The Talk Show

193: ‘Crack Marketing Team’ — Live From WWDC 2017 With Phil Schiller and Craig Federighi


00:00:00   [ Background Sounds ]

00:00:03   >> Ladies and gentlemen.

00:00:05   [ Cheering ]

00:00:09   Thank you for joining us in San Jose's gorgeous California

00:00:13   Theater for tonight's live presentation of the talk show.

00:00:18   At this time, won't you please take a moment

00:00:23   to silence your cellphones and other electronic devices.

00:00:27   And then won't you please welcome to the stage my friend and yours John Gruber.

00:00:34   [Applause]

00:00:54   Hello and welcome to the talk show live from WWDC 2017.

00:01:02   We have a good show, I think. We have a beautiful theater, I know.

00:01:10   So something to get out of the way up front, this show would not have possibly happened without our sponsors.

00:01:18   We have three perfect sponsors for this show.

00:01:21   Our first sponsor is Jamf.

00:01:23   J-A-M-F.

00:01:26   Jamf specializes in Apple device management.

00:01:30   If you have mobile devices and a small company

00:01:33   and you need them, the MDM management type stuff,

00:01:38   go to them, they support Apple stuff.

00:01:40   They're Apple only.

00:01:42   They support Apple stuff before it's out.

00:01:43   They're already probably working on the betas for iOS 11.

00:01:47   It's absolutely great.

00:01:49   Go to jamf.com/talkshow, and you will find out more.

00:01:57   Our second sponsor, another great sponsor,

00:01:59   perfect for this audience, is MacStadium.

00:02:03   [APPLAUSE]

00:02:05   It is time to get your build server out of your office

00:02:08   closet or out from under your desk

00:02:10   and put it in the hands of some professionals.

00:02:12   Get a real build server.

00:02:13   You can't just go and get some no-name Linux server

00:02:17   and have Xcode running on it, you need a Mac.

00:02:19   Go to that Mac Stadium, knows how to professionally host

00:02:24   Mac OS X server in a great colocation environment.

00:02:28   All of their cloud accounts start with a 30-day free trial

00:02:35   that is production ready, so when your 30-day free trial

00:02:38   is over and you're like, I wanna pay, I wanna go with this,

00:02:41   you just roll right over and you're already got it.

00:02:44   visit maxstadium.com/df for more information.

00:02:51   [LAUGHTER]

00:02:55   Are you guys dying?

00:02:56   You guys probably want to know who's the guest?

00:02:58   [LAUGHTER]

00:02:59   Should I just tell you that and then I'll do the other sponsor?

00:03:01   [AUDIENCE REACTS]

00:03:03   Our third sponsor, great sponsor.

00:03:06   Are there any Indiv developers here in the audience?

00:03:09   [APPLAUSE]

00:03:12   Our third sponsor is SetApp.

00:03:14   SetApp, if you haven't heard about it, is a new subscription service for indie Mac apps.

00:03:20   You pay $9.99 a month as a user, and you get access.

00:03:23   They have over 70 apps in the service.

00:03:27   Think of it as like Netflix for apps.

00:03:30   It is quite a deal, and they've got a ton of great apps in there.

00:03:34   It's brought to you by MacPaw, a long-time independent Mac developer, the makers of CleanMyMac

00:03:42   and Gemini and a bunch of other stuff.

00:03:44   So they know the indie Mac market.

00:03:47   They are tried and true indie Mac developers.

00:03:50   It's a great service.

00:03:52   If you're a user, which is probably more people

00:03:54   than developers, go there and check it out.

00:03:56   And if you're a developer, certainly go there

00:03:58   and look at it and see if it makes sense for your app

00:03:59   to be part of their service.

00:04:01   Go to setapp.com and you can find out more.

00:04:06   And last but not least, one of the great traditions

00:04:11   of this show is I think every single time I've done it,

00:04:14   I don't know, I lost count,

00:04:15   I think it's like the sixth one, maybe seventh,

00:04:17   but I think every single time we've had an open bar

00:04:20   and that open bar has been sponsored by the same company,

00:04:23   the great people at MailChimp.

00:04:25   (audience cheering)

00:04:29   So if you enjoyed a few beverages before the show,

00:04:35   you can thank MailChimp for that and I thank them.

00:04:37   It's a great company.

00:04:38   If you have email marketing needs, go check out MailChimp.

00:04:40   It's terrific.

00:04:42   And I thank them for sponsoring the open bar here

00:04:46   'cause I certainly wasn't gonna pick up the bill.

00:04:48   (audience laughing)

00:04:50   I know how some of you drink.

00:04:51   I don't know what that was.

00:04:56   We'll get past it.

00:04:58   So yesterday when I was at the,

00:05:01   the last, there's one last group of people I wanna thank

00:05:05   and that's you, all of you.

00:05:07   Those of you in this room,

00:05:09   those of you listening at home,

00:05:11   yesterday during the press event,

00:05:13   there's all sorts of stuff backstage,

00:05:14   and there's briefings, and you have schedules,

00:05:16   and there's people in the press,

00:05:17   you have to sit around all day.

00:05:19   While you guys are in the State of the Union,

00:05:21   people like me are sitting around talking,

00:05:22   and I got to talk with Walt Mossberg.

00:05:25   (audience cheers)

00:05:29   But yeah, it was a great talk, it always is.

00:05:32   He's exactly what you think he's like in person.

00:05:35   He's full of very strong opinions.

00:05:38   (audience laughing)

00:05:40   But he was talking about what we do,

00:05:42   and I was telling him how, you know,

00:05:44   what I loved about him was that he was a columnist,

00:05:46   and that to me there was something in my head

00:05:48   about the style of writing that's writing a column,

00:05:50   and that's what I always wanted to do at Daring Fireball,

00:05:52   and he was very complimentary.

00:05:53   I said, "Yeah, I can see that.

00:05:54   "I think you'd do a great job."

00:05:55   But then he said, "You know what?

00:05:58   "I've really enjoyed these last few years is I've,"

00:06:01   he said, "I've really enjoyed podcasting."

00:06:03   And I said, "You know, I have too,

00:06:06   But from when I was a senior in high school until I actually got Daring Fireball as a

00:06:12   job off the ground, I thought I want to be a columnist.

00:06:16   I want to be a writer and I want to write something like this.

00:06:18   I never in a million years thought that I would be a broadcaster as something that I

00:06:23   should put on my tax return.

00:06:27   But I am.

00:06:28   There's no question that the podcast is half my job.

00:06:31   And I enjoy it and it is totally unexpected.

00:06:35   And for some reason, I love everybody who reads Daring Fireball.

00:06:39   If you read it and you don't really listen to the show, that's fine.

00:06:45   But for some reason, I feel like I have more of a connection with the people who listen

00:06:48   to the show, like bumping into people on the sidewalk here in San Jose.

00:06:53   And if you do see me anytime this week, please say hello.

00:06:56   But the people who say, "I love your show," seem more connected personally than the people

00:07:02   who say, "I love Daring Fireball."

00:07:04   I love both of them and it's all complimentary,

00:07:06   but I just thank all of you for being here.

00:07:09   I really cannot believe that I'm here

00:07:12   in this theater for this show.

00:07:14   (audience applauding)

00:07:18   I also feel like maybe listeners of the show

00:07:27   have more of a sense of my actual personality.

00:07:30   Like maybe my writing kind of conceals it

00:07:32   and it's a little bit more formal.

00:07:34   But if you do know me at all, you know that I,

00:07:38   when I find something I like, I like to keep going with it.

00:07:42   And so if you liked last year's show, I think you'll like,

00:07:48   (audience cheering)

00:07:51   I think you're gonna like this year's show too.

00:07:58   Let me introduce Phil Schiller and Craig Federighi.

00:08:03   [APPLAUSE]

00:08:26   Nice little place you have here.

00:08:29   So Craig, did you think my intro went on too long?

00:08:31   Is that why you ran?

00:08:32   Like, it's a habit.

00:08:34   All right, big secret thing here.

00:08:37   So whenever we do keynotes, Craig always bolts on stage.

00:08:42   He's full of energy and he runs out there.

00:08:44   And it's really impressive, right?

00:08:46   Because it just kicks things off, just like that.

00:08:49   And the rest of us don't.

00:08:52   And rehearsing for this keynote, someone who remained nameless

00:08:59   said, it's so great when Craig does that.

00:09:01   You should all run on stage.

00:09:04   And I said, no, because I'll trip and fall and be an idiot.

00:09:09   And then I'll regret it.

00:09:11   So it's his thing.

00:09:13   And it's a great thing.

00:09:14   And it's impressive.

00:09:16   Wow, thank you, Phil.

00:09:17   Thank you.

00:09:18   [APPLAUSE]

00:09:21   All right.

00:09:25   I don't know about you guys, I thought yesterday's keynote was longer than most podcasts.

00:09:34   Right about the same.

00:09:35   But we covered more stuff.

00:09:45   You covered a lot of stuff, which is going to make tonight very difficult.

00:09:49   - It also seemed as though there could have been more.

00:09:53   What was the first draft of the keynote?

00:09:57   - Three and a half hours.

00:09:58   (audience laughing)

00:09:59   Truth.

00:10:00   - It's true.

00:10:01   - So we had to cut.

00:10:03   In a perfect world, we would have liked to get to two hours.

00:10:06   It's a goal that we think it's kind of hard

00:10:08   to hold your bladder for some people longer than two hours.

00:10:12   But that's as close as we got.

00:10:13   We got close, but not there, and we cut a lot.

00:10:17   And we kind of ran out of this stuff.

00:10:20   We just don't want to cut any of this stuff.

00:10:22   We love it.

00:10:22   - I thought, running through stuff from the keynote,

00:10:27   I want to talk about the Mac first.

00:10:29   (audience cheers)

00:10:34   Because not just that yesterday was a very strong day

00:10:37   for the Mac on software, on laptop hardware,

00:10:41   on desktop hardware, but in the recent months

00:10:44   with discussions we've had and your announcements

00:10:47   about the Mac Pro, I feel like the Mac,

00:10:50   not that it's in a different place,

00:10:51   but it's certainly in a different place

00:10:55   perception-wise now than a year ago.

