon Mac from scratch,

00:06:24   even with a lower security configuration.

00:06:26   The only way I could think of

00:06:28   that could potentially be used

00:06:31   to achieve such a thing

00:06:32   would be to hand-assemble

00:06:34   the entire volume scheme

00:06:35   that's required by a Mac OS installation

00:06:37   on Apple Silicon

00:06:37   using the file system components

00:06:38   from the older OS

00:06:40   and the firmware or kernel cache

00:06:42   from the newer OS.

00:06:43   Doing that is far from trivial,

00:06:45   and even then,

00:06:46   the result would likely be a machine

00:06:47   that kernel panics very often,

00:06:48   even if not before it even finishes booting up.

00:06:51   Yeah, so basically,

00:06:52   if you're thinking like,

00:06:54   should I attempt some kind of Mac OS hack

00:06:58   to get something working,

00:06:59   if Guy Rambeau says

00:07:00   that's going to be really difficult,

00:07:01   don't.

00:07:03   Just don't.

00:07:03   It's not going to happen

00:07:04   for the rest of us.

00:07:06   Basically, it's impossible.

00:07:07   All right, and then we talked last episode

00:07:10   about uninterruptible power supplies,

00:07:13   I believe this was NASC-ATP,

00:07:15   and how they deal in volt amps,

00:07:17   and the two or three of us,

00:07:20   but particularly John and I

00:07:21   probably should have known this

00:07:22   or did know this at some point,

00:07:23   kind of shrugged.

00:07:24   Well, Marshall C. writes,

00:07:26   I have my Bachelor's of Science

00:07:27   in Electrical Engineering,

00:07:28   and I'm a licensed professional engineer

00:07:29   in the EE power discipline.

00:07:31   EC power is tricky,

00:07:32   but here's my layman's explanation.

00:07:34   Real power, or P,

00:07:36   which is measured in watts,

00:07:37   this is the electric power

00:07:39   used to run your computer,

00:07:39   heat up your toaster,

00:07:40   and what your utility bill bills you for.

00:07:43   Reactive power,

00:07:45   which is represented by Q or V-A-Rs,

00:07:47   is often called fake power.

00:07:49   This is the electric power

00:07:50   used to form the necessary magnetic fields

00:07:53   for the transmission of AC power.

00:07:55   Then finally, there's apparent power,

00:07:57   which is represented by S,

00:07:59   and that's volt amps,

00:08:00   a combination of real and reactive power.

00:08:02   Imagine an X-Y axis

00:08:03   where the real power is on the X axis

00:08:05   and the reactive power is on the Y axis.

00:08:06   Apparent power is the distance

00:08:09   from the origin to the coordinate.

00:08:10   Non-utility people will never need

00:08:13   to think about reactive power

00:08:14   unless their job involves industrial equipment

00:08:15   like electric arc furnaces,

00:08:17   furnaces,

00:08:18   I can get there, I swear,

00:08:19   large motors, etc.

00:08:21   and they're responsible

00:08:22   for the engineering or bills

00:08:23   to the power company.

00:08:24   Apparent power is how devices

00:08:25   that generate power

00:08:26   have to be rated

00:08:27   because they're responsible

00:08:28   for providing the real

00:08:29   and reactive power for the system.

00:08:30   For most consumer situations,

00:08:32   you can assume that the watt rating

00:08:34   will be roughly the same

00:08:35   as your V-A rating.

00:08:36   Look, AC power is insanely complicated.

00:08:40   Like what I said last,

00:08:41   DC power, fairly simple to conceptualize.

00:08:44   AC power, it's so complicated.

00:08:46   One thing,

00:08:48   I went to a family member's house once

00:08:51   and they had plugged into the wall

00:08:53   these like things

00:08:55   that claim to be energy savers

00:08:57   and they like lit up

00:08:57   and I was curious

00:08:58   so I looked it up

00:08:59   and it's basically like

00:09:02   it's some weird Amazon device

00:09:04   from God knows where in China

00:09:06   that bridges the two AC prongs

00:09:09   with a capacitor.

00:09:10   I think it was the gist of it

00:09:11   and at first I was like

00:09:13   there's no way that could possibly work

00:09:14   and I was mostly right

00:09:17   but not entirely right.

