Up next in 10
This week, with special guest Christina Warren: Actual real-life renders of iOS 27, a big fix coming for AirPods users, Jony Ive’s bizarre Ferrari Luce design… and that’s about it.
Produced by Extra Ordinary for Cult of Mac
Music composed by Will Davenport, arranged by D. Griffin Jones
Chapters:
0:00 - Intro
3:09 - iOS 27 renders
34:36 - iOS 27 AirPods UI
46:12 - Ferrari Luce design revealed
1:01:52 - More on GitHub and vibe coding
1:15:32 - Aftershow
Follow us!
Threads: https://www.threads.net/@cultofmac
Mastodon: https://mastodon.social/@cultofmac
Instagram: https://instagram.com/cultofmac/
X: https://x.com/cultofmac
Facebook: https://www.facebook.com/cultofmac
Show More Show Less View Video Transcript
0:00
Coming up, actual real life leaked
0:04
renders of iOS 27, a big fix coming for
0:07
AirPods users, Johnny Ives bizarre
0:10
Ferrari luche design, and that's about
0:13
it.
0:16
Good evening and welcome to the Cult of
0:18
Mac podcast. I'm your host today, D.
0:20
Griffin Jones. Uh, last week, Lander was
0:23
out because one of his many children was
0:24
graduating some level of school. We
0:26
still haven't figured out what. I don't
0:28
know why he's not here today. He didn't
0:30
tell me. But uh don't worry, we've still
0:32
got Lewis Wallace. Hello, Lewis.
0:34
>> Hey, it's great to be here even if
0:36
Lander can't be.
0:37
>> Mhm. It's great to be here because
0:39
Lander.
0:40
>> Oh, I didn't say that. I did not say
0:42
that.
0:44
>> Also joining us returning, we have
0:46
Christina Warren. Welcome.
0:47
>> Hey, thanks so much for having me back.
0:50
>> I think you're first uh repeat guest.
0:52
>> Yes. I love that. That makes me feel
0:54
great. I love that. I love that. Uh my
0:57
mom listens to the show now and uh she
0:59
she she really liked you last time.
1:01
She'll be she'll be glad to hear that
1:02
you're back.
1:02
>> Oh, awesome. Well, well, hi. Hi,
1:04
Griffin's mom. Great. Great. Uh thanks
1:06
for liking me.
1:08
>> Every a few times I I go over there,
1:09
she's like, "Hey, when are you going to
1:10
have that chick on again?"
1:12
>> She really?
1:14
>> Wow.
1:15
>> Come on. It's that chick. Her name is
1:17
Film Girl.
1:18
>> Exactly. Exactly. It's is that film
1:20
girl. No. Exactly. No.
1:22
So, this week's episode of the podcast
1:25
is brought to you.
1:27
>> Oh, no.
1:28
>> By the number five.
1:31
>> You need to zoom out. There we go. Okay.
1:33
>> Five.
1:34
>> Uh, I don't know why. I I ordered some
1:36
uh furniture a while ago, like one of
1:38
those long, skinny, tall tables that
1:40
like goes behind a sofa. And uh you
1:42
know, I went out on the front porch, saw
1:44
the package there, I was like, "Okay,
1:45
that's got to be it." Uh, and then there
1:47
was another envelope, like a a bubble
1:49
sort of envelope, really skinny, from
1:51
the same like return address. Uh, and I
1:55
wasn't really sure what that was. Maybe
1:56
they shipped it into like two separate
1:57
packages. I opened up the second one and
1:59
it is a uh gold foil balloon of the
2:04
number five.
2:07
I'm not entirely sure why. Uh, so this
2:09
week's episode is brought to you by the
2:10
number five. Uh,
2:11
>> that is bizarre.
2:13
>> That's bizarre.
2:13
>> Mhm. Somebody's
2:15
>> It's not your fifth birthday, right?
2:16
>> This is what I was going to say. I was
2:18
going to say my my nephew just turned
2:19
five actually. Yeah. And a few weeks ago
2:22
and I was like maybe maybe there's a
2:24
mixup, right? Maybe something was
2:25
supposed to go to like a fifth birthday
2:27
house and like you got the balloon
2:28
instead. But you know what? It's a good
2:29
balloon. And uh
2:30
>> it is. Yeah. It'll be really good next
2:32
year which will be my my fifth
2:34
anniversary at Cult. But uh Oh, nice.
2:36
>> Today it's just fine. It's well beyond
2:38
the fifth episode of the podcast uh as
2:40
well. But here we are. Did it have your
2:42
name on the package? For sure. I mean,
2:45
>> it was addressed and and spelled to like
2:48
the same address that, you know, the the
2:50
the table arrived at. Same return
2:52
address spelled exactly the same way.
2:54
Maybe this outfit sells both uh
2:57
furniture and uh gold foil balloons.
3:00
>> I wouldn't pop it, man. Might be
3:01
anthrax.
3:03
>> Oh, yeah. I'll keep it inside. Lewis,
3:06
why don't you tell us about iOS 27?
3:09
>> What a thing to wake up to this morning.
3:11
Holy moly. Uh Bloomberg's out with this
3:14
big post about, you know, the redesign
3:17
of iOS 27 and it's got I I I don't ever
3:21
remember seeing anything this uh you
3:23
know, extensive from Bloomberg. They've
3:24
got all of these um renders showing
3:28
basically what Apple's going to show off
3:30
in a week, a week and change. Uh big
3:34
changes coming to Siri. A lot of this is
3:36
based on reporting that we've already,
3:37
you know, read and talked about at
3:40
infant item or however in the world you
3:41
pronounce that. Uh but this shows how
3:46
it's going to look. It, you know, Siri
3:49
popping out of the dynamic island and
3:52
which is apparently going to be the
3:53
hidey-hole where Siri lives in iOS 27.
3:58
uh the implementation of the Siri
4:02
uh visual intelligence stuff inside the
4:04
camera app. All these really and they're
4:06
actually I mean honestly they're they're
4:08
gorgeous looking renderings.
4:10
>> They did a very good job. Yeah,
4:12
>> it's very impressive. But it's it's a
4:14
ton of it. You know the it shows what
4:16
that uh dynamic island is going to look
4:18
like. It kind of expands out. Although
4:21
it's got that weird liquid glass
4:23
warping, which I know I'm already tired
4:26
of that. Uh, you know, the the search or
4:31
ask box, which has been reported before,
4:33
shows how that works. Uh, it's just it's
4:35
really it's kind of shocking how how
4:38
in-depth these things are and how it
4:40
really brings to life all these rumors
4:41
we've heard. It it's a pretty big
4:44
change, you know. Uh, and as we've been
4:46
discussing for weeks, months, etc., Um
4:49
AI is just like the main thing in iOS
4:52
27. You know, this redesigned Siri
4:54
that's, you know, going to work,
4:57
absolutely going to work work
4:59
beautifully. Um
5:02
there's like all these different ways
5:03
that they're implemented. You know, the
5:05
like a Siri app that works like a
5:07
chatbot. A lot of really rich looking um
5:11
images and stuff. Like I saw there's one
5:12
like a search for a basketball player
5:14
shows a a picture of him which is you
5:16
know this looks nice. Um tons of AI
5:21
powered information that you can get you
5:22
know doing searches in Siri and uh I
5:26
don't know. Have you guys had a chance
5:28
to look at it yet?
5:29
>> Yeah. Yeah. I mean like it's always hard
5:31
to tell with these things when it's a
5:33
render. Um even though and in this case
5:35
I'm I'm assuming and we we talked about
5:37
this a little bit pre-show. Looks like
5:38
the renders were done by by Bloomberg's
5:40
in-house team and and I don't know, you
5:43
know, um which means that, you know,
5:45
they were able to do it to whatever
5:46
specifications
5:48
um Mark and I guess the other editors
5:50
wanted, but it's always hard to kind of
5:52
predict based on on a render if you
5:54
know, okay, what is this actually going
5:57
to look like? Is this really going to be
5:58
what we're going to see or is this, you
6:00
know, um the the um outside artists
6:02
interpretation based on on the facts
6:04
that that they've heard? Um, but just
6:07
what what we're seeing, I mean, you
6:08
know, it makes logical sense. It looks
6:10
really good. And I I I like the more I
6:13
think about it, and I know this was
6:14
reported before, I do like the idea of
6:16
using uh the dynamic island as kind of
6:20
the the place to access Siri, assuming
6:23
Siri's actually going to be useful this
6:25
time. I actually do like that as uh kind
6:28
of an always on kind of place, right?
6:30
Like I think that that is a good way of
6:32
not having to take up a button, which is
6:35
kind of what's happened in the past. You
6:36
know, people have had, you know, oh,
6:38
we'll make the action button Siri. No,
6:39
but people want that for other types of
6:41
things. Um, and and I think that
6:44
especially if you're going to be able
6:46
to, you know, have some of the visual
6:48
intelligence and other aspects built
6:50
into it. I like that. I I think that
6:52
could be kind of a good UI paradigm to
6:54
be like, okay, if I don't have another
6:57
thing that's using my dynamic island at
6:59
this time, I can just tap on this space
7:01
and and pull up Siri and then interact
7:03
with it and either ask a question that
7:05
might be unrelated to what's on my
7:06
screen or maybe it is related to what
7:08
I'm doing and what what I'm in the flow
7:09
with. I like that a lot. So, I I like
7:12
kind of that perspective and and I think
7:13
that um yeah, I mean, these renders are
7:15
are very are very good. good. I mean, it
7:18
looks at least from what the renders are
7:20
showing like a much more refined kind
7:22
of, you know, iOS 26 style, which is
7:25
obviously what we've been hearing for
7:27
the last few months. That's what they're
7:28
going to be doing with the overhaul with
7:29
iOS 27 is is it's not going to be a
7:33
complete rethinking of anything, but
7:35
just more refinements of what we've had
7:37
before. But I think it looks good and
7:39
and I I'm I'm excited to see how they
7:42
will integrate these features. Again,
7:44
the big caveat is will it actually be
7:46
good this time into the other
7:48
applications and into the other, you
7:49
know, parts of the OS because that's the
7:51
thing that if this can actually be
7:53
useful, I think that a lot of people
7:55
would really enjoy this because that
7:57
that gives you the sort of integration
7:59
that you don't have with existing AI
8:01
apps right now. um no matter what, you
8:04
know, access they can have, it's not
8:05
going to be the same as being kind of a
8:07
system level integration where I can
8:08
pull it up anywhere anytime and I can
8:10
have it built into the photos app and I
8:12
can have it built into to notes and I
8:14
can have it built into to other aspects.
8:16
Um and and I I think that they if
8:18
they're able to pull this off, this
8:20
approach strikes me as being better than
8:22
just having the the Apple intelligence
8:24
kind of button right now that gets in
8:26
your way that you accidentally invoke
8:28
more often than not and you're like,
8:29
"Please, please go away. I don't
8:31
actually want you to summarize anything
8:32
for me because you'll take five minutes
8:34
and then be terrible. Um, so you know,
8:37
>> I almost wonder if they actually mocked
8:40
it up as like a little Swifty demo, a
8:42
demo app because I mean the people have
8:45
always done like concept renders of, you
8:47
know, this is what I dream iOS 16 will
8:49
be like. And they always don't look very
8:51
good. They even the best ones like they
8:53
get a bunch of like little appley design
8:55
details wrong because they're just
8:57
trying to build it from scratch in like
8:58
you know Adobe After Effects and and
9:00
that only gets harder with like with
9:01
glass because it's no longer just like a
9:03
plane of translucent glass. It's got to
9:05
have like the lensing effects and all
9:07
that. But I really wonder if they like
9:09
mocked these up in Swift UI because
9:11
there it goes beyond like the the few
9:13
images that we can uh show in the
9:14
article. Like if you click if you have
9:16
access to Bloomberg and you can look at
9:17
the full story like they've got the a
9:19
mocked up like Siri app interface that
9:21
shows what that's like and it's got like
9:23
you know it looks like liquid glass. I
9:25
mean the other part of this is that now
9:27
now that we've heard this rumor for a
9:28
while, we can really sink into the fact
9:29
that like having it in the dynamic
9:31
island and also merging it with
9:32
spotlight does two things. I mean, it it
9:35
gives the iPhone universal access to
9:36
Spotlight, which it has never had
9:38
before. You always have to, you know,
9:40
swipe up, swipe up, swipe up like
9:41
sometimes three times to get to your
9:42
home screen and then swipe down to get
9:44
Spotlight. But having it everywhere on
9:46
the iPhone would actually be excellent.