00:10:57   I thought, here's one thing,

00:10:58   I would just throw this off the bat,

00:10:59   I thought that the Mac stuff alone yesterday

00:11:03   would have made a pretty good Mac World Expo keynote

00:11:06   back in the day.

00:11:06   (audience laughing)

00:11:09   - We used to find a way to stretch things out.

00:11:12   (audience laughing)

00:11:14   - I don't know about Mac World,

00:11:15   - At least at Apple Expo Paris or Japan.

00:11:17   (audience laughs)

00:11:20   - So starting with Mac OS, so when you,

00:11:24   Craig, when you introduced Mac OS High Sierra.

00:11:28   - Uh-huh.

00:11:28   (audience laughs)

00:11:32   Where you going with this?

00:11:33   - I was seated in the press area with friend of the show,

00:11:36   sometimes guest of the show, Serenity Caldwell

00:11:38   was right next to me, and you were 30 seconds into it

00:11:42   after the name and she says to me,

00:11:45   "I can't believe it, I don't think he's gonna make

00:11:46   "a high joke."

00:11:48   - She doesn't know Craig that well.

00:11:51   - As soon as she said that to me, you made the joke.

00:11:55   It's a tradition.

00:11:58   I was very happy about this announcement

00:12:00   because, if people listen to the show,

00:12:03   my hope for the Mac is, look, the Mac has all the major

00:12:06   tent pole style features that it needs.

00:12:08   I think what it just needs is sort of refinement.

00:12:12   And is that, that's what the name High Sierra

00:12:15   sort of sets the bar for?

00:12:17   - It's a proud, it could have been Snow Sierra or something.

00:12:20   (audience laughing)

00:12:21   We've done that.

00:12:22   There's a proud tradition of Mac releases

00:12:24   that I think some of our most loved releases sometimes

00:12:27   are when we take a year to refine and perfect

00:12:30   and we wanted to do it again.

00:12:31   - Mountain Sierra wouldn't have worked.

00:12:34   - That's the thing.

00:12:35   It was Snow Sierra, Mountain Sierra,

00:12:36   where do we go with this?

00:12:38   our crack marketing team.

00:12:40   (audience laughing)

00:12:43   - First of all,

00:12:46   (audience cheering)

00:12:49   first of all, I don't appreciate when you use

00:12:52   the crack word and point to me.

00:12:54   (audience laughing)

00:12:57   But second of all, it is a sad state

00:12:59   when you're naming products and you're more afraid

00:13:02   of what your head of engineering is gonna say about it

00:13:04   than all the press and all the customers.

00:13:07   But truth be told.

00:13:09   - Now, it's one of those things,

00:13:15   and it happened repeatedly through the keynote,

00:13:17   and this is how you can tell it was a JamPad keynote,

00:13:19   is one of the things that you guys do in a keynote

00:13:21   is when you get to the end of a segment,

00:13:23   and you still have 20 or 30 other things

00:13:26   that are new about this subject,

00:13:28   you'll put up like a catch-all slide.

00:13:30   And in the press area, people like tend to

00:13:33   take their phone out or if they have a standalone camera,

00:13:35   and then they'll point up and they'll take a picture of that

00:13:37   and then in their notes at the end they have these,

00:13:39   go through all those catch up slides,

00:13:41   see if there's anything interesting.

00:13:42   And it went by so fast where it kept happening

00:13:45   where people would be like, oh, it's gone.

00:13:47   (audience laughing)

00:13:48   But one of the little things you called out

00:13:50   was a 35% improvement in the efficiency of storage of mail.

00:13:55   - Yeah.

00:13:56   - And that's the sort of thing

00:13:58   that's like music to my ears about this

00:14:00   because to me it's not that there's all these people

00:14:02   out there whose MacBooks are stuck to the limits

00:14:05   of the drive with mail, but that's the sort of thing

00:14:09   that only happens when you really go through the code

00:14:12   and do a refactoring.

00:14:16   - Yeah, yeah, well, and really starting out this year,

00:14:19   every team went and said, hey, what do we wanna make faster

00:14:24   in ARIA, the code, the Finder guys were like,

00:14:26   hey, it should be a little faster to open a Finder window.

00:14:29   So they put some focus on that.

00:14:32   Every demo of the Photos team,

00:14:34   they bring in the slowest Mac they could find

00:14:35   and show me how fast it was launching.

00:14:37   It's like that, here's 100,000 photo library,

00:14:42   let's see how fast it launches, right?

00:14:44   So yeah, when you start seeing stats like that,

00:14:47   it is a sign of everybody in engineering

00:14:52   putting a focus and going deep in their area.

00:14:54   - And you're saying that this is the type of release

00:14:57   where teams can say, here's what we would like

00:15:01   to throw effort at, 'cause we're not happy

00:15:05   with the performance of this part of the system.

00:15:09   If you can give us the time and let us do it,

00:15:11   we can really get that going.

00:15:12   - Yeah, absolutely.

00:15:13   I mean, certainly at Apple, there's a real blend of saying,

00:15:17   hey, we're coming out with a new machine, a new iMac Pro

00:15:21   with a really interesting architecture.

00:15:22   We all have to do our part to make that possible,

00:15:26   or if you look at the iPad Pro and what it took

00:15:28   to do promotion, huge effort.

00:15:30   So this, yeah, it's awesome.

00:15:33   - You're skipping ahead, iPad.

00:15:34   - Sorry, sorry Mac.

00:15:36   But so there's definitely a fair amount

00:15:41   where we have goals as a company and as a release

00:15:44   where we ask all teams to pull in what's so awesome

00:15:47   about Apple is the teams will all rally to the cause.

00:15:51   But at the same time, this release, we said,

00:15:53   listen, here's 50% of the time off the top.

00:15:57   Tell us how you just want to make your stuff better.

00:15:59   And teams took to it.

00:16:01   It's great to have great people.

00:16:04   Safari got a pretty good chunk of that.

00:16:09   And it seemed like you were-- just flat out said,

00:16:14   Safari is faster than Chrome.

00:16:17   You helped prompt me to do that.

00:16:19   Yeah?

00:16:19   Yeah.

00:16:20   Yeah.

00:16:21   It was a few weeks ago that you wrote something about Safari

00:16:25   and you were, you complimented Safari in one regard

00:16:29   and then you said, yeah, and it's okay

00:16:31   that Safari's not the fastest.

00:16:32   And I'm like, what?

00:16:34   (audience laughing)

00:16:36   And I realize when every time one opens their browser

00:16:39   and goes to a particular search engine

00:16:41   that there's an ad that says get a faster browser

00:16:44   that eventually it seeps in and people stop

00:16:46   questioning where that's coming from.

00:16:50   And so we thought we'd bring some knowledge.

00:16:55   And it's all true, man.

00:16:57   I mean, that team is unbelievably obsessive

00:17:01   about performance.

00:17:01   And they're absolutely the best in the industry.

00:17:05   And the Safari team rules, WebKit team,

00:17:08   the combination of them.

00:17:09   And they build the fastest browser on the planet.

00:17:12   And honestly, I'm getting sick of people not giving them

00:17:16   their due.

00:17:17   So we had to say it.

00:17:19   (audience cheering)

00:17:21   Go Safari team.

00:17:22   - Beware a pissed off engineer.

00:17:29   - And I thought that the Safari news yesterday

00:17:36   hit on the major themes that Apple has been,

00:17:40   I mean performance is obviously something

00:17:42   you guys always care about, so there's that,

00:17:43   we just covered that.

00:17:44   But then the other angle is the privacy angle, right?

00:17:49   And so there's a new intelligent tracking prevention, and you talked about it, but can

00:17:56   you nerd out on us a little more here about what's intelligent about it?

00:18:00   Yeah, you bet.

00:18:02   You know, actually years ago, Safari was the first browser to have these mechanisms to

00:18:09   try to prevent cross-site tracking.

00:18:13   And there became a point where the tracking industry, the tracking industrial complex

00:18:20   is pretty inventive.

00:18:23   And they came up with some pretty wild mechanisms.

00:18:27   I mean, for a while it was how can we use storage and flash and how can we use any API

00:18:31   you can imagine to try to maintain a tracking cookie effectively across everywhere you go

00:18:37   on the web.

00:18:39   And the Safari team has been really drawing deep to try to solve this problem.

00:18:46   And we figured out that through a combination of essentially double keying the cookie, saying

00:18:53   if you pulled a resource and you were on -- I don't want to pick a particular publication

00:18:59   -- you were on publication A and it included some JavaScript that tried to pull content

00:19:03   from this tracker, that instead of allowing that to be keyed

00:19:09   by the tracker's domain, which would be the same domain when

00:19:11   you pulled that tracker from another site,

00:19:14   we'd segregate it.

00:19:14   We'd say, oh, well, this tracker's

00:19:16   going to see a different cookie when you're on site A

00:19:18   than when you're on site B than when you're on site C.

00:19:19   So we started siloing it.

00:19:21   But these guys are even more insidious than that.

00:19:24   And so we had to actually use some local machine learning

00:19:27   to figure out which are trackers,

00:19:30   to partition their data, to delete their data.

00:19:33   But all of this is really just to meet,

00:19:36   I think, what is many reasonable people's expectation, right?

00:19:39   - Right. - When I'm on this site,

00:19:40   I know I'm seeing their ads,

00:19:41   I don't expect that there's any record anywhere else

00:19:44   about what I was looking at.

00:19:46   And we think that's the way the web should be,

00:19:51   and the Safari team had to work incredibly hard

00:19:52   to make that so.

00:19:53   (audience applauding)

00:19:56   (Applause)

00:20:02   I use the word creepy a lot talking about the tracking,

00:20:05   and I do think it's creepy, and I feel like somehow

00:20:08   because it's online on a screen and it's computerized

00:20:12   and nobody really knows what happens,

00:20:13   that people aren't creeped out in the way that they would be in the real world.