00:09:19   Like I looked up like you know

00:09:22   how do these purport to work

00:09:24   and does it actually achieve

00:09:25   some kind of power savings

00:09:26   somehow magically

00:09:27   and the answer basically is like

00:09:29   it plays a trick

00:09:30   with the way those AC like phases

00:09:34   and pulls and induction motor loads

00:09:37   like it pulls some kind of trick with that

00:09:39   that like if it did a really big version of it

00:09:43   it might slightly save you some

00:09:45   of your billing of your power

00:09:47   but of course these little tiny things

00:09:49   were scams

00:09:50   and didn't actually do anything

00:09:51   like notable

00:09:52   but like the principle

00:09:53   upon which they claim to work

00:09:55   was not entirely BS

00:09:57   it was just mostly BS

00:09:58   but yeah they are regardless scams

00:10:01   but I thought that was interesting

00:10:02   it's like oh

00:10:02   there's other weird complexity

00:10:04   to AC power

00:10:04   that like

00:10:05   it's sort of

00:10:06   kind of

00:10:08   might enable weird tricks

00:10:10   to happen

00:10:10   on a much larger scale

00:10:12   I mean

00:10:13   tampering with your electrical meter

00:10:15   will also save you money

00:10:16   yes

00:10:17   probably counts as fraud

00:10:18   and so does

00:10:19   definitely

00:10:19   modifying the electrical system

00:10:20   at your house in some way

00:10:21   to make the electrical meter

00:10:22   or read incorrectly

00:10:23   so yeah

00:10:24   all right

00:10:25   Jared Nichols writes

00:10:27   with regard to smoke alarm

00:10:28   remote notifications

00:10:29   so Marco

00:10:31   I believe you were talking

00:10:32   last episode

00:10:33   about how

00:10:34   you've got your

00:10:35   you know temperature sensors

00:10:36   and we were talking about

00:10:37   leak sensors

00:10:38   but then what about

00:10:39   smoke detecting

00:10:42   and so Jared writes

00:10:43   listen to Overtime

00:10:44   and Marco mentioned

00:10:45   smart smoke alarms

00:10:46   by Yolink

00:10:46   if you've got HomePods

00:10:48   in the house

00:10:48   you've effectively now

00:10:49   got smart smoke alarms

00:10:50   I've got regular

00:10:51   residential smoke alarms

00:10:52   and when my son

00:10:53   home from college

00:10:53   burns bacon

00:10:54   while my partner

00:10:55   and I are not home

00:10:56   we get a home app alert

00:10:57   about the smoke alarms

00:10:59   going off

00:10:59   burn the bacon again

00:11:00   yeah

00:11:01   and then back to normal business

00:11:02   yeah so

00:11:04   I thought this was interesting

00:11:05   so I actually

00:11:05   I have gotten one of these alerts before

00:11:07   because I do have HomePods

00:11:08   and one time

00:11:10   the smoke alarms

00:11:11   did go off

00:11:11   and I was

00:11:12   I was not home

00:11:13   but my wife and son were

00:11:14   and I

00:11:16   and it comes in

00:11:16   as one of those

00:11:17   like high priority alerts

00:11:18   that breaks through

00:11:19   do not disturb

00:11:19   and so like

00:11:20   it will sound

00:11:21   no matter what

00:11:22   like it's one of those

00:11:23   kind of you know

00:11:23   urgent alerts

00:11:24   on the iPhone

00:11:25   and it sounds

00:11:26   you know like

00:11:27   it sounds like

00:11:27   an urgent alert

00:11:28   you know it's like

00:11:28   a very alarming sign

00:11:29   like oh god

00:11:30   and you look

00:11:30   and it's like

00:11:30   you know

00:11:31   it gives you a notification

00:11:32   saying like oh yeah

00:11:33   there's

00:11:33   it sounds like a smoke alarm

00:11:34   is going off in your house

00:11:35   and like oh my god

00:11:36   and in that case

00:11:39   it turned out

00:11:39   they had burned something

00:11:40   in the oven

00:11:40   but I have seen

00:11:42   these notifications

00:11:42   and so it is better

00:11:44   than nothing

00:11:44   in the sense that like

00:11:45   if you want to be

00:11:46   remotely notified

00:11:47   of your smoke alarm

00:11:47   going off

00:11:48   that's good

00:11:49   that's one way to do it

00:11:50   another way

00:11:51   is to just get a smoke alarm

00:11:52   like nests

00:11:53   or you know

00:11:54   or these Yolink ones

00:11:55   or there's a bunch

00:11:55   of weird Amazon brand ones now

00:11:56   and the reason why

00:11:59   I kind of smiled

00:12:00   when I got this

00:12:01   when we got this feedback email

00:12:02   is like yes

00:12:03   the HomePod can do it

00:12:05   but do you trust

00:12:08   your HomePod

00:12:08   to do anything important

00:12:11   reliably

00:12:11   like I like as much

00:12:14   I'm glad I had it

00:12:15   and it did indeed

00:12:16   work for me once

00:12:16   but my HomePod

00:12:18   does not reliably do

00:12:21   literally any of the things

00:12:22   that it is marketed

00:12:23   as doing

00:12:23   like it doesn't do

00:12:24   a single one reliably

00:12:26   I would say

00:12:27   I have

00:12:27   maybe a 60%

00:12:29   success rate

00:12:30   with the things

00:12:30   I try to have happen

00:12:32   on my HomePod

00:12:32   so it kind of points out

00:12:35   like if Apple wants

00:12:36   these features

00:12:37   to be taken more seriously

00:12:38   there's so much work

00:12:40   to do with the reputation

00:12:41   and the reliability

00:12:42   of Siri

00:12:43   of these products

00:12:44   of their backing services

00:12:45   it has to be

00:12:47   way better than this

00:12:48   like right now

00:12:48   the idea

00:12:49   that I would rely

00:12:50   solely on this

00:12:51   it literally made me

00:12:53   laugh out loud

00:12:54   earlier today

00:12:54   when I saw this email

00:12:55   so it's nice to have it

00:12:57   as a backup

00:12:58   like oh that's

00:12:58   that's cute Apple

00:12:59   nice job

00:13:00   you know that's nice

00:13:00   but I'm not going

00:13:02   to rely on this

00:13:02   I would rather

00:13:03   also have smart smoke

00:13:04   alarms that notify me

00:13:05   through any

00:13:06   literally any other means

00:13:07   besides hoping a HomePod

00:13:09   will do one of its

00:13:10   many accessory jobs

00:13:12   I had the same thought

00:13:13   when I saw this email

00:13:14   I got the original HomePod

00:13:17   when it was first released

00:13:18   and had it for a long time

00:13:20   recently Marco sent me

00:13:22   two of his original HomePods

00:13:24   like the big ones

00:13:25   the original big HomePods

00:13:26   because he just didn't

00:13:27   want them anymore

00:13:28   and around that time

00:13:30   my original HomePod

00:13:31   started being a little bit wonky

00:13:32   like it wouldn't do OS updates

00:13:34   it would

00:13:35   if you saw the

00:13:36   in the Home app on the phone

00:13:37   it would show the little spinner

00:13:38   for an OS update

00:13:38   but then it would just disappear

00:13:39   and it kept saying

00:13:40   we're having trouble

00:13:41   signing into your Apple ID

00:13:42   bring your phone close

00:13:43   to the HomePod

00:13:44   and do blah blah blah

00:13:45   and I refactory reset it

00:13:47   and reset it up

00:13:48   a million times

00:13:49   did all sorts of stuff

00:13:50   and it was just

00:13:50   it just wasn't working

00:13:51   and eventually

00:13:53   it stopped showing up

00:13:54   as an airplay target

00:13:56   I was like

00:13:56   well whatever

00:13:57   I'll swap it down

00:13:58   for as the other two

00:13:59   were in the basement

00:13:59   I'll swap it down

00:14:00   for one of the

00:14:00   Marco's ones in the basement

00:14:02   so I put mine down

00:14:03   into the basement

00:14:04   it was in like

00:14:05   a stereo pair

00:14:07   but mine was like

00:14:07   just dead

00:14:07   so I just put it there

00:14:08   brought one up

00:14:09   from the basement

00:14:10   put it upstairs

00:14:10   it started doing

00:14:12   the same thing

00:14:12   swapped it again

00:14:13   for the third one

00:14:14   and now

00:14:15   the situation I'm in

00:14:16   is the third

00:14:17   and last one

00:14:17   that still

00:14:18   somewhat functions

00:14:20   and somewhat does

00:14:20   something like

00:14:21   you can talk to it

00:14:22   and turn the lights off

00:14:22   still tells me

00:14:24   that I was having

00:14:25   trouble shining

00:14:25   to my Apple ID

00:14:26   still won't do

00:14:27   an OS update

00:14:28   I think

00:14:30   sometimes it appears

00:14:31   as an airplay target

00:14:32   for audio

00:14:32   sometimes doesn't

00:14:33   and then my

00:14:34   original HomePod

00:14:35   now when you plug it in

00:14:37   doesn't do anything

00:14:39   doesn't turn on

00:14:40   nothing lights up

00:14:41   like you know

00:14:41   that's all you can do

00:14:42   with these HomePods

00:14:42   is plug them in

00:14:43   there's no on and off

00:14:44   switch

00:14:44   so my original HomePod

00:14:45   is dead dead

00:14:47   because there's that

00:14:47   whole issue that

00:14:48   you can find YouTube's

00:14:49   videos about

00:14:49   where there's some

00:14:50   component on the board

00:14:51   that burns out

00:14:52   because they have these

00:14:52   people who just repair

00:14:53   them anyway

00:14:53   so what I'm saying is

00:14:55   I have three first

00:14:56   generation HomePods

00:14:57   and they're all sort of

00:14:59   on their way out

00:14:59   I feel like that was

00:15:00   just a flawed hardware

00:15:01   product

00:15:02   and I considered

00:15:04   you know like maybe

00:15:04   I should just get a

00:15:05   HomePod mini

00:15:05   if I just want to use

00:15:06   it to turn the lights

00:15:07   on and off

00:15:07   and use it for my

00:15:09   testing against other

00:15:10   things or whatever

00:15:10   but reading this

00:15:12   thing of like using

00:15:12   it as a smoke alarm

00:15:13   like I have that

00:15:14   feature enabled

00:15:14   on all these things

00:15:16   just because why not

00:15:17   but absolutely would

00:15:18   not use it as my

00:15:19   primary means of

00:15:20   anything

00:15:21   I don't even use it

00:15:22   as my primary means

00:15:22   of turning the lights

00:15:23   on and off

00:15:23   because there's a

00:15:24   Google device in the

00:15:25   same room and most

00:15:25   of the family uses that

00:15:26   one because it works

00:15:27   every time

00:15:27   all right so Dion

00:15:30   Garrett writes with

00:15:30   regard to the limits

00:15:31   of LLMs and LRMs

00:15:33   the app paper

00:15:34   discussed in the

00:15:35   overtime segment of

00:15:36   ATP 643 mentioned

00:15:37   LRMs complete collapse

00:15:39   on out of distribution

00:15:41   problems this is not

00:15:43   the same as outside

00:15:44   of the training data

00:15:45   which is what we had

00:15:46   said a few times

00:15:47   neural nets generalize

00:15:49   really well on inputs

00:15:50   that are not in the

00:15:51   training data but are

00:15:52   drawn from the same

00:15:52   distribution as the

00:15:54   training data I think

00:15:55   you know this already

00:15:56   but there's a huge body

00:15:57   of people who believe

00:15:58   that LLMs cannot

00:15:59   reproduce anything that

00:15:59   wasn't literally in the

00:16:00   training data that's

00:16:02   certainly not true

00:16:03   yeah so that is a

00:16:04   fine distinction it's

00:16:05   worth making although

00:16:06   the fact that LLMs can

00:16:08   reproduce things that

00:16:09   were not in the

00:16:10   training data but just

00:16:11   are in the distribution

00:16:11   of the training data

00:16:12   that's where hallucinations

00:16:14   so-called hallucinations

00:16:15   come from like all of

00:16:16   it is just statistical

00:16:17   stuff based on the

00:16:18   training data so there's

00:16:19   the distribution of

00:16:20   the stuff in the

00:16:21   training data and

00:16:22   there's the things that

00:16:22   can be reproduced and

00:16:24   yeah it'll it'll put

00:16:26   so if it looks like

00:16:27   something would fall

00:16:27   naturally within the

00:16:28   distribution it will

00:16:29   produce that is that

00:16:30   valid or true or

00:16:32   or you know or

00:16:33   are the things in the

00:16:34   training data true or

00:16:35   are the things that like

00:16:36   it synthesizes because

00:16:37   they're within the realm

00:16:38   of the stuff that's in

00:16:39   the training data that

00:16:40   may not necessarily be

00:16:41   true either and that's

00:16:41   you know that's why you

00:16:42   will get LLMs to say

00:16:43   things that you said

00:16:44   this information this

00:16:45   fact this answer was

00:16:47   not in the training data

00:16:48   no it wasn't in the

00:16:49   training data but it

00:16:50   was apparently in the

00:16:50   quote-unquote

00:16:51   distribution of the

00:16:52   training data and

00:16:52   that's why it got

00:16:53   spat out of the

00:16:54   machine but anyway

00:16:55   important important

00:16:56   distinction this is from

00:16:57   a follow-up from many

00:16:58   episodes ago so we're

00:16:59   sorry for the delay but

00:17:00   WWD kind of sidetracked

00:17:02   it and speaking of

00:17:02   that right so we

00:17:04   talked about that paper

00:17:05   that Apple had put up

00:17:06   the illusion of

00:17:07   thinking understanding

00:17:08   the strengths and

00:17:08   limitations of reasoning

00:17:09   models via the lens of

00:17:10   problem complexity and

00:17:12   there was a rebuttal

00:17:13   about it that made a

00:17:15   bunch of waves and we

00:17:16   again we just didn't

00:17:17   have the chance to talk

00:17:18   about it the rebuttal

00:17:19   was called the illusion

00:17:20   of the illusion of

00:17:21   thinking and it was by

00:17:22   Alex Lawson and

00:17:23   Claude from

00:17:24   Anthropic reading from

00:17:26   there Apple's paper says

00:17:27   that large reasoning

00:17:27   models or LRMs

00:17:28   exhibit accuracy

00:17:29   collapse on planning

00:17:30   puzzles beyond certain

00:17:31   complexity thresholds we

00:17:33   demonstrate that their

00:17:34   findings primarily

00:17:34   reflect experimental

00:17:35   design limitations

00:17:36   rather than fundamental

00:17:37   reasoning failures

00:17:38   then there was a

00:17:39   whole bunch of talk

00:17:40   about that there was a

00:17:41   rebuttal to that

00:17:42   rebuttal by Victor

00:17:43   Martinez Victor writes

00:17:45   putting aside the stunt

00:17:45   of having Claude Opus as

00:17:47   a co-author yes I'm not

00:17:48   kidding the paper is in

00:17:49   itself is a poor

00:17:50   rebuttal for many

00:17:51   reasons we shall which

00:17:52   we shall explore but

00:17:53   mainly for missing the

00:17:54   entire point of the

00:17:55   paper in prior research

00:17:56   by AI researchers such

00:17:58   as professor some K

00:18:00   based name I'm not

00:18:00   going to try to

00:18:01   pronounce because I

00:18:01   will butcher

00:18:02   oh you gotta attempt

00:18:03   it you gotta attempt

00:18:03   it come combahampti

00:18:05   combahampti I'm gonna

00:18:08   say compound potty yeah

00:18:10   I think you're yeah I

00:18:11   think you're closer then

00:18:13   Gary Marcus chimed in

00:18:14   again seven replies to

00:18:16   the viral Apple

00:18:17   reasoning paper and why

00:18:18   they fall short and

00:18:19   we're gonna link all

00:18:20   these in the show notes

00:18:20   of course and then

00:18:22   finally it turns out

00:18:23   that Alex Lawson's

00:18:23   rebuttal was a quote

00:18:24   unquote joke so

00:18:26   reading from Alex

00:18:27   people were treating

00:18:28   my elaborate poopoo

00:18:29   post as real science

00:18:31   and while I'd aim to

00:18:33   point out some real

00:18:33   issues with the post I

00:18:34   hadn't really expected

00:18:35   anyone to take it

00:18:35   seriously look I should

00:18:37   have been more careful

00:18:37   the original version I

00:18:38   uploaded had some

00:18:39   genuinely terrible

00:18:40   sections stuff Claude

00:18:41   had written that I

00:18:41   hadn't bothered to

00:18:42   check properly I had

00:18:44   vibe coded if you will

00:18:45   the whole thing and

00:18:46   quite frequently asked

00:18:47   for an entire rewrite if

00:18:48   I wanted a small

00:18:48   section to be different

00:18:49   there was a

00:18:50   computational

00:18:51   complexity analysis

00:18:51   that was just

00:18:52   complete nonsense

00:18:53   if I'd been doing my

00:18:54   actual job or if I'd

00:18:55   been working on an

00:18:56   actual paper I'd have

00:18:58   been more careful

00:18:59   my thinking was that

00:19:00   the whole point was to

00:19:01   show that even Claude

00:19:02   could find problems with

00:19:03   the original paper I

00:19:03   wasn't trying to write

00:19:04   something good I've

00:19:05   updated the paper to

00:19:06   fix the biggest mistakes

00:19:07   from the first version

00:19:07   but the lesson's been

00:19:08   learned once something's

00:19:09   out there it's out of

00:19:10   your hands

00:19:11   yeah there's kind of

00:19:12   reaction of like boy my

00:19:13   thing kind of went

00:19:13   viral but you know

00:19:14   haha it wasn't serious

00:19:16   so okay I guess lesson

00:19:18   learned but anyway that

00:19:19   all that drama

00:19:20   essentially happened in

00:19:21   the time between when

00:19:21   this uh rebuttal of the

00:19:23   paper was put into our

00:19:25   show notes and then

00:19:25   WWC happened so we

00:19:27   were looped around to

00:19:28   find it and you know

00:19:29   there's Apple's paper

00:19:30   then there was this

00:19:31   thing rebutting it that

00:19:32   turned out to just be a

00:19:33   I love how we're using

00:19:34   vibe coding for things

00:19:35   that are not coding

00:19:35   a vibe coded quote

00:19:38   unquote joke response

00:19:39   that was kind of

00:19:40   serious because it was

00:19:41   like I do think there's

00:19:41   problems but I'll have

00:19:42   Claude help me write it

00:19:43   and won't that be fun

00:19:44   turns out not so fun

00:19:45   Lister Periato wrote

00:19:48   with regard to HDR on

00:19:49   the uh 26 OS's I've

00:19:51   noticed that Apple is

00:19:52   using way brighter HDR

00:19:54   effects on interface

00:19:54   elements in the betas

00:19:55   for iOS 26 but I

00:19:57   haven't read or heard

00:19:58   anyone discussing that

00:19:59   including Apple I think

00:20:01   that they started playing

00:20:02   with this with the new

00:20:03   Siri animations on

00:20:04   iOS 18 and now with

00:20:05   26 it's sort of

00:20:06   everywhere is this

00:20:06   available for developers

00:20:07   to play around with

00:20:08   well it scratches my

00:20:09   nerdy itch for novelty

00:20:10   more chroma more luma

00:20:11   more power I can see

00:20:13   it risks getting old

00:20:14   and gauche gauche is

00:20:15   that gauche thank you

00:20:17   at least I knew enough

00:20:18   to know it was wrong

00:20:18   uh anyway uh old and

00:20:21   gauche very fast so I

00:20:22   hope Apple's designers

00:20:23   thread carefully it

00:20:25   could be used the same

00:20:25   way that game developers

00:20:26   use lights to guide

00:20:27   players attention or it

00:20:28   could end up blinding

00:20:28   them out of their train

00:20:29   of thought I will

00:20:30   note that we did

00:20:31   mention this when we

00:20:31   were discussing the uh

00:20:32   x86 OS's or I

00:20:34   I keep writing it

00:20:35   xos 26 the OS 26 is

00:20:38   um yeah I think uh

00:20:40   Marco noted it and I

00:20:41   think maybe I did it

00:20:41   when we were talking

00:20:42   about in the past but

00:20:42   yeah I've been noticing

00:20:43   it more and more it's a

00:20:44   thing they're doing um

00:20:45   you know so they did it

00:20:47   with Siri with the if

00:20:49   you've noticed on your

00:20:49   phone when the new

00:20:50   Siri animation came

00:20:51   one of the reasons you

00:20:52   might have noticed it

00:20:53   is that it literally is

00:20:55   brighter if you have an

00:20:55   HDR screen on your phone

00:20:56   which most of us do

00:20:57   these days uh and

00:20:59   they're using it in all

00:21:00   sorts of places like uh

00:21:02   I think like in

00:21:02   tooltips and popovers

00:21:04   on macOS they'll they're

00:21:06   briefly in and when we

00:21:08   say HDR we just mean

00:21:09   brighter than the

00:21:10   brightest white uh that

00:21:12   your screen is set to

00:21:13   display so uh as we

00:21:15   discussed before if you

00:21:15   have a mac with HDR

00:21:16   there is the maximum

00:21:18   brightness that macOS

00:21:20   will let you set the

00:21:20   screen so if you just

00:21:21   make like a window like

00:21:22   a text edit window how

00:21:24   bright will the white in

00:21:25   that window get and it

00:21:26   used to be like three or

00:21:27   four hundred nits and

00:21:28   now I think they'll let

00:21:28   you go up to 700 or

00:21:29   whatever but max with

00:21:31   HDR screens often go

00:21:32   up to as high as

00:21:32   1600 nits but they

00:21:34   don't draw the text edit

00:21:35   window at 1600 nits

00:21:36   because a they would

00:21:37   draw a lot of power and

00:21:38   b would be blinding in

00:21:39   many environments um so