9:49
It's with with iPhone mirroring on the
9:51
Mac.
9:52
>> You can access Spotlight. I think it has
9:54
like the command three shortcut for it.
9:57
>> It's like, why can't you get that on my
9:59
iPhone? As soon as I played around with
10:00
that, I was like, this would be really
10:01
great to just have on my iPhone. Yeah.
10:03
No, I fully agree because that is one of
10:05
the and because it's annoying. You know,
10:06
you kind of have kind of the spotlight
10:08
like feature in iOS, but it doesn't work
10:10
the same way as it does on a Mac and
10:11
it's inconsistent and sometimes it will
10:14
actually find things otherwise. So,
10:15
yeah, I'm with you. I think I would love
10:16
to have, you know, that kind of
10:18
functionality that you get the on the
10:20
iPhone mirroring.
10:21
>> And on the iPad, you have Spotlight, but
10:23
only if you have an external keyboard.
10:24
It doesn't have a gesture on on its own.
10:26
So, it sounds like, you know, the
10:27
gesture is going to I mean, the top of
10:29
the screen is going to be a very busy
10:30
space now, at least on a non-folding
10:32
iPhone. On a portrait iPhone, like you
10:34
can swipe down from the left for
10:35
notification center, the middle for
10:37
spotlight/ Siri, and now the right for
10:39
control center. That's a that's a busy
10:42
space.
10:43
>> Yeah, that that's a good point. That is
10:44
a busy space. And I do wonder I mean I
10:46
think this will be the interesting thing
10:47
is is how they will deal with I guess
10:49
like conflicting dynamic island things
10:51
because already right now um if you have
10:53
something like now playing but if you
10:55
have another thing like there are two
10:56
things you can kind of see it once and
10:58
you can kind of swap between them but
11:00
usually once you've dismissed one of the
11:02
app intents or whatever it's called from
11:04
the dynamic island it's gone right like
11:06
you've got to kind of like pull it back
11:07
up again. So, I'm thinking like if you
11:09
have I run into this when I travel
11:11
sometimes where I'll have two competing
11:12
apps. Um like my my my uh like the Delta
11:15
app and FlightD will both be trying to
11:18
tell me when I'm landing and um and you
11:21
know and you can kind of get around that
11:23
a little bit. I am curious if you're
11:24
going to be kind of reserving some of
11:27
this space for Siri. Maybe it is just
11:29
the fact that it's a tap gesture. Maybe
11:30
that's all it is is just if you tap in a
11:32
certain area, it pulls up Siri and it'll
11:35
handle the dynamic island otherwise the
11:37
same way that it has. Um, but I that'll
11:40
be the only interesting thing. But you
11:42
you bring up a good point. It's it'll be
11:44
it's you now have like three different
11:45
touch points at the top of the screen.
11:47
You know, one for, you know, kind of
11:49
notification center, one for, you know,
11:51
uh, Siri or for uh, you know, um, uh,
11:54
Spotlight, and then another for um,
11:57
control center.
11:58
>> I mean, that's a good point as well. You
11:59
can have up to two things in the dynamic
12:00
island at once. Yeah.
12:02
>> So then it becomes five different tap
12:05
areas in a very narrow space. I mean
12:07
>> I we were also talking about this like
12:09
in our in our Slack. Like how is this
12:10
going to work on the iPad? Like maybe
12:12
you just have to drag down from an
12:15
invisible area at the top. I mean on the
12:18
iPad there's also like the
12:19
multi-windowing mode where dragging from
12:20
the top of the screen moves windows
12:22
around. So that conflicts as well like
12:25
>> but you know they also have to have it
12:27
because I mean when you have a folding
12:29
iPhone and you have it unfolded to the
12:31
center and there's no cutout in the top
12:33
middle of the screen. I mean the gesture
12:35
has to work the same there.
12:36
>> Yeah.
12:37
>> At least it'll be really easy to find
12:38
the center of the screen because it'll
12:40
have like a big crease going down a
12:41
little bit. But
12:43
>> uh German's post today uh actually says
12:47
there's two starting points for Siri.
12:48
One is, you know, classic speaking the
12:51
the word. And the second method is
12:53
entirely new. Quote, "Apple plans to let
12:56
users swipe down from the top center of
12:58
the iPhone anywhere in the system to
13:00
launch a new search or ask interface."
13:01
>> And then you swipe down even further and
13:04
then it opens the Siri app, which is a a
13:07
clever way to do it. I mean,
13:08
>> yeah.
13:08
>> I mean, is it going to be an app on your
13:10
home screen, do we think? Is it going to
13:11
have an icon?
13:13
>> That's what I assumed before I read
13:14
this.
13:15
>> Sounds like it.
13:16
>> I think so. I mean, I I think that they
13:18
kind of h I mean, I don't know. I guess
13:19
they could go one of two ways. I would
13:21
like them to do that just because I
13:23
think that it would be nice to have a
13:25
dedicated app where you can have all of
13:27
your, you know, conversation history and
13:29
other stuff and just go into it. But I
13:32
would also like it to be able to be
13:33
invoked, whether it's through voice or
13:35
through a gesture anywhere you are,
13:38
right? I I I would I would like both.
13:39
And and I I you know um I wouldn't
13:43
necessarily want like and I think this
13:45
would take you out of it to be honest if
13:47
I'm in another application. I don't want
13:49
to be ported from my browser or what I'm
13:52
writing or my email or anything else
13:54
into the Siri app just because I've
13:55
invoked that. Right? Like I think that's
13:57
a bad experience. I think instead I
13:58
might just want to bring it up and say
14:00
okay I'm going to ask a question about
14:01
this and it can do it in the background
14:03
and maybe alert me when it has um
14:05
something that's done right. Um, or you
14:08
know, if I'm using a foldable phone,
14:10
maybe I have a separate window open,
14:11
right? Like that that would be that that
14:13
that sure would be would be great.
14:15
>> I mean, there are there are like a lot
14:16
of hidden apps. Like visual intelligence
14:18
is an app, but it doesn't have an icon.
14:20
>> Well, I was going to say I was going to
14:21
say like, yeah,
14:22
>> the magnifier was like that for a while.
14:24
>> Um, I can see a situation where Apple
14:27
thinks, well, you can always open it
14:29
just by swiping down and then swiping
14:30
down further. And I can also foresee
14:33
putting an icon on the home screen might
14:35
send the wrong message. were like, "We
14:37
don't want people to think that you have
14:38
to tap this icon on the home screen,
14:39
open it if we want to train you to swipe
14:42
down."
14:43
>> If we want you to Yeah, that's a really
14:44
good point. And I don't know how you
14:45
handle that because on the one hand,
14:46
you're right. You don't want people to
14:47
think this is the only way you interact
14:49
with it. On the other hand, you have
14:51
people who have four years of builtup
14:53
muscle memory of opening up the Chat GPT
14:55
app or the Claw app or the Gemini app or
14:57
Perplexity or anything else. And so
15:00
there's
15:00
>> Yeah, but Apple doesn't want to be like
15:02
those people.
15:02
>> Oh, I know. Apple.
15:03
>> Well, they want to be Apple. But but but
15:05
what I'm saying is you have like users
15:06
who it's like this is not the first time
15:08
they've used you know ver an AI
15:11
assistant like that that's that's the
15:13
thing. I mean, if if Apple wants to come
15:15
at this pretending that this is all
15:16
brand new, net new information for its
15:19
user base, I think they're going to not
15:21
do well, right? Like two years ago, they
15:24
could pretend like everything they were
15:25
announcing was brand new and and magical
15:27
and no one else had done before. And
15:29
then it turned out, well, a
15:31
>> you didn't you didn't land it. And and
15:33
B, what
15:34
>> that was another two years ago.
15:35
>> That was another two years ago. And two
15:36
years ago, that wasn't even an accurate
15:38
statement, right? two years ago. It it
15:40
to me like my my my big critique of WWC
15:43
of 2024 is I was like I I work in AI. I
15:46
know what's happening in the space.
15:48
What's being talked about and announced
15:50
does not align with my reality of what
15:52
I'm seeing and what I'm hearing and
15:54
what's happening here. Um putting any of
15:56
the, you know, claims they made that now
15:58
they're going to have to, you know, they
16:00
settled class action over aside like it
16:02
just the the messaging didn't align.
16:05
Now, I feel like again to your point,
16:07
like it's two years later, you're even
16:09
I'm not going to say further behind
16:10
because I think that they weirdly timed
16:12
this in a in a really good way um for
16:14
for themselves. But I mean, people have
16:18
different I guess built-in um
16:22
expectations and use cases and muscle
16:24
memory for how they've already been
16:25
using with these apps. And so, yeah, I I
16:28
have a feeling you're probably right
16:29
that they won't make it an app on a home
16:30
screen. I feel like they will be like,
16:31
"Oh, no, we don't need this." But
16:33
there's part of me that hopes that it is
16:34
there. Just kind of like, you know,
16:36
passwords was was not an app for a long
16:38
time. It was hidden in the settings
16:39
screen and like you had to create um you
16:42
know, a shortcut or at least I did to go
16:44
directly to that section before they
16:45
released the passwords app. And and I
16:48
don't I I don't know. I just feel like
16:50
if there's going to be if there's going
16:51
to be like a level where I can handle,
16:53
you know, like my, you know, my
16:55
conversations and stuff like that, I
16:57
guess that's fine. if I either open up
16:59
the full app going all the way down or
17:01
if I go into settings to to to drop into
17:03
it. But I wouldn't mind if there was an
17:05
application. Maybe Apple doesn't put it
17:06
on the home screen. Maybe they put it in
17:07
one of the folders, you know, hidden
17:09
away like like like the measuring app or
17:11
something.
17:12
>> It's not that Apple was behind. It's
17:14
that two years ago they were two years
17:16
ahead.
17:18
>> Yeah. Yeah. Okay.
17:20
>> Hire me, Greg Jwak.
17:23
Okay. Lewis, tell us about the camera
17:24
app.
17:25
>> Yeah. So, we've heard tons about this
17:27
customizable camera app that's coming in
17:29
iOS 27. Now, we get to see what it's
17:31
probably going to look like. By the way,
17:32
I I just went and checked. I mean,
17:33
Bloomberg said this is based on
17:36
information that they have seen and
17:38
information from people familiar with
17:39
the work, which, you know,
17:43
oddly enough, those people didn't want
17:45
their names in the story. And uh also,
17:47
oddly enough, Apple refused to comment
17:49
on the whole situation. But uh yeah,
17:51
this looks pretty slick, too. you know,
17:52
like showing you actually how you'll be
17:54
able to
17:56
uh manipulate the the camera app so you
17:59
can put the the settings that you want
18:00
that you use, make them easy to find on
18:04
the screen.
18:05
>> It looks very similar to I don't know
18:07
like what is a thing called uh
18:09
>> putting widgets on the screen
18:11
>> control center stuff. Yeah, it's a
18:13
similar interface and you know if you
18:14
got an iPhone that's running iOS 26,
18:17
you're gonna say, "Oh, this makes
18:19
perfect sense." And it does make sense.
18:20
I mean, there's so many. The camera app
18:23
is so I mean, it's not as, you know,
18:26
amazing as some of the third party ones,
18:27
but it's got a lot of settings and stuff
18:30
that are important. And if you actually
18:32
care about taking pictures and you want
18:34
to do things a certain way, it's pretty
18:37
great that you can surface the the
18:40
toggles and switches and settings that
18:41
you particularly like to use. Last year
18:44
they they they cleared out all of the
18:45
buttons and now there's just a lot of
18:47
blank white space and now they're
18:48
letting you put whatever you want back
18:49
in there. So
18:51
>> it's I mean I actually I like the the
18:53
way it is now. I I prefer that over the
18:55
way it was before.
18:56
>> It's
18:57
>> Yeah, I'll probably I mean you still
18:58
have all the functionality. It's just
19:00
hidden. You have to like tap the photo
19:01
button or the video button to get to the
19:03
settings. And honestly, I might I might
19:05
keep it in that state. You know, as much
19:07
as I fiddle with those every every time
19:08
I open the camera app, I kind of like
19:10
the clean look as well. As we mentioned
19:12
before, baking in Apple intelligence. So
19:13
like now when you swipe through the the
19:16
standard things, I think right now it's
19:17
just video and photo is is the stock
19:19
thing, right? But now if you slide over
19:22
one more, maybe it'll be in the center,
19:24
whatever. I don't think it actually says
19:25
here, but there's going to be a little
19:27
Siri looking icon that uh launches Apple
19:30
Intelligence. So that I mean, the goal
19:32
there is to make that actually visible
19:34
to people so they don't have to
19:35
remember, oh, I have to press the camera
19:37
control and I have to hold it a certain
19:39
amount of time.