00:20:17   Like if I go to Macy's and I'm looking at jeans,

00:20:23   And I decide, eh, I don't like them.

00:20:26   I'm not going to buy any jeans.

00:20:27   And then I go into another store and somebody comes up to

00:20:29   me and says, would you like some jeans?

00:20:32   Right.

00:20:33   That's creepy.

00:20:36   I'm immediately thinking, how do you know this?

00:20:40   What is going on here?

00:20:41   Let me see the jeans, though.

00:20:42   But I am creeped out.

00:20:46   And that's why it works.

00:20:47   Because maybe, obviously sometimes it works.

00:20:50   But I do think that there should be a

00:20:53   a reasonable expectation of privacy in that regard.

00:20:55   And just putting your browser in private mode

00:20:59   isn't a great solution to that because then you

00:21:01   lose all sorts of helpful user features that are based on,

00:21:05   hey, we're remembering stuff.

00:21:07   Yeah, and that was actually the hardest part about it.

00:21:10   It's easy to do if you're willing to completely break

00:21:12   the web.

00:21:13   Right.

00:21:14   But what's the point of that?

00:21:18   But the kind of thing that-- it's not just I go here

00:21:22   and I look at the jeans, I mean it's,

00:21:24   you go here and then they say, hey, we noticed that your

00:21:28   car, your lease on your car is nearly up.

00:21:32   And they don't say that, but they know that, right?

00:21:34   They're like, we know you own, I don't know what you own

00:21:37   if you own a car, but, do you own a car?

00:21:40   - I was gonna buy the Apple car.

00:21:41   - Okay, so, at the time,

00:21:45   (audience applauding)

00:21:47   well done.

00:21:51   But we're going to have to work on that then.

00:21:53   If-- you know.

00:21:55   [LAUGHTER]

00:21:57   They know.

00:21:59   They're correlating, hey, we know what you bought

00:22:02   at the grocery store, and we know this,

00:22:03   and they tie it together.

00:22:05   And it's like, yeah, it's creepy.

00:22:07   Well, it's great that Safari is working on it.

00:22:08   And then the other one, too.

00:22:10   And it got a huge--

00:22:11   I never know what you guys can tell.

00:22:12   Here, we get pretty good audience feedback.

00:22:14   But--

00:22:15   [LAUGHTER]

00:22:15   It's very direct.

00:22:17   The auto block feature got a huge applause,

00:22:22   and it's because people are so,

00:22:24   people use their MacBooks all over the place,

00:22:27   whether it's work or a library or something,

00:22:29   and when all of a sudden you click a link,

00:22:30   somebody's like check this out,

00:22:31   and you get all this sound, it's embarrassing.

00:22:34   - It is a great way to tell who's not

00:22:36   paying attention in a meeting.

00:22:38   (audience laughing)

00:22:41   I kid you not, literally, while we were reviewing

00:22:46   this feature, the slides for this feature in one of our rehearsals, somebody's browser

00:22:51   goes off playing an ad.

00:22:53   And this is someone who shall not be named who is not yet running the beta.

00:22:57   And it was a perfect proof point.

00:23:00   >> I'm going to make a bet.

00:23:01   I'm going to bet it was Eddie.

00:23:11   Eddie is one of our best beta testers.

00:23:14   He was already on the beta.

00:23:17   Eddie is aggressive with the betas.

00:23:20   In fact, in most of our executive team meetings,

00:23:23   Eddie is updating two or more devices on the boardroom table.

00:23:27   He's going to spread out that morning's update.

00:23:30   It's true. This is really true.

00:23:33   Is there anything in macOS High Sierra

00:23:37   that either didn't get time on stage

00:23:39   on stage or didn't get enough that you would like to talk about?

00:23:42   Oh boy. Yeah, you know, on the way over here I was tapping out notes to myself on all the

00:23:50   things I was going to forget to say when I got here. I feel like I'm not going to pull

00:23:54   my phone out right now and look at that list.

00:23:57   Oh, that's all right. What about the face syncing across devices? Now that's something

00:24:06   that we talked about last year.

00:24:10   And it was per device.

00:24:14   And so you'd get a new device.

00:24:16   And it even-- again, throughout the year,

00:24:19   it played into when the new MacBook Pros came out last fall.

00:24:25   And some people were saying, eh, my battery life

00:24:28   isn't that great.

00:24:28   And you guys looked into it.

00:24:29   And part of it is the first run experience.

00:24:31   And it's redoing spotlight indexes and this stuff.

00:24:36   - Clearly you guys, it wasn't there last year

00:24:40   simply because it wasn't done.

00:24:41   It wasn't like you didn't have it on the.

00:24:44   - No, we've been working on it.

00:24:45   And it's actually something that's harder to get right

00:24:48   than you think because when you go through the process

00:24:51   of classifying your photos, we offer up and we say,

00:24:56   hey, is this John?

00:24:57   And then, you know, is this John?

00:24:58   And you say yes, yes.

00:24:59   But when you say yes to one face,

00:25:02   we're actually saying, okay, well there are a thousand

00:25:04   of the photos we saw that we think are the same person

00:25:07   as this, so we're going to count those all as John.

00:25:09   But all you really told us is about that one photo.

00:25:13   And so what we actually sync-- and it's important

00:25:16   that we get this right-- is we sync only the photos,

00:25:19   the specific photos you told us about.

00:25:21   And then each-- because that's really the true data, right?

00:25:24   That's what the user really told us.

00:25:26   And then we sync that information.

00:25:28   And then each device, because they're independently

00:25:30   doing photo recognition, are taking that

00:25:33   and re-deriving the coming to the common result

00:25:36   on all your machines to synchronize.

00:25:37   So we synchronized actually very little data

00:25:39   and it's all just the truth data

00:25:41   and then we're able to pull back the right solution.

00:25:44   But getting that architecture completely right

00:25:46   was, took a little longer than we would have liked

00:25:50   and so we're super excited to have it done now.

00:25:52   - But do you think it's safe to say,

00:25:53   I think that in broad strokes there are people

00:25:56   who think that while Apple places a very high priority

00:26:01   on privacy and there's features that are useful

00:26:04   to have going through the cloud

00:26:07   and Apple therefore misses out on some features

00:26:10   because of their stance on privacy

00:26:12   because they don't want it to go through the cloud.

00:26:15   I get the impression that it's more that sometimes

00:26:17   it takes longer to do it because you're only gonna do it

00:26:20   in a way that is very, very,

00:26:23   it takes longer to do it with privacy.

00:26:25   - I think it takes longer to sometimes to do things right

00:26:27   and I think we've come up with architectures

00:26:30   and we started to talk about some today,

00:26:33   or I mean it's a day, God, it's been a blur,

00:26:35   but yesterday I guess it was, wasn't it?

00:26:37   That about how we're actually able to sync some information

00:26:42   through the cloud like what Siri derives about you

00:26:46   in a way that's completely secure and then encrypted.

00:26:49   So we're starting to build some of the tools

00:26:52   that enable us to keep all the privacy properties we want

00:26:55   but have some of those cloud

00:26:59   characteristics. And so I think we'll be getting faster and faster in this regard, but we've got to preserve privacy along the way.

00:27:07   [Applause]

00:27:15   It was announced last year, APFS, but it's out now. It's on my phone. It's been on my phone for a couple of months.

00:27:24   What was that like when that iOS update rolled out?

00:27:29   To have a .3 update that changed the file system.

00:27:34   And the other thing too is you guys,

00:27:41   Tim had the slide where 90% of iPhone users

00:27:45   are on the latest.

00:27:46   So they're gonna get upgraded.

00:27:48   - Our file system team is unbelievable.

00:27:51   What they, yeah, they deserve.

00:27:55   (audience applauding)

00:27:58   What they pulled off in a couple of years,

00:28:03   I think any comparable benchmark of file systems

00:28:07   in the past has probably taken a decade.

00:28:08   And when, I mean their degree of automation and rigor,

00:28:13   I mean they're coming in saying,

00:28:15   okay we've gone from five nines to six nines

00:28:17   on the reliability of this process.

00:28:20   We actually had this process running for earlier iOS updates

00:28:25   where when you updated to 10.1 or 10.2,

00:28:30   we were trial migrating your whole file system,

00:28:33   consistency checking it, reporting back to us

00:28:36   whether the upgrade was 100% clean,

00:28:38   and then rolling it back.

00:28:40   (audience cheering)

00:28:49   - There's the old carpenters saying,

00:28:51   measured twice, cut once.

00:28:52   (audience laughing)

00:28:53   With a file system change for a,

00:28:56   Phil, how many iPhones are in use?

00:28:59   - Over a billion.

00:28:59   (laughing)

00:29:02   - So it's more like measured--

00:29:05   - So remember that video at the beginning?

00:29:06   - Right. - The apocalypse?

00:29:07   - Right, yeah.

00:29:09   - Yeah, it got it wrong.

00:29:10   - That's what it could have been had it not worked.

00:29:12   But it was absolutely the,

00:29:15   I mean, never in the history of file systems

00:29:18   Has there been a growth from zero to hundreds of millions,

00:29:23   almost literally overnight?

00:29:25   - All right.

00:29:28   (audience applauding)

00:29:31   The other half of the Mac story is hardware,

00:29:37   and yesterday was, in my opinion,

00:29:40   a huge day for Mac hardware.

00:29:42   Now, you spoiled a bit of it a couple of weeks ago.

00:29:47   (audience laughs)

00:29:48   But I actually feel like that round table discussion

00:29:52   and the stories that came out of it

00:29:54   actually made yesterday's keynote

00:29:57   a little bit more dramatic,

00:29:58   because it seemed to me that as John Turnus was up there

00:30:01   introducing the new iMacs,

00:30:05   and talking about the higher end configs

00:30:09   that are available,

00:30:10   it sounded to me like he might be describing

00:30:12   the iMac configurations more like pro configurations

00:30:17   that you openly discussed at that meeting.

00:30:20   And I was like, well maybe that's,

00:30:21   maybe they're not gonna call it like iMac Pro.

00:30:24   They're just going to keep beefing up these systems.