00:21:41   that's what we say by

00:21:42   HDR brightness something

00:21:43   that is brighter than

00:21:44   the cap on the

00:21:46   brightness of plain

00:21:47   white in just like part

00:21:49   of the user interface

00:21:50   something that's not an

00:21:51   HDR photo or whatever

00:21:52   and what they're doing

00:21:53   in Tahoe and a bunch of

00:21:54   places is making things

00:21:55   like popovers briefly

00:21:57   appear when they first

00:21:59   like pop over when they

00:22:00   first pop up briefly

00:22:01   appear at a much higher

00:22:02   brightness and then fade

00:22:04   back down to like the

00:22:05   full white level

00:22:05   brightness and it's an

00:22:07   interesting effect but

00:22:08   I'm not sure it fits in

00:22:10   because it's so it's not

00:22:12   it's not ubiquitous it's

00:22:13   just a few elements do

00:22:14   this and they do it

00:22:15   briefly so much so that

00:22:17   you might think am I

00:22:18   seeing something you're

00:22:19   not it's it's a real

00:22:20   thing that's happening

00:22:21   but you could go hours

00:22:23   without seeing this

00:22:24   effect and then have it

00:22:25   appear on the phone

00:22:26   obviously the Siri

00:22:27   animation does it every

00:22:28   single time so you kind of

00:22:29   get used to it there

00:22:30   but on the Mac you might

00:22:32   think you're seeing things

00:22:33   you're not it's not a ghost

00:22:34   inhabiting your computer

00:22:35   those things really are

00:22:36   briefly a little bit

00:22:37   brighter than you would

00:22:37   expect them to be

00:22:38   yeah I'm noticing this all

00:22:40   over the iPhone beta as

00:22:41   well even even mostly in

00:22:43   like edges of things like

00:22:45   when you know with with

00:22:46   part of the glass effects

00:22:48   here and there there's a

00:22:49   lot of a lot of controls

00:22:50   have like little glints

00:22:51   you know on the edges of

00:22:53   like oh there's this is

00:22:54   reflecting light and it's

00:22:56   hard to tell whether it's

00:22:58   passing that HDR threshold

00:23:00   or not but I'm pretty sure

00:23:01   it is and it's it's doing

00:23:03   it right now it is doing it

00:23:05   kind of you know occasionally

00:23:07   and subtly and and in small

00:23:09   areas of the of the bit

00:23:11   maps and stuff in small areas

00:23:12   of the of the screen but it's

00:23:14   definitely there like it sure

00:23:16   looks like it's there and I

00:23:18   think this is you know on one

00:23:20   hand it kind of it kind of

00:23:23   feels like cheating because

00:23:25   apps can't really do their

00:23:28   own UI very easily this way

00:23:30   like like in in a custom way

00:23:32   like for apps to use the the

00:23:35   HDR range would require it

00:23:37   requires a lot of hacks and

00:23:38   has a lot of limits and it's

00:23:39   it's not it's not easy for

00:23:41   apps to do so basically what

00:23:42   this will do is it will make

00:23:44   Apple's stock controls and

00:23:47   Apple's built-in system apps

00:23:48   that use this kind of

00:23:49   technique or system features

00:23:51   it'll make them look way

00:23:54   brighter and bolder and

00:23:55   better and easier to notice

00:23:57   than what most apps are

00:23:59   displaying most of the time

00:24:01   so this can be used to a

00:24:03   useful effect like I do think

00:24:05   the Siri animation being

00:24:07   brighter than than 100% you

00:24:09   know like that helps it jump

00:24:11   out and and show like okay

00:24:12   this is a different level of

00:24:13   the system that that is above

00:24:15   what everything else is going

00:24:17   on my screen I also think it

00:24:19   it's yet another way though

00:24:21   like as they're expanding it

00:24:22   now into a lot of these like

00:24:23   you know glass effects that

00:24:24   we can use but we can't

00:24:25   really necessarily you know

00:24:27   customize or make our own

00:24:28   that what that will do is

00:24:31   that will make apps that use

00:24:33   the stock controls stand out

00:24:36   even more than using custom

00:24:37   controls now as talked about

00:24:40   with the introduction of this

00:24:41   stuff the developer ecosystem

00:24:43   around you know custom designs

00:24:45   right now Apple is basically

00:24:47   like putting out putting a

00:24:49   shot out there that says

00:24:50   like look we're gonna make

00:24:51   some really cool stuff and if

00:24:53   you use our stock controls

00:24:54   mostly unmodified your stuff

00:24:57   will look like our stuff and

00:24:58   it'll look cool and it'll look

00:24:59   different and now it'll look

00:25:00   even you know even better with

00:25:02   these like little you know

00:25:02   little specular highlights on

00:25:03   the edges of glass and stuff

00:25:04   that are even brighter than what

00:25:05   you can do in your app very

00:25:06   easily but that's not what most

00:25:09   apps do most apps are made by

00:25:10   big companies who want all

00:25:11   their apps look the same and

00:25:12   they use something like

00:25:13   electron or you know some

00:25:14   kind of cross-platform toolkits

00:25:16   and stuff so that all their

00:25:17   apps you know look the same

00:25:18   everywhere and the same on the

00:25:19   web and everything is flat and

00:25:21   everything will look old

00:25:21   forever after this that's the

00:25:24   ecosystem they're playing into

00:25:25   so I think it's really it's it's

00:25:27   a bold move and it's it might

00:25:30   be a little bit of hubris but I

00:25:31   think it's kind of in the best

00:25:32   way that like Apple is gonna be

00:25:34   like you know what if you want

00:25:35   to do your own custom thing good

00:25:37   it's gonna look drab and flat

00:25:40   compared to what we are doing

00:25:41   it's an interesting you know

00:25:44   carrot to bring people over to

00:25:45   using stock controls I don't

00:25:48   think it will work for big

00:25:49   companies at all I don't think

00:25:51   the big companies care I don't

00:25:52   even know how much Apple cares

00:25:53   about them caring but I think

00:25:55   for for you know for the built-in

00:25:57   apps and for developers who use

00:26:00   the built-in you know UI widgets in

00:26:03   a less customized way I think it'll

00:26:05   be a nice little polish and it'll

00:26:06   it'll make our apps stand out

00:26:07   even more for the for those of

00:26:09   us who do you know stick to the

00:26:10   stick to the stock stuff I feel

00:26:13   like it's kind of like using

00:26:14   exclamation point important in a

00:26:15   CSS document in style sheets

00:26:18   that's a way for you to get a

00:26:20   property to take precedence over

00:26:22   one that would otherwise override

00:26:23   it due to like selector

00:26:24   specificity or something but you

00:26:26   only get one of those you get

00:26:28   one exclamation point important

00:26:29   there's no like exclamation point

00:26:31   important X2 X3 X4 all those I'm

00:26:34   sure there's CSS acts to get that

00:26:35   so the analogy may not be

00:26:37   perfect but anyway how many

00:26:39   chances do you get to say now

00:26:42   there's a new layer of the UI and

00:26:44   we're going to make that one

00:26:45   brighter probably only the one or

00:26:48   maybe two because there's only so

00:26:50   much of a brightness delta that

00:26:52   eyes can comfortably tolerate and

00:26:54   it's great that screens are getting

00:26:55   brighter but they're still kind of

00:26:57   like the regular brightness and

00:26:58   then the maximum brightness and

00:27:00   may like I said maybe you go with

00:27:01   the regular brightness a medium

00:27:02   one and then the highest one if we

00:27:04   get something and do like 4000 nits

00:27:05   on a phone or something but I'm

00:27:07   not sure that's the best way to add

00:27:11   emphasis to a user interface it is

00:27:14   possible to make a visual and

00:27:17   functional and conceptual hierarchy

00:27:19   without relying on large brightness

00:27:23   deltas as a differentiator I'm not

00:27:25   saying it's a bad idea to do it but I

00:27:26   am saying that Apple should you know

00:27:29   think twice about making this a

00:27:32   foundation of its visual hierarchy I

00:27:34   think it's fine for like we have a

00:27:35   new design and this is a flourish that

00:27:37   we can add to it but you really don't

00:27:39   like they like this is why I keep

00:27:40   going back to dollar sign important

00:27:41   you really can't only get to do this

00:27:42   once and you should maybe for the

00:27:46   next big redesign I guess you could

00:27:47   keep doing this but maybe it's better

00:27:49   to think of another way to show this

00:27:52   layering again we've done it in

00:27:54   computer interfaces for years and

00:27:56   years without HDR so it's you know

00:27:58   now that we have HDR this is another

00:27:59   tool that we can put in the tool belt

00:28:00   but I feel kind of like this is a an

00:28:02   exuberant explosion based on like well

00:28:06   now we have this new tool in our

00:28:07   quiver we've never had this before we

00:28:08   can use HDR to differentiate things

00:28:10   differentiate things in the

00:28:11   interface it's not great for battery

00:28:13   life it is sometimes a little bit

00:28:15   jarring I think it will be fun in the

00:28:17   short term but I hope this does not

00:28:19   last more than 10 years this is the

00:28:22   mindset I'm like probably only have to

00:28:24   worry about this for like 10 years

00:28:25   it'll be okay yeah if Apple still

00:28:26   around then which it probably will be

00:28:28   so I'm not a giant fan of this I think

00:28:30   it will be fun for a short period of

00:28:31   time but I think there are better ways

00:28:34   to do what it's trying to do with

00:28:36   regard to iPad OS 26 John Edwards

00:28:39   right my hypothesis on the 12 window

00:28:42   limit in iPad OS 26 is that it is a

00:28:44   safeguard against unresponsiveness on

00:28:46   the Mac if the window server stops

00:28:48   responding you can still move the mouse

00:28:49   cursor itself around even if it's

00:28:51   beach balling which is an indicator that

00:28:52   the system isn't completely hosed if

00:28:55   window drag is slow like John

00:28:56   encountered when more than one user

00:28:58   was logged in a few months back too

00:29:00   soon too soon the cursor stays locked

00:29:02   to the drag it just feels like pulling

00:29:04   the window through molasses on a

00:29:06   touch-first device there's no

00:29:07   guaranteed analogous affordance if

00:29:09   the windows on screen aren't

00:29:11   responsive to touch the system feels

00:29:12   broken as Marco pointed out even on

00:29:14   the latest and greatest iPhone

00:29:15   hardware liquid glass and the overall

00:29:16   performance of the current beta feel

00:29:17   slow beginning with the last few

00:29:19   releases of iOS 18 my iPhone mini

00:29:21   locks up in Apple's own apps and

00:29:22   becomes unresponsive to touch I can

00:29:24   tell you from firsthand experience

00:29:25   that this behavior sucks this is an

00:29:27   answer that Federighi and others

00:29:28   actually gave when asked the same

00:29:30   question I forgot the timing on this

00:29:31   one know whether John Edwards said

00:29:32   this before those interviews were

00:29:33   released or after or if he'd seen

00:29:35   them or not but either way we're like

00:29:37   they weren't talking about the 12

00:29:38   window limit which they didn't talk

00:29:39   about too much where they were

00:29:39   mostly talking about like why is it

00:29:41   taking so long to get windowing on

00:29:43   iOS and they gave the power as a

00:29:45   possible thing and then we talked

00:29:46   about the 12 window limit and said

00:29:47   well maybe that's why they're doing it

00:29:48   because as Federighi said in his

00:29:50   interviews like we want it to always be

00:29:51   responsive to touch because if it

00:29:53   isn't it feels broken and I'm here

00:29:56   to say the Mac also feels broken when

00:29:59   stuff stops working I know what he's

00:30:00   saying about like oh but like if the

00:30:01   drag is slow both the cursor that the

00:30:04   mouse is controlling and the place

00:30:05   where it is on the title bar will stay

00:30:07   together so it'll be slow together it

00:30:09   still feels awful first of all and

00:30:10   second of all there are plenty of

00:30:12   ways that the Mac OS can and has

00:30:14   locked up such that you can't drag

00:30:17   windows or the mouse doesn't move

00:30:19   anymore if you look back to the classic

00:30:21   Mac OS days and even if that's not

00:30:24   true even if like oh I'm as long as

00:30:26   the the cursor stays on the thing I'm

00:30:27   dragging I'm fine that it's updating

00:30:29   at three frames every 10 seconds no

00:30:31   you're not fine it feels totally

00:30:32   broken like there are just so many

00:30:33   ways or like the beach ball the beach

00:30:35   ball appears and now you can't click

00:30:37   anything you're like well these the

00:30:38   beach ball is still spinning I guess

00:30:39   the computer is all right no so this

00:30:41   notion that like and Federighi was

00:30:42   pointing out it's like well these touch

00:30:43   devices they always have to feel

00:30:45   responsive the Mac should also

00:30:46   always be responsive why is the bar

00:30:48   lower from the Mac like when terrible

00:30:50   things happen to the Mac user

00:30:51   interface it also feels terrible it's

00:30:53   not like we're fooled by like well

00:30:54   it's just a Mac it's fine if this

00:30:55   was an iPad I'd be annoyed and so I

00:30:58   you know there's I don't think this

00:30:59   is a you know I was discussing past

00:31:02   episodes about exactly how much power

00:31:04   modern iPads have compared to Macs

00:31:06   that ran Mac OS 10 with no window

00:31:07   limit on them I still feel like the

00:31:09   main reason for the limit has to be

00:31:11   either RAM limits and their refusal to

00:31:13   go to swap without a again I don't

00:31:15   know the details of what the

00:31:16   situation is like or just the plain

00:31:18   fact is it's a small screen and

00:31:20   they're trying to keep it simple I do

00:31:21   wonder if this limit will increase as

00:31:24   as time marches on but you know 10

00:31:26   years from now when iPads come with

00:31:28   64 gigs of RAM standard to be like

00:31:29   well they can't handle more than 12

00:31:30   windows and to be like really Mac

00:31:32   was 10 ran 128 megabytes but this 32

00:31:34   gig RAM iPad Pro can't have more than

00:31:36   12 windows on screen at once so that's

00:31:38   iPad OS for you it's moved by leaps

00:31:42   and bounds this year but it is still

00:31:44   holding itself back from being the

00:31:46   unlimited bonanza that is Mac OS for

00:31:49   some good reasons because again that's

00:31:51   the selling proposition it is simpler

00:31:52   it has it gives you less ability to get

00:31:55   yourself into trouble and making lots

00:31:56   of one windows is one way you can do

00:31:58   that but 12 feels weird and somewhat

00:32:01   arbitrary maybe in iPad OS 27 we'll get

00:32:04   a baker's dozen Wow all right then

00:32:08   John you had asked for not really asked

00:32:10   for but had talked about the idea of a

00:32:12   glass iPhone can you recap what that

00:32:14   was about please yeah listener actually

00:32:15   sent this in we were talking about

00:32:16   like the rumors of liquid glass and the

00:32:18   rumors of like using the the light

00:32:20   sensors or the camera to figure out

00:32:21   where the glint should be on the UI this

00:32:23   is back when we just had the rumors of

00:32:24   what it's supposed to look like we

00:32:25   didn't know yet and the listener wrote

00:32:26   it and said well you know I'm also

00:32:29   talking about it making an OS that

00:32:30   would fit for the rumored all glass

00:32:32   iPhone and the rumors the patent they

00:32:33   had for making a thing literally

00:32:34   entirely out of glass including the

00:32:36   case the front the back everything

00:32:38   just all glass and Alyssa wrote and

00:32:40   say well you could have a the

00:32:42   equivalent of that today if you just

00:32:44   turned on the back camera and had the

00:32:46   back camera show what was behind the

00:32:48   phone on the phone screen so when you

00:32:50   were looking at your phone any area

00:32:52   that wasn't content you'd see quote

00:32:54   unquote through the phone to

00:32:55   whatever's behind it and guess what

00:32:57   somebody did it artisanal asparagus

00:33:00   water on Mastodon has a cool demo movie

00:33:04   they just mocked up an app doing this

00:33:06   exact thing which is turn on the back

00:33:09   camera and show the image but then put

00:33:10   some UI on top of it it's really hard

00:33:12   to do this you know this makes you

00:33:14   appreciate Vision Pro but also

00:33:15   understand why Vision Pro has to

00:33:16   strap itself to your face you can't

00:33:19   actually do this press in a

00:33:21   perspective correct way without

00:33:22   knowing where your eyes are all the

00:33:24   time so the illusion is broken if you

00:33:26   were not lined up it's kind of like

00:33:28   the what are called the volume that

00:33:30   big LED screen that wraps around

00:33:33   actors that they use at IOM to do

00:33:35   like the Mandalorian and a bunch of

00:33:36   other shows that only looks good from

00:33:38   the perspective of the camera if

00:33:40   you're standing not where the camera

00:33:41   is the background looks all wonky

00:33:43   same deal with this like it does it's

00:33:45   not perspective correct but anyway if

00:33:46   you want to see what that looks like

00:33:47   there's a video we will link it in the

00:33:48   show notes kudos to artisanal

00:33:50   asparagus water for mocking this up

00:33:52   also like isn't it weird if you're

00:33:55   simulating the idea of the phone being

00:33:57   transparent isn't it weird that you

00:33:58   don't see your hand yeah I mean well

00:34:01   they don't have a camera they can

00:34:02   show that but they could they could

00:34:03   do a 3d model of the hand like they

00:34:04   do like the what are you the

00:34:05   personas there you go see you can't

00:34:08   see your face but we you just put

00:34:09   your hand in front of it'll model your

00:34:10   hand and then yeah it'll show your

00:34:12   fingers yeah we scanned your hand and

00:34:13   now here is a hand that looks almost

00:34:14   like your hands nobody will notice

00:34:16   probably asterisk and then finally for

00:34:19   tonight Joseph Humphrey writes the

00:34:21   name of the app you mentioned in an

00:34:23   earlier episode that can translate

00:34:24   natural language terminal commands is

00:34:26   sub stage and it's made by me who

00:34:29   Joseph self self proclaims himself as

00:34:31   friend of the show Joseph Humphrey to

00:34:34   be honest that's not really how that

00:34:35   works but I appreciate the hustle to be

00:34:37   honest I get where you're coming from

00:34:38   when you express doubts about using it

00:34:40   as I was saying to my wife the other

00:34:41   day if someone else other than myself

00:34:42   had made the app I genuinely don't

00:34:44   know if I'd trust it and want to

00:34:45   install it either the thing that makes

00:34:47   the difference for me is that one I

00:34:48   made it so I know exactly what it's

00:34:49   doing and two I added risk assessment

00:34:51   as a feature so it assesses commands

00:34:53   instead of blindly running them and

00:34:55   three you can inspect and edit the

00:34:56   commands before running them too I'm

00:34:58   also always finding new use cases

00:35:00   daily that let me do things that I'm

00:35:01   not sure I'd even think to do with

00:35:03   chat GPT just because sub stage is

00:35:05   right there for example my most

00:35:07   recently quote trim the sides of this

00:35:09   video to make it 16 by 10 instead of

00:35:12   16 by 9 quote something like that

00:35:14   would be a huge faff I guess this is a

00:35:16   Brit in after effects and I'm not sure

00:35:18   how how long I'd have to poke around

00:35:20   and handbrake to figure out whether

00:35:22   it's even possible or not there I did

00:35:24   actually I have like I said I have

00:35:25   this app and saw I couldn't remember

00:35:26   the name of it at the time so I wanted

00:35:28   to put this in here so to tell people

00:35:29   what it was called and I have actually

00:35:31   used it recently I don't use it much

00:35:32   because like I said on the show I'm a

00:35:33   little nervous about even assessing the

00:35:36   commands myself by looking at them

00:35:37   like it's just I don't know it's but

00:35:39   anyway I had occasion I forget what I

00:35:41   was using it for but it was something

00:35:42   very straightforward that I trusted it

00:35:44   to do and I you know it was it was

00:35:47   faster than going to my web browser

00:35:49   window that or it was already open to

00:35:51   Gemini or chat GPT or something and

00:35:53   asking it the ffmpeg command and then

00:35:54   copying and pasting it sub stage does

00:35:57   what you wanted to do which is let me