19:41
I I I it because that is a pretty useful
19:44
feature. It and it's it's actually I
19:46
don't know I find it kind of tricky to
19:47
invoke it. I mean a lot of times I
19:49
accidentally end up opening the camera
19:50
instead of doing it. So the whole camera
19:53
control has just been a a fussy thing.
19:56
It's like I guess shoehorning visual
20:00
intelligence into the screenshot
20:02
functionality. But I think all that has
20:03
done is annoyed people.
20:04
>> Yes. Because you because again you hit
20:06
it like when you're not trying to I'm
20:08
taking a screenshot. I'm trying to crop
20:10
it and instead now I've invoked visual
20:12
intelligence and it's now trying to edit
20:14
my screenshot for me and I'm like
20:16
>> and if you accidentally tap on the
20:17
screenshot then it'll bring it up again.
20:19
>> Exactly.
20:20
>> You don't know how to go away.
20:21
>> I'm glad I'm not the only one that's
20:22
annoyed by that.
20:23
>> Do you ever use visual intelligence?
20:25
>> No. Absolutely not. No. No. Not not
20:29
unless like like Lewis because I also
20:31
take screenshots like 50 times a day and
20:33
I invoke it accidentally and then I'm
20:35
and then I'm annoyed, right? Like this
20:36
is this is kind of my whole beef, I
20:38
guess. like I'm I'm sorry to be like a
20:40
hater, the last like two years of like
20:41
Apple intelligence is that I only use it
20:43
when it is an accident and then I get
20:45
annoyed with it which then makes me not
20:47
want to use it again. So, I'm I'm
20:49
genuinely trying to go into the new Siri
20:51
area with like a very open mind and like
20:52
I I have like you know um from you know
20:56
the reports that are coming out like I
20:58
and I know people who work at Apple I
20:59
know they're working really hard to
21:01
deliver a good experience but I have to
21:03
be honest like the last couple of years
21:04
I've just been like no this is you know
21:07
when this gets in my face it's it's not
21:09
it's not what I want at all. So no I
21:12
don't use the visual intelligence.
21:13
>> Do you feel like having it right in the
21:14
camera app would make you use it more? I
21:17
would actually and that's the thing
21:18
because because to lose this point I
21:20
remember when I tried to use it before
21:22
it was finicky and I was like okay well
21:23
this is a pain and then you just
21:24
accidentally invoke it and you're like
21:26
well now I'm annoyed but if it was
21:28
something where especially if it works
21:29
and again like I keep saying as as the
21:31
caveat but it's important if it's
21:32
actually useful and this time I think
21:34
they're actually using a frontier model
21:35
that will be good so I I I have hopes
21:37
for that then I do feel like that could
21:40
be interesting where because there are
21:42
these scenarios um this was a couple of
21:44
years ago now but there was like a a
21:46
weird um I guess a a pan or or or or or
21:52
like some sort of container um that
21:53
someone gave my my mom and I was like
21:56
what is this for? and I took a picture
21:59
of it and I gave it to Chad Gabbt and
22:00
then they gave me a result and I don't
22:02
know if it was exactly right but it was
22:03
kind of close enough and then I found
22:04
something on Reddit and it turned out it
22:06
was you know either like it was it was
22:09
just some sort of weird kind of platter
22:11
thing and I would love to have like
22:15
because you have the scenarios you're
22:16
like what type of plant is this or or
22:17
what is this object or you know um what
22:20
what um kind of does this signify it
22:23
would be much easier for me to pull up
22:25
the camera app which you already use for
22:27
QR codes and for other things to do that
22:30
than having to think about, okay, well,
22:31
now I've got to pull up like a different
22:33
application and I have to open up the
22:34
camera there and then I have to go
22:36
through this whole process. Like I I
22:38
think that yeah, if it was built in
22:40
directly to the camera, I I do feel like
22:42
that would be something that I would use
22:44
a lot more, especially if it works.
22:45
>> Especially if it works is uh carrying a
22:48
lot of weight in that sentence, but
22:49
we'll see.
22:50
>> I mean, right now, visual intelligence
22:51
is not bad. Uh I use it all the time to
22:54
identify plants. seems to be perfectly
22:57
fine for that. I don't know that I've
22:59
ever used it to like add a concert to,
23:03
you know, ticket to my calendar or
23:05
anything like that. All the things they
23:06
say.
23:06
>> I've only been to California once, but
23:08
it are there just like concert posters
23:10
on every like street corner. How often
23:13
is that the case instead of just house?
23:15
So, how would I know?
23:16
>> Okay. Well, in in in Seattle, we're kind
23:19
of famous like people literally plaster
23:21
um I guess like some of like the
23:22
telephone poles and other things with,
23:24
you know, posters for concerts and stuff
23:25
like it's it's kind of like a known
23:27
thing and sometimes they're old,
23:28
sometimes they're not. So, I guess
23:29
that's maybe a thing, but yeah, I mean I
23:31
I I don't know how how common it would
23:33
be. I mean, you see it around college
23:35
campuses and things like that where
23:36
people put, you know, signs up of
23:38
posters of of events and I guess that
23:40
could be useful um if if you're going to
23:42
use it that way. I don't know how many
23:44
people I'd be curious to know how many
23:46
people just in general like I think
23:47
that's one of those use cases that the
23:49
AI companies have kind of invented to be
23:51
like oh this is how people will use the
23:54
fact that the the models have vision
23:56
support that they'll that they'll do
23:58
this but I think it's much more likely
24:00
you know things like Lewis like I want
24:01
to identify this plant or I want to
24:03
identify you know something about like
24:05
this this weird object in my house or
24:07
maybe you know this type of bird or
24:08
something else like that that seems like
24:10
a much more likely thing just because
24:13
you've been able to do for years, even
24:15
pre this wave of Apple intelligence, you
24:18
know, there have been things built in
24:19
like, okay, if I take a a photo of
24:21
something, you know, um the OCR features
24:23
would let me search, you know, um in in
24:26
my photos and and pull up things that
24:28
might have been in that text. Um, and so
24:30
I can see that being useful. Maybe like
24:33
again like it just feels like such a
24:35
planned demo be like, "Oh, I'm gonna
24:36
take this photo and then say add this to
24:38
my calendar."
24:39
>> When I I think that the more likely
24:41
thing would be to somebody to see, oh,
24:43
this thing is on, you know, uh, June
24:45
8th. And instead, I would just like to
24:47
speak to my assistant and say, "On June
24:49
8th, add, you know, this this to my
24:52
calendar."
24:53
I think even if I did like run into a
24:55
poster and decide I want to go to this
24:57
event, I would at least like go to the
25:00
website first anyway to make sure it's
25:02
>> exactly
25:03
>> see what it's about. I guess I'm never
25:05
going to decide solely on the poster
25:07
alone. But
25:08
>> Right. Right. Exactly. And and in that
25:10
case, it's like if I'm at the website,
25:11
what I would really like would be for
25:12
the website to have support where okay,
25:14
once I've, you know, signed up or
25:16
whatever, like it's going to send a
25:17
thing to my calendar, right? or or or my
25:19
calendar is going to automatically see,
25:20
you know, the email, which already is a
25:22
feature that that that exists now, and
25:23
say, "Okay, we see that you you know,
25:25
you have this. Can I add this, you know,
25:26
to to your calendar for you?" Yeah, I'm
25:28
with you. Like, that would be I'd much
25:29
rather have the website have an easy way
25:31
to like add it to my wallet, you know,
25:33
or than than like, oh, I I see this
25:36
great poster. I'm I'm definitely going
25:38
to take this photo and and that's going
25:40
to be my method of adding it to my
25:41
calendar.
25:42
>> The problem with visual intelligence is
25:43
that it's ephemeral. Like, yeah, I
25:45
could, even if I wanted to do that, I
25:47
could snap a snapshot of the poster with
25:50
visual intelligence. But then if my
25:51
friends are still walking down the
25:53
sidewalk and I leave and then I don't
25:55
know, something weird happens and it
25:56
goes away, then I have to run back and
25:59
take a picture. Like, I'm just going to
26:00
take a picture in the room.
26:01
>> I'm just going to take the picture
26:01
anyway, right? Because the picture is
26:02
going to let me do the same thing. Um,
26:04
and and yeah, I I think I think that's
26:07
true. So, I think that where where it
26:08
would make more sense would be if you
26:12
have some sort of wearable device that
26:15
has a camera on it that is always able
26:18
to kind of capture things for that for
26:19
that ephemererality that's maybe not
26:21
your phone, right? So, I'm I'm walking
26:23
by and I'm I'm wearing a pen or maybe I
26:25
have something built into my headphones
26:26
or whatever and I'm able to just kind of
26:28
invoke from there. Okay, you know,
26:31
capture this or or add this, you know,
26:33
to my to-do list or something that I
26:35
could see maybe being useful.
26:36
>> Or maybe if you know, visual
26:39
intelligence is inside the camera, like
26:41
taking it with visual intelligence also
26:42
takes a picture at the same time.
26:44
>> Yes, absolutely. Yeah. I mean, that
26:45
that'd be a thing, too, right? Like if
26:47
it's also going to save it to your
26:48
camera roll, maybe a different part of
26:49
the camera roll, right? Maybe you don't
26:50
want it to take up everything else, but
26:52
maybe it's organized the the same way
26:53
like screenshots are or whatever. Like
26:55
you'd be like, "Okay, maybe there's a
26:56
separate section that just says, you
26:58
know, visual intelligence." I don't
26:59
know.
27:00
>> All right. So, we have we have one more
27:01
detail. Lewis,
27:02
>> the uh the grammar stuff. Yeah. Uh so,
27:05
that actually looks okay. I mean, they
27:07
have a a interface here where has has
27:10
the original thing that you wrote long
27:12
hyphen time and then the suggested thing
27:14
that they want you to change long time
27:16
no hyphen. And it thankfully it spells
27:19
out there that you you know this removes
27:21
the hyphenation
27:22
>> just in case you happen to figure that
27:24
out on your own. But, uh, you know,
27:27
okay, accept all suggestions. I guess I
27:29
guess it starts to make sense if you
27:31
have like a list of five things it wants
27:33
to change in your, you know, email or
27:35
whatever it is that it's checking. Uh,
27:37
but the way it is right now, there's
27:39
like, you know, radio buttons that you
27:41
have to tap and say, so I guess if
27:43
there's five and you say, "Okay, I want
27:44
three." And you ex, you know, accept the
27:48
one, you know, that selection of of
27:50
things that you have. So, that could be
27:52
fine. Although I got to say that
27:54
particular screen looks a lot clunkier
27:56
and less uh
27:57
>> Mhm.
27:58
>> I just I don't know. It just looks very
27:59
basic. Um
28:00
>> the the current writing tool is how it
28:01
already works which already kind of does
28:03
this but it's only like a panel on like
28:05
the lower half of the screen.
28:06
>> This fills the entire screen. So then
28:08
you lose the context of what
28:10
>> exactly you're writing,
28:11
>> right? Which which which even in their
28:13
suggestion like long dash time or long
28:15
time that's completely context dependent
28:17
on what that's going to be, right? like
28:19
one of those like I mean there are
28:21
stylistic choices there too, but they
28:23
mean different things. And so I would I
28:26
would prefer if you're going to show it
28:28
at least the way that this render does
28:30
if I had the full sentence and then you
28:32
bolded or highlighted the word that
28:35
you're wanting to change. But I would at
28:36
least like to see like the, you know,
28:38
the the full sentence so I can see the
28:40
context because for grammar especially
28:42
like that matters.
28:44
>> Yeah, that's that's a good point. I mean
28:45
it is it's it feels strangely
28:47
disconnected from the process of you
28:49
know writing and editing. Uh
28:52
so I don't know that that's the one
28:54
thing that I look at these and it stands
28:56
out as like h I don't know how much I
28:58
like that but uh
29:00
uh and I guess we should at some point
29:02
say you know obviously this is you know
29:05
based on reporting by German and leaks
29:09
and whatever and he even says in this
29:11
post obviously this could change. So,
29:13
who knows? Maybe this is just uh the the
29:16
first shot at this and maybe it'll
29:17
actually look better. I do like the way
29:19
that the camera editing tools look. You
29:21
know, the smart editing stuff which you
29:23
the AI editing stuff,
29:25
>> clean up, which already exists, but then
29:28
they're going to have extendrame.