00:30:27   And there were other people in the press,

00:30:29   you know we don't talk a lot, we're trying to listen,

00:30:31   but it seemed like everybody was like, is that the,

00:30:33   that's the--

00:30:33   - Yeah, it was like 64 gigs of RAM and 5.5 teraflops.

00:30:36   I mean it was a meaningful, it is a pretty awesome machine.

00:30:42   The VR demo on the planet Mustafar,

00:30:46   where the woman was meeting Darth Vader,

00:30:49   that was running on the iMac

00:30:51   that you could go to the Apple Store and buy today.

00:30:53   - That's right. - Right?

00:30:54   - Yeah.

00:30:55   - And then came the iMac Pro.

00:30:59   (audience laughing)

00:31:01   - Is that the name of this podcast?

00:31:03   And then came the iMac Pro.

00:31:04   - And everybody immediately in the press started saying,

00:31:09   well this, wait, are they announcing the Mac Pro?

00:31:12   that they were going, this is the Mac Pro.

00:31:16   I got a text from someone, I got a text from Bill Evans

00:31:22   at Apple PR, like 10 seconds after this happened,

00:31:25   he goes, this is not the Mac Pro, by the way.

00:31:27   (audience laughing and applauding)

00:31:31   But, talk to me about the iMac Pro,

00:31:36   and who do you think it's for,

00:31:39   Where is this, who is this for that the already pretty pro

00:31:44   Core i7 27-inch iMac isn't enough for?

00:31:50   - Well, I think that there's a bunch of customers already

00:31:55   that use the iMac Pro for pro work.

00:31:57   In fact, we talked about this in the round table,

00:31:59   that if you look at the total universe of pro users on Mac,

00:32:03   the biggest group is on MacBook Pro.

00:32:06   The second biggest group is on an iMac, the current iMacs,

00:32:11   and third biggest is on Mac Pro.

00:32:14   And so we have a bunch of people already happy

00:32:17   with what iMac does today, and they use their iMacs,

00:32:21   and so now they get more with the faster iMacs.

00:32:25   The hope is, and the belief is, that there are customers

00:32:29   across all of those who want more performance,

00:32:33   and they like all-in-ones.

00:32:35   I mean, that's the Mac started, that's its life.

00:32:38   It began, it's all in one is the beauty of it,

00:32:41   the integration of it.

00:32:42   And if we can give them the kind of performance

00:32:44   you expect from a tower desktop in an iMac,

00:32:48   there's a base of customers who want that

00:32:50   more than the current iMac has.

00:32:52   And they're pushing the boundaries of video editing

00:32:56   and 3D rendering and doing AI work

00:33:01   and want massive fast build times.

00:33:04   and to them there isn't enough power.

00:33:06   As much as you can give them, they can use.

00:33:08   And that there are customers for that.

00:33:10   How many, we don't know, do we start shipping and see,

00:33:12   we think we have a handle on what that is,

00:33:14   we forecast what we planted,

00:33:16   but it's significant to the Mac business,

00:33:20   and we'll see to the universe of computing,

00:33:22   but already the response have been getting emails

00:33:25   and tweets and posts from people who are saying,

00:33:28   I'd never imagined I could get an iMac like this,

00:33:30   this is what I want, so we'll see.

00:33:32   - Yeah.

00:33:33   The other part of the hardware story,

00:33:35   and there's also new MacBook Pros,

00:33:36   and the new MacBook,

00:33:39   and a megahertz improvement to the MacBook Air.

00:33:44   (audience laughing)

00:33:47   I think you said it gets more megahertz.

00:33:52   - It does, we put several more in the box.

00:33:54   (audience laughing)

00:33:57   - But I think a big part of the hardware news

00:34:00   that came out yesterday is the eGPU story,

00:34:04   external GPU.

00:34:05   (audience applauding)

00:34:08   - Got some Thunderbolt fans in the audience.

00:34:10   - Talk to me about the work that goes into that,

00:34:15   where it's not,

00:34:16   it must have a big part to do with your team,

00:34:20   where it's not just find some of the electrical engineers

00:34:24   and say, give us a GPU and we'll plug it in to Thunderbolt

00:34:28   and graphics will get better.

00:34:30   It doesn't work like that.

00:34:33   Yeah, that's pretty much how it went.

00:34:37   No, I mean, there are a whole bunch of things.

00:34:39   I mean, certainly, at the level of-- a Mac has never

00:34:44   had a GPU disappear on it.

00:34:46   I mean, not certainly in the way that it was going to recover from.

00:34:50   And so at a whole system architectural level

00:34:53   and a software architectural level,

00:34:54   the idea that a GPU becomes something

00:34:57   that could appear on the system and could disappear

00:34:59   on the system and that that's okay,

00:35:01   (audience laughing)

00:35:04   is a very big thing to sort out at many levels.

00:35:11   And so we had to work on that.

00:35:13   The other thing that was true not just in our stack,

00:35:16   but will be true in third party apps as well,

00:35:20   is Thunderbolt has actually

00:35:22   really pretty substantial bandwidth,

00:35:25   but the latency is higher

00:35:26   than if you're right on the local bus.

00:35:28   And so certain graphics workloads

00:35:32   have done a really good job of managing.

00:35:34   I'm going to move all my textures over there.

00:35:36   I'm going to do a relatively few number of round trips

00:35:39   between that GPU that's doing a ton of work on the data

00:35:41   that has cached over there and the CPU.

00:35:44   And actually, a lot of the VR engines, the big gaming

00:35:47   engines like the ones from Epic and Unity,

00:35:49   actually perform pretty great on any GPU.

00:35:51   It's almost as if it's right on the bus

00:35:53   because they're so well-tuned for how they move the data.

00:35:56   But there are a lot of apps that are not nearly so disciplined

00:36:01   in how they manage those round trips.

00:36:02   And so that latency can be problematic.

00:36:05   Now, when we work with them, we can find ways

00:36:07   to tune that out of the system.

00:36:10   But out of the gates, we know that's

00:36:12   going to take some time.

00:36:13   And so that's why we wanted to first get it out there,

00:36:16   have a developer kit, have the developers who

00:36:18   have pro apps that really could scale with the GPU,

00:36:22   and give them time to work with us

00:36:24   and use the tools we provided to optimize their apps

00:36:27   so that when customers get any GPU,

00:36:30   they actually experience the benefits.

00:36:32   - The other thing is that,

00:36:36   we talk about GPUs and it's obviously graphics

00:36:41   is the G in GPUs and you think of things like AR and VR

00:36:44   and it's no surprise that they were

00:36:46   major parts of the demos yesterday,

00:36:52   'cause they demoed well.

00:36:54   and they are graphics.

00:36:56   But the other thing that's going on

00:36:58   in the world of computer science at large

00:37:01   is that all of this machine learning work

00:37:04   is going through GPU processing, not CPU processing,

00:37:08   'cause that's just, I don't know.

00:37:11   But it's beyond, it's over my pay grade

00:37:14   of how I understand how computers work.

00:37:16   But the eGPU is gonna be a major factor in that too, right?

00:37:19   - That's right.

00:37:20   I mean, I put up a chart of the growth and performance

00:37:24   of GPUs, we didn't show the line for CPUs,

00:37:26   but it wasn't quite as impressive.

00:37:28   (audience laughs)

00:37:29   Right?

00:37:30   Because GPUs are a case where as we've been able to shrink,

00:37:34   increase transistor density, you can essentially

00:37:37   just throw more and more transistors at the problem

00:37:40   of graphics processing and it pretty much scales up.

00:37:43   It's just a very parallelizable task.

00:37:45   And it turns out that if you want teraflops of performance

00:37:49   to run a machine learning model,

00:37:52   it's you can parallelize that on a GPU

00:37:54   and you can get tremendous wins.

00:37:56   And the CPU just can't compare.

00:37:59   And it kind of, if you look at the trajectory

00:38:01   on the road forward, you're gonna continue to see

00:38:03   that kind of path, just because the nature

00:38:04   of the parallel computation of a GPU.

00:38:07   And so an eGPU is absolutely an awesome solution for that.

00:38:12   And that's also why it's important that we're putting

00:38:14   these kind of big GPUs in systems

00:38:16   like the high-end iMac and iMac Pro.

00:38:19   For some tasks, it's absolutely the best tool for the job.

00:38:21   And I think providing APIs like Metal and now Metal 2,

00:38:28   for any of you developers out there,

00:38:30   if you have a computationally intensive task,

00:38:32   the absolutely the rocket fuel you have

00:38:36   is to move that task onto the GPU if at all possible.

00:38:39   And I know you don't want to skip to iPad,

00:38:41   but if you saw that Affinity Pro demo, stunning what they've

00:38:45   been able to achieve compared to, you know,

00:38:47   they're blowing away what people are doing

00:38:48   in many cases on desktops because they're making

00:38:50   such efficient use of metal and the power of the GPU

00:38:54   to do work that in years past,

00:38:56   people tried to write that code on the CPU

00:38:59   and they're just seeing unbelievable acceleration.

00:39:01   - Also, I don't wanna lose the fact that

00:39:04   when you think about how we're doing the rollout

00:39:06   and the plan of the eGPU to developers first

00:39:09   and then customers later, this isn't about just,

00:39:12   hey, get some great Twitch gaming on Mac.

00:39:14   This is about a pro strategy.

00:39:17   And so just like we launched the iMac Pro,

00:39:19   and we're doing work on the Mac Pro,

00:39:21   and we're doing Metal 2 work,

00:39:23   the eGPU was about what would a pro user want?

00:39:25   How do they want to get it?

00:39:26   And what apps do they need,

00:39:28   and how long is it gonna take for those to be tuned

00:39:29   so they can count on them to get their job done?

00:39:32   This isn't just a quick throw it out there thing.

00:39:35   So we did it as another example

00:39:36   on the list of things we're doing for pro customers.

00:39:39   - Yeah.

00:39:39   (audience applauding)

00:39:43   (audience applauding)

00:39:46   One of the things I loved about the Star Wars demo,

00:39:51   well I mean number one, I loved that it was John Knoll.