00:36:00   just do all that for you right here

00:36:01   you know in the finder right click the

00:36:04   thing or you know invoke the command or

00:36:05   whatever and then just tell me what

00:36:08   you want I'll do it right here so it's

00:36:09   it's cutting out steps and in the same

00:36:11   way that a lot of these LMs save you

00:36:14   time by doing something that you would

00:36:16   have to do a bunch of web searches for

00:36:18   first they'll just do it right in the

00:36:19   window you can get even faster by

00:36:21   having it integrated directly and the

00:36:22   chat GPT app for mac os does very

00:36:25   similar things but it unfortunately

00:36:27   does not run an Intel Mac so I'm not

00:36:29   running it on my main machine so I don't

00:36:31   get a chance to use it we are sponsored

00:36:34   this episode by Squarespace the all-in-one

00:36:37   website platform for entrepreneurs to

00:36:39   stand out and succeed online whether

00:36:41   you're just starting out or managing a

00:36:42   growing brand Squarespace makes it easy

00:36:44   to create a beautiful website engage

00:36:45   with your audience and sell anything

00:36:47   from your products to your content to

00:36:49   your time all in one place and all on

00:36:51   your terms there are so many great

00:36:53   features for Squarespace for

00:36:55   businesses these days I've highlighted

00:36:56   before the design intelligence the

00:36:58   Squarespace payment system that lets you

00:37:00   take payments from all sorts of

00:37:01   sources and new methods as they come up

00:37:03   also they have a great invoicing

00:37:05   product this is an easier way to

00:37:06   collect payments so you can focus on

00:37:08   growing your business invoice clients

00:37:10   and get paid for your services turn

00:37:11   leads into clients with proposals

00:37:13   estimates and contracts and simplify

00:37:15   your workflow and manage your service

00:37:16   business all on one platform they also

00:37:18   have great analytics you can measure

00:37:20   your end-to-end online performance

00:37:22   with powerful website and seller

00:37:24   analytics get insights on top traffic

00:37:26   sources understand how your reach is

00:37:28   growing track sales metrics and learn

00:37:30   where to focus new engagement it's all

00:37:32   the data you need to scale your brand

00:37:34   or business fully integrated and clearly

00:37:36   displayed and they even have domains

00:37:39   now every dream needs a domain

00:37:40   Squarespace domains makes it easy to

00:37:41   invest in your dream confidently with

00:37:43   premium privacy and security features

00:37:45   included in every purchase so

00:37:47   Squarespace is just a great all-in-one

00:37:49   platform anybody can use it there's no

00:37:51   coding required there's no expertise

00:37:53   no technical knowledge required so if

00:37:55   you know somebody comes to you in your

00:37:56   life and like hey can you help me make

00:37:58   a website for me or my business you can

00:38:00   say I you don't need my help go to

00:38:02   Squarespace and it empowers them to do

00:38:04   it themselves which is way better than

00:38:06   the help you could possibly provide as a

00:38:07   nerd so have them or have you go to

00:38:10   squarespace.com and build the site in

00:38:12   trial mode you get a free trial when

00:38:14   you're ready to launch go to

00:38:15   squarespace.com slash ATP and I bet you

00:38:17   will because I bet after the trial

00:38:19   you're going to see how great it is

00:38:20   you'll save 10% off your first

00:38:22   purchase of a website or domain with

00:38:23   that link so once again squarespace.com

00:38:25   slash ATP for 10% off your first

00:38:27   purchase thank you so much to

00:38:29   Squarespace for sponsoring our show

00:38:31   all right there was a rumor that broke a

00:38:37   few days ago Apple is planning this

00:38:39   reading from Mac rumors Apple is planning

00:38:41   to launch a low-cost MacBook powered by

00:38:43   an iPhone chip according to Apple

00:38:44   analyst Ming-Chi Kuo in an article

00:38:47   published on X quote complain explain not

00:38:49   complain explain that the device will

00:38:51   feature a 13-inch display and the a18

00:38:54   pro chip making it the first Mac powered

00:38:56   by an iPhone chip the a18 pro chip

00:38:58   debuted in the iPhone 16 pro last year

00:39:00   the more affordable MacBook could come

00:39:02   in silver blue pink and yellow finishes

00:39:04   the new MacBook is expected to hear

00:39:07   Eric to enter mass production late in

00:39:09   the fourth quarter of 2025 or early in

00:39:11   the first quarter of 2026 Apple's

00:39:13   hoping that it will drive MacBook

00:39:14   purchases and account for 20 to 28% of

00:39:16   sales next year additionally a different

00:39:19   post on Mac rumors Mac rumors has

00:39:20   spotted evidence of such a device in

00:39:22   back-end code related to Apple

00:39:24   intelligence last summer and

00:39:25   subsequently confirmed its use of the

00:39:26   a18 pro chip the machine features the

00:39:29   identifier Mac 17 comma 1 so I will talk

00:39:33   about this rumor for in more detail in a

00:39:35   second but the headline here I believe

00:39:38   this is straight from Mac rumors also from

00:39:39   the story yes straight from Ming-Chi Kuo

00:39:41   low-cost is the adjective modifier on

00:39:45   this MacBook Apple is planning to launch a

00:39:47   low-cost MacBook and the only thing we

00:39:50   know about this other than the size of

00:39:51   the screen which is the same as the

00:39:52   current 13 inch MacBook Air and some

00:39:55   colors is that it comes with a phone

00:39:57   chip how much cheaper do you think the

00:40:01   a18 pro is than the m4 hmm not that

00:40:07   much I can't imagine I mean I think it's

00:40:09   cheap I agree that it's probably

00:40:10   cheaper not the least of which is

00:40:11   because they make a bazillion of them

00:40:12   because it's a phone chip and it is

00:40:14   physically smaller so just by going by

00:40:16   the math of like die area like the

00:40:18   bigger you make the chip the more

00:40:19   expensive it is it makes sense to me

00:40:21   that it would be cheaper but I don't

00:40:24   especially on low-end devices I don't

00:40:28   think the SoC is usually the most

00:40:30   expensive component in the machine and

00:40:34   the prices involved here especially since

00:40:36   Apple isn't paying somebody else's

00:40:37   profit margins other than TSMC's on these

00:40:40   chips like I'm thinking like maybe like

00:40:45   50 bucks 30 bucks less expensive like do

00:40:50   you think that's the ballpark yeah I'm

00:40:52   guessing it is I'm guessing the price

00:40:54   difference is less than a hundred

00:40:55   dollars yeah and so this rumor is that

00:40:58   it's a low-cost MacBook now you you know

00:41:01   we'll get to what Marco has here in a

00:41:03   second you save money lots of different

00:41:04   places but I haven't seen anyone really

00:41:06   mention this is like wow a low-cost you

00:41:09   know a cheaper computer would be a great

00:41:11   idea and you could use a phone chip in

00:41:13   it and that would make it cheaper and

00:41:15   this is great I'm like okay but if that

00:41:16   was the only difference if it's the M4

00:41:18   MacBook Air with the M4 swapped out for

00:41:21   the A18 Pro I'm thinking best case

00:41:25   scenario this is a hundred dollar cheaper

00:41:27   computer which is not nothing because

00:41:28   the thing is a thousand dollars so if you

00:41:29   have one for nine hundred dollars like

00:41:30   that's good but is that enough to

00:41:34   justify this thing's existence or does

00:41:35   this replace the M4 MacBook Air well I

00:41:37   think the the even if it is only 50 or

00:41:40   hundred dollars different cost to Apple

00:41:43   when you multiply that a million times

00:41:46   two million times five million times

00:41:47   that adds up real quick and so I don't

00:41:49   think you're wrong but it wouldn't

00:41:50   surprise me if they wanted to go after

00:41:52   what seems to be a not very large Delta

00:41:54   just because on the scale in which they

00:41:57   operate that can make a huge difference

00:41:58   yeah so that leads to to Marco's idea

00:42:01   for a game here what could be removed or

00:42:04   limited to make this cheaper than a

00:42:06   MacBook Air other than obviously the

00:42:08   A18 Pro versus the M4 so let's just

00:42:11   that just take that as a given because

00:42:12   all the rumors seem to agree that

00:42:13   that's that's the deal with this

00:42:15   including Mac rumors finding a thing

00:42:17   that seemed to identify that chip

00:42:19   number so we've swapped out the M4 for

00:42:21   the A18 Pro how else do you get money

00:42:23   out of this machine let's let's do a

00:42:26   round robin and everyone and you got to

00:42:28   pick you know like a giraffe like until

00:42:30   we don't we run out of things that we

00:42:31   can think of yeah because I have a list

00:42:33   here of like remote the keyboard some

00:42:34   of them are pretty crazy and and you

00:42:36   know and just for context all right the

00:42:38   MacBook Air currently starts at a

00:42:39   thousand dollars isn't like a 900 for

00:42:42   education or something yes something like

00:42:43   that and then there's the Walmart M1

00:42:46   MacBook Air still for sale for 650 yeah

00:42:49   also for reference the cheapest iPad

00:42:52   they sell is 350 and it has a 250

00:42:57   keyboard option so a combined if you want

00:42:59   the cheapest iPad with their keyboard and

00:43:02   trackpad that go with it it's six

00:43:04   hundred dollars and what what chip does

00:43:06   that one come with that's the a 17 I

00:43:08   believe so that so the base model iPad

00:43:11   currently for sale does not run Apple

00:43:13   intelligence the iPad mini has the 17

00:43:16   pro so it can but that's more expensive

00:43:19   but so just the cheapest iPad they make

00:43:21   and I think you can look at some of the

00:43:24   things they do with that cheap iPad to

00:43:26   get that iPad down to 350 without the

00:43:27   keyboard and I think we're looking at a

00:43:30   similar likely product here now obviously a

00:43:35   laptop requires a keyboard and a trackpad

00:43:37   so I'm guessing I'm guessing this laptop

00:43:41   includes you know it the rumor already

00:43:44   says a 13 inch screen so a 13 inch screen

00:43:46   keyboard and trackpad now the details on

00:43:49   those can matter so who'd like to go

00:43:50   first I'll go first and this is the thing

00:43:53   that I hope they don't do I seriously

00:43:55   hope they don't do but you can't rule it

00:43:58   out remove RAM

00:43:59   at the Apple 8 gigs Apple just bumped all

00:44:03   of its max 16 gigs and we celebrated how

00:44:06   can Apple screw that up you know how oh our

00:44:08   new low-cost model and of course because

00:44:09   it's the lowest cost model only comes to

00:44:11   the 8 gigs I really hope they don't do

00:44:12   that but that would save them at least

00:44:14   five dollars yeah and look when you're

00:44:17   trying to hit a low probably because you

00:44:19   know like you know you look at the maybe

00:44:21   they're trying to replace the Walmart

00:44:22   MacBook Air like and to with like a newer

00:44:25   product that can actually run Apple

00:44:26   intelligence and that they can you know

00:44:27   keep making new models of over time but

00:44:31   still so like if they're trying to come in

00:44:33   around you know 650 600 700 range they're

00:44:38   gonna have to cut a decent amount of stuff

00:44:40   because here's the one thing they will

00:44:42   never cut is margin you know that Apple

00:44:44   will never cut margin so they're gonna

00:44:47   make the same margins they're making on

00:44:49   that $1,000 MacBook Air so how how do

00:44:53   they get there how do they shave off like

00:44:55   a third of the price well one of you go

00:44:57   so I did Ram what's the next one Casey

00:44:59   what about the the fidelity and of the

00:45:03   display and I don't just mean going

00:45:05   non-retina worse well no I'm serious

00:45:08   like not even just non-retina but what

00:45:10   if the backlight gets worse and weaker

00:45:12   what it only does RNG not be hey that's

00:45:15   it you got it save savings of one-third

00:45:18   no like a like a crummier backlight or

00:45:20   again maybe non-retina I would be stunned

00:45:23   if they went non-retina but it is a

00:45:24   possibility you have to pick you have to

00:45:26   pick one I wouldn't go with non-retina

00:45:28   but a cheaper screen yeah like and I'm

00:45:30   specifically thinking about like you know

00:45:31   one of the things they tout in the

00:45:33   MacBook Pros is oh it's 11 billion nits

00:45:36   well what if this is three nits right to

00:45:39   speak you know well the current MacBook

00:45:41   Air screen and the M1 MacBook Air

00:45:43   screen are also already worse this could be

00:45:46   you know one thing to consider is that

00:45:47   that cheap iPad lacks P3 wide color and an

00:45:53   AR coating anti-reflective coating yeah when

00:45:55   you get down to these low-end screens it's

00:45:57   difficult to like but yeah cheaper screen

00:45:59   is yeah is reasonable but I think I think

00:46:01   the way they would do that would not be

00:46:02   resolution because resolution is not that

00:46:04   expensive these days like I think it would

00:46:06   be no display P3 but should probably save

00:46:08   them on some you know calibration also and

00:46:11   you know some certain like you know chip

00:46:12   support good call so no P3 and then you know

00:46:15   different coatings or enclosure so maybe it

00:46:17   doesn't have the AR coating because again

00:46:18   like that's what they did to make that

00:46:20   cheap iPad screen well the thing about

00:46:22   screens and a cheaper screen is that given

00:46:25   the age of like let's just say again the

00:46:27   M1 MacBook Air given the age of that

00:46:29   machine at a certain point or even just

00:46:32   the M4 one like the cost of the the screen

00:46:35   that's in the M4 MacBook Air is going down

00:46:38   all the time because it is so old and just

00:46:40   like so far away from the cutting edge it's

00:46:43   kind of like I don't know this shows a name

00:46:44   for this parts curve but like we've joked

00:46:46   about it many times in the past but it's a

00:46:47   real thing something that's like older

00:46:49   technology gets cheaper and cheaper as

00:46:51   it's supplanted by newer technologies but

00:46:53   then all of a sudden it starts getting more

00:46:54   expensive again because it's so old that

00:46:55   nobody wants it and people stop making it

00:46:57   and I do wonder if Apple walks that line

00:46:59   with their screens but like yeah they get

00:47:01   cheap because the the the machine they sell

00:47:04   doesn't get cheaper and cheaper over time

00:47:06   but the parts that go into it do and so I do

00:47:08   wonder if when the a18 pro MacBook whatever

00:47:12   comes out it might have a better screen

00:47:14   than the M1 MacBook Air simply because you

00:47:16   can't buy screens that are as bad as the

00:47:18   MacBook Air anymore and it will still be

00:47:20   cheaper just like Casey saying it will

00:47:22   actually be a cheaper screen than the M1

00:47:24   MacBook Air screen was when that machine

00:47:26   was introduced but now a cheaper screen

00:47:28   actually ends up better so that I could

00:47:30   see them saving costs there Marco what's

00:47:32   your what's your pick all right my my I have

00:47:35   a lot of others but my first pick is you

00:47:37   got to think like in addition to using

00:47:40   the low-end iPad as inspiration for some

00:47:42   of these changes you also got to think

00:47:44   how do they keep people who just want the

00:47:47   cheapest MacBook who from buying this

00:47:50   instead of like they want to preserve

00:47:52   MacBook Air sales to no SSD so how do

00:47:56   they most importantly how do they keep

00:47:59   businesses from buying this one they want

00:48:02   businesses to keep buying the Airs and

00:48:03   the pros so my number one pick is no

00:48:07   external monitor support and also think

00:48:11   about it the a18 Pro probably doesn't

00:48:14   support external monitors it might but

00:48:16   yeah that's that's a good call that you

00:48:17   know I'd just be part of the if it's just

00:48:19   part of the SoC then they kind of get that

00:48:21   for free and it's not really a cost

00:48:22   savings but yeah I don't I'm not not

00:48:24   sure that would save them that much

00:48:26   money I don't know if it would save the

00:48:27   money but I could see them doing it on in

00:48:29   software because the 18 Pro absolutely

00:48:31   does support an external monitor because

00:48:33   you can do that with your phone right I

00:48:34   mean I'm sure I've done this yeah I think

00:48:36   so oh I'm pretty darn sure you can and

00:48:39   so I don't think that that would be a

00:48:40   savings but I do agree with you they

00:48:42   would absolutely I could see them doing

00:48:44   this as a feature gate just to get people

00:48:46   to upgrade like you had said yeah because

00:48:48   they they want no businesses to buy this

00:48:50   computer they want businesses to keep

00:48:51   buying the Airs and the pros all right

00:48:54   John I mean I don't this is kind of in

00:48:56   the same spirit as what you were saying

00:48:58   with your thing but use an existing like

00:49:01   use the M1 case design like I'm assuming

00:49:04   the M1 case design is cheaper than the

00:49:05   M4 one but like use a case design they've

00:49:07   been doing for so long that it's just

00:49:09   incredibly cheap for them to do that they

00:49:10   have all the machines to do it you know

00:49:12   so no new case design and it would also

00:49:14   differentiate it from the M5 MacBook Air

00:49:17   because it would literally still be the

00:49:18   wedge shape like that older thing so yeah

00:49:20   use use the M1 case design and you think

00:49:23   they would do that instead of making a

00:49:24   new cheap case oh yeah because like

00:49:26   mate there's no such there's no such

00:49:28   thing as that when it comes to cheap like

00:49:29   a new cheaper to make case yeah maybe

00:49:31   cheaper to make but the cost of of

00:49:34   developing that new thing it would take

00:49:36   a while to recoup that I've the Tim

00:49:38   Cook way is take some existing thing

00:49:40   that we've already been making for I

00:49:41   mean that's part of the reason the

00:49:42   Walmart one still uses the wedge it's

00:49:44   like look we already we already know how

00:49:45   to make that case we've mastered it we

00:49:47   have the machines for it and

00:49:49   presumably it is cheaper to make them

00:49:50   the newest and latest and greatest

00:49:52   case so yeah I think they would just

00:49:53   keep using that one well what if like

00:49:55   you know there there are cases of all

00:49:57   in in you know the last for a decade or

00:49:59   more they've all been unibody case

00:50:00   designs where you basically have like

00:50:02   you know big blocks of aluminum and

00:50:04   you have these machines that like carve

00:50:06   out sections of it a cheaper way to

00:50:09   make metal cases is to have them be in

00:50:11   different like flat sheets that are

00:50:13   like screwed together or something what

00:50:15   if like if they're actually designing a

00:50:16   case from scratch to be inexpensive I

00:50:19   would think that maybe that might

00:50:21   have some significant savings of going

00:50:22   like a multi-part case I think I know

00:50:25   what you're talking about because they

00:50:25   used to make that before the original

00:50:27   MacBook Air came out yeah I think

00:50:28   that's actually more expensive to

00:50:30   manufacture due to the fastening issue

00:50:33   like I really do feel like the M1

00:50:34   MacBook Air case given how slim it is

00:50:37   and how long they've been making it may

00:50:39   actually be the cheaper option cheaper

00:50:41   in terms of repairs even cheaper in

00:50:42   terms of manufacturability for all the

00:50:44   parts don't have to be aligned and

00:50:45   connected and precision made it's just

00:50:47   one top and one but you know what I

00:50:49   mean like you know there's there's

00:50:50   there are pieces but there's so there's

00:50:52   far fewer pieces and far fewer

00:50:53   fasteners so I I do wonder if that

00:50:55   would actually be a cost savings and I

00:50:57   think honestly I think they would

00:50:58   never do that they've been they've had

00:50:59   unibody again since the original

00:51:01   MacBook Air I don't think they're ever

00:51:02   going back until there's a new

00:51:03   materials breakthrough and they start

00:51:04   making them out of like transparent

00:51:06   aluminum or graphite or something

00:51:08   all right Casey what's next back on the

00:51:11   chopping block here we go

00:51:12   uh may this might be silly but it's

00:51:15   one of the first things that jumps to

00:51:17   mind is what about crappier radios

00:51:19   crappier bluetooth radios older

00:51:21   bluetooth radios older wi-fi radios

00:51:23   they already do that for all their