29:33
>> You know, those might be awesome. You
29:34
know, extend the picture. You know, like
29:36
you got a picture of a uh I don't know,
29:38
a house or something, but you didn't get
29:40
the the bottom of it. you say, "Okay,
29:42
extend it to get the bottom of the
29:43
picture or reframe it so it gets a
29:46
different uh, you know, part of the the
29:48
frame." I I of all these things that to
29:51
me like I could see this being working
29:53
horribly cuz I've never been that
29:56
excited by the cleanup. If you remember
29:58
the very early days or people using it
30:00
and you know, these horrible things
30:01
happening and I don't use it hardly at
30:03
all. I and I mean I guess or actually
30:06
may like maybe never do I use it anymore
30:08
and maybe it's gotten a lot better than
30:10
it it was at the very beginning but uh
30:12
you know I have other tools for editing
30:14
images. I don't do it on my iPhone.
30:17
>> The example that they that they show
30:18
here in this in this image is a photo of
30:21
like a plant sitting on a table and a
30:23
bunch of other plants and you know those
30:26
are framed in front of like a wall with
30:27
a bunch of shadows falling on it and
30:29
they're shadows of plants so they're
30:31
very complex shapes. This would not be
30:33
easy to extend or reframe.
30:36
>> No, it wouldn't. But which is why like I
30:39
I'm I'm I think it's it's notable that
30:42
that is the example that they use. I bet
30:43
that came my guess is that that came
30:46
from reporting, you know, showing that
30:48
that would be the exact sort of thing
30:49
you'd want to show off in a demo
30:50
scenario to be like this is how good it
30:52
is is that you notice that these shadows
30:54
exist in these places and that's a
30:56
difficult thing to computationally
30:58
recreate and reframe. Um I I to to
31:01
Louis's point you know the the stuff
31:04
that is ex existed before which was
31:06
based you know on the the ondevice model
31:08
I would agree like is not good. um
31:11
Gemini which you know they're obviously
31:13
you know doing their own controls and
31:14
things on top of uh but you know that's
31:16
if that's going to be powering this
31:18
level of Siri and I don't know if it
31:20
will or won't is a really really good
31:24
large language model when it comes to to
31:26
image like it's exceptional with image
31:28
not just with creating images but with
31:29
other stuff I mean and Google has
31:31
frankly been ahead of the game on Apple
31:33
with this sort of thing for a decade or
31:35
more right and and I I I certainly don't
31:38
expect that that suddenly like Siri's
31:39
going to get all the capabilities of
31:41
Google Photos like overnight because I I
31:44
don't think that that's how that will
31:45
work. But it does give me hope that
31:48
maybe they would be able to not have it
31:51
be terrible, right? Because there are
31:53
those scenarios I think for a lot of
31:55
people if it works well where yeah,
31:57
okay, I just got the slightly off center
31:58
and I want this to be refframed or I
32:00
need to, you know, expand this a little
32:02
bit because I I missed, you know, uh
32:04
part of the the the side of this or
32:06
that. And and these are, you know, tools
32:08
like like, you know, Adobe has had like
32:10
um you know, contentaware fill for for
32:12
15 16 years. It's not like this is new
32:15
technology per se. It's just that the
32:16
the way that LLM's work mean that you
32:19
can be even more efficient and you can
32:20
do even more of it uh more quickly. Um
32:23
and potentially, you know, even account
32:26
for for shadow stuff, which would be
32:28
wild if if true. Um, but but I can see
32:31
that that being the sort of feature that
32:32
for like regular people who are just
32:35
wanting to get the best photo of their
32:37
family or, you know, their dog or
32:39
whatever, be like, "Oh, yeah, you know,
32:42
this was such a great shot, but now I've
32:44
had to, you know, go into my tools and
32:46
try to, you know, edit it and and and
32:48
frame it in such a way, but now it's
32:49
cropped in a in a different gulp
32:51
position. Oh man, if only I'd had, you
32:52
know, that that extra, you know, little
32:54
bit of space um above or below or or to
32:57
the side." To be able to do that, I
32:58
think, would be really cool.
33:00
I mean, especially when you when you
33:02
want a picture to be your lock screen.
33:03
Like the the the iPhone lock screen is a
33:06
very weird shape. Yeah. Especially
33:08
because like the upper, you know, 40 of
33:09
it is taken up by the clock. Like unless
33:12
you intentionally back up really far
33:14
away as you're taking like a picture of
33:16
somebody to leave like a lot of headroom
33:17
above the photo, which would not make
33:19
for a good composition, but that's how
33:21
you have to make a good lock screen,
33:23
>> you know? It's it's so true. And it's
33:24
the same thing like when you look at
33:25
like contact posters for people in
33:27
contacts, right? And like it'll show you
33:28
kind of like the post like it'll show
33:30
you the photos that will look good um
33:32
when when you're setting that like for
33:33
yourself or something else and it's
33:35
difficult to do that because it's just
33:37
as you said it's just a weird ratio and
33:38
so even for things like that I think it
33:41
would be great. There there was a
33:42
mention I think in the report of AI
33:44
created wallpaper which whatever I mean
33:46
that it just seems like such loweffort
33:48
stuff but the only
33:49
>> who loves image playground that much
33:50
that they want to see it every time they
33:52
get a notification or look at their
33:54
phone. No, absolutely right. But but but
33:56
but where I kind of go with that is
33:58
okay, if you took kind of maybe that
34:00
approach and you're saying, okay, maybe
34:02
there's a way where we can to your point
34:04
like take photos or think things you
34:06
want and use Apple intelligence to
34:10
um make it iPhone wallpaper ready,
34:13
right? Like that would be I'm not going
34:14
to lie, like if if you could do that in
34:16
a way that was good enough, which it's
34:18
your it's your iPhone's wallpaper. It
34:20
doesn't have to be perfect. I think a
34:21
lot of people would be really like, "Oh,
34:23
okay. This this this foot of my of my
34:24
baby now looks slightly better, you
34:26
know, than than the weird crop job that
34:28
I that I did before.
34:30
>> Yeah. Now it has three arms.
34:31
>> Yeah.
34:36
>> We've also got some details on AirPods.
34:39
No, no pictures for these. You'll just
34:41
have to use your imaginations. But, uh,
34:43
Apple continuously stuffs more and more
34:45
features into the same menu that debuted
34:47
with AirPods back in 2016. hearing aid
34:51
mode, head gestures, spatial audio,
34:52
conversation awareness. Each feature
34:54
comes with its own toggle or submen,
34:57
everybody's favorite part of a new
34:58
feature, new settings. The result is a
35:01
messy toggle filled screen that can be
35:02
overwhelming. Uh, well, Mark German said
35:06
in his newsletter that iOS 27 will
35:08
introduce a more functional, better
35:10
organized, and more streamlined
35:12
experience, making it easier for iPhone
35:14
users to toggle AirPods features on and
35:16
off. The redesign should make its way to
35:19
iPad OS and Mac OS 27 as well. You know,
35:23
they have all these extra settings for
35:24
these features, but then they also have
35:25
like the control center settings where
35:27
you can change the listening mode or you
35:29
can turn conversation awareness on or
35:31
off, but they only have like a few of
35:33
the settings there. For the other ones,
35:34
you have to go into the settings app and
35:37
that's not easy to do, you know, at the
35:40
in in the blink of an eye of, you know,
35:42
you want to suddenly change something.
35:44
So, it sounds like they'll be they'll be
35:45
redesigning that whole interface. I
35:47
mean, the pairing interface as well. I
35:49
like it was cool 10 years ago, but I
35:51
don't need to see a giant thing slide
35:53
up, take over the bottom half of my
35:55
screen that says AirPods every time I
35:57
put them in.
35:58
>> Right. No, exactly. I mean, it like like
36:00
it was 10 years ago. It was it was
36:02
frankly magical when you would open them
36:03
up for the first time and and you would
36:05
see, you know, kind of like that that
36:06
thing come up and and now like, okay,
36:08
it's not bad, but I'm like, okay, I've
36:10
seen it once. I don't necessarily
36:12
especially if you if you pair and unpair
36:13
a lot which I I I don't know how common
36:16
that is. Um I think most people probably
36:18
have their you know devices uh connected
36:21
most of the time but if you are ever in
36:22
those situations where something is
36:23
paired or unpaired a lot you're like
36:25
okay that this animation is now not not
36:27
the same thing that it used to be. Um,
36:29
the other benefit, frankly, of changing
36:31
how this looks is that it will take
36:33
longer. I don't know how long it will
36:35
take, but it will take longer for the
36:36
the counterfeits to to copy this
36:39
potentially, right? because that that's
36:42
that's kind of the problem that's that's
36:43
happened so far and this has been a
36:45
thing I guess for the last 5 years or so
36:47
is that you know people can get really
36:48
screwed buying AirPods Max especially um
36:51
because those are expensive and and they
36:53
can make really good-looking you know
36:55
fake boxes and then you you'll look at
36:57
the headphones and they'll they'll seem
36:59
good enough and you know you have them
37:01
in a parking lot behind a store that
37:03
you've met somebody from Craigslist or
37:04
or eBay off of and and that little you
37:08
know pair um animation comes up and
37:10
you're Great. This is legit. It is not
37:13
legit. You know, this is this is like a
37:15
a very bad knockoff that you've now paid
37:17
too much money for. And so, I don't
37:18
know. I I don't I doubt that this that
37:20
that had any, you know, play on them
37:22
redesigning things. But maybe
37:24
>> you're even more screwed if you buy
37:25
authentic AirPods Max because then you
37:27
spent $550 for that experience.
37:29
>> Don't even get me started cuz I've done
37:30
that twice. So, don't even don't even
37:32
get me like mistakes were made not doing
37:35
it a third time. Um, and uh, but the uh,
37:39
but yeah, but I mean, you know, I'm sure
37:40
that that wasn't part of why they went
37:42
into this, but that might be for a brief
37:45
window. I have no doubt that someone
37:47
will reverse engineer the setting screen
37:49
and whatnot, but but for for a window of
37:50
time, I think that'll make it harder for
37:52
the the the non Airpods to to, you know,
37:55
maybe appears themselves.
37:57
>> I'm honestly afraid to pair my AirPods
38:00
with anything else because I don't want
38:02
it I don't want it to be ruined. like
38:03
they work they work great for me right
38:04
now and I Bluetooth has burned me so bad
38:07
that I I don't want to try connecting it
38:09
to a switch or anything. In fact, what I
38:11
want to do when I want to listen like
38:12
both to a podcast and you know play
38:15
video games is I put in the AirPods and
38:17
then I have a separate pair of
38:18
headphones that I put over top of them.
38:20
>> Oh my god. I've done the same thing.
38:22
Honestly, I've done the same thing and
38:24
which is so dumb and and um and I do
38:26
have uh but what I also do and this is
38:29
also dumb but like I have number of
38:31
pairs of AirPods. I have AirPods Max and
38:33
then I have my Sonos headphones and I
38:34
have some Sony's and other things and
38:36
then it's just almost like I have like
38:38
you know device determinant headphones.
38:40
Um and like when I travel especially I
38:43
don't think AirPods Max are good travel
38:45
headphones at all um for many many
38:47
reasons. Um and you know that is when
38:50
like but they are still remembered.
38:53
They're not not actively paired with my
38:54
phone but like my phone remembers them.
38:56
Sometimes I have to go through the whole
38:57
song and dance again. But it's usually
38:59
one of those things where I'm like,
39:00
"Okay, I can use this. I have the
39:02
Airfly, you know, set up if I need to
39:04
use it, you know, um, and on on the
39:06
plane and and we're good." But yeah, I
39:09
mean, it is it is kind of silly. Uh
39:11
Apple really has nailed the their
39:13
workarounds for Bluetooth so well that
39:15
we're willing to either have them in our
39:18
ears and then put other headphones on
39:19
top of it to multitask or
39:22
>> buy completely separate headphones that
39:23
we're not pairing with our our you know
39:25
just with our iPhone products just to
39:27
avoid
39:29
>> the Bluetooth pain.
39:31
>> It never got better than that.
39:33
>> Yeah, it's true. It's true.