00:39:54   I was like why isn't the audience storming the stage,

00:39:57   it's John fucking Knoll.

00:39:59   The guy invented Photoshop.

00:40:01   Holy shit, here he is.

00:40:03   - Thank you all for holding back on that.

00:40:05   (audience laughing)

00:40:08   - But the thing I loved about the demo is sure,

00:40:11   I like Star Wars, I can imagine it.

00:40:13   It'd be very fun to put some VR on

00:40:16   and have a fight with Darth Vader in a game or something

00:40:19   with this high graphic showing off the graphics.

00:40:22   But what I liked about the demo was

00:40:24   that she wasn't playing a game, she was doing creative work.

00:40:28   It was creative professional work of moving stuff.

00:40:31   And you can see, it's like what PageMaker or QuarkXPress

00:40:36   was to doing graphic design on screen.

00:40:40   this is what it's like to do 3D creation.

00:40:42   It makes sense that you'd want to create this 3D stuff

00:40:46   in a 3D view like this.

00:40:47   So I love that the demo was really more of

00:40:50   like a really cool creative app.

00:40:54   - Yeah, you're totally right.

00:40:55   I mean, it is funny that we've been doing

00:40:57   almost our 3D creation in 2D.

00:41:00   - Right. - Right.

00:41:01   And so to actually put yourself as in the real world

00:41:04   in a 3D environment in order to look at what you've created,

00:41:07   I mean, instead of this cycle where I'll try to manipulate

00:41:09   it in 2D and then I'll try to render on it to now be able to interact with it in 3D and

00:41:17   to get that stereoscopic sense of it live as you're moving.

00:41:22   It's instant feedback.

00:41:24   It's such a productive way to work.

00:41:26   And this is great for VR.

00:41:28   It's also actually, you know, it's one of these hard things to get through in a demo

00:41:32   on a flat screen, but also in AR when you can have an object and walk around it and

00:41:38   get closer and do these kinds of things in real space,

00:41:42   it gives you this sense of what you're working on

00:41:47   that's really tough to capture in a traditional environment.

00:41:50   And so as Phil says, this really,

00:41:52   our big inspiration in supporting VR

00:41:55   is about content creation.

00:41:57   - I think the other thing I got out of yesterday's keynote,

00:42:01   and it's not, you know, every year,

00:42:02   it depends on what you have to announce.

00:42:04   And sometimes it's more developer-oriented,

00:42:06   and sometimes it's not,

00:42:07   just because that's what's ready in early June.

00:42:10   But I thought yesterday's was a very, very developer

00:42:13   first keynote.

00:42:15   In terms of-- there's less of an emphasis,

00:42:18   especially on the Mac side, I think, in some ways,

00:42:21   of tent pole features for users.

00:42:24   And it was more about new frameworks for developers,

00:42:26   like ARKit.

00:42:27   I mean, ARKit's a big one.

00:42:29   Is it called-- is there a VRKit or no?

00:42:31   It's just ARKit.

00:42:32   But--

00:42:33   It's Metal.

00:42:34   Right.

00:42:34   And the Metal 2.

00:42:35   And Metal 2.

00:42:36   Right.

00:42:37   In other words, and what I heard was that

00:42:39   that when the hands-on area opened up at four to attendees,

00:42:44   the number of developers who were trying to, you know,

00:42:48   get, see what this is like,

00:42:50   the enthusiasm was just through the roof

00:42:52   and the line was out the door down the hall.

00:42:55   - Yeah, well I mean, it was evident really in presenting it

00:42:58   just, you know, when we got to the slide with AR on it,

00:43:01   I mean, the room started to lift a little bit

00:43:05   And it's clear there are so many creative minds out there

00:43:10   that are just itching to take advantage of what we've done.

00:43:13   And we really saw that with the,

00:43:16   sometimes you work on something and you know,

00:43:18   you have a feeling how important it can be,

00:43:20   how big it can be,

00:43:22   but then you start bringing a couple early developers in

00:43:24   and they start taking a look at it and you're like,

00:43:27   wow, this is gonna be even bigger than we thought.

00:43:29   And that's the reaction.

00:43:31   So that's what's so exciting about getting it out here now,

00:43:35   suddenly millions of developers really are gonna have

00:43:38   an opportunity to kick the tires on this

00:43:42   and do a whole bunch of things with it

00:43:44   that we really couldn't even imagine.

00:43:46   - And to tie a knot on the Mac segment,

00:43:48   I really felt like yesterday's combination

00:43:50   of software announcements, developer announcements,

00:43:53   not just features in Mac OS, but developer frameworks

00:43:58   and things that developers can create in Mac OS

00:44:00   combined with, I think, a very, very impressive

00:44:03   hardware story that still doesn't involve the Mac Pro

00:44:08   that you talked about two months ago, really should put

00:44:13   to rest the notion that Apple is losing interest in the Mac

00:44:18   and that the Mac is some team that's sitting there with cob

00:44:22   webs and dust gathering on them.

00:44:24   Glad to hear that came through.

00:44:33   That's the truth.

00:44:33   - We'll say it over and over and over again.

00:44:37   We said it before and at least now I think

00:44:39   there's tangible reasons for people to trust it,

00:44:41   that the future of the Mac is really strong.

00:44:44   We have a long roadmap ahead, this complete commitment

00:44:47   from the whole company that we're dedicated to it

00:44:50   and the investment's really strong there

00:44:53   and we think we're gonna keep the Mac at the forefront

00:44:55   of what makes the best personal computing

00:44:57   and that's what we believe, that's what we want to do

00:44:59   and I'm glad people saw some of that this week.

00:45:03   I am.

00:45:04   (audience applauding)

00:45:07   All right.

00:45:11   iPad, iPad Pro.

00:45:13   Another one, and this is a recurring theme,

00:45:15   it was with the HomePod speakers,

00:45:17   hard to demo in a big cavernous room

00:45:20   to see how this is gonna sound in your living room.

00:45:22   VR is kind of hard to demo on a 2D screen.

00:45:28   The 120 megahertz refresh rate of scrolling on the thing,

00:45:32   It's hard to show on a, I don't know, 30 frames per second video, I don't know.

00:45:38   But when you get your hands on this thing, I don't know, has anybody here seen the new

00:45:42   iPad Pros yet?

00:45:48   It feels like...

00:45:49   It's awesome.

00:45:50   It's all new.

00:45:51   It's sort of almost like a step like going from non-retina to retina, because it's like

00:45:56   this, I don't want to go back.

00:45:58   It really is.

00:45:59   - There, it was actually, I don't know,

00:46:01   it was probably three years ago

00:46:02   when we put together prototypes

00:46:05   and we were using a Mac Pro to drive a custom display

00:46:10   and get it running, a special board to run at 120 hertz

00:46:16   and we had a Safari webpage that we'd all pre-rendered

00:46:20   and we could do it at 120 hertz

00:46:22   and we were just one after another gathering around it.

00:46:25   Right?

00:46:26   (audience laughing)

00:46:27   And everyone was just, you know, oh my God,

00:46:31   we have to do this.

00:46:32   And, but it took that kind of dedication to it

00:46:37   because then it's like, okay,

00:46:39   let's talk to the silicon team, right?

00:46:41   Let's put this, let's go build some silicon

00:46:44   that we can ship in a few years

00:46:45   because it really took that.

00:46:47   I mean, the whole, it required custom silicon

00:46:50   and then it required big changes to our graphics drivers,

00:46:54   our core animation, now schedules,

00:46:56   animations at variable frame rates.

00:46:58   You can say, well, my animation needs to run at 30,

00:47:02   or mine needs to run at 60 or 120,

00:47:05   and it knows how to then drive the display appropriately.

00:47:08   Obviously, we had to manage power.

00:47:11   And then we'd optimize every app,

00:47:12   'cause they had half as much time

00:47:14   to get the next frame ready.

00:47:16   - Pencil team, take advantage.

00:47:18   - And the pencil team, so now it's great,

00:47:21   because we have twice as many opportunities

00:47:23   to draw as you're moving your pencil,

00:47:25   but then we also upped the scanning rate.

00:47:28   We doubled that to 240 to capture where the pencil was.

00:47:33   And so this was just a massive, one of these things

00:47:36   that's just a massive effort across silicon hardware

00:47:39   and software to pull something off where you really,

00:47:42   you know, this early glimpse of how great it can be

00:47:45   and then to just have it come together.

00:47:48   And I'll just, you know, Phil was in a meeting,

00:47:50   he had one and I was out of my corner of the eye.

00:47:53   I'm like, oh my God, that's such smooth scrolling

00:47:55   happening over there, you know, if it was like pages

00:47:57   or something.

00:47:58   - I was using the new iPad Pro without others

00:48:01   knowing I had it yet in our own internal meeting.

00:48:04   And so I figured I'm being really subtle.

00:48:06   I get in a case nobody can tell,

00:48:07   and I'm just using, and Craig's doing this double-taking.

00:48:11   What?

00:48:12   Yeah.

00:48:13   - Yeah, it's great.

00:48:16   And I think some others have pushed on resolution

00:48:19   as the big thing.

00:48:20   Actually, I think our focus on color depth, on refresh rate,

00:48:25   I mean these are where the really big wins are,

00:48:31   on brightness, true tone display.

00:48:35   I mean I think our display team is just doing unreal work

00:48:39   and it's great to-- (audience applauding)

00:48:48   I thought that that really showed too,

00:48:50   just to go back as I skipped it,

00:48:52   but I thought that the new iMac displays in person,

00:48:56   it's like this seems impossible.

00:48:58   It really looks more like a piece of film

00:49:03   with a bright projection,

00:49:06   like rear projection coming through.

00:49:08   It's incredible.

00:49:09   And I have to say, somebody who owns the first 5K iMac,

00:49:13   when the second one came out

00:49:14   and it had the wider color gamut,

00:49:16   It's like, that looks nice, but I don't need to upgrade.

00:49:19   And now I feel like my iMac at home looks dim.

00:49:22   I mean, it's--

00:49:23   I suggest you upgrade.