00:51:23   computers yeah yeah you know what I

00:51:26   mean but you know where where the

00:51:28   MacBook Pro the newest MacBook Pro

00:51:30   maybe has wi-fi seven I don't even

00:51:31   know if that's true anymore I can't

00:51:32   keep up but it doesn't yeah I think

00:51:34   we're still on 6e although funny you

00:51:37   mentioned this Casey the base model

00:51:38   iPad does not have wi-fi 6e it has

00:51:41   wi-fi 6 there you go and so you know

00:51:44   could you go even older would that be

00:51:46   an advantage I don't know but you know

00:51:48   some sort of crippled bluetooth radio

00:51:51   wi-fi radio something along those lines

00:51:54   where it's just not a modern chip it's

00:51:57   sufficient but not stellar all right my

00:52:01   next pick the ports one USB-C port and a

00:52:06   headphone jack and that's it Casey will

00:52:09   love it oh stop it so no thunderbolt

00:52:12   support no mag safe no HDMI because

00:52:15   again we have no external monitor

00:52:16   support in my theory here so one USB-C

00:52:19   port and a headphone jack just like the

00:52:21   old MacBook yeah actually the jumping

00:52:24   slightly ahead here Stephen Hackett

00:52:25   wrote about this and he was musing about

00:52:28   the possibility of an 18 pro and a Mac and

00:52:30   he says the immediate downside to the

00:52:32   18 pro is that only supports USB 3 at

00:52:34   10 gigabits per second yep not

00:52:35   thunderbolt that would make any Mac

00:52:37   with an 18 at its heart only capable of

00:52:38   USB-C and that's fine that's just that's

00:52:41   what they used to do with their long

00:52:42   computers all the time anyway I think

00:52:43   then there yeah I think I mean yeah

00:52:46   having one port is crappy but having

00:52:48   no thunderbolt I think is a given and so

00:52:50   the 18 is not a limiting factor there

00:52:52   mine I expected I thought you're gonna

00:52:53   steal this from me is a pretty easy

00:52:56   one smaller battery mmm smaller

00:52:58   batteries cost less money yeah I was

00:52:59   about to say the same thing you're

00:53:00   stealing all mine y'all think of the

00:53:01   iPad the iPad has a smaller battery

00:53:04   and the phone certainly has a smaller

00:53:06   battery like can't how small a battery

00:53:09   can you get away with and still have

00:53:10   accessible battery life because for

00:53:12   every amount that you make that

00:53:13   battery smaller first of all makes the

00:53:14   laptop lighter but second of all makes

00:53:15   it cheaper because smaller batteries

00:53:17   cost less money yeah that's pretty

00:53:18   good all right Casey I'm out of any of

00:53:22   the ones that I brainstormed when we

00:53:23   started the game but I think did we

00:53:25   already talk about a teeny tiny SSD did

00:53:27   somebody no we did not we did I said

00:53:28   remove the SSD I think a teeny tiny

00:53:32   SSD to find that as you will I think

00:53:33   you know 128 gigs or something like

00:53:35   that I don't even know what the we're

00:53:36   just making the world's most terrible

00:53:38   computer here because I mean this is

00:53:39   what happens if your job is to remove

00:53:41   costs you can remove costs when it's

00:53:43   like we're rapidly getting to the

00:53:44   point now where we would not

00:53:45   recommend this computer I think we

00:53:47   may have passed the point where we

00:53:48   would not recommend it oh and

00:53:49   actually come to think of it a

00:53:50   corollary to what I just said is

00:53:52   whether or not the SSD is is smaller

00:53:55   in terms of capacity what about

00:53:57   making it considerably slower you know

00:54:00   where SSDs are still lightning fast

00:54:02   even crappy SSDs by and large are

00:54:03   lightning fast it might cost them more

00:54:05   money to get a slower SSDs it's fair but

00:54:09   I mean again I'm just brainstorming

00:54:10   what if what if it's either smaller or

00:54:12   slower and that that low-end iPad does

00:54:15   start with a 128 gig model and so yeah

00:54:18   like and I did I looked for for instance

00:54:21   though like as a point of comparison

00:54:23   like what are the cheapest PC laptops

00:54:25   like in this price range like you know

00:54:27   in like the 600 ish dollar price range

00:54:29   what do you get on a PC laptop these

00:54:30   days and you get 16 and 512 like

00:54:33   that's that's a pretty pretty

00:54:35   universal across the board yeah I know

00:54:37   yes Apple lives in a different

00:54:38   universe that's the other thing about

00:54:39   this cheap laptop I know we're like

00:54:41   talking about in the context of the

00:54:42   Walmart one which I think has no

00:54:43   configuration options but like can you

00:54:45   imagine them making a low-cost

00:54:47   MacBook like it's like you said with

00:54:48   the minis the lower cost of Apple makes

00:54:50   a Mac you'll double the price when you

00:54:53   try to add RAM to it if that's even an

00:54:54   option like if there are any options to

00:54:56   configure the RAM or the SSD and this

00:54:58   thing is cheaper than any existing Mac

00:55:01   you will literally double you almost

00:55:02   double it on the Mac mini you said it's

00:55:04   like a $600 computer add RAM it adds

00:55:06   $400 to a $600 thing so if there was a

00:55:10   $400 Mac mini adding any amount of

00:55:12   additional RAM that Apple offers you

00:55:13   would double the price of this so it's

00:55:15   they're they're getting real close

00:55:16   there with their obsession with you

00:55:18   know ridiculous margins I mean so one

00:55:21   idea I had is what if they don't offer

00:55:24   any options except color I was you to

00:55:27   keep stealing all my ideas that's the

00:55:30   nature of the nature of a draft yeah

00:55:31   that's that was what I was gonna say

00:55:33   no options but like the colors are in

00:55:34   the rumor but you're saying no options

00:55:36   except for color yeah like no no

00:55:38   storage no RAM option just like here

00:55:40   here's the MacBook that's what it is

00:55:42   that and you know that's it's you can

00:55:44   pick you know yellow or blue or

00:55:46   whatever on that that's it yeah I mean

00:55:48   again I think maybe the M1 Walmart one

00:55:50   is already like that is that the case

00:55:51   I believe yeah you can't pick storage I

00:55:55   don't think on it it's just like there's

00:55:56   one configuration and you pick color

00:55:57   that's it all right well I'm glad you

00:55:59   didn't pick my one exists last

00:56:01   legitimate one that I could think of no

00:56:03   charger in the box oh that's a good one

00:56:06   that's a real good one thinking like

00:56:08   tim cook now what can we remove no box

00:56:11   some assembly required it just comes

00:56:14   shrink wrapped loose here it's just

00:56:17   wrapped in a piece of paper it's

00:56:18   biodegradable that's amazing yeah but

00:56:20   no no charger they've been ripping that

00:56:21   out of our products like everyone's got

00:56:22   like it'll charge with any phone

00:56:24   charger like no charger in the box all

00:56:26   right one I got a couple of the

00:56:28   miscellaneous ones and then a big one

00:56:29   so my miscellaneous ones are if you

00:56:32   look at the keyboard and trackpad the

00:56:36   base iPad that that $250 keyboard folio

00:56:40   that it has it advertises is first of

00:56:43   all non-backlit keyboard I think that's

00:56:45   a given this thing will not have a

00:56:47   backlit keyboard I also it says there in

00:56:50   the in the in the you know the keyboard

00:56:52   folio thing for the iPad they call the

00:56:54   trackpad a quote click anywhere trackpad

00:56:57   but if you do the tech compare with other

00:57:00   iPad stuff it it does not list it as

00:57:03   saying haptic feedback so like I'm

00:57:06   thinking it's actually not a taptic

00:57:09   engine trackpad but they have somehow

00:57:11   engineered the mechanics that you can

00:57:12   push anywhere and the whole thing does

00:57:14   indeed click down I don't know I've

00:57:16   never used this keyboard folio so I don't

00:57:17   know that but that's how that's how it

00:57:18   reads in the tech specs so I'm guessing

00:57:20   non-backlit keyboard and a non-force

00:57:22   touch trackpad yeah I think that makes

00:57:24   sense sounds terrible it does sound kind

00:57:27   of terrible but it makes sense all right

00:57:28   what are your other ones only two

00:57:29   speakers the MacBook Air has four with

00:57:32   no spatial audio support I think

00:57:33   that's probably somewhat of a given and

00:57:37   now the here the big resistance I said

00:57:42   they didn't they probably really don't

00:57:43   want businesses to buy these what if it

00:57:47   only runs app store apps I don't think

00:57:51   they'll do that I don't think so either

00:57:52   and I don't think that honestly I don't

00:57:54   think that would actually make it any

00:57:55   cheaper and honestly businesses might

00:57:57   like that more they'll only run apps

00:58:00   their apps and that they love distributing

00:58:01   stuff from their own enterprise app

00:58:02   stores you know what I mean like that

00:58:03   solves problems for businesses maybe

00:58:05   all right so if they did all this first

00:58:08   of all I think iBook has got to be the

00:58:10   name like that would be such a great

00:58:12   name right now whenever you try to think

00:58:14   of something obvious and good regarding

00:58:15   naming it's probably not what Apple do so

00:58:17   yes iBook has come up a lot people have

00:58:19   fondness for the name it would make

00:58:22   sense lots of things make sense yeah

00:58:24   especially when it comes to naming I

00:58:25   mean I remember maybe Apple made the iBook

00:58:27   store that was not a store that sold

00:58:29   laptop computers no so yeah we'll see

00:58:33   what they do with that I mean I think

00:58:36   you know they one of their past not so

00:58:39   great naming things required us to give

00:58:41   a name to the MacBook adorable because

00:58:43   what was his actual name MacBook that

00:58:46   didn't catch on the computer didn't catch

00:58:48   on and the name didn't catch on so all

00:58:50   right slow down no the computer did sort

00:58:51   of catch on that's the problem well so

00:58:53   that that's part of the dream of this

00:58:54   one is like oh it could be really thin

00:58:55   like that one but like if if they're

00:58:57   going to reuse the m1 case so much for

00:58:59   that dream you know what I mean I don't

00:59:00   know we'll see if they I guess they can

00:59:02   reuse that really thin case of that if

00:59:04   they still have the like the machining

00:59:05   tools and everything for making that but

00:59:08   yeah it just like people are attaching

00:59:13   their hopes and dreams this they're like

00:59:14   oh thinner lighter cheaper lower power

00:59:18   like that's great and they just start

00:59:20   envisioning the MacBook one or the

00:59:21   MacBook adorable as we used to call it

00:59:23   because the thing we described is

00:59:25   actually kind of like that like that

00:59:26   was so old that the screen probably is

00:59:28   worse right and I did have a backlit

00:59:31   keyboard I think Casey yes I'm pretty

00:59:34   I'm pretty darn sure but that wasn't

00:59:36   like that wasn't a Walmart special

00:59:38   machine it was actually you know it was

00:59:40   priced about what you would think it

00:59:42   would be priced at it wasn't like the

00:59:43   cheapest no it was priced quite a bit

00:59:45   more than what you would think it would

00:59:46   well you would think because it was so

00:59:48   slim and you know it's like this the

00:59:49   thinnest computer ever made like it was

00:59:51   a breakthrough at the time from Apple it

00:59:53   was it was outside the realm of

00:59:54   normalcy and Apple's product lines

00:59:55   whereas I feel like this one will be

00:59:56   well within the realm of normalcy and

00:59:59   Apple's product line to hit a cheaper

01:00:01   price point so the final question what

01:00:04   do we think that if they if they do this

01:00:06   like you know this like really

01:00:07   decontented laptop what do we think

01:00:09   that price point is I think they'll just

01:00:11   sell it for exactly the same price as

01:00:13   they're selling the M1 MacBook Air so

01:00:15   650 ish yeah I think that's I mean

01:00:17   regardless of how much it costs to make

01:00:18   I think that's what they'll sell it

01:00:20   for like it will slot into that and

01:00:22   honestly as the years go on 650 is worth

01:00:24   less now than it was when the M1

01:00:26   MacBook Air was first introduced so

01:00:28   yeah I think they'll hold the line

01:00:30   maybe maybe 599 you know if we're lucky

01:00:33   but I feel like that it's going to

01:00:34   replace that computer yeah I think I

01:00:37   would go again like the the base iPad

01:00:40   plus its keyboard and trackpad case is

01:00:42   600 the Walmart MacBook Air is 650 the

01:00:45   real MacBook Air is a thousand I think

01:00:47   it should be around 600 but it will be

01:00:51   like 750 like I think they're not I

01:00:53   think they're not going to go as far

01:00:54   down as the as we all want them to

01:00:57   because they never do and I think it's

01:00:58   going to be closer to the MacBook Air's

01:01:02   price than we feel that it should be

01:01:04   I think I think that what they end up

01:01:07   doing is they basically replace the

01:01:10   Walmart one and say sorry Walmart you're

01:01:13   cut off and this is the new quote

01:01:15   unquote Walmart you know MacBook or

01:01:18   MacBook iBook MacBook Air whatever they

01:01:20   call it and then they're getting all of

01:01:23   the proceeds from it rather than

01:01:24   Walmart taking a little cut and I think

01:01:26   650 bucks or very very close to so

01:01:30   here's here's my other angle on this

01:01:31   the rumor of this computer and the idea

01:01:33   of this computer so we'll link to

01:01:35   Jason Snell's graphs and six colors

01:01:37   about the speed of the 18 Pro and how

01:01:40   compare compares to the M1 a summary is

01:01:42   that the 18 Pro is 46% faster than the

01:01:47   M1 in single core tasks and pretty much

01:01:50   the same in multi-core but single core is

01:01:51   all that really matters for just like

01:01:52   sort of day-to-day tasks or whatever and

01:01:55   this highlights the interesting fact and

01:01:57   this is one of my hobby horses that I'm

01:02:00   always bouncing on on the show usually I'm

01:02:03   talking about it in the realm of media

01:02:05   catching up to the limits of human

01:02:07   perception which audio has and video is

01:02:11   always racing towards as computational

01:02:15   power becomes cheaper and as it increases

01:02:17   the maximum computational power available

01:02:19   for a given you know a number of watts

01:02:21   increases over time it starts to race

01:02:26   past the needs for doing some basic stuff

01:02:29   long ago computing power raced past the

01:02:32   point where if you want to type words

01:02:34   into a document it can be handled it used

01:02:36   to be that you needed a good computer to

01:02:38   do that to like one word star or whatever

01:02:40   on DOS like you need a you need you

01:02:42   couldn't buy a bargain basement one you

01:02:45   needed a computer that could keep your

01:02:46   whole document in memory and maybe one

01:02:48   that could have proportional fonts so

01:02:49   you need even a bigger fancier computer

01:02:51   like a Mac which costs so much money

01:02:53   but now you know a little kids toy that

01:02:57   comes in a cereal box could has more

01:02:59   computing power than the early PCs like

01:03:00   typing we have passed by web browsing is

01:03:04   a surprisingly complex and computationally

01:03:06   intensive thing and seemingly only getting

01:03:08   more so as web apps become more complex

01:03:11   but I think an a18 pro has no problem

01:03:16   browsing the web or doing spreadsheet stuff

01:03:20   or anything like you're doing word processing

01:03:21   like that level of computation is now

01:03:25   satisfied by essentially phone chips we do

01:03:28   it on our phone people use their phones

01:03:29   as their main computers they do incredibly

01:03:31   complicated things on their phone you

01:03:32   basically run Photoshop on your phone your

01:03:34   phone has so much more computing power

01:03:35   than whole swaths of computers that we

01:03:38   use as the highest of the high-end thing

01:03:39   so as this continues to happen like the

01:03:43   things people regular people need to use

01:03:45   their computer for like going to a web

01:03:48   using a couple of simple apps and doing

01:03:50   word processing only become easier and

01:03:53   easier to accomplish so it makes perfect

01:03:55   sense to me for phone chips to make their

01:03:57   ways to the Mac because I don't know the

01:04:01   right word for this but like the available

01:04:03   computing power is now more than sufficient

01:04:06   to do the things that most people need to

01:04:08   do with their computers not saying it

01:04:10   couldn't be faster people do occasionally

01:04:11   have to wait for their computers in

01:04:13   particular storage speed and stuff like

01:04:14   that so that it will continue to get

01:04:16   better over time but this machine makes

01:04:18   so much sense to me not just as a way to

01:04:22   save money but recognizing that most people

01:04:27   don't need this much power out of their

01:04:29   Macs that the baseline power especially

01:04:31   with the advent of Apple silicon is so

01:04:32   massive that it's like an M1 is actually

01:04:34   overkill for a lot of people and this is

01:04:36   actually faster that 46% faster than an M1

01:04:39   in the ways that counts which brings me to

01:04:41   my other idea about this computer which the

01:04:44   rumors don't support but I think would be

01:04:45   valid is what if this wasn't a low-cost

01:04:47   machine what if instead this was just the

01:04:50   bottom of the MacBook Air range but they

01:04:52   didn't take every ounce of cost out of it

01:04:54   and instead just added it as essentially

01:04:57   the get way more battery life option it

01:05:01   would have the same size battery as like

01:05:03   the M5 and M6 MacBook Air in the same

01:05:05   case with all the same features the same

01:05:07   screen like everything the MacBook Air gets

01:05:08   this gets the only difference is hey do you

01:05:11   not need an M5 for what you do with your

01:05:14   MacBook Air get the one with the A20

01:05:18   whatever processor or A18 Pro in it same

01:05:21   size battery but boy you'll get even more

01:05:23   battery life and no it's not that much

01:05:25   cheaper but like instead of thinking this

01:05:27   is like oh this is the way we're gonna get

01:05:29   cost out of it think of it as just a

01:05:30   different trade-off between computational

01:05:32   power and battery life essentially I could

01:05:36   see that being a perfectly valid product it's

01:05:38   not what's rumored here to be clear but I

01:05:40   think the day may come when we see that as

01:05:43   an option because if the if Apple keeps

01:05:46   doing the whole thing where it's like M

01:05:47   something and then M something Pro M

01:05:49   something Max and something else or

01:05:50   whatever the M something I feel like

01:05:52   already with the M4 is more power than

01:05:56   like a student needs to use web apps and

01:05:58   word processing and spreadsheets for their

01:06:00   college stuff like it's overkill if you

01:06:04   you could get something that is less

01:06:05   powerful but also uses less battery power

01:06:09   yeah that would be great and so I hope

01:06:12   that either either they they stop doing

01:06:16   the M M4 M4 like that they add something

01:06:18   below that and I don't think honestly the

01:06:20   answer isn't just do M4 and then M5 M5 Pro

01:06:22   M5 Max or whatever something like you know

01:06:25   using a phone chip maybe they've got the

01:06:27   unified design maybe they unify the naming

01:06:29   between the A chips and the M chips and

01:06:31   just have a line of chips that they use

01:06:34   throughout their products and somehow name

01:06:37   them in a sensible way where it's a scale

01:06:38   up from the smallest and least powerful to

01:06:40   the biggest and most expensive and they

01:06:42   just are and most powerful and just

01:06:44   distribute that across their line of

01:06:46   products without this artificial

01:06:48   bifurcation between both the A series are

01:06:50   for iPads and phones and the M series are

01:06:51   for Max unless it's for Vision Pro but

01:06:53   also some iPads of M's you see what I'm

01:06:55   getting out there so that's not today but

01:06:57   that's just something I'm thinking about

01:06:58   that a time may come when the M5 M6 or

01:07:02   M7 is too powerful has it's what the

01:07:05   power of it is wasted on a MacBook Air and

01:07:08   you can make that a better computer for

01:07:09   people by putting in a quote-unquote phone

01:07:12   chip and giving it way more battery life

01:07:13   hmm this has got me thinking of a of a

01:07:17   dark possibility here what if this is not

01:07:20   really that much cheaper like you were