39:35
>> You know, there's a lot of stuff in
39:36
there. There's a lot of stuff that's
39:37
like to it's fun, you know, a lot of
39:39
features and fun and useful to to go in
39:41
and change things. And lately, I've been
39:43
uh experimenting with the hearing aid
39:45
mode, which I put off doing for ever and
39:48
ever. But I finally took the what it say
39:50
like four minutes or something to do the
39:52
little hearing test, which always makes
39:53
my the hair on the back of my neck
39:56
>> uh rise because I hate those hearing
39:58
tests.
40:00
Start to notice the tinidis from years
40:02
of rock and roll. Anyway, uh I so I'
40:06
I've started using that a little bit and
40:08
it is. By the way, if you haven't tried
40:10
it, I mean, it's
40:11
>> great.
40:12
>> It's freaking weird. It's It's like
40:15
you're Spider-Man or something. You can
40:16
hear things that you never heard. I I
40:18
was in Ohio, like I said, uh and I was
40:20
listening to a podcast or something and
40:21
I forgotten to turn off that hearing aid
40:23
mode and I I was lying in bed and and I
40:27
with the AirPods with the hearing I was
40:29
I heard this noise. I'm like, "What is
40:31
that noise?" just little like, "Oh my
40:33
god, that's rain." And I took that thing
40:36
out and I couldn't hear at all.
40:37
>> Wow.
40:38
>> And uh, wow. I guess I do have some
40:40
moderate hearing loss. Anyway, uh, but
40:42
it's it's also I think it actually makes
40:45
you hear things that like maybe normal
40:47
people can't hear. I mean, it's it it
40:49
feels like
40:50
>> I was telling Griffin this, I feel like
40:52
I'm on a like a in one of those videos
40:54
where they're doing, you know, Foley
40:56
jobs for a movie, you know, like every
40:58
step is like a a creek that's not
41:01
actually a shoe on a floor. It's like,
41:03
you know, whatever it is to make it
41:04
sound like a really impressive sounding
41:06
shoe on a floor. Everything is it's like
41:09
this hyper awareness. Um, so I I I just
41:12
want to mention that if people have
41:14
those things and they have any sort of,
41:16
you know, hearing deficiency, hearing
41:18
loss, it's a pretty great feature. It's
41:21
pretty amazing. And I tried to uh on
41:23
this last trip just, you know, like at
41:25
uh, you know, going up to like rental
41:27
car counters or something, right? Cuz I
41:29
mean, I don't know if it's just me and
41:31
it might just be my hearing loss, but I
41:33
mean so many people these days are like,
41:35
"Hi, to help you."
41:38
Worse yet, if they have like a a face
41:39
mask on and they're doing,
41:42
you know, it's just impossible to tell
41:44
what people and and I find myself like
41:45
angry old man like, "I'm sorry, what?
41:48
What? I can't hear you." So, uh, that
41:51
helped. It's it's uh, you know, I
41:53
slightly less angry.
41:55
>> Yeah. My my mom uses that that feature a
41:58
lot and um because she's you know she
42:00
gets her hearing tested every year and
42:02
they say you know you don't need hearing
42:04
aids um or whatever but she has you know
42:06
some hearing loss but she knows that her
42:08
hearing is not as good as it used to be
42:10
and she already wears AirPods all the
42:12
time anyway because she's always
42:13
listening to podcasts like my mom
42:15
literally has like three pairs of
42:16
AirPods I'm not joking that she cycles
42:18
through every day and then she sleeps in
42:20
them too and then we'll get up in the
42:21
middle of the night and we'll like
42:22
change out like the AirPods and and it's
42:24
it's a whole thing and
42:25
smoking
42:26
>> genuinely. Genuinely, she's she's with
42:28
them all the time and but she uses the
42:30
hearing aid mode more and more
42:32
frequently and and um you know it's
42:34
great for watching TV and for other
42:36
stuff and yeah I mean even if you don't
42:38
like I I can't take advantage of it
42:40
because unfortunately like the hearing
42:41
test is like oh no your hearing is is
42:43
great. I'm like, that's awesome. But
42:45
sometimes, you know, you would like to
42:46
maybe, like you said, like you're you're
42:48
at a rental car, you know, counter and
42:51
um even if you have great hearing, that
42:52
can be a scenario where there's a lot of
42:54
noise happening coming from a lot of
42:55
places. People might not be speaking
42:57
super loud. Um and and you've got, you
42:59
know, things coming in from other
43:00
places. That's nice. But the the
43:03
downside of all this, and I think this
43:05
is kind of what like German had pointed
43:06
out, is that they've added all these
43:08
features to AirPods over the years, and
43:11
it's just really difficult to find it,
43:13
right? Like I I wonder how many more
43:15
people would even be aware of like I
43:17
think Apple's done a pretty good job um
43:19
socializing the the hearing test
43:21
feature, but like how many more people
43:23
would be aware that yeah, this can
43:25
actually now be used as a hearing aid um
43:29
if it were a little bit easier to to
43:31
manage the the settings. Um, and I think
43:34
though the reporting also said, and this
43:36
I think this goes to maybe your theory,
43:38
uh, Griffin, about them not having a um,
43:40
Apple or a separate Siri app is that,
43:44
um, you know, apparently like they
43:45
they've heard calls that people want a
43:47
separate AirPods app, but a companion
43:49
app and they're like, "No, we're going
43:50
to keep it in in settings for now."
43:52
Which, I mean, fair enough. That's fine.
43:55
Um, as long as it's it's laid out
43:57
better. I I think that's that's the
43:59
biggest thing. you know, they have like
44:00
an API where, you know, you can add a an
44:03
AirPlay button to the bottom of your,
44:05
you know, podcast or music app to to
44:07
control your audio destination, um, an
44:10
AirPlay if you want.
44:11
>> Maybe maybe they'll have like a um
44:14
>> an a similar API for that to like evoke,
44:17
you know, bring up AirPod settings
44:18
because right now, I mean, you might not
44:20
even know that you can do this like you
44:21
have to bring up control center
44:24
on the volume level thing and then you
44:27
see the AirPod settings. But I mean, if
44:28
that were like built, if they're
44:29
redesigning this while they're at it,
44:31
>> they build in like a system API where
44:33
you can add in AirPod settings directly
44:35
from Overcast. I mean, that would be
44:36
excellent.
44:37
>> I have to I have to check. Can you do
44:39
the hearing aid thing on off from the
44:42
control center?
44:43
>> I don't think so.
44:44
>> I don't think you can for that.
44:45
Although, I wish that you could, but um
44:47
but I'd have to check. But I wouldn't be
44:50
surprised if they added that though,
44:51
right? Like if you could turn that on or
44:53
off that would be awesome because I
44:55
frequently use that setting you you
44:57
mentioned that. Yeah, I don't think a
44:58
lot of people know it's in control
44:59
center. If you you know tap and you
45:01
press down and hold you can adjust the
45:03
settings and you can say okay I want it
45:04
you know to be um you know dynamic or do
45:07
I want it to be you know noise
45:08
cancelling or or whatever. And you can
45:09
also you know change your volume and and
45:11
other stuff. That would be I think
45:13
fantastic if you could do the the
45:16
hearing aid mode that way too. You know
45:18
how it when you're setting up a custom
45:20
control center, there's like the two
45:22
special ones you can do that fill up a
45:23
whole screen. Like you can make the
45:25
connectivity widget full screen or the
45:27
um
45:28
>> what's the other one? Or the like the
45:30
you know the smart home stuff I think is
45:32
the other one.
45:33
>> Um maybe maybe that's how it'll work.
45:36
Like it'll be a control center screen
45:37
that's like all AirPod settings.
45:40
>> See, I think that would be great. I
45:42
would love that because already I have,
45:44
you know, I'm I'm dealing with, you
45:46
know, going into that setting area a
45:49
lot. I would love that. And then if if
45:51
and then at that point, I think if it
45:53
was in control center, then it it should
45:56
be shortcutable. And so then if you
45:58
could like create an automation, right,
46:00
where you could then maybe say turn on,
46:02
you know, um hearing aid motor, turn on
46:04
this or that, like that, that could be
46:06
really cool.
46:06
>> And then it'll be Apple intelligenceable
46:09
as well.
46:10
>> Yeah. There you go.
46:12
Christina, why don't you tell us about
46:14
um not not the Apple car, but uh next
46:17
best thing.
46:18
>> I was gonna say there's no Apple car,
46:19
but Johnny IV did in fact design a car u
46:22
or or at least worked on a car. So, um
46:24
Ferrari officially pulled off uh the
46:27
wraps off of a luch, I guess that's how
46:29
you say it, which is it's its first ever
46:31
all-electric vehicle. It's also their
46:32
first five-seater and their first sedan.
46:35
And this uh um was uh designed or I
46:39
guess worked on from um Love from which
46:42
is the collective that Johnny IV does
46:44
along with um Mark Newsen and and they
46:48
work together on the Apple Watch and and
46:49
I think uh you know the the trash can
46:52
Mac Pro and Mark News has done a lot of
46:54
other projects too. work together very
46:56
closely. And so this came out and you
47:00
know the the kind of the joke is like
47:03
other than the color and the fact that
47:04
it has like the the Ferrari badge like
47:06
this does not look like a Ferrari which
47:09
is kind of what and this was the article
47:11
that you guys wrote on on cult of Mac
47:13
kind of said that like the memes race to
47:15
folk fun at Johnny Ives's new Ferrari
47:17
Luché design because when you look at it
47:20
and I'm not even a car person but you
47:22
have an idea in your mind of like what
47:23
is a Ferrari and then you look at this
47:25
and you're like this is not a Ferrari.
47:27
Sorry. What is this?
47:28
>> The the memes are pretty hilarious. And
47:30
oh my god, it's like the California or
47:32
the Los Angeles mayor's race. There's so
47:36
many AI like long AI videos like
47:40
basically dragging Ferrari through the
47:43
coals for this, you know, like h this
47:45
horrible thing. You know, I can't even
47:47
say what they say because they're pretty
47:50
there's a lot of swearing,
47:52
but uh you know, pretty hilarious stuff.
47:54
People did people did not react well.
47:56
There's I think Ferrari stock tanked
47:58
like seven points that or 7% that day. I
48:01
don't know if it's back up, but uh yeah.
48:03
Oh my god.
48:05
>> I got to say though,
48:06
>> sorry. Go on.
48:07
>> I really like it.
48:08
>> I think it's a good car. Yeah. Like
48:10
there's the jokes that um Oh, you know
48:14
it it looks so much like the Nissan Leaf
48:16
and like Yeah, sure. And like the
48:19
grandmother accidentally buys you like a
48:21
Fun Tendo Whiz instead of a Nintendo Wii
48:23
kind of way. Like you don't have to be,
48:25
you know, Doug Demiro car level expert
48:28
to be able to tell a sports car apart
48:30
from a midsized SUV. Like, yeah, they're
48:35
the same color and I guess car- shaped,
48:37
but I I think it looks good. It looks
48:39
very distinctive. Like, it's got I
48:41
didn't realize this until like I I
48:42
really looked at the design for a little
48:45
bit, but and you know, some some
48:47
electric cars since they don't have like
48:48
the engine under the hood, they have
48:49
like the the frunk, you know, the empty
48:51
sort of area. What they've done instead
48:53
is they don't have a hood. It It's not
48:55
like a solid piece in the front that's
48:57
like a crumple zone. It's like imagine
48:59
like a magic mouse how it's sort of like
49:01
surfboard shaped. Uh it's like that all
49:04
the way from the tip of the front bumper
49:05
and then it's like a continuous glass
49:07
piece that goes all the way to the tail
49:09
and then they just have like an overhang
49:11
that's hoodshapedish. It's like an
49:14
illusion. I think it looks really cool.
49:16
>> It looks like a good place to for leaves
49:18
to accumulate. Well, they would just
49:20
they would just blow right through
49:22
because there's nothing there's nothing
49:23
there. I mean,
49:25
>> yeah, there's nothing for really shape.
49:26
>> Yeah, it is. It because it kind of looks
49:29
odd, right? Like you've got like this
49:30
this whole kind of like top piece of
49:32
glass which is which is essentially, you
49:34
know, kind of kind of your your your
49:35
roof and your hood thing that kind of
49:37
then, you know, cascades into I guess
49:39
like the the front of the car.
49:41
>> I mean, it's close. I mean, how do you
49:42
make an electric car that's not just
49:44
like I mean, the peak aerodynamic shape
49:46
is just like a curve, like an egg, and
49:48
nobody wants to drive an egg. Uh, you
49:50
know, one of my favorite cars is the
49:51
Honda IonX 6, but I like it because it
49:54
looks so bizarre and weird. Uh, it has
49:56
like that weird sloping sloping back.