00:49:25   [LAUGHTER]

00:49:28   What-- tell me about everything that

00:49:31   is encompassed by ProMotion.

00:49:35   What is that-- it blurred-- I don't-- what?

00:49:38   Like, is the 120 hertz thing part of ProMotion?

00:49:41   Yeah, it's 120 hertz.

00:49:43   It's the whole variable refresh rate

00:49:46   that allows us, 'cause when you're driving at 120 hertz,

00:49:48   of course, that's more computationally intensive, right?

00:49:50   You're rendering twice as many frames,

00:49:52   you're driving the display twice as fast,

00:49:54   so that's inherently consuming more power,

00:49:56   and so the key about ProMotion is both,

00:49:59   from a power savings point of view,

00:50:01   a lot of the time things aren't moving,

00:50:02   and we can take the refresh rate way down,

00:50:04   which has a commensurate savings in power,

00:50:06   and so net it's a win, but also there's content

00:50:10   like that you're watching movie content,

00:50:12   and maybe it's shot at 24 or 48,

00:50:15   and normally you're doing this three two pull down

00:50:17   to hit 60 and it's actually slightly uneven

00:50:20   and now on a promotion display it's perfect, right?

00:50:24   And so you get none of that.

00:50:25   (audience applauding)

00:50:28   - On the software side for iOS,

00:50:36   iPad in particular got a huge amount of attention

00:50:41   this year in iOS 11 with significant improvements,

00:50:46   expansion to multitasking, multiple things on screen at once

00:50:51   with drag and drop.

00:50:54   Is that, on iOS, was that,

00:50:58   am I making it harder than it seems,

00:51:02   but is the sandbox nature of iOS make it harder

00:51:06   for the drag and drop because it seems like

00:51:09   Anything you can drag, you should be able to drag now,

00:51:12   but is that, are there implications for that

00:51:15   with sandboxing that--

00:51:16   - Oh, there totally are.

00:51:18   The team has gotten, has built up quite a security

00:51:23   architecture to make these kinds of things

00:51:25   actually pretty natural for us to do,

00:51:27   but when you're starting to drag,

00:51:30   you're conceivably dragging that content

00:51:33   over many apps on which you don't actually intend to drop.

00:51:37   you can be starting a drag and then swiping over

00:51:41   multiple applications and so our interaction is such

00:51:45   that we tell the app about what kind of thing

00:51:50   it might be going to receive, but we don't actually

00:51:53   let it manipulate the data and even if it wants

00:51:56   to tell us that oh, because you're hovered over me

00:51:59   and the user might drop over me, I want you to scale

00:52:02   the photo a little differently because this is how

00:52:04   it would look here, we do that, we let it tell us

00:52:06   without ever giving it the data.

00:52:08   And the system will render without giving it the data.

00:52:11   And then when you drop, OK, now you've given permission

00:52:14   to have access to the data.

00:52:15   And we're actually able-- sometimes

00:52:17   the data that's being dragged, it

00:52:18   could be some giant image or movie that the source app has.

00:52:24   And when you drop it on the destination,

00:52:25   we're able to instantly transfer it to the destination

00:52:28   because we use APFS instant snapshots to hand it

00:52:33   a copy of the file.

00:52:36   and it can start consuming it,

00:52:38   and the other app can even then go ahead and modify it,

00:52:40   it doesn't matter, we've taken an instant cheap snapshot

00:52:43   to transfer the data securely.

00:52:44   So there's absolutely a lot of thought

00:52:48   that goes into making that secure,

00:52:50   but in the end, it all comes together.

00:52:53   - You can give the example also in files,

00:52:55   how you give access to reading a specific file to a app

00:52:59   without giving everything open to it.

00:53:01   - Yeah, that's right, I mean, in the traditional

00:53:04   PC architecture, every app, if it wanted to be able to

00:53:08   browse your documents folder or your desktop,

00:53:11   well it can see all those files, it can read those files,

00:53:14   and even modify all those files.

00:53:16   On iOS, of course, everything exists in that sandbox,

00:53:19   but what happens is with the Files app,

00:53:21   if you drag a image out of any folder you have

00:53:26   into Keynote or into whatever third party app,

00:53:31   we're just punching a very small hole in the sandbox, right?

00:53:34   to get it just that access.

00:53:36   So even a rogue app can't go and look at all your files.

00:53:41   - And the path to that pinprick is through user actions,

00:53:45   which is an implicit granting of permission.

00:53:48   - Yeah, there's no dialogue popping up saying,

00:53:51   do you want to grant this app the access to this file?

00:53:53   Because your action is, you knew,

00:53:58   you indicated that you knew you wanted that app

00:53:59   to see that file and only that file.

00:54:01   We don't want to grant any more than what you

00:54:03   and your gesture indicated you were trying to accomplish.

00:54:06   Huge, huge applause line.

00:54:08   I mean, I think people have been dying for this.

00:54:12   But the Files app.

00:54:13   Yeah.

00:54:14   [APPLAUSE]

00:54:18   So it's not just files on your iPad.

00:54:21   And it works with third parties.

00:54:25   The demos included Box and Dropbox.

00:54:28   But there's also APIs that apps can use.

00:54:32   So that like, if somebody had like a chat app,

00:54:37   like a WeChat or one of those type things,

00:54:40   and you wanted to be able to have the user pick any file,

00:54:45   not just an image from your image library,

00:54:49   they don't have to write the code for that.

00:54:50   There's a standard with the files app.

00:54:52   It's sort of like the iOS equivalent

00:54:54   of open and save dialog box.

00:54:56   - That's right, and in fact,

00:54:58   one of our real test cases for this

00:55:02   were the iWork apps, Pages, Keynote, and so forth.

00:55:04   They traditionally had their own file browser.

00:55:06   But of course, their own file browser

00:55:07   could only browse the files inside the apps container.

00:55:12   And we were able to make the, essentially,

00:55:16   as you say, the sort of open and save panel, the library view

00:55:19   that we provide to third party apps, a level of fidelity

00:55:24   that we could just replace the entire implementation

00:55:26   inside of those apps with the system panel

00:55:29   and make it extensible enough that they could do,

00:55:32   at the app level, everything they wanted

00:55:33   while giving them this full access.

00:55:35   So we think this is gonna drive a lot of functionality

00:55:39   across all these apps, 'cause they're gonna get access

00:55:41   to safe browsing for the user, the whole file system,

00:55:45   but also consistency.

00:55:47   So you'll have actually a good file browser built

00:55:50   to all these apps.

00:55:51   - The Files app is also just a fascinating topic

00:55:55   because of the path it took to get to where we are.

00:55:59   As you recall, when we started with iOS,

00:56:02   one of the goals was if you could recreate a computer

00:56:04   post-internet age, what are the problems that occurred

00:56:08   that were good things went awry,

00:56:10   and how do you fix them for the future?

00:56:12   And one of the areas was file management.

00:56:15   That on a PC, it's great that we've got our file systems

00:56:18   and we launch apps by launching files,

00:56:21   but for most consumers, they actually don't manage

00:56:24   the files with hierarchies that are logically organized.

00:56:27   They actually just go to the internet and say,

00:56:28   "Oh, download this, download this,"

00:56:29   and it spreads everywhere,

00:56:31   and desktops are filled with a complete mess

00:56:33   and nobody can find anything.

00:56:35   So one of the goals with iOS from the beginning was,

00:56:37   how do you fix that so the average consumer using it

00:56:40   never runs into that problem?

00:56:42   And so we had the document model of managing files

00:56:45   just built in automatically,

00:56:47   which is great for the billion users,

00:56:49   but of course through the years,

00:56:51   pros and advanced users have said,

00:56:53   "No, I like to create a hierarchy.

00:56:55   "I need to create workspaces

00:56:56   "where I manage all my documents together

00:56:58   And in the beginning we thought, we don't know how we make these two things make sense

00:57:02   together.

00:57:03   And that's what's happened with the files.

00:57:04   There's years of investigation and figuring out how to do that and solve for that.

00:57:09   And so for most consumers, they will never need to use the files out.

00:57:14   They work just as they did already on iOS.

00:57:16   They use their phone with one hand, they get to the documents, they get to what they want.

00:57:20   They don't have a desktop cluttered with files and they don't get confused by it.

00:57:24   But anyone else who wants it, it's there and it works within it.

00:57:27   So it's very consistent between the two models.

00:57:29   - I mean it speaks to the difference between

00:57:31   iOS and Mac as platforms that the Mac,

00:57:33   and it's just the way it was designed,

00:57:35   and I think it's too late to change,

00:57:36   but when you boot up a new Mac,

00:57:40   where do you wind up after you go through the first run?

00:57:43   You wind up in the Finder.

00:57:45   You're in the file system.

00:57:45   That's the starting point.

00:57:47   And on iOS, you've got,

00:57:50   now you've got Dropbox, you've got Box,

00:57:52   you've got your iCloud Drive,

00:57:53   so if you have a shared folder on Dropbox,

00:57:55   You can get it from any app that supports it,

00:57:57   but when you start up, what do you see?

00:57:59   You see the same thing you saw

00:58:00   when the iPad shipped originally.

00:58:03   Here's your apps.

00:58:04   - Yeah, yeah, no, and I mean, part of the wonder

00:58:07   of the iPad is that it's brought computing

00:58:10   to both the very young and the very old

00:58:13   in a way that computers weren't able to reach,

00:58:16   and we didn't wanna break that,

00:58:20   and we feel great that we've been able to have,

00:58:23   many of us who want to take advantage

00:58:26   of this additional power, but do it in a way

00:58:28   that we can have that and not break

00:58:30   what has made the iPad so special.

00:58:33   And I think we did that with files.

00:58:34   And I think we also, I think I've been pretty,

00:58:37   I feel really successful in preserving that

00:58:39   with how we've handled the doc and multitasking.

00:58:42   If you don't flick from the bottom of the screen

00:58:45   and you just hear that little kid

00:58:47   who just hits the home button,

00:58:48   'cause you know you wanna go back and see your stuff,

00:58:51   that just works.