01:07:23   saying what if this is basically the new

01:07:25   MacBook Air maybe it goes from a thousand

01:07:28   to nine hundred dollars and they just

01:07:31   push the price of the Air up because

01:07:34   here's the thing if you look right now

01:07:35   you know in the current lineup they sell

01:07:37   the Air M4 for a thousand dollars you

01:07:40   know there's like you know disabled

01:07:41   cores probably but like there's a thousand

01:07:42   dollars then they sell the Pro the 14

01:07:46   inch Pro with the same chip the M4 but

01:07:49   with some MacBook Pro niceties you know

01:07:50   the screen the speakers everything

01:07:52   sixteen hundred dollars that's the

01:07:55   cheapest Pro with the same chip the

01:07:57   screen is so much more expensive than

01:07:59   the computer though it is true but it's

01:08:01   not six hundred dollars more expensive

01:08:02   like and there's other changes too but

01:08:04   like so the MacBook Air to the MacBook

01:08:06   Pro with the same M4 chip is a six

01:08:09   hundred dollar difference although

01:08:11   again there's probably a binning thing

01:08:12   yeah there's a binning thing going on

01:08:13   with without binning it's a four hundred

01:08:16   dollar difference so I wonder if this

01:08:19   is really just a way for them to close

01:08:20   that gap and Apple never closes gaps in

01:08:22   making things cheaper I'm guessing

01:08:24   maybe maybe the the cheap laptop is

01:08:27   really only nine hundred dollars and

01:08:30   then they push the Air up to like twelve

01:08:32   hundred dollars yeah we'll see like it

01:08:34   is kind of a shame that these rumors

01:08:35   seem so concretely set on the 18 Pro

01:08:37   because that chip does not have a

01:08:40   unified name like they've unified the

01:08:41   name of their OS's but their chips are

01:08:43   still weirdly bifurcated between A and

01:08:45   M's and one R and a C for a different

01:08:50   thing and speaking of don't forget the

01:08:51   H's yeah yeah that's right speaking of

01:08:53   C Lima Lum asks do you think the new

01:08:55   phone app in Tahoe might mean cellular

01:08:58   functionality will be coming to max and

01:09:00   I would say putting an 18 Pro into a Mac

01:09:03   also makes people think that lots of

01:09:05   things make people think about perhaps

01:09:06   cellular being in max honestly there is

01:09:09   so much in common between Mac and phone

01:09:11   hardware and has been since the advent of

01:09:13   Apple Silicon that there's nothing new

01:09:16   that they do including adding the phone

01:09:18   app or making their own C1 chip or

01:09:20   putting a literal phone chip inside a

01:09:22   Mac that to me makes it any more or less

01:09:25   likely because it has been so far over

01:09:27   the edge of a thing they should

01:09:28   obviously obviously do for so long

01:09:30   nothing else needs to change nothing

01:09:33   else needs to happen it's like well

01:09:34   previously I didn't think it would or

01:09:35   should happen but now that they put the

01:09:37   phone app in Tahoe they're definitely

01:09:38   going to do it Apple you need to do it

01:09:40   no matter what it's it's like there's

01:09:42   there's so much it's like the scales are

01:09:45   like a million pounds on one side and a

01:09:47   feather on the other and we're all just

01:09:48   sitting where their arms are going well

01:09:49   where is it yeah I mean and again I will

01:09:52   remind everyone every iPad since the very

01:09:56   first iPad including all of the cheap

01:09:59   models of the lowest end models of iPad

01:10:01   every single one has had a cellular

01:10:04   option since day one including the ones

01:10:06   with an M chip in the Marco surprisingly

01:10:09   yes including the ones I didn't think

01:10:11   anything with an M chip could do

01:10:12   cellular that's impossible all this Mac

01:10:14   hardware is in all the iPad pros now and

01:10:17   yet or maybe it's the way around but

01:10:19   still like all this all this iPad it's

01:10:21   all in Macs now and yet we still don't

01:10:24   know like that yeah it is not a hardware

01:10:26   consideration and it I'm guessing

01:10:28   realistically it never has been like

01:10:30   the reason we don't have cellular Macs

01:10:32   you know whatever their reasons are

01:10:34   it's not hardware or software yeah or

01:10:37   services at this point because yeah so

01:10:39   I don't think the phone app makes it

01:10:41   any more likely I think that is just a

01:10:42   continuing thing where they take apps

01:10:44   that previously were on the phone and

01:10:45   bring them to the Mac and the phone

01:10:47   app is just one more of them great I

01:10:48   like it it's a good thing it doesn't

01:10:50   make me think that cellular on the Mac

01:10:52   is any more likely than it should have

01:10:53   been before because I feel like the

01:10:54   likeliness is already pressing up to 80

01:10:57   90 percent and holding steady there

01:10:58   that was a fun game I have no idea what

01:11:02   the heck is gonna happen but that was a

01:11:03   fun game I appreciate it honestly this

01:11:05   is I think this is a very fun rumor like

01:11:07   I love thinking about the idea of like

01:11:09   well why did they take this other

01:11:10   combination of hardware and make you

01:11:12   know make an even smaller or lower end

01:11:14   Mac with it like that's interesting to

01:11:15   me or with way more battery power

01:11:17   cheaper gets people excited more

01:11:18   battery power gets people excited I'm

01:11:20   excited about the whole the whole

01:11:22   concept of computing power surpassing the

01:11:24   needs of very common applications and I

01:11:27   think that's just about to happen to

01:11:29   web browsing which is a pretty amazing

01:11:30   yep I agree a hundred percent a couple

01:11:33   other quick thoughts I was on upgrade

01:11:36   excuse not upgrade clockwise this week

01:11:39   we briefly talked about this and let me

01:11:42   tell you if this portends if this is if

01:11:45   this is leading to a reincarnation of the

01:11:48   MacBook adorable holy mother I'm so in

01:11:52   because if you remember the MacBook

01:11:53   adorable leaving aside the fact that it

01:11:54   only had one port which kind of sucked

01:11:56   one of the worst parts about it was that

01:11:59   the processor was just dog slow even on a

01:12:02   good day that process processor was dog

01:12:05   slow and the idea of having something

01:12:07   that was that thin and that light but

01:12:10   not utterly crippled by a crappy Intel

01:12:13   CPU oh man does that sound good oh and

01:12:17   also no no butterfly keyboard sorry trigger

01:12:20   warning Marco but I know butterfly keyboard

01:12:22   that sounds incredible so I mean I don't

01:12:27   expect that's what's gonna happen but

01:12:29   man that would get me really really

01:12:30   interested so proud of your

01:12:32   pronunciation of dog Casey yeah I would I

01:12:35   would love to see that because like

01:12:37   ever since Apple Silicon like we've been

01:12:39   saying like wow we can't wait for them

01:12:41   to make you know a good version of the

01:12:43   12 inch MacBook with this with these this

01:12:45   news you know tech that they have and they

01:12:47   haven't like you know the the the M2 and

01:12:50   forward MacBook Air is great and you know

01:12:54   like with its with its newer case design

01:12:55   and everything it is super slim super

01:12:57   light but it's still not the same

01:12:59   portability class as the ancient 11 inch

01:13:02   MacBook Air or the 12 inch MacBook like

01:13:04   it's still significantly larger and heavier

01:13:07   you know for the modern day they like the

01:13:10   12 inch MacBook was I believe two pounds

01:13:13   right like 2.0 or like close to 2.0 uh what

01:13:16   was it I think it was yeah around two

01:13:18   pounds something like that yeah I think

01:13:19   and I think we're currently at 2.4

01:13:21   something for the 13 inch Air yeah so

01:13:22   like you know we're we're we're still

01:13:25   bigger and heavier we still don't have

01:13:27   that magical like you know airplane

01:13:29   tray table computer that we used to have

01:13:31   I know the rumor says a 13 inch screen

01:13:32   so it sounds like if that kind of

01:13:35   computer is coming this probably isn't

01:13:36   it but this does show maybe a path they

01:13:39   can get there like if they if they decided

01:13:42   for whatever reason that like that they

01:13:44   can't do it with the full suite of M you

01:13:47   know M chip series features although they

01:13:49   do it in the 11 inch iPad Pro and the 11

01:13:53   inch iPad Air so I I don't know why they

01:13:56   couldn't do it but like a proper you know

01:13:59   11 or 12 inch laptop that's actually you

01:14:02   know Mac OS not just an 11 inch iPad I

01:14:05   think would be so cool and you know in

01:14:08   that two pound range I really hope they

01:14:10   make that and I'm not I haven't given up

01:14:12   hope yet same I don't know if this is it

01:14:14   as I said but I haven't given up given

01:14:16   just to be clear based on the game

01:14:17   before the price I was giving it was for

01:14:20   the price of the a team based laptop

01:14:22   computer that Apple would put out if it's

01:14:24   going to be a low-cost model it is not

01:14:26   the price of what I think after we

01:14:28   removed everything in our game the

01:14:29   thing would cost because I don't think

01:14:30   Apple would remove all the things to be

01:14:31   listed we really gutted this thing to

01:14:33   the point where it's a bad product

01:14:34   real-time follow-up the weight of the

01:14:38   MacBook adorable 2.03 pounds or

01:14:40   insensible units just a shade less than

01:14:42   a kilogram oh no Jason wants us to

01:14:44   point out the tariffs may throw a

01:14:45   monkey wrench into this whole thing

01:14:46   what else is new oh and by the way the

01:14:48   and the 13 inch MacBook Air is 2.7

01:14:50   pounds oh wow 2.7 versus 2.0 is a

01:14:53   big difference yeah it really really

01:14:54   so like we are not in that same class

01:14:57   with the new MacBook Air it the new

01:14:59   MacBook Air is a wonderful machine in

01:15:00   lots of ways but we are not in that

01:15:02   same class we are sponsored this

01:15:05   episode by delete me right now the

01:15:07   headlines are chock-full of data

01:15:08   breaches and regulatory rollbacks

01:15:10   making us all vulnerable but you can

01:15:12   do something about it delete me is here

01:15:14   to make it easy quick and safe to

01:15:15   remove your personal data online

01:15:17   delete me is knows your privacy is

01:15:19   worth protecting so here's what they do

01:15:21   they go to hundreds of data broker

01:15:23   websites and they opt you out and file

01:15:25   takedowns on your behalf so that the

01:15:27   data brokers have to remove your data

01:15:29   from their databases and they show you

01:15:31   regular personalized privacy reports

01:15:33   showing what they found where and what

01:15:35   they were able to remove always working

01:15:36   for you constantly monitoring and

01:15:38   removing whatever you want not to be

01:15:40   in those data broker sites on the

01:15:41   internet and you know there's there's

01:15:43   all sorts of news that shows data

01:15:44   breaches and stuff there's a recent

01:15:46   coinbase data breach where a hacker

01:15:48   obtained names addresses phone numbers

01:15:50   partial socials and bank account

01:15:51   numbers data brokers can vacuum up a

01:15:53   lot of that data from the hack they can

01:15:55   create a searchable profile about you and

01:15:57   sell it to whoever wants it so delete me

01:15:59   helps keep you out of those sites so take

01:16:01   control of your data and keep your

01:16:04   private life private by signing up for

01:16:05   delete me now at a special discount for

01:16:08   our listeners get 20% off your delete me

01:16:10   plan when you go to join delete me dot com

01:16:13   slash ATP and use promo code ATP at

01:16:16   checkout the only way to get 20% off is to

01:16:18   go to join delete me dot com slash ATP and

01:16:21   enter code ATP at checkout that's join

01:16:23   delete me dot com slash ATP code ATP thank

01:16:27   you so much to delete me for sponsoring

01:16:28   our show

01:16:29   all right let's do some ask ATP and neil

01:16:35   writes with machines like the mac mini

01:16:36   having such expensive storage many people

01:16:38   look at adding external storage with an

01:16:40   external NVMe drive via USB 4 or a

01:16:44   thunderbolt enclosure are there reliability

01:16:46   issues with having something permanently

01:16:48   attached like this for example do you

01:16:49   regularly regularly run into it randomly

01:16:52   being unmounted is there more risk to file

01:16:55   corruption if you're using it for a

01:16:57   backup is Thunderbolt more reliable than

01:17:00   USB 4 I don't have any strong answers

01:17:03   for this with a Mac mini or something

01:17:05   stationary I suppose there's no reason you

01:17:07   couldn't do something like this but I

01:17:10   don't know if it were me I would try my

01:17:11   darndest to find a storage solution

01:17:15   that's internal that works both from a

01:17:16   cost and space perspective but yeah I'd

01:17:19   get why you wouldn't want to do that for

01:17:20   sure as I think the one of us who is

01:17:22   probably using has used and probably still

01:17:25   is using the most external storage because

01:17:26   I'm only using desktops and it's convenient

01:17:28   to use for desktops I will give a Casey

01:17:30   style intervention here and say if you are

01:17:34   using external storage connected to your

01:17:36   Mac and it's randomly unmounting itself

01:17:40   that is not a way to live you need to stop

01:17:45   and either figure out how to fix it or get

01:17:49   rid of that hardware I know too many

01:17:51   people who say like oh this happens all

01:17:53   the time it's not a big deal it just you'll

01:17:55   plug it in and remount it or it'll come

01:17:56   back in a second that's not acceptable for

01:17:59   storage that you plan on using like it just

01:18:02   you can't have that and the reason people

01:18:04   live with that is because the problem that

01:18:06   Neil was getting at here is it's not

01:18:08   actually that easy to find reliable

01:18:12   inexpensive external storage like it's it

01:18:16   used to be easier when it was much more

01:18:17   common for people to do that but most of

01:18:19   these days most people get by with what's

01:18:20   inside their computers so it becomes this

01:18:22   narrower and narrower market and if you go

01:18:25   on like Amazon or Newegg or whatever you

01:18:27   can find tons of really really cheap

01:18:28   external storage from name brands that is

01:18:30   nevertheless incredibly unreliable and

01:18:32   flaky and so everyone always wants to

01:18:34   know what's the good one what's the one

01:18:36   that I should buy and then you point out

01:18:37   some Thunderbolt enclosure like yeah no

01:18:39   seriously I'm not paying that much for

01:18:40   that thing and it's like so to answer

01:18:42   the other question is Thunderbolt more

01:18:44   reliable than used before usually but that

01:18:47   may be because the Thunderbolt stuff

01:18:48   costs seven times as much and just just

01:18:51   the cost difference means that using more

01:18:52   reliable components all around and the

01:18:54   companies that make them cater to

01:18:56   customers who demand more reliability like

01:18:58   it's not just that you get what you pay

01:19:00   for it's like it's if you don't need

01:19:02   Thunderbolt you would you would like to

01:19:03   not pay for it but every time I try to

01:19:06   and I have so many external enclosures

01:19:08   here I try to find like an enclosure to

01:19:10   fit like an SSD or back in the day a

01:19:11   spinning drive or whatever I always felt

01:19:13   like I was rolling the dice and if any of

01:19:15   these external you know devices whether

01:19:18   they come with a mechanism inside them

01:19:19   or whether it's a shell that I plug

01:19:21   something into if they have any kind of

01:19:23   reliability things or it's like

01:19:24   randomly unmounting that's the end of

01:19:26   that thing in my life like I stop you

01:19:28   but you can't you can't have it you

01:19:30   will corrupt your data best case you'll

01:19:32   just lose some data worst case you'll

01:19:33   corrupt the thing and you know I

01:19:35   need to loosen data like oh something

01:19:36   didn't get written completely worst case

01:19:38   you'll corrupt the entire thing so if

01:19:40   you're ever in that situation and the

01:19:42   drive keeps unmounting you need to fix

01:19:44   that ASAP treated as a five alarm fire

01:19:46   and I wish I had better advice on how to

01:19:49   find something reliable but

01:19:50   the only I mean the best you can do is

01:19:53   more expensive ones tend to be more

01:19:55   reliable in my experience Thunderbolt

01:19:57   things both enclosures and drives tend

01:20:00   to be more reliable they cost so much

01:20:01   more money but I you know it's painful

01:20:03   but I do it like for example my my one

01:20:05   terabyte external drive that I currently

01:20:07   have Tahoe on is a Thunderbolt drive

01:20:11   which makes no sense because like I

01:20:13   don't even know what the speeds the

01:20:14   internal storage can do it's so old but

01:20:16   it's just it's just so much more

01:20:18   reliable for me than than my other USB

01:20:20   based things that I have but yeah don't

01:20:24   like and and if it is reliable yeah you

01:20:26   can use it for backup it's fine like

01:20:27   there's especially if it's on a

01:20:29   desktop computer it's always plugged in

01:20:30   it's always powered up especially if

01:20:31   it's bus powered because then it's

01:20:33   meant to run on the bus don't put it

01:20:34   through 17 hubs like directly connected

01:20:36   if you can or make sure you know like

01:20:38   if it is reliable by all means use it

01:20:40   as backup it'll be fine if it's not

01:20:42   reliable you need to stop using it yeah

01:20:44   definitely I have I use a bunch of

01:20:47   external drives for things like time

01:20:49   machine and occasionally like an

01:20:51   archive drive here and there it

01:20:53   obviously is harder to make reliable

01:20:57   when you are constantly moving the

01:20:59   drives you know so like in the in the

01:21:02   context of like a laptop it's a lot

01:21:03   harder to make that reliable over time

01:21:05   like you know the cable will keep

01:21:06   bending back and forth maybe you might

01:21:08   slowly wiggle it out of the port the

01:21:09   port might loosen like you know there's

01:21:11   all sorts of stuff like that but in the

01:21:12   context of a desktop we're just sitting

01:21:14   around I have found for the most part

01:21:16   you're you're fine getting any modern

01:21:19   USB thing and it's fine now I have

01:21:23   found there is basically no correlation

01:21:26   between the reliability of USB devices

01:21:30   and their price or their brand name for

01:21:33   that matter I have some rock-solid USB

01:21:37   hubs and cables and driving closures or

01:21:41   external drives external SSDs I have some

01:21:43   rock-solid ones that are just like no-name

01:21:45   brands and I've bought some great ones

01:21:48   from like you know like there were like

01:21:50   some some high-end brands or like you

01:21:52   know well-respected names in the

01:21:53   business that were unreliable pieces of

01:21:55   crap and the main reason behind a lot of

01:21:57   this is like when you're making like a

01:21:59   drive enclosure or a USB hub or

01:22:01   something there's only so many actual

01:22:03   like controller chips that are

01:22:05   manufactured that will power something

01:22:07   like that so almost any brand that you

01:22:09   get is going to be using the same you

01:22:11   know one of a very small number of

01:22:13   controller chips as every other brand is

01:22:15   using so I have found that like the

01:22:18   sometimes you just luck out and you get

01:22:20   a really good one sometimes you get a

01:22:21   bad one like John I think if it starts

01:22:24   randomly unmounting it is done like you

01:22:26   got to replace it like that that will

01:22:27   never get better for that one but again

01:22:28   I've had I've had most most of the ones

01:22:30   I've had are just fine and they last

01:22:33   you know effectively forever and by the

01:22:35   way on the chip thing one of the things

01:22:37   you can do if you do want to try to

01:22:38   research this is certain chip sets get a

01:22:41   reputation for reliability or unreliability

01:22:43   and if you can just look at the fleet of

01:22:45   no-name brands or the fleet of name

01:22:47   brands and find someplace that will list

01:22:49   oh this one has this chip in it the all

01:22:51   these have this chip and all these have

01:22:52   that chip like even if there's only three

01:22:54   possible chips that are being sold on the

01:22:55   market maybe one of those is the chip

01:22:57   you don't want and the two other ones are

01:22:58   good you can find that information out