49:58
Um, but you know what what they've done
50:00
here is they they they have an illusion
50:02
because like they have a thing in front
50:04
that adds like an extra level, but it's
50:07
like still equally aerodynamic.
50:09
>> Yeah. Yeah. Like I think they they claim
50:11
that like it's the lowest drag
50:13
coefficient in Ferrari history achieved
50:16
through a styling, conversions, active
50:18
air shutters, and ride height logic that
50:20
lowers the front by 10 mm even while
50:22
cruising. I mean, okay, you're spending
50:25
$650,000 on the car. Let's also point
50:28
that out. Um I I I I hope I hope you
50:30
really en enjoy it. Um but no, it's it's
50:33
an interesting it's an interesting
50:34
design. It just it's it doesn't look
50:36
like a Ferrari. At the same time, I'll
50:39
defend them a little bit. Like this this
50:41
is not a sports car. This is their first
50:43
sedan. This is a five-seater. This is
50:45
not the typical thing that you would
50:48
have a Ferrari for. Does Ferrari need to
50:51
have something that's not you know the
50:52
typical sports car? That I'm not sure
50:54
about. Um but you know that that that's
50:57
that's an interesting challenge. I do
50:59
think the interior at least the interior
51:01
shots like it looks it looks great. I
51:04
think the interior looks looks really
51:05
fantastic. Oh yeah, it looks gorgeous.
51:08
You know, even the rear looks really
51:09
good, too. It has sort of like those um
51:12
uh Chevy Cobalt rear tail lights on
51:15
where it's like the black panel with
51:17
like the circle donut shaped lights in
51:18
the middle. But I I think they're
51:20
they're trying to evoke like the Ferrari
51:22
F40 or the other ones that have like the
51:23
circular tail lights. But
51:25
>> I I've seen more Chevy Cobalts than I
51:26
have Ferrari F40s. So that's what I
51:28
think it looks like. But you know what?
51:30
I thought that was an attractive rear
51:32
tail light design. uh you know where
51:34
those cars are all like 20 years old now
51:36
they're sort of disappearing from the
51:37
roads but you know we we've got a
51:40
Ferrari doing it now like it I think
51:42
it's good look and the interior yeah
51:44
it's it's gorgeous like they they've
51:46
developed like a bunch of custom OLED
51:48
displays that wrap like around and
51:51
inside the physical controls that you
51:53
still have like they wax poetic about
51:56
how how precisely they've they've
51:58
engineered these parts and focused on
52:00
like the button feel and the switch feel
52:02
of all of these things. Like one of the
52:04
first cars I drove was like a 1991 Honda
52:06
Accord. And like every single button
52:08
switch was had like micro switches in it
52:10
that had like a really good thunk and
52:12
click feeling. Like I would just, you
52:14
know, press a bunch of the AC buttons
52:15
and like switch them all on and switch
52:17
them all off like one at a time just so
52:19
I could like press all of those buttons
52:21
all in a row.
52:22
>> And you know what? Yeah, the buttons
52:24
broke. Uh I have a new car, a newer car
52:26
now that has more reliable buttons that
52:28
haven't broken, but they don't feel as
52:29
good. They're mushy. It's like pressing
52:31
a calculator. Who wants to do that? I I
52:33
I'd love to touch the buttons in that
52:34
interior. They they look really nice.
52:36
>> No, I just say, did you see the video
52:37
where Johnny IV was talking about
52:39
touchscreens in cars and basically
52:40
saying it's it's a horrible interface
52:42
for a car. Fantastic for an iPhone,
52:43
terrible for a car. He's absolutely
52:45
right. I mean, anytime I've tried to use
52:46
like CarPlay, like I I constantly like,
52:49
well, I just want to, you know,
52:51
>> fast forward the podcast for 60 seconds
52:54
and am I going to die as I do it?
52:56
Because, you know, you're you got to
52:57
look at the screen. And and frankly, I
52:59
mean, that would be less troublesome if
53:02
their touch targets were uh larger,
53:05
>> but uh I thought it was interesting. He
53:07
he went on and on about, you know, the
53:08
the beautiful way that those those
53:10
things you're talking about, the
53:11
switches and all that and and yeah, I I
53:13
think, you know, I think this car looks
53:15
fine. If it was a $60,000 electric car,
53:18
I'd be like, "Yeah, it's awesome."
53:20
$640,000,000.
53:23
But it's a Ferrari. I mean like we we
53:26
look none I I don't think that there are
53:28
many people who are kind of in our kind
53:30
of audience who are the actual market
53:31
for this car. I think it's like do you
53:33
live in Dubai?
53:35
>> Um and and and and that that's kind of
53:37
like your you know your plot thing. If
53:38
the answer to that is no, then it's like
53:40
you are not like the target for for for
53:42
this car more than likely. But what
53:44
happens with these sorts of things um is
53:47
that you will see kind of the and this
53:50
happens historically is that and you see
53:51
this in racing too is that you see like
53:53
the the things that start out kind of
53:55
the high-end that the supercar market
53:57
will trickle down into into other car
54:00
designs too. And and I don't know if
54:01
that's going to be a scenario that will
54:03
happen here because it's not as if
54:05
Ferrari is somehow breaking new ground
54:06
by having an electric sedan. Like it's
54:09
come on guys that that that's kind of
54:11
been a known quantity for a long time
54:13
now. Um but it is interesting I think to
54:15
kind of see their approach at that and
54:17
and yeah I I agree like I think that
54:19
Johnny I was completely correct in
54:21
talking about the importance of having
54:22
physical controls in cars. That said, I
54:25
also think that the way they are using
54:27
screens in this interior is really
54:30
smart, right? I think it's kind of the
54:31
combination of the two things where
54:32
you're expected at this point to have
54:35
more screens that are probably going to
54:37
be LED powered and not, you know, like
54:39
like like physical, you know, like uh
54:41
you know, mechanical um dials or
54:42
whatnot. Um but if you can pair that
54:45
with physical buttons, I think that
54:47
maybe is is kind of the the best of both
54:49
where you get things that can kind of be
54:51
dynamically updated as as they need to.
54:53
Um, but also I'm not having to take my
54:56
eyes off the road. I can, you know,
54:58
scroll through something with a button
55:00
or a knob. Um, and and also feel more
55:03
confident in what I'm doing rather than,
55:05
um, okay, now I'm playing the iPad, you
55:08
know, as I'm driving. You're like,
55:09
"Okay, well, no wonder you're you're
55:12
getting in a car accident." because like
55:14
it it is sort of weird that it is
55:16
illegal in most states um in the United
55:19
States anyway to like be on your phone
55:21
or have your phone in your hand um while
55:24
you're driving, but it's completely fine
55:27
to be interacting with the giant iPad
55:31
that is in in in the front of of of your
55:33
Tesla or you know any other modern car
55:35
at this point. Like that's completely
55:36
fine, but oh no, if you got if if you
55:38
got the phone in your hand, you're going
55:40
to get a ticket and points on your
55:41
license. The worst part about a screen
55:43
is that you can't like rest your finger
55:45
on a button and wait to press it, right?
55:47
Like as soon as your your finger touches
55:48
it, it it immediately activates. So like
55:50
you have to be extremely precise. You
55:52
can't even if you know where all the
55:53
buttons are, like you're I mean you're
55:55
driving down the road if it's a bumpy
55:57
road, like your finger's like bouncing
55:58
around. Like it's not a precision
55:59
instrument.
56:00
>> Yeah. You got to like rest another
56:02
finger on like a dead spot of the screen
56:03
and like to like make sure you can align
56:05
it up right. Like in in this Ferrari
56:07
there are all of the there are the
56:08
screens but then under each like thing
56:10
that you might adjust like air speed or
56:12
temperature it's got it has a physical
56:14
switch underneath it and you can use
56:16
either.
56:17
>> Um I mean in terms of the price like
56:19
yeah $640,000 that's an insane amount of
56:21
money but
56:22
>> you know that's not like the cheapest
56:23
Ferrari like there are cheaper Ferraris
56:25
that are in like the $200,000 range but
56:27
maybe this will eventually get there.
56:29
>> Maybe. I mean usually I think Ferraris
56:31
to start in kind of the 300,000 range
56:32
and you have to be on a waiting list.
56:34
So, even if the the list price is a
56:36
certain thing, I looked into this a long
56:37
time ago, so if I'm wrong, listeners, I
56:39
apologize. Um, but I think typically how
56:41
it works is like you've got to be on a
56:42
waiting list
56:43
>> and and many times those lists are years
56:45
long and so people will spend more money
56:47
or dealers will spend more money to to
56:49
get people off of lists faster. But
56:51
yeah, I mean, this is obviously you
56:52
could get a Ferrari for less money, you
56:54
know, spend only $300,000 on a supercar.
56:58
Um, I mean, and I feel like this is part
57:00
of part part of the reason for this
57:02
price is this is the first sedan, right?
57:04
This is the first like one that they're
57:06
kind of doing. But yeah, there's there's
57:10
nothing to be said that you might not
57:11
see these kind of, you know, design
57:14
features or motifs be um, adopted or or,
57:18
you know, inspired by, you know, future,
57:21
you know, car makers who are maybe, you
57:23
know, selling $100,000 cars instead.
57:26
>> It's the same thing that that works with
57:27
Apple. like this is the first time they
57:28
had to develop an electric drivetrain
57:31
from scratch. You know, the first time
57:33
Apple like redesigns a MacBook Pro, that
57:34
one's that one's going to be more
57:35
expensive. They raise the price for that
57:37
and then it the prices go down over time
57:39
and the features trickle down to the
57:40
cheaper models.
57:41
>> So, how how do the specs uh check out?
57:44
You're a car freak, Griffin. Did you see
57:46
this?
57:47
>> 122 kilowatt hour battery. That's that's
57:50
like standard for a big battery that the
57:53
Tesla Model S has had like 100 kilowatt
57:55
hour batteries I think going on like 10
57:58
years now. You're you're really limited
58:00
by like volume and space. Like cars are
58:02
only so big. That's about as big as you
58:05
can fit a battery underneath a car
58:07
unless you make it like extremely long
58:08
like a limousine. Um that's pretty
58:10
standard. Four electric motors, like one
58:12
motor per wheel. That's something that
58:14
you only really see on like highest
58:16
performance ridiculous supercars. You
58:19
know, it's usually one motor in the rear
58:22
or if you if you want to go ridiculous,
58:24
you have one in the front and in the
58:25
back for performance models. Four motors
58:28
is like that supercar territory. 0 to 62
58:31
miles per hour in 2 and a half seconds.
58:34
I know I think some Teslas are down to
58:37
like 1.9 seconds, but those don't have a
58:39
Ferrari badge on the front. Yeah. And
58:42
and those don't have any physical
58:43
controls. So,
58:45
>> yeah, I I bet you that those physical
58:46
controls are something else. I mean, I
58:48
would like to just sit in one of these
58:49
if somebody would ever let me into one
58:51
just to, you know, see what it feels
58:53
like to shift the gear or turn the, you
58:55
know, volume on the on the radio.
58:58
>> I mean, here in San Francisco, you can
58:59
pretend to be rich and just go to a
59:01
dealership.
59:02
>> Yeah. Go to a dealership.
59:03
>> Yeah.
59:04
>> Yeah. 192 m per hour top speed. Uh,
59:07
that's that's pretty high. Not a lot of
59:08
Teslas go that fast. 320 mi 9 mile
59:11
range. I bet you don't go that far if
59:13
you're going 192 miles per hour.
59:15
>> No, but you know what? Like you don't
59:16
get that sort of range like in a regular
59:17
Ferrari, like a gas powered Ferrari.
59:19
Like they they have like
59:20
>> also true
59:21
>> awful terrible gas mileage. So I feel
59:23
like I mean this is slightly different
59:25
because again like it's it's a sedan. So
59:27
it's not necessarily like a a a supercar
59:30
type of experience even though it was
59:31
priced like one. But, you know, that I
59:35
think is the interesting balance that
59:36
they have to strike here is like, okay,
59:37
how do we have a a big enough battery
59:40
that will, you know, power this for what
59:42
we want, but how fast do we go even
59:44
though it maybe isn't as fast as our
59:45
topline um cars, you know?