00:58:52   it still works just as it did before.

00:58:54   But if you're someone who is more sophisticated

00:58:57   and it's very natural,

00:58:58   flick up into the app, switch it,

00:58:59   pull something down, move it around,

00:59:01   like it scales up to all of that

00:59:03   in a way that I think really gives a more

00:59:06   pro user more capability.

00:59:10   - And there's like a path.

00:59:11   It's not like you have to jump off a cliff to go there.

00:59:13   It's like you can--

00:59:15   - Pretty continuous experience, yeah.

00:59:16   - And it reminds me of like features that are there

00:59:20   for quote unquote power users in the Mac

00:59:22   where most people, if they wanna find a folder,

00:59:24   find a folder, double click it,

00:59:26   and then double click the next thing.

00:59:28   And people who, like us, who use it all the time,

00:59:31   know you can use the arrow keys and command down,

00:59:33   command down to go down the file hierarchy.

00:59:36   But that's not there for the regular user who just needs,

00:59:39   they don't have to go down to eight levels of hierarchy

00:59:41   'cause they only have two.

00:59:42   - That's right, yep.

00:59:43   - And they could just double click.

00:59:45   iOS software features, some of them.

00:59:50   One of them that struck me were the new Siri voices.

00:59:53   - Yeah.

00:59:54   - And it seems to me like in this verbal communication,

00:59:59   the AI assistant computing, that the voice of the device

01:00:04   is sort of like the equivalent of the typeface in a GUI

01:00:08   where you want it to be nice.

01:00:11   It seemed very impressive.

01:00:14   I didn't get to play with it yet,

01:00:15   but it seems like it's really, really much more realistic.

01:00:18   - Oh yeah, it's a really, it's a big improvement.

01:00:21   I completely agree with the metaphor.

01:00:23   I mean, the voice is the interface, it is the typeface.

01:00:27   It is here, and it colors the whole experience.

01:00:31   And this voice is, and the synthesis that goes into it

01:00:36   is completely Apple developed, and we really were able

01:00:42   to build something starting from the state of the art

01:00:46   in using deep learning to understand

01:00:49   how to inflect the voice,

01:00:51   how to, you know, where to place the pauses,

01:00:54   how to stitch together all of these pieces of natural speech

01:00:58   and it makes a big difference.

01:01:00   You know, we're still not done.

01:01:01   I mean, speech, the human voice is an incredibly

01:01:05   sophisticated thing and we are incredibly sensitive

01:01:07   about how we interpret it.

01:01:11   But this is a big step forward and yeah,

01:01:14   I had an incident, I guess I'm a little,

01:01:16   should be a little ashamed of it, but I was driving with a friend who is not a disclosed

01:01:21   Apple employee. Like a month ago with my phone and I forgot it had the new voice on it and

01:01:28   I was hooked up to CarPlay and Siri started talking and he's like, "Wow, that sounds really

01:01:36   good." Like, "Really? I don't notice anything." It's the speakers.

01:01:43   >> It's the speakers.

01:01:45   >> Yeah, must be.

01:01:46   I don't know what to say.

01:01:46   Yeah.

01:01:47   >> I'll tell you one other thing that happened throughout many

01:01:49   of the Siri topics through the keynote.

01:01:51   I'm going to say something unbelievably risky.

01:01:55   Hey, Siri.

01:01:57   And it wasn't that long ago where I would now literally get 10,000 emails.

01:02:03   What did you do?

01:02:04   You just set off my device and blah, blah, blah.

01:02:06   And very few people had that experience throughout the show.

01:02:11   and for good engineering reasons of a lot of advancements

01:02:16   done in the recognition of your own voice

01:02:18   versus other people's and multiple devices

01:02:20   not all going off at the same time

01:02:22   and a lot of really great improvements there.

01:02:24   - And that seems like the type of feature

01:02:25   that takes a lot of work and then gets no credit at all

01:02:29   because when you say hey Siri, here in this theater

01:02:31   and people's phones aren't lighting up,

01:02:34   that's just the way it's supposed to be

01:02:36   because clearly that's not what I was trying to do

01:02:38   but when they do, people notice.

01:02:40   You can stop using dingus, maybe.

01:02:42   [LAUGHTER]

01:02:44   [APPLAUSE]

01:02:50   Yeah, I don't know if people--

01:02:52   We might have a new title for the episode.

01:02:53   There we go.

01:02:56   Yeah, I don't know if people are really conscious of it.

01:02:58   And we've made improvements on it this year as well.

01:03:01   But we built a mechanism into Siri,

01:03:06   where if you have multiple devices of your own,

01:03:10   and this was a problem certainly at Apple

01:03:12   where a lot of us have a lot of devices around our office,

01:03:16   and you'd say the Siri trigger phrase

01:03:19   and every one of your devices,

01:03:21   which would correctly recognize you,

01:03:23   but they'd all independently correctly recognize you

01:03:25   and all start responding together in this chorus.

01:03:29   (audience laughing)

01:03:30   And that wasn't great.

01:03:32   And we knew, of course, coming out with HomePod as well,

01:03:35   we'd only be adding to the chorus.

01:03:38   And so we built something where actually

01:03:41   all of your devices are coordinating wirelessly.

01:03:46   And so the moment they all hear you,

01:03:49   they all broadcast to each other

01:03:52   which one they think is closest to you,

01:03:56   like a sense of how close,

01:03:58   which one they think you're talking to,

01:03:59   and they also will sense which one

01:04:01   was most recently used by you,

01:04:03   'cause maybe you're using your phone,

01:04:05   and the fact your iPad's close,

01:04:06   but you're probably thinking about your phone,

01:04:08   weighs into it, and they essentially instantly vote

01:04:10   on who should respond, and the others all quiet down,

01:04:13   and one of them speaks up.

01:04:15   But this, I mean, it's a really subtle thing,

01:04:17   it involves all of this networking,

01:04:19   and it's so important to having a great experience

01:04:22   with something like using your voice to ask a question.

01:04:25   - And some other companies might do that on their servers,

01:04:27   but we do it locally.

01:04:28   - We do, yeah.

01:04:29   (audience applauding)

01:04:32   (audience applauding)

01:04:35   So on that point, I think that there is,

01:04:43   consensus is the wrong word, but the conventional wisdom,

01:04:47   the conventional wisdom is that the doing it

01:04:51   on your device's first strategy that Apple has with this

01:04:56   is flawed and that the strategy of doing it

01:05:00   in a monstrous thing like the opening apocalypse movie

01:05:05   in a cloud server is going to leave you behind.

01:05:09   And I got the sense in the keynote yesterday

01:05:12   that you guys are confident of your chances

01:05:15   in the Siri versus the other companies' products battles

01:05:20   coming already here and coming in years to come.

01:05:24   - Yes.

01:05:25   (audience laughing)

01:05:28   (audience applauding)

01:05:31   One more iMessage, or not iMessage thing,

01:05:34   one more iOS thing was the iMessage syncing,

01:05:36   and that ties in with the Mac as well.

01:05:39   This seems like the type of thing that's gonna make

01:05:44   a lot of people very happy, and I think it really helps,

01:05:48   maybe with you guys, there's Apple employees who test,

01:05:51   like you said, disclose hardware,

01:05:55   but for someone like me who, again, I'm not complaining,

01:05:58   I love my job, this is amazing that I get review units

01:06:00   of hardware, but then I wind up with seven of these things

01:06:05   and I set up a new iPhone and if I don't download

01:06:08   a whole backup from the cloud,

01:06:10   I don't have my iMessage there, but I will now.

01:06:15   - Yeah, that's right.

01:06:17   I think, you know, yeah, it's good.

01:06:20   (audience applauding)

01:06:22   - But in a way, the main point though,

01:06:24   but in a way that the thing that would get pulled down

01:06:27   is end-to-end encrypted, meaning it was encrypted

01:06:30   on the device when it went to Apple,

01:06:33   and the encrypted format, it's on the servers there,

01:06:36   can't be opened by Apple.

01:06:38   - That's right, we've--

01:06:40   - Yeah.

01:06:41   (audience applauding)

01:06:44   - Yeah, our security and encryption team

01:06:50   has been doing work over a number of years now

01:06:52   to be able to synchronize information across

01:06:56   what we call your circle of devices,

01:06:59   all those devices that are associated with a common account

01:07:02   in a way that they each generate and share keys

01:07:06   with each other that Apple does not have.

01:07:09   And so even if they store information in the cloud,

01:07:13   it's encrypted with keys that Apple doesn't have.

01:07:16   And so they can put things in the cloud,

01:07:18   they can pull stuff down from the cloud,

01:07:20   so the cloud still serves as a conduit

01:07:24   and even ultimately kind of backup for them,

01:07:27   but only they can read it.

01:07:30   And we built on that this year,

01:07:33   and we're using it for messages in the cloud,

01:07:37   and they also will roll keys automatically

01:07:41   so that you have that kind of,

01:07:42   and exchange them amongst each other.

01:07:44   And we use that to also be able to take

01:07:47   what Siri is learning about you locally,

01:07:50   and make sure that you want really one Siri, right?

01:07:53   to your mind there's your Siri and your Siri

01:07:56   knows about you and you don't wanna have to teach

01:07:59   every time you get a new device or use a different device

01:08:03   that it's as if you're talking to a different assistant.

01:08:05   You want one.

01:08:06   And so Siri's now able to exchange that information

01:08:09   between your own devices but in a way that's private to you.

01:08:14   And so yeah, this is an example of where we're really

01:08:17   investing in the technologies necessary

01:08:19   to both deliver the capability and preserve privacy.

01:08:23   Excellent.

01:08:25   [APPLAUSE]

01:08:33   The iOS App Store is pretty popular.

01:08:36   [APPLAUSE]

01:08:40   I don't know.

01:08:40   I mean, I think it gets used a couple of times a day.

01:08:43   10 apps daily, I think.

01:08:45   Total redesign.

01:08:47   I mean, it does look like a new App Store.

01:08:50   Is there anything on the thinking of that

01:08:53   that didn't make it into the keynote?