01:23:00   it's annoying it's hard to find because

01:23:01   the manufacturers probably aren't going to

01:23:03   tell you but someone on YouTube will have

01:23:04   cracked the thing open and said oh this

01:23:06   cheap enclosure uses the good chip and

01:23:08   this cheap enclosure uses the bad chip

01:23:09   they're exactly the same price you

01:23:12   wouldn't know this unless you found the

01:23:13   review like that's why it just feels

01:23:14   like rolling dice sometimes because you

01:23:15   don't know like it's not even that cost

01:23:17   difference between these cheap

01:23:18   controls just sometimes there's good

01:23:19   ones and bad ones or sometimes there's

01:23:20   controllers that like mac os likes and has

01:23:23   a good driver for and ones where mac os

01:23:25   has a credit controller like it's not

01:23:26   even always the drive's fault so that's

01:23:28   one thing you can do if you really want

01:23:29   to dig into it yeah one other thing

01:23:32   about Thunderbolt though Thunderbolt is a

01:23:36   higher-end spec it has more requirements a

01:23:39   long time ago at some media briefing I

01:23:42   was talking to Greg Jaws back Jaws about I

01:23:45   believe I was complaining about USB C back

01:23:47   in like you know the bad old days of the

01:23:48   MacBook Pros with the touch bars and

01:23:50   everything and one thing that Jaws said

01:23:52   has stuck with me he said he said like

01:23:55   Thunderbolt hubs and Thunderbolt devices

01:23:58   are held to a way higher standard than

01:24:01   USB devices because every Thunderbolt device

01:24:04   has to be certified and it goes through a

01:24:06   more rigorous process than most USB devices

01:24:08   do now again that was many years ago I

01:24:11   don't know you know to what degree that

01:24:13   is still true but I have so far like in

01:24:16   my computing life I have had so I've had

01:24:19   some flaky USB devices I've never had a

01:24:22   flaky Thunderbolt device they've all been

01:24:25   rock-solid now there are downsides in

01:24:28   addition to them costing usually at least

01:24:31   three times as much as the USB equivalent

01:24:33   they also you know they tend to you know

01:24:36   use more power that you know there's

01:24:38   there's almost nothing that's Thunderbolt

01:24:40   bus powered like it's you know you're

01:24:42   looking at external you know external

01:24:43   power supplies much of the time usually

01:24:46   Thunderbolt chips are bigger and hotter so

01:24:48   the Thunderbolt even something as simple

01:24:50   as like a drive enclosure that's gonna

01:24:51   have a big hot Thunderbolt bridge chip in

01:24:54   it probably and so that whole thing is

01:24:55   gonna be hotter and maybe it'll need a

01:24:57   fan or at least it'll just be sitting

01:24:58   there like melting your desk by the way

01:25:00   my one terabyte external Thunderbolts

01:25:02   drive is bus powered does not have a fan

01:25:06   it doesn't get hot yeah because they

01:25:08   get hot anyway so Thunder so Thunderbolt

01:25:10   stuff it is better but it is also so much

01:25:13   more expensive that for most people and

01:25:15   most needs it's not worth it I find

01:25:18   Thunderbolt is best for hubs like if you

01:25:21   want to like have like a port expander

01:25:23   that you're plugging your laptop into

01:25:24   with one cable and you want a bunch of

01:25:25   ports that are gonna be reliable

01:25:26   Thunderbolts great for that for just like

01:25:29   an external SSD my needs don't need

01:25:34   external SSDs to be massively fast I'm

01:25:37   mostly using them again in like a time

01:25:38   machine or an archive context so I just

01:25:41   need my SSDs my external SSDs to be

01:25:43   large in capacity you know I need like

01:25:46   the four terabyte the eight terabyte and

01:25:48   like so space is more important and then

01:25:50   performance so I just go USB on those

01:25:53   and it's been fine Thunderbolt would have

01:25:55   been even more expensive than they

01:25:56   already were and it's just not worth it

01:25:58   for me but again for hubs definitely

01:26:00   worth it and my my backup drive like my

01:26:02   eight terabyte SSD it's connected with

01:26:05   SATA oh my god because it's internal

01:26:08   but like somewhere in there is some

01:26:10   crappy ancient old ship that is

01:26:12   translating whatever that SSD is so it

01:26:15   can speak SATA to the internal SATA bus

01:26:17   on my Mac which is not a thing that

01:26:18   most people have to worry about but like

01:26:20   it drives home the point that for

01:26:22   things like backup drives or like clone

01:26:23   drives or time machine drives if you

01:26:26   want them to be SSDs don't worry so much

01:26:30   about the bus speed because it's just

01:26:31   running in the background at low priority

01:26:33   tasks and slowly siphoning bits over and

01:26:35   the main advantage to it being SSD is the

01:26:37   fact that it doesn't make noise

01:26:39   yeah the but the actual like protocol

01:26:41   maximum speed for drives doesn't really

01:26:46   matter to almost anyone almost any of the

01:26:49   time so yeah get the USB one it'll be

01:26:52   fine you know it's funny you were talking

01:26:55   about Thunderbolt hubs just earlier

01:26:57   today the Caldigit TS5 plus is in stock

01:27:01   enough that they're letting you order

01:27:02   them from Caldigit directly and so I've

01:27:06   placed my order for my five a hundred

01:27:08   dollar oh my god as much as your

01:27:11   computer if you have a Mac mini yeah

01:27:13   seriously but I'm excited to have it so

01:27:16   I have an m3 max so I don't have

01:27:17   Thunderbolt 4 5 I don't Thunderbolt 5 but

01:27:20   as far as I can tell this should still

01:27:21   work it just won't be as quick and I'm

01:27:23   very excited to have an on-board 10

01:27:26   gigabit ethernet port because the one on

01:27:28   the on the TS4 that I have is two and a

01:27:30   half gig and honest to goodness maybe

01:27:33   this is a not a placebo but maybe I'm

01:27:35   convincing myself of something that really

01:27:37   isn't true but I feel like when I'm

01:27:39   moving video files like onto or off of

01:27:41   the Synology it is noticeably quicker I

01:27:44   mean it's two and a half times quicker to

01:27:47   do that with the existing Thunderbolt 4 now

01:27:49   that I have hardware my ubiquity

01:27:51   hardware that supports you know faster

01:27:53   than gigabit ethernet and so I can only

01:27:55   imagine how amazing it's going to be to

01:27:56   have 10 gigabit ethernet because I

01:27:58   already put a daughter board I think we

01:27:59   talked about this put a daughter board in

01:28:01   the Synology to have a 10 gigabit

01:28:02   ethernet port in that and then now in a

01:28:04   few weeks when the ships I'll have 10

01:28:06   gigabit ethernet on my desk and I am

01:28:09   very very excited all right Patrick

01:28:12   Harms writes I remember Marco talking

01:28:15   about block storage or iSCSI for

01:28:17   something similar he experimented

01:28:20   experimented with to increase the

01:28:22   internal storage on his Mac what I want

01:28:24   to achieve is reliable protective

01:28:25   protected storage on a NAS unraid or

01:28:27   true NAS that can be used most

01:28:30   efficiently by the Mac single user only

01:28:32   for storing the photo library or bigger

01:28:34   files the connection speaking of is

01:28:36   10 gigabit ethernet between both

01:28:37   devices I found this post on reddit

01:28:39   describing a workaround with a mounted

01:28:40   disk image on an SMB share any advice

01:28:43   so what I used iSCSI for was to get

01:28:47   around a limitation of backblaze or a

01:28:51   feature of backblaze so so backblaze you

01:28:53   know the online cloud backup that they've

01:28:54   been a sponsor before I they're still my

01:28:56   preferred cloud backup provider and the

01:28:58   end backblaze will back up at no

01:29:00   additional charge any external drives that

01:29:03   are connected to the Mac I had gotten this

01:29:06   huge Synology NAS a million years ago with

01:29:09   everybody else on the show and and I was

01:29:12   looking for a way to like let me have

01:29:13   cloud storage of what I think it was

01:29:16   something like you know eight or 12

01:29:18   terabytes of archive data without having

01:29:21   you know that be directly connected to

01:29:22   my computer in the form of an external

01:29:24   drive I wanted the you know this the

01:29:25   NAS to be doing that and so I did and

01:29:28   and I went through two so iSCSI is not

01:29:31   natively supported by macOS you have to

01:29:32   buy a an iSCSI initiator software I went

01:29:36   through two of them I believe the last one

01:29:38   was the was the ato extend san family I

01:29:42   think that's still the only one that

01:29:43   works I could be wrong this is terrible

01:29:46   mac software like the the the iSCSI

01:29:48   initiating process is horrendous and what

01:29:51   you get is what you expect it to be like

01:29:54   it behaves like a disk directly connected

01:29:57   however the NAS can't see it because

01:30:01   what you're doing is you're having the

01:30:04   Mac have its own in its own file system

01:30:07   its own you know block storage

01:30:08   abstraction so all those files that are

01:30:11   on that giant that giant share being

01:30:13   hosted by the NAS the NAS doesn't see it

01:30:15   as files the NAS sees it as a volume of a

01:30:18   format it doesn't understand oh I don't

01:30:20   think I knew that that's fascinating so

01:30:22   this it basically it removes a lot of

01:30:25   the reason to have a NAS because you know

01:30:28   its applications can't really interact with

01:30:31   the contents of that drive so it's it's

01:30:34   fairly limiting again the software on the

01:30:36   Mac is terrible it did work and backblaze

01:30:40   did happily back it up as though it was an

01:30:42   external drive but I feel like any pretty

01:30:45   much any other solution would be better so

01:30:47   in this case like you know having some

01:30:49   kind of like you know mounted disk image

01:30:50   on other what was otherwise just a

01:30:52   regular file share that's probably not

01:30:54   going to be as fast for the Mac like

01:30:56   there might be more overhead in terms of

01:30:58   shuffling those blocks back and forth and

01:30:59   like into the disk image but that's

01:31:02   probably the better approach I disagree

01:31:04   with that I think that would be much

01:31:07   worse because I think any Mac application

01:31:10   that sees your iSCSI initiated NAS storage

01:31:14   thing as just a volume like it's that's

01:31:16   going to work the way it expects because

01:31:18   you formatted it and at the time probably

01:31:20   HFS plus or whatever it was like it's

01:31:22   just treating it like a drive and it works

01:31:24   like a drive and it you know any any kind

01:31:26   of thing that's like oh you need to have

01:31:29   you boot it off your internal drive for

01:31:30   apple intelligence or photo libraries

01:31:32   can only be on a volume of this type

01:31:34   or whatever it's straightforward now you

01:31:35   can make a disk image format in any

01:31:37   format you want but it's not like it's

01:31:40   it's a disk image there's all sorts of

01:31:41   stuff going on behind the scenes to make

01:31:43   that work that especially when it comes

01:31:45   to things like file locking and the

01:31:47   atomicity atomicness whatever the word is

01:31:51   of operations in the file system all of

01:31:53   that can potentially change when dealing

01:31:55   with disk images so i would absolutely

01:31:59   recommend against taking a disk image

01:32:02   putting it on a NAS and mounting that

01:32:04   disk image on your Mac and then putting

01:32:05   your photo library on it like the that

01:32:07   stack of things is a not a good idea if

01:32:10   you're not putting your photo library on

01:32:11   and you just want to have a disk image on

01:32:13   your NAS and put files in it that should

01:32:15   work but it's an incredibly inefficient use

01:32:17   of your NAS you know as you break down the

01:32:21   layers it's like look just just mount

01:32:22   your NAS and use it as a network attached

01:32:25   storage and unless you have some other

01:32:26   need like I want backblaze to back this

01:32:28   up or I want to put my photo library on

01:32:29   it and in either of those cases you really

01:32:32   have to look at what you're aiming for if

01:32:33   it's just a bunch of archive data and you

01:32:35   want to get it backed up as part of your

01:32:37   low one price thing backblaze iSCSI is

01:32:40   probably the only solution but if you want

01:32:41   to put your photo library on it I would

01:32:43   absolutely not put a disk image on your

01:32:45   NAS and mount it even if it appears to

01:32:47   work it's just it's too many layers it's

01:32:50   too slow and I don't know if that's even a

01:32:52   supported configuration for apple photos

01:32:54   and I wonder if apple would even know

01:32:56   because they would be asking you well

01:32:57   what disk image format did you use did

01:33:00   you use a sparse bundle did you use a

01:33:01   read write disk image did you use the

01:33:03   new as if whatever that format was that

01:33:05   we talked about on the past show like and

01:33:07   then how is it mounted and what are the

01:33:08   samba options on the serving side and on

01:33:11   the client side and what version of mac os

01:33:12   you were doing too many questions for

01:33:14   something as important as a photo

01:33:15   library

01:33:16   then finally for today matt jorgensen

01:33:19   writes my son recently graduated with a

01:33:21   degree in computer science what do you

01:33:22   recommend for new graduates looking for

01:33:24   their first job well first of all

01:33:25   congratulations

01:33:26   second of all I would say and I think

01:33:30   we've talked about this several times in

01:33:31   the past but it's one of those things we

01:33:32   should do periodically like John talks

01:33:34   about if you can get them involved in

01:33:37   some sort of open source I think that

01:33:39   that's very healthy either that could be

01:33:41   open sourcing their own thing it could

01:33:44   be participating in bug fixes or features

01:33:46   in some other open source project that

01:33:48   especially I think would be great

01:33:49   because that teaches you a lot about

01:33:50   working with other people working with

01:33:52   other people's code and so on and so

01:33:54   forth so I think the the first thing I

01:33:57   would look at is getting involved in some

01:33:58   sort of open source and trying to make a

01:34:00   name for yourself within a project you know

01:34:02   be be more than just an anonymous lurker in

01:34:05   some project and see if that can help

01:34:07   my advice is probably more general and not

01:34:11   like as specific as Casey's was if you

01:34:15   have if you're just graduated with a

01:34:17   computer science degree and you have any

01:34:19   interest whatsoever in being involved in

01:34:24   any kind of like startup or you know

01:34:28   leading-edge technology or like be one of

01:34:31   a few number of people in a small company

01:34:34   with a lot of responsibility now is the

01:34:36   time to do it when you're young if you

01:34:39   don't have any interest in that don't do

01:34:40   it like don't do it just as you think it's

01:34:41   what people do but if you're like I always

01:34:43   wanted to be like one of five programmers

01:34:44   working on a project or something I always

01:34:46   wanted to be involved in a startup or

01:34:47   whatever assuming your son graduated with a

01:34:50   degree and is like you know young like in

01:34:54   the mid-20s or whatever the expected age

01:34:57   someone graduates with that degree now is

01:34:59   the time to do it because it will only get

01:35:01   harder to do that and also having a job like

01:35:06   that early in your career where you are one

01:35:08   of a small number of people will force you to

01:35:11   learn how to do a whole bunch of stuff and

01:35:14   that will make you a much more valuable

01:35:16   employee when you get tired of the startup

01:35:18   world or when you want to go in a company

01:35:20   that's not going to go under later you will

01:35:23   have you will have so much more real world

01:35:26   experience and knowledge than people who

01:35:29   went to work for IBM right out of school

01:35:31   to throw IBM under the bus but like went to

01:35:33   work for some big company so when you go for

01:35:35   a big company you do learn things on the job

01:35:36   but it's a much more stable environment

01:35:39   where what's expected of you you're not you

01:35:41   don't go in there and it's like now you have

01:35:42   17 jobs and you got to do all these things

01:35:44   you got to learn them all now they're gonna

01:35:45   it's going to be more sustainably paced

01:35:48   let's say but you will learn less it will

01:35:51   take you longer or your your your skills

01:35:53   will be more narrow because there's 75,000

01:35:55   other people who do their own specialized jobs

01:35:57   this is not to say that like do startups when

01:35:59   you're young because that's the time to burn

01:36:02   you out you shouldn't have burnout even

01:36:03   when you're young but what I am saying is

01:36:05   if you're in a company with a small number

01:36:06   of people it can still be a healthy work

01:36:08   environment and also you will still be

01:36:10   required to learn how to do way more jobs

01:36:12   because there's just not enough people so

01:36:13   someone's got to figure out how to

01:36:14   administer this Linux machine congratulations

01:36:17   you're the sysadmin now someone's got to

01:36:19   learn this new API or this new language

01:36:20   congratulations it's you and one other

01:36:22   person like you will learn so much and you

01:36:25   will be battle tested and when you your

01:36:28   company inevitably goes under because that's

01:36:29   what happens to most startups when you apply

01:36:31   for that job let's say Apple or Google or

01:36:34   whatever you should look a lot better at that

01:36:36   than the other candidates because you will

01:36:38   actually literally know how to do more stuff

01:36:40   also you know just money wise a pretty good

01:36:46   career strategy is go work in the startup

01:36:49   scene and get as many shares of as many

01:36:52   companies as you can because you know over

01:36:55   time some of those will end up being worth

01:36:58   money not all of them maybe not even most of

01:37:01   them or in my case none of them but like the

01:37:04   way venture capital works is VCs make a whole

01:37:07   bunch of very risky bets and they tend to come

01:37:10   out ahead because a few of them pay off big you

01:37:14   can do that as an employee getting shares along

01:37:17   the way you're playing your own version of that

01:37:19   game so if you work for a startup you know the

01:37:23   early the smaller and earlier the startup the

01:37:25   more shares that you will get and the more they

01:37:27   will be worth over time as a percentage of the

01:37:29   entire company then secondly once you are in the

01:37:34   startup scene in your city or in your part of the

01:37:38   industry when one startup dissolves you tend to get

01:37:42   picked up by other ones or it tends to be easier

01:37:44   for you to stay in that community because startup

01:37:46   founders and their investors all talk to each

01:37:49   other and so they know like oh hey this this startup

01:37:53   just just collapsed over there but oh they had a

01:37:55   really good CTO I'm going to try to hire them or

01:37:57   whatever like there's this whole network that you

01:37:59   get by being in the startup scene that like people in

01:38:03   your city will know you and so even if a particular job

01:38:08   you know ends up going under or not working out that

01:38:11   well usually it's easier for you to get more because

01:38:14   you are in that that like social click really and

01:38:17   obviously it depends like you know what city you are in

01:38:20   like what region what type of what part of the

01:38:22   industry are you in that that all matters but it's like

01:38:26   like swimming to the professions it's very much based on who you know and what

01:38:30   your track record is and the startup scene is small and they talk a lot and

01:38:35   their investors talk a lot it worked out well for me but you know a lot of that

01:38:39   was luck obviously but like I didn't just have shares in tumblr I had shares in a

01:38:43   few different things and not all of them worked some of them I lost money most of

01:38:48   them I came out ahead or or at least they at least it was like a push you know so

01:38:52   make a lot of bets and I know a lot of it will you know this isn't just me like a

01:38:56   lot I know a lot of people who have done this like work for startups see what you

01:38:59   can see you know how many how much you know little slices of equity you can get and

01:39:04   I'm not saying don't also work for a salary everyone there is getting a salary you

01:39:09   know that's what the investors money is for so so get your salary and also get a get as