59:48
>> Battery drain 1% every second,
59:50
>> right? That's the thing, right? like
59:52
that that that I mean because because
59:53
that would literally I mean that's been
59:54
one of the things um I think correct me
59:56
if I'm wrong on this that has kind of
59:58
really held back uh kind of the the
1:00:00
higher end kind of manufacturers from
1:00:01
going into electric is the fact that the
1:00:03
power that would be required to you know
1:00:05
from from from electric to you know do a
1:00:08
lot of the things they do just you're
1:00:10
you're you know you're you're there's
1:00:12
just not a battery big enough and the
1:00:14
space constraints maybe if we had a
1:00:16
different type of of material maybe if
1:00:18
we weren't using lithium ion we were
1:00:20
using you know nuclear or maybe that
1:00:22
would be different, right? But but but
1:00:24
putting a different energy, you know,
1:00:26
thing thing aside, there are just like
1:00:28
some like hard limits there. So
1:00:30
>> I feel like from an engineering
1:00:31
perspective, it's impressive that
1:00:33
they're able to kind of fit all of this
1:00:35
in this type of, you know, size car with
1:00:39
these types of features.
1:00:40
>> Unlike a gas car, it can control each of
1:00:42
those motors incredibly fast. It says
1:00:45
that it updates the actuation 200 times
1:00:47
per second. So, you know, this is the
1:00:50
car you want to be in. if you're you're
1:00:51
driving over icy roads, you know, I
1:00:53
mean, the traction control is insane on
1:00:55
electric cars. Um, each wheel has
1:00:57
independent torque vectoring, active
1:00:59
suspension control, independent steering
1:01:01
on the rear axle. Uh, they said for the
1:01:04
noise, you know, some of the
1:01:07
some some of the traditional car
1:01:10
companies moving into electric cars for
1:01:11
the first time, like Dodge, they'll give
1:01:13
you like a fake V8 sound. Uh but in a
1:01:17
very Johnny Iive way they say a
1:01:19
precision accelerometer mounted at the
1:01:20
center of the rear axle captures the
1:01:23
actual vibration of the rotating
1:01:24
electric components. That signal is
1:01:26
filtered, equalized and amplified,
1:01:29
working like an electric guitars
1:01:30
amplifier. The result is a sound rooted
1:01:33
in the real physics of the machinery,
1:01:35
not synthesized. Ferrari spent 5 years
1:01:37
and 40,000 km of dedicated testing
1:01:40
developing it. That's I mean you can see
1:01:44
you can just imagine Johnny IV reading
1:01:46
that aloud hear his voice.
1:01:48
>> I was going to say I I can hear it too.
1:01:50
No, it's it's uh it's incredible.
1:01:52
>> Christina, last time you were on I think
1:01:54
you were you you're on medical leave.
1:01:56
>> I was on medical leave. Yeah, I I had
1:01:57
had surgery um only a couple of weeks I
1:01:59
think uh prior to that and so I'm now
1:02:01
back from medical leave. So, yay.
1:02:03
>> Had you already started on at GitHub
1:02:05
before that?
1:02:06
>> I had I had um Yeah, I'd I'd already
1:02:08
started back at GitHub. So, long story
1:02:10
short, my background is that I used to
1:02:12
be a journalist, then I went into um
1:02:14
engineering, and um I was at Microsoft,
1:02:16
then I was at GitHub, then I was at
1:02:18
Google Deep Mind, and now I'm back at
1:02:20
GitHub.
1:02:21
>> So, now now you're back there again
1:02:22
full-time. How how long have you been
1:02:24
back?
1:02:25
>> Um I guess I've been back from medical
1:02:26
leave since for about a month and for
1:02:28
about six weeks now, I guess.
1:02:30
>> Okay. Uh lot of get that that's been a
1:02:33
busy busy month and a half, has it?
1:02:35
>> It's been a real busy month, month and a
1:02:36
half. I don't know. We've been in the
1:02:37
news a little bit. I don't know if
1:02:38
you've seen it, but yeah. I know a lot
1:02:39
of people are using GitHub and which is
1:02:42
great but can also uh can also lead to
1:02:44
some problems as uh as as you know any
1:02:47
anytime you have your your you know
1:02:48
traffic volume you know ramp up 15 times
1:02:52
um you know unexpectedly uh that that
1:02:55
can that can lead to some lead to some
1:02:57
uh performance and and rel reliability
1:03:00
issues which the team is like very much
1:03:02
dedicated and and trying to you know
1:03:04
improve and and and get under control.
1:03:07
I'm I'm sure this is very low on your
1:03:09
list of priorities, but my my personal
1:03:10
website is hosted on GitHub pages.
1:03:12
>> Yeah.
1:03:12
>> As a matter of fact, uh is that at risk
1:03:15
if
1:03:15
>> I don't think so. No, look, we we we use
1:03:17
GitHub pages all the time too, right?
1:03:19
Like obviously if if there are
1:03:22
performance like issues where you know
1:03:24
something goes down like GitHub pages
1:03:25
can also be one of the services that can
1:03:27
go down. But no, I mean like get get
1:03:29
it's it's uh we're actively working on
1:03:32
trying to improve the reliability of the
1:03:34
site all the time and um
1:03:36
>> here first the the the three readers of
1:03:37
my blog extraordinary will be we'll be
1:03:39
happy with that.
1:03:40
>> Well no but continue to use GitHub pages
1:03:41
because it's it's a great resource. I
1:03:42
mean and honestly it is and it's one of
1:03:44
those things that um I mean but but
1:03:45
we've seen tremendous amount of of usage
1:03:47
of GitHub page we've seen tremendous
1:03:49
amount of usage of everything basically
1:03:50
since um end of November. Basically the
1:03:53
the longest story short of it is that
1:03:55
the capabilities of the frontier models
1:03:57
changed in a very important way um where
1:04:00
like the the claude and the the open AI
1:04:02
models got really really good in ways
1:04:05
that they weren't before and so you
1:04:07
started seeing agenta capabilities that
1:04:09
you know had kind of been talked about
1:04:12
as types of things that executives like
1:04:14
brag about oh this will change this and
1:04:16
that to actually being like oh no this
1:04:18
is actually good enough um now um for um
1:04:22
a lot use cases. And what happens with
1:04:25
that is that when things are really
1:04:27
good, people use them a lot. And where
1:04:29
do people tend to store their source
1:04:31
code? GitHub. Where do people tend to
1:04:33
deploy their code? GitHub. And so if you
1:04:36
have like this kind of ramp up in model
1:04:38
capabilities, ramp up and more and more
1:04:40
people using the models because they've
1:04:41
become so good, more and more people,
1:04:43
you know, exploring coding for the first
1:04:45
time because now they can get really
1:04:47
good results from um, you know, their
1:04:49
own steering. Even people who had been
1:04:51
professional developers who might not
1:04:52
have been buying into the AI uh uh hype
1:04:56
around like the productivity gains you
1:04:58
could you could get are now going, "Oh,
1:05:01
okay. Now the cap the capabilities are
1:05:03
actually good enough that I would use
1:05:06
this and and and would um you know, let
1:05:08
let this do work for me." More and more
1:05:11
people are going to be building things
1:05:13
more than ever before. And I'm sure that
1:05:14
you've seen that just in the the number
1:05:16
I mean Apple's even seen this like the
1:05:17
the app store review times have gone
1:05:20
through the roof. And that's not just
1:05:22
because of Vibe coding, but that's
1:05:24
definitely part of it is that more and
1:05:26
more people now have easier ways to
1:05:28
build apps that are are performant and
1:05:31
are good and and there is I think some
1:05:32
for some people a little bit of kind of
1:05:34
a derogatory uh kind of a you know
1:05:36
association with hearing the term vibe
1:05:38
coding but at this point I don't know
1:05:40
any professional developer. I know there
1:05:41
are people who are out there, but I I
1:05:43
personally don't know anyone who is not
1:05:45
using AI um you know assistance in in
1:05:48
their tooling of some sort um or or
1:05:50
agents in some way and and it's it's
1:05:53
just part of how things are built. And
1:05:54
and what that does though is that
1:05:57
obviously the the AI can go much faster
1:06:00
than you know a human would. And so you
1:06:02
have like a we have a you know slew of
1:06:04
new Mac and iOS apps that are available
1:06:06
from people who might have great ideas
1:06:08
and and would say okay well this was
1:06:10
something I wanted to build but it would
1:06:12
take me you know six months to a year
1:06:14
and as a side project and now they're
1:06:16
like okay I can do this in an afternoon
1:06:18
and then continue to iterate and refine
1:06:20
it um and and and ship it. So right now
1:06:24
a lot of that cost is amortized because
1:06:26
all of these AI company every single AI
1:06:28
company is losing money I think
1:06:29
sometimes by a ratio of like 20 to1. Uh
1:06:33
>> if the how do you see that going forward
1:06:36
because obviously it's an untenable
1:06:37
situation. OpenAI can't keep burning
1:06:40
money forever. Uh if if they were fairly
1:06:43
priced, how do you see the economics of
1:06:45
that changing? Like would these products
1:06:46
still be good enough that people would
1:06:48
spend 20 times more on it just for the
1:06:50
same thing? Well, and I think that's
1:06:51
actually kind of what we're seeing right
1:06:52
now. I mean, this was an instance that
1:06:54
we we've had at at GitHub. We have a
1:06:56
product called GitHub Copilot, which was
1:06:58
one of the very first AI, you know,
1:06:59
coding assistant products, and it also,
1:07:02
you know, works similarly to the way
1:07:03
that Cloud Code um or or OpenAI Codox
1:07:06
does. Um, but you can choose your AI
1:07:08
models. So, like it's not like you can
1:07:10
use either either model actually or you
1:07:12
can even use models from other providers
1:07:14
too um with it. And we'd had kind of a a
1:07:18
flat pricing where you get a certain
1:07:20
number of requests for a certain amount
1:07:22
of money a month. Um, and we've had to
1:07:24
uh to to change that. And we had to
1:07:26
change that just because the a it was
1:07:29
difficult for us because again as the
1:07:31
model capabilities increased. Our own
1:07:34
costs went through the roof in terms of
1:07:36
like what we would pay for requests from
1:07:37
the model providers went way up.
1:07:40
>> People can ask for more complex
1:07:42
questions.
1:07:43
>> Right. Exactly. because it used to be
1:07:44
one of those things where you'd be like,
1:07:45
"Oh, okay. Well, a request will cost
1:07:47
about this many tokens and tokens are
1:07:49
how the um the you know um AI kind of
1:07:52
economy is is kind of mediated." That
1:07:55
changed when suddenly your cost could be
1:07:58
okay. Well, it could be this many tokens
1:07:59
or it could be this many and and some of
1:08:01
the newer models also use a lot more
1:08:03
tokens and and can um also store more
1:08:06
context and more information about your
1:08:07
codebase, which means it's going to
1:08:09
potentially cost more. And so we've
1:08:11
moved from you know like kind of a flat
1:08:13
like um you know fee to usage based
1:08:16
billing which is where the rest of the
1:08:18
industry has moved unless you're in the
1:08:20
consumer space like uh I think Cloud
1:08:22
Code uh for Cloud Code Max and and uh
1:08:25
Codeex like they still subsidize for
1:08:27
their individual users if you pay $100 a
1:08:30
month or $200 a month you might get more
1:08:32
than $200 a month in in you know actual
1:08:36
usage. But if you're a business Yeah. uh
1:08:39
what you've seen happen from from all
1:08:40
the major providers and and that
1:08:42
includes GitHub is that we're moving to
1:08:43
these usage based models. Now, what will
1:08:46
that mean for how people will use these
1:08:48
tools? That's what we're figuring out.
1:08:50
And I think that there will be some
1:08:52
circumstances where you'll go, okay, you
1:08:54
know what, the results are still good
1:08:56
enough and I'm still so much more
1:08:58
productive that it's worth me paying
1:09:00
whatever this costs because this is so
1:09:02
good. But I think in some scenarios it
1:09:04
becomes okay well maybe I don't need to
1:09:06
use the latest greatest most high-end
1:09:09
cutting edge frontier model that costs a
1:09:11
bajillion dollars. Maybe I can use
1:09:13
something else or maybe I can even use
1:09:16
local models for for certain types of
1:09:18
tasks right which is I think really
1:09:20
great and this is where you know um
1:09:22
Apple silicon is is really ahead of the
1:09:24
pack and in some regards because of its
1:09:26
unified memory and because of the fact
1:09:28
that so many people so many developers
1:09:29
use Macs. there's good tooling built in.