01:39:14   many shares as you can place a lot of bets on that on those tables so to speak and

01:39:19   chances are one of them will pay off eventually or more than one and you know

01:39:24   that that's a nice way to get you know a lot of you know bonus money and that's

01:39:27   very much a a good path when you are young because like John was saying like you

01:39:32   know you have you can take more risks when you're young startup founders also tend

01:39:37   to be young and there is insane ageism in this industry so by the time you're in

01:39:43   your 40s startups won't hire you anyway that's terrible but it's the reality it

01:39:48   shouldn't be that way but it is that way so while you're young go play that game

01:39:53   collect your shares and see what happens later in life as a result you can always

01:39:57   go work for the big tech companies later but the startup startups are very much a

01:40:02   young person's game and then finally I would say for you to stand out when

01:40:07   looking at employers or when having employers look at you one way to do that is

01:40:12   to be in their boys club so you know if you are also like a Stanford graduate or

01:40:17   whatever I don't know I was not any of those things so I couldn't play that game

01:40:21   the version of the game I played was I cared a lot and I could program now I don't

01:40:29   mean I was a great programmer I wasn't and I'm still not I meant I could I can

01:40:34   program you'd be surprised if you haven't seen the high the like applicant pool for a programming

01:40:42   job out there and if you haven't interviewed programming job candidates you may be surprised

01:40:49   how relatively few of them both care and can program now over time this has gotten better there's more

01:40:59   qualified programmers now than there used to be and you're entering right now what's probably a pretty

01:41:03   crappy job market because there were all these tech layoffs recently and so there's a lot of

01:41:07   unemployed programmers right now who do know how to program that's temporary you know that will change

01:41:11   when you are applying for jobs if you show that you care and if you can demonstrate even it doesn't

01:41:18   have to be anything like massively impressive if you can just point to any code you have written

01:41:23   whether it's like here's a here's a github project for a simple app I made you will stand out from that

01:41:30   and again it doesn't have to be anything incredibly impressive or complicated just if you can point to a github

01:41:36   repo that you made and say I completed an app that does a simple thing I guarantee you almost none of the

01:41:44   competing candidates for that job will be able to point to any code they wrote that works that kind of

01:41:51   thing makes you stand out more than you think and if you go into the interview and you show that you care

01:41:55   that you're into programming and that you you are interested in this company that you know do some

01:42:00   research about the company you're going into and and figure out what's important to them and you can

01:42:05   you can show them like hey this this problem that it looks like you're working on this sounds really

01:42:09   cool and I hear some thoughts I might have on it like show that and and you'll be fine I feel like

01:42:14   you're overestimating the expected return on working for lots of companies in terms of equity because I think

01:42:19   the answer is it's basically zero like it rounds very quickly to zero because it's just the odds are

01:42:25   so low and you'd have to have so many different jobs to try to increase those odds and having a lot

01:42:31   of jobs itself exhausting I mean even if only just dealing with the the the 401ks that you have to roll

01:42:37   over into each other and dealing with that hassle like I can tell you that all the companies that I had

01:42:41   equity in in my startup years netted me their losses or zero dollars the only company that gave me

01:42:47   any equity that was ever worth anything was my very last job before I quit to do this podcast

01:42:52   and that was a job that I got when I was in my 40s for a very established very big non-startup company

01:42:57   that you know and my equity wasn't worth that much but it was worth more than zero or more than negative

01:43:02   so I think that's the expected experience not saying you shouldn't do it like by all means do it but like

01:43:07   it can be exhausting to try to like I got to put lots of bets down because then you're just having a lot

01:43:11   of jobs and on the flip side of this on a non-startup side that's why I preface this by saying if

01:43:15   you're interested in this if you're not interested in that world and don't want to be it at all and

01:43:18   just want to get a good established job you can do that and it's fine and like I would actually

01:43:23   recommend that to a lot of people it's what I hope my children do I hope they don't go into the startup

01:43:27   world I hope they get a job with a big established company but if you do that my advice to someone like

01:43:32   applying to Apple or Amazon or Google or something is what your goal should be when getting that job

01:43:39   especially if you're just out of school is you need to learn how to be a worker essentially learn

01:43:45   how to be an employee learn how to be on a team you're going to be learning how to to interact

01:43:51   with people and work as a team to accomplish a goal and this sounds dumb and you think you did it all

01:43:55   you have we had group projects in school it's exactly the same thing it's not that's the main skill that

01:44:00   you should concentrate on learning I know you're thinking about all you know what kind of programming

01:44:04   things we need to know to get past the interview and that's all true and everything and like what am I

01:44:07   going to do in my job or whatever but the main thing you're going to do in your job I almost

01:44:10   guarantee for the first five to ten years is learn how to be an employee learn how to be a co-worker

01:44:16   learn how to work in a team learn how to be successful in that environment success defined as learning how

01:44:23   to accomplish the goals set before you as a team which sounds so dumb and simple but it is literally

01:44:28   the hardest thing and they don't teach you that in college and they don't teach you that in high school

01:44:32   really they try to give you good projects they try to have that one course about teaching you that like

01:44:36   agile methodology or something that's not what the it's not it right it is a social skill that is

01:44:42   integral to being successful in big companies with lots of people working as a team in a startup you

01:44:48   can get away with being a you know just a renegade lone wolf and doing your own thing and having 50

01:44:53   jobs and everything's all crazy anyway whatever that's the environment where you can get away with

01:44:57   that that's the environment where you can you know flourish without those skills but you go work for a big

01:45:03   company no one wants that person there they need someone who can work as a team together on you know

01:45:10   on goals that are agreed upon like it's a totally different environment and as you get older you're

01:45:15   not going to want to work in those chaotic startup type things you're going to want to work for one of

01:45:19   those wealthy companies that can afford to give you a good salary and good vacation and good benefits

01:45:23   and a stable job and everything and those companies require you to be a good employee and worker

01:45:29   and again in all these cases is the main thing i'm gonna hammer home to all my children once they get

01:45:34   off into the real job market do not let the job exploit exploit you don't let it burn you out there's

01:45:38   all sorts of things you have to be wary of to be because your employer is not there to take care of you

01:45:42   or to love you or to make sure your life is great so you have to be on lookout for this all the time

01:45:46   but also you have to learn how to be a worker in these big companies and so

01:45:49   i mean the ideal thing is if you really really want to do startups do them when you're young

01:45:53   and then transition over but when you make that transition to over to the big companies you're still

01:45:57   going to have to learn all that stuff about how to work as a team how to work together how to work in a

01:46:01   more professional environment um and i've seen a lot of people have very long careers and really struggle

01:46:08   with either with that transition to a big company or maybe they start off in big companies and just never

01:46:12   learned how to work with people and they don't succeed as much so there you go that's a new computer

01:46:18   science graduate employee advice 2025 one more rebuttal of john's you know being down on uh on getting a bunch of

01:46:26   um when i say place a bunch of bets i'm not saying change jobs every six months i'm saying change jobs

01:46:31   every few years like the way most people in tech do that's not enough bets then well to change the odds

01:46:37   from zero well but also it matters obviously what bets you take it you know you can't necessarily know

01:46:44   ahead of time what will pay off but you can be in areas that are more or less likely to pay off

01:46:50   typically um you know the way it has gone for the last you know 30 years in tech is if you want a

01:46:56   higher chance of building you know meaningful wealth from stock shares and stuff in early companies

01:47:03   the earlier you're in the company the better the bigger role you have the better obviously but also

01:47:10   what area of the industry the company is in and what phase that's in in its maturity matters a lot

01:47:16   you want to be in a part of the industry that is young and experiencing a whole bunch of growth and

01:47:22   change and flux so you know obviously like you know when i started with tumblr in 2006 that was

01:47:28   consumer web apps like that was very much still a big thing then you know later on like you know

01:47:33   iphone apps mobile apps have come in now that area is ai if you want to maximize your odds try to get

01:47:40   into an ai startup that's where all the action is if you're listening to this for some reason

01:47:44   five or ten years from now that answer might be different and you know you got to look around like

01:47:49   where is the action in the industry right now what is like new and growing and changing and lots of flux

01:47:54   go there and get some bets placed on that table and your odds are not not too bad that like you know

01:48:01   but one of those could work out pretty well instead of zero the point zero zero zero zero one i think you'd be

01:48:07   surprised and again it's but and you know place a few bets and see what happens and the whole time

01:48:11   get salaries anyway and uh and you know they're you know then you aren't you know you aren't out any

01:48:17   money you know so to speak but it it's not a bad strategy and again like those people will all then

01:48:23   promote you to the other startups in that community so you know if you if you show that you're a great

01:48:28   programmer uh or even just a good worker who like does their job and you know and is not like causing

01:48:33   problems for for people two or three years from now you're looking for a job and your old boss might

01:48:38   be a might be a good referral to tell you know tell some new company that they're friends with like oh

01:48:42   yeah i know i know this person they're a really good programmer that works in big companies too by the

01:48:45   way that's just the general advice on and when you have any kind of job is your next job will probably

01:48:50   come from somebody you knew at your first job or like one or two degrees separated so even if you go

01:48:54   work for a big company right out of school it's another reason to learn how to be a good worker is

01:48:58   because some person that you work with is going to be at some new company and that's the way to get

01:49:02   your foot in the door there you decide you don't like this company you want to go to a different one

01:49:05   you get laid off whatever you're looking for another job you need that leg up it's it's it's who you

01:49:10   know um and it's and it's what they think of you if they know you but hated your guts probably not

01:49:14   going to hire you yeah maybe don't ask them for a reference all right also authority is taken not

01:49:18   given all right thank you very much for listening everybody uh our sponsors this episode were

01:49:22   squarespace and delete me and you can of course sponsor us directly by becoming a member thank

01:49:28   you to our members as well one of the many perks of membership is atp overtime our weekly bonus topic

01:49:34   this week in overtime we're gonna be talking about apple's newest app store changes for the eu

01:49:39   if these are even still the current ones by the time this episode is released in a few days we'll find

01:49:45   out um you can hear us talk about that in overtime and many other member perks by joining us at

01:49:49   atp.fm slash join thanks everybody and we'll talk to you next week

01:49:54   now the show is over they didn't even mean to begin because it was accidental

01:50:03   oh it was accidental

01:50:06   john didn't do any research marco and casey wouldn't let him because it was accidental

01:50:14   and you can find the show notes at atp.fm and if you're into mastodon you can follow them at c-a-s-e-y-l-i-s-s so

01:50:32   That's Casey Liss, M-A-R-C-O-A-R-M-N-T-M-A-R-C-O-R-M-N-T-M-A-R-C-O-R-M-N-S-I-R-A-C-U-S-A-C-R-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C-U-S-A-C

01:51:02   started paying for youtube so that his family won't see ads on youtube which is the thing that

01:51:06   both marco and i recommended as a no-brainer thing that you should do because kids watch so much

01:51:11   youtube and we don't want our kids to see ads and let's stop that from happening by paying a small

01:51:16   amount of money per month it's well worth it to protect your children from ads casey has already

01:51:20   exposed his children to too many ads so he's better better now than a better light than never

01:51:24   but then during some conversation that you have like a mastodon i heard you say that you were you

01:51:30   mentioned on the show you're like oh i started paying for like youtube light or something i

01:51:33   don't know what the name of the product so i'm like well whatever he found the whatever thing

01:51:36   that he could pay for that was uh gave him what he wanted and nothing else but someone was someone

01:51:41   saying like how do you get your maybe your kids are still seeing ads maybe that's why declan didn't

01:51:46   mention to you that he that he thanks you for getting rid of ads because he's still seeing ads

01:51:49   because the thing you paid for is not a family plan but only allows one person one account to not see

01:51:56   ads and your answer to this person was oh i just let declan use my account in youtube is this all

01:52:02   correct so far yeah yeah yeah yeah i'm here to tell you not to do that okay why not i mean lots of

01:52:09   reasons first of all you're being cheap but let's set that aside yes second of all your son is logged

01:52:16   into your google account right now well no this is only on the shared apple tv but yes i mean strictly

01:52:22   speaking that is correct that's not good he can read your gmail he can delete your google account

01:52:27   he can read everything in your google drive he can't accomplish any of that on the shared apple tv

01:52:32   can't accomplish kids are very clever second of all setting all of that aside do you want your son's

01:52:41   videos and your video history mingling with each other i have long since given up on that and the

01:52:47   recommendations affecting each other oh yeah my recommendations are trash i can tell you that

01:52:51   pretty soon declan's not going to want you to see what he's looking at well that's also fair and then

01:52:56   i'll have a different problem and then i'll probably have to spring for the family plan but for now

01:53:00   i didn't want to pay for it at all but it became unbearable so i wanted to pay for the crappiest

01:53:05   cheapest thing i could to get away with it and that's youtube premium light which is eight bucks a month

01:53:11   and it gives you um you know an ad free experience with a couple of asterisks i will say that if you

01:53:17   read because somebody else that reached out on maston was like or maybe it was email and said hey

01:53:21   does this actually work because if you read their documentation about youtube premium light it sounds

01:53:27   as though you're getting one fewer ad as opposed to no ads and uh i think the the real brass tax of it is

01:53:35   is that for watching music videos or music things i believe i still do get subjected to ads but for

01:53:42   pretty much anything else i don't or at least that seems to be what's going on so first of all you need

01:53:46   to pay for the better one second of all you need to not let your son be logged into your google account

01:53:50   is the keys to everything how long do you think it'll be before he figures out exactly how much power he

01:53:54   has now that he can get like password reset emails to your account like this is this is just i'm not

01:54:00   saying he's a bad person i'm saying kids and eventually teenagers this is an irresistible

01:54:05   amount of power that you have given him and it's a bad idea i guess also he doesn't want all of your

01:54:12   videos in his browsing history and recommendations yeah he's not going to want that pretty soon either

01:54:16   i mean yeah that that time will come but it's not here yet and so i'll cross it when i get there

01:54:20   don't let your kids log into your google account like it's your main account for all like your stuff

01:54:25   well like it's because logged into youtube means logged into gmail which logs into google drive

01:54:30   being logged he he could be in our show notes right now deleting everything

01:54:33   that is true um well that's true asterisk again the only mechanism he has to do this is the apple tv

01:54:41   and there's only but so much damage he can do on the apple tv you have you seen you you already haven't

01:54:45   you detailed how kids get past all the things like marco has so kids get past all the thing on the

01:54:49   chromebooks kids will find a way life will find a way casey they're like the dinosaurs in jurassic park

01:54:54   you cannot give them this access i i understood the reference god damn it uh no i i i understand

01:54:59   what you're saying also consider though that you are all in on google stuff and i am pretty much all

01:55:04   out i do have a gmail account that i don't use isn't that where your casey list stuff is all no no

01:55:09   remember i'm on fastmail baby oh wow right you have some protection but still it just it's this is not a

01:55:15   good idea pay the extra seven dollars a month like give your whole family ad free youtube on their own

01:55:20   accounts but there are no other accounts but mine at the moment there will come a time there will be

01:55:25   there rapidly will be declan should probably already have one someone asked about this like oh they don't

01:55:29   let you have a google account unless you're x years old or whatever if you have to lie about their age to

01:55:33   give them an account you probably should but i think they're a kid account and if you can tell the truth

01:55:37   about their age do so and then eventually their kid accounts will transition to adult accounts yeah my uh

01:55:42   my son recently turned 23 according to google because he he uh he has a like because you know

01:55:51   we had the same idea like when we first i got the youtube premium thing for the family you know years

01:55:55   ago and i created the account for him with his actual birthday and actual age and then it's like oh

01:56:01   he can't do anything on youtube because he's below 13 and so now he's actually legally above 13 but

01:56:08   for for all those intervening years he like you know if i wanted him to have an ad for youtube

01:56:13   experience i had to create a second child who didn't exist who was above 13 and have him log into

01:56:19   that account it's it was a mess but i'm sure everybody does that yeah i mean i'm not necessarily

01:56:25   arguing with anything that you're saying but i think the the priorities that your family have both your

01:56:32   families have i suspect my family will get there but we aren't there today and when we get there

01:56:37   i'll reevaluate but for now it's fine so when he's watching youtube on his ipad he's seeing it he doesn't

01:56:42   have an ipad well he doesn't have any kind of computing device doesn't have a phone doesn't have an ipad

01:56:46   he has he currently he has i forget which one it is but he has an old hand-me-down iphone for me that

01:56:53   doesn't have any service so it only works where there's wi-fi which for him basically means it only

01:56:58   works in the house and he doesn't watch youtube on that it's not installed on that it will come

01:57:02   very soon oh i'm sure it will i again i'm not trying to say that you're wrong or that that that

01:57:09   this isn't my future but it's just not something that happens now and honestly our house today again

01:57:15   i know this will change i absolutely understand this will change but today i have a handful of channels i

01:57:22   watch on youtube but not many and that's about it i don't ever typically just like let the algorithm

01:57:26   wash over me i don't watch very much youtube declan will occasionally put on uh preston plays which is

01:57:33   a minecraft youtuber that he absolutely adores oh yeah we we had a big preston phase over here

01:57:38   uh but i mean he seems fairly innocuous um yeah it was fine does he shout all the time i mean i don't

01:57:46   know but suffice to say that he'll you know dual screen playing minecraft on the switch as he's watching

01:57:51   preston sort of in the background and also sometimes watching whatever uh michaela is

01:57:55   doing on her um fire tablet thing um but uh right now erin never watches youtube pretty much ever so

01:58:04   again it's it's not a priority for us today i cannot stress enough because i can hear the feedback emails

01:58:11   flooding in i understand this is my future i get that but it's not my present well don't make this like

01:58:18   the glass of water repositioning episode i i understand what you're saying it was a warning

01:58:24   that was not relevant at the time but became relevant so very very similar

01:58:27   Beep, beep, beep.