1:09:32
Um, which I never thought I would say
1:09:33
for AI. Um, that AI has always kind of
1:09:36
been the home of like like Windows and
1:09:37
to a lesser degree Linux. That's where
1:09:39
the development has always happened
1:09:41
historically. Um, it's always been
1:09:43
NVIDIA chips and NVIDIA's programming
1:09:45
interface and um, and now that's, you
1:09:48
know, uh, changing um, with with with
1:09:50
Apple Silicon and with the tooling
1:09:52
that's happened in the last few years.
1:09:53
you can do a lot with local models using
1:09:56
you know things like LM Studio or Olama
1:09:58
and other applications on your Mac and
1:10:00
you can even integrate that into you
1:10:02
know your your um coding applications
1:10:04
and so I think what we'll see is a mix
1:10:07
because you're right I don't think that
1:10:09
these labs can continue to subsidize
1:10:12
these things forever but the question
1:10:14
will become okay so this is giving me
1:10:17
great results and this is doing so much
1:10:19
for me is it so much better that I'm
1:10:21
that I can afford to spend
1:10:23
what this is costing or am I just going
1:10:26
to maybe go back to you know using
1:10:28
humans for some things maybe using um
1:10:31
less expensive models right I mean that
1:10:33
that is kind of the irony is I think
1:10:34
people are looking at the cost and going
1:10:36
okay I can just pay a person to do this
1:10:39
>> the problem is at least is what I
1:10:42
personally run into is in the last you
1:10:44
know six months like you become addicted
1:10:45
to the speed you become addicted to how
1:10:48
much more you can accomplish and I
1:10:50
sometimes have to kind of set myself out
1:10:52
of it and go no I don't have to use AI
1:10:55
for this. I can actually do this, you
1:10:56
know, myself, too. Maybe it takes
1:10:58
longer, but I can get a result and it's
1:11:01
it's going to be just as good or maybe
1:11:02
in some cases better. Um, I think it
1:11:05
just comes down to expectations. But
1:11:06
yeah, I think that that's going to be
1:11:07
the challenge that a lot of businesses
1:11:09
really face over the next year is okay,
1:11:12
we know that our developers finally are
1:11:14
liking and using these tools. How do we
1:11:16
manage the costs and maybe how do we
1:11:18
find like the right balance of like what
1:11:20
models we use for certain purposes? And
1:11:22
I really do feel like like local um um
1:11:26
models um have a great opportunity here.
1:11:29
The one caveat is that as all the money,
1:11:32
you know, that's being spent on data
1:11:33
centers happens, all the RAM, all the,
1:11:35
you know, uh SSDs, everything is going
1:11:38
into that. And so, you know, it's more
1:11:40
expensive than ever before to buy a
1:11:42
really beefy, you know, laptop or or or
1:11:45
Mac mini. We can't even get, you know,
1:11:47
Mac minis anymore. Um, and and like that
1:11:50
would be a great, you know, maybe
1:11:51
machine like the the Mac studios that
1:11:53
have the highest brand capacities were
1:11:55
the first ones that they cut. They were
1:11:57
like, "No, no, no. We're not going to
1:11:58
sell those anymore."
1:11:59
Um and and so you know like it it's kind
1:12:03
of like this this I guess at least until
1:12:07
something either more capacity breaks or
1:12:09
there's some other sort of change
1:12:11
there's just kind of this tension where
1:12:12
it's like okay everybody wants more
1:12:14
compute and we're still constrained by
1:12:15
those things and as a result the prices
1:12:17
are what the prices are.
1:12:18
>> Everybody everybody benefits from more
1:12:19
efficient models. I mean the the data
1:12:21
center has become less expensive now
1:12:22
that the game is disappearing from the
1:12:24
world like those won't need as much and
1:12:27
it'll power like I mean I think pretty
1:12:29
powerful things like if you can run all
1:12:31
these things on device then you know I
1:12:33
mean suddenly it's a software issue
1:12:35
where Apple the main limiting factor is
1:12:37
Apple just doesn't let you build apps
1:12:39
directly on an iPhone and iPad
1:12:41
>> right no absolutely absolutely and I
1:12:43
will say this too you know the prices of
1:12:44
the models right now are really high and
1:12:46
that is ah historic usually the prices
1:12:48
go down these have gone But I will say
1:12:50
like if you look and granted we only
1:12:52
have about five years of data, but if
1:12:53
you look at kind of what the progression
1:12:55
has been, you know, they tend to the
1:12:59
capabilities of the higherend models do
1:13:00
tend to kind of trickle down. And so I'm
1:13:02
hoping that we'll see that. There are
1:13:03
also some models like DeepSeek, you
1:13:05
know, um which uh their pricing is very
1:13:09
very competitive. their capabilities are
1:13:11
not going to be as good as what you
1:13:13
would get from the highest end, you
1:13:14
know, models from OpenAI or um uh
1:13:17
anthropic, but it the the pricing is is
1:13:20
is very efficient. Now, there are I know
1:13:22
people have their own kind of concerns
1:13:24
geopolitically about that, but those
1:13:26
models can be run on um you know uh you
1:13:30
don't have to get them directly from
1:13:31
deepseek. They can be hosted on other
1:13:33
you know like American or or European um
1:13:36
you know cloud hosts too. And I think
1:13:38
that that's that's also interesting to
1:13:40
see that there's like almost the these
1:13:42
kind of two kind of waring like factions
1:13:45
where you have the major frontier labs
1:13:47
that are getting, you know, more and
1:13:48
more powerful but more and more
1:13:49
expensive and then you do have like the
1:13:50
Chinese labs that are doing some really
1:13:52
interesting things with efficiency and
1:13:54
and with price.
1:13:55
>> I could pick your brain on this for
1:13:56
another hour but I know you've got to
1:13:57
get going. You've got a you've got a
1:13:59
real job on
1:14:00
>> Yeah. Yeah. Unfortunately. Yeah. So, uh,
1:14:02
uh, it's not WWDC, but Microsoft Build
1:14:04
is, uh, is next week. And there will
1:14:06
obviously be a lot of AI stuff, and
1:14:08
there might even be some, you know, um,
1:14:09
there's going to be definitely some
1:14:10
GitHub stuff, too. So, that's that's why
1:14:12
I have to unfortunately, uh, depart
1:14:14
because I have to have to go to a
1:14:15
meeting for that. But,
1:14:16
>> imagine you've got things you've got a
1:14:18
busy schedule. So, I guess if if you'd
1:14:20
like more Christina Warren, uh, you can
1:14:22
find her on Mastadon and Blue Sky at
1:14:24
Filmgirl and on the Mac Break Weekly
1:14:27
podcast uh, and at GitHub. Yeah,
1:14:30
>> people can find you. I don't know. Just
1:14:32
walk over to the GitHub headquarters,
1:14:33
knock on your door.
1:14:34
>> You could. Yeah. Or I'm film girl on
1:14:36
GitHub. I'm on film girl. One word on on
1:14:38
on on um Blue Sky and GitHub. Film_girl
1:14:42
on Twitter and uh Mastadon. I know it's
1:14:44
confusing. This is my bad. Um I again
1:14:48
like I didn't know when I created my
1:14:50
handle in high school that this would be
1:14:52
my professional identity. Mistakes were
1:14:54
maybe made. I don't know. But yeah. I
1:14:56
know. But if but but I think if you
1:14:57
Google film girl, you'll you'll find
1:14:58
most of those most of my things. I think
1:15:00
even like a christina.dev is a ancillary
1:15:03
website that I have set up. Um and I
1:15:06
need to add more things to it. And yeah,
1:15:08
I'm uh I'm on Mac break weekly every
1:15:10
Tuesday.
1:15:11
>> Okay, you heard that, Mom. Text us on
1:15:13
iMessage at [email protected]
1:15:15
to send in questions, comments, and
1:15:17
feedback for the show. You can send an
1:15:19
audio message or a short video for us to
1:15:20
play, too. Thank you all for listening,
1:15:22
for watching. We will see you all next
1:15:24
time. Have a great weekend. See you.
1:15:26
>> Bye.
1:15:35
>> Computers.
1:15:36
>> They were a mistake.
1:15:38
>> There's It's a bicycle for the mind.
1:15:41
>> Exactly.
1:15:44
>> There's the curb.
1:15:50
>> Then what's the what's the tandem
1:15:51
bicycle for the mind?
1:15:53
>> Oh god. That's the Vision Pro, isn't it?
1:15:55
Because you've got to be connected to
1:15:56
the battery pack and everything.
1:15:59
>> I think you're right. Yeah. I don't know
1:16:01
why they why they call it a spatial
1:16:02
computer and not a tandem bicycle for
1:16:04
the
1:16:11
wow. Is this the first time German has
1:16:14
ever done mock-ups like this based on
1:16:17
his
1:16:18
>> I was trying to think that. I I don't
1:16:19
think it's common. I think he's done it
1:16:22
like once before, but not He didn't make
1:16:24
like a big feature out of it.
1:16:26
>> Yeah. I mean, I I don't know. It just,
1:16:29
as people say, it just hits different.
1:16:33
>> It really does. I mean, I wonder if
1:16:34
Apple's going to go after him.
1:16:36
>> Well, it doesn't look like Well, they're
1:16:38
also Yeah. As Bloomberg, they're not
1:16:39
going to go after Bloomberg, but they're
1:16:40
crediting illustration 731. So, I wonder
1:16:43
if they used like a if another, you
1:16:46
know, party
1:16:47
illustrated that for them because that's
1:16:49
who the credits are to.
1:16:51
>> Well, uh, AI told me that that is
1:16:56
Bloomberg's in-house design.
1:16:58
>> Oh, amazing. Okay, great. Even better.
1:17:01
All right, that Okay. Yeah.
1:17:03
>> Uh, because they're located at 731
1:17:05
Lexington.
1:17:06
>> Oh, that's right. They are. Yeah. Yeah.
1:17:07
Yeah. Yeah. Yeah. Yeah. Yeah. I've I've
1:17:09
been I've been to the Bloomberg offices
1:17:10
a number of times. They uh they have a
1:17:11
really great cereal bar.
1:17:13
>> Oh yeah.
1:17:14
>> No, genuinely. Genuinely. Yeah.
1:17:16
Genuinely like like um in in like uh one
1:17:18
of the the main kind of prominent things
1:17:19
because I I used to go on Bloomberg
1:17:21
quite a lot and I would love like when I
1:17:23
would leave doing especially an early
1:17:24
morning hit, they'd have like this
1:17:26
massive cereal bar that was just kind of
1:17:27
you know take what you want of like
1:17:29
every type of cereal you could imagine,
1:17:30
every type of milk you could imagine.
1:17:32
You're just like okay cool.
1:17:35
>> Rat milk just like on the Simpsons.
1:17:39
>> Mulch. Mulch. Milk. Milk is what they
1:17:41
have now. Now, now now with like vitamin
1:17:43
in or whatever it was on the Simpsons,
1:17:46
but yeah, Rat Milk. Exactly.
1:17:49
>> I like the Do they Do they have like the
1:17:51
Moltome Meal like knockoff uh cereals as
1:17:54
well if you prefer the knockoff instead
1:17:56
of the original?
1:17:57
>> You know, I don't know. I I because I I
1:17:59
have a feeling they were probably all
1:18:00
the the brand names, but they weren't
1:18:01
they were it was kind of like in college
1:18:03
where like they're in like kind of the
1:18:04
dispensers. So I
1:18:06
>> it's not the same, but it doesn't come
1:18:07
out of a bag,
1:18:08
>> right? Well, yeah, but but but it's much
1:18:10
easier, you know, to to pour it into a
1:18:11
bowl or or a cup, right? Because that
1:18:13
was the thing, too, is like you could
1:18:14
have it to go, right? And so many times
1:18:16
I would like have like, you know, um
1:18:18
like, okay, I've got to get back, you
1:18:20
know, um lower um uh Manhattan, so I'm
1:18:24
just going to stop on the floor, grab
1:18:26
myself a little um cup full of uh you
1:18:29
know, Cheerios or something and some
1:18:31
milk.
1:18:32
>> Frosted mini spooners.
1:18:33
>> Correct. And then just be like and and
1:18:35
now I'm going to get on the subway and
1:18:36
and go a few stops. But yeah, I mean
1:18:39
>> we could not I mean look I think between
1:18:41
that and like the Bloomberg terminal
1:18:42
access I was like oh I this would be an
1:18:45
okay place to to to work. Um
1:18:49
clearly my standards are very low, but
1:18:51
yeah.
#Science