01:08:56   - Yeah, yeah, there's a lot.

01:08:58   I'll pick on one, I'll just pick out one thing to mention.

01:09:02   So the, on every product page is app reviews and ratings,

01:09:07   and there are ratings for every app.

01:09:10   One of the things I learned talking to developers

01:09:12   over the last year or so is that some developers

01:09:16   don't like submitting their updates

01:09:18   because it resets the rating.

01:09:20   And so they would get upset saying, oh man, I have a choice.

01:09:24   Fix some bugs or blow away my ratings

01:09:26   or keep my high rating.

01:09:27   I have a 4.7 and I don't want to submit it.

01:09:30   And I thought that was kind of stupid.

01:09:32   So--

01:09:32   [LAUGHTER]

01:09:34   So with this--

01:09:35   [APPLAUSE]

01:09:40   And in deference to the team, this is logic

01:09:43   to why it's a tough trade off of what to do for users,

01:09:46   what users want to see.

01:09:49   So what we decided is with this turn of the App Store,

01:09:53   now that that rating will not reset.

01:09:57   When the developer submits the app,

01:09:58   they don't be afraid of that, it can stay as it is.

01:10:01   But you can imagine when you're thinking through that,

01:10:03   there are some developers that have bad ratings

01:10:04   and they thought they nailed it with some new update

01:10:06   and they wouldn't like that, so it'd be 50/50.

01:10:08   So there's a switch when they submit an iTunes Connect

01:10:11   that says, no, I wanna blow away my rating

01:10:12   and start over with this update.

01:10:14   And so we'll let them do that.

01:10:15   So developers now in control of that situation.

01:10:18   (audience applauding)

01:10:21   I thought one of the understated aspects

01:10:29   of the design change is the separation of games from apps.

01:10:34   And so at a technical level, a game is an app.

01:10:38   It's an app that you launch and the thing you do in it

01:10:40   is play the game, but I think having the App Store

01:10:44   treat them as separate entities

01:10:47   is a big change because I don't think,

01:10:50   the games are so popular that you look

01:10:53   at the top selling charts, I mean, I don't know,

01:10:55   you probably look at 'em since you're in charge

01:10:56   of the store. (audience laughs)

01:10:58   They're all games, you can't get a break.

01:11:01   And so the most popular apps for like productivity

01:11:05   or something like that are very popular,

01:11:07   but never get a chance to crack the charts.

01:11:11   - Over and over again, every meeting with developers,

01:11:13   what you hear is discovery, discovery, discovery.

01:11:15   Like please help us with discovery.

01:11:17   The rest of the stuff we can deal with,

01:11:18   we need more discovery.

01:11:19   And so there are a number of things

01:11:21   in the new design for that.

01:11:22   This is one of those, which is separating the two.

01:11:25   In addition, we have the most popular gaming platform

01:11:28   in the world now, by some, I don't know what order

01:11:31   of magnitude, and so it's nice to say,

01:11:33   let's create a section dedicated to games,

01:11:36   because we love games and we wanna be able to express

01:11:39   that more and more, and now having a place

01:11:42   is gonna give us the freedom to start being more

01:11:44   more gamer friendly and have a dialogue with gamers

01:11:48   and have the ability for app developers

01:11:50   not to feel buried in.

01:11:51   And the number one was charts, was,

01:11:53   hey, I wanna show up on the charts,

01:11:55   I can't get discovered there because all the hot games

01:11:57   are showing up, by separating them,

01:11:59   the one question was would people understand

01:12:02   games and apps, they're all apps, of course they will.

01:12:05   One of the really important tenets we always have is,

01:12:08   treat the customers as smart, not stupid, they're smart.

01:12:10   They'll understand it.

01:12:12   (audience applauding)

01:12:15   Yeah, I really don't think anybody's confused

01:12:20   about what a game is, right?

01:12:21   I really don't, you know, but it's easy

01:12:23   to think yourself in a circle and think,

01:12:25   well, it's technically an app, you know.

01:12:27   So what about the Mac App Store?

01:12:30   Is the Mac App Store--

01:12:32   (audience applauding)

01:12:35   - So I will say this until you're all satisfied one day.

01:12:41   Just like the discussion we had earlier about the Mac,

01:12:43   which is we care about it, we believe in it,

01:12:46   we're invested in it, and you need proof to see the actions.

01:12:51   We care about the Mac App Store, we believe in it,

01:12:55   it matters, and over time you'll see proof of that.

01:13:00   But first, as you point out, this new design is,

01:13:03   I think it's almost on the order of magnitude

01:13:06   of a challenge to change as doing the file system

01:13:10   underneath everybody.

01:13:11   Changing the App Store completely was a daring, bold,

01:13:15   dare I say, courageous action.

01:13:18   [APPLAUSE]

01:13:21   So I know.

01:13:26   I'm going to regret that.

01:13:28   And that was a huge thing to ask of the team, the design

01:13:34   team, the engineering team, everyone who works on it.

01:13:37   And so doing it first for the iOS App Store

01:13:40   is absolutely the right thing to do,

01:13:43   and then other things over time.

01:13:45   - Good to hear.

01:13:48   (audience applauding)

01:13:51   Last major topic on my list is the HomePod.

01:13:59   And I,

01:14:05   There's a certain language I'm allowed to use

01:14:07   and how I can say it.

01:14:08   I had the opportunity to listen to HomePod

01:14:11   in a side-by-side demo with a Sonos Play 3

01:14:17   and an Amazon Echo playing the same songs

01:14:20   in a reasonably simulated living room size type thing.

01:14:25   And they sound great.

01:14:28   They really, really do.

01:14:30   Well, it sounds great.

01:14:32   And home pod, a home pod sounds great,

01:14:36   but you close your eye and it's really kind of,

01:14:39   that's why I pluralized it without thinking.

01:14:41   You close your eyes and it's hard to think

01:14:44   that that's coming from a single source.

01:14:45   Like the multi-directional stuff

01:14:48   that you describe about it is true.

01:14:51   It is, there's no doubt about it.

01:14:53   It does not sound like sound coming from a point.

01:14:56   It sounds like, I don't know how many,

01:14:58   I don't have the best ears,

01:14:59   but it doesn't sound like one speaker.

01:15:01   And it was like, there's these rules about

01:15:06   how you're allowed to talk about what the experience was.

01:15:09   I wasn't allowed to touch it, no photos or video,

01:15:12   but you were definitely allowed to say

01:15:16   which devices it was next to

01:15:17   and how much better it sounded than that.

01:15:19   (audience laughing)

01:15:22   And I thought in your introduction,

01:15:28   I thought it was very fair,

01:15:29   where you said here's a product

01:15:30   and you had a picture of a Sonos

01:15:31   and it's, this sounds good, but it's not smart.

01:15:34   Here's the Amazon Echo.

01:15:36   It is smart and it doesn't sound good.

01:15:40   And in the demo, it was like that.

01:15:42   I think we played about, we heard about six songs,

01:15:44   and I would say on at least four out of the six,

01:15:46   I thought that the HomePod sounded clearly better.

01:15:48   There were two where I thought it was neck and neck

01:15:51   with the Sonos.

01:15:52   They both just sounded good to me.

01:15:56   And the poor Echo (laughs)

01:15:59   really sounded bad in every comparison. So what I really want to talk about is the iPod

01:16:09   speakers from 2005. [ Applause ]

01:16:11   Or I don't know what you're -- it was around -- it was --

01:16:18   >> iPod Hi-Fi. >> iPod Hi-Fi.

01:16:20   [ Applause ]

01:16:25   >> You make it better.

01:16:29   >> I circle back to that though because I feel

01:16:34   like it's a continuous thread.

01:16:36   Like in some ways the introduction was the same.

01:16:39   It's we love music and our customers are playing music

01:16:43   and we want to make your music sound good.

01:16:46   And it seems like something that you guys have been

01:16:48   after for a while.

01:16:49   Yes, and sometimes you get to try again to get it right,

01:16:54   and that's okay.

01:16:55   Different thing, very different scenario, different time,

01:16:59   different use case.

01:17:00   I will say there is one, at least one gigantic difference,

01:17:06   which is our audio engineering team now is remarkable.

01:17:11   It's incredible.

01:17:12   You've seen it in everything from what they've done

01:17:15   with iPhone 7 and stereo speakers and that,

01:17:19   I think they did a phenomenal job.

01:17:20   The four speakers in iPad Pro and that and on and on,

01:17:25   they're really talented and incredibly capable,

01:17:28   the AirPods and--

01:17:31   (audience applauding)

01:17:34   So we have a depth of audio engineering talent,

01:17:39   far, far, far greater now, hardware and software,

01:17:42   than we had in that day.

01:17:45   And there were things that were good about the HiFi,

01:17:47   but I won't go into it because time will run out.

01:17:50   But there were some really good things about it,

01:17:52   but it doesn't matter.

01:17:53   The team now is amazing.

01:17:55   And I do think that the AirPods are an example

01:17:59   of conventional wisdom.

01:18:01   When people saw it was,

01:18:02   oh, these are gonna not be popular,

01:18:03   these don't make sense, blah, blah, blah.

01:18:05   And we said, no, they're really magical

01:18:08   and the team's done an incredible job.

01:18:11   I believe, and time will tell,

01:18:12   that it's the same case here,

01:18:14   that just like AirPods, HomePod is incredibly magical.

01:18:17   It will change people's thinking of what can be done

01:18:20   with a product in this category for music.

01:18:22   And we'll see if we're right, but I think we are.

01:18:26   - Well, the difference I see is that today,

01:18:30   you guys are able to make things into,

01:18:32   more and more of the things you make are computers.

01:18:34   AirPods are computers.

01:18:36   - Two computers.

01:18:37   - Right, two computers that communicate simultaneously

01:18:40   without either one being the main one

01:18:44   and the other one being the other one.

01:18:45   - Across a big bag of water, which is very hard.

01:18:48   (audience laughing)

01:18:50   - And that to me is the difference.

01:18:52   - Thank you.

01:18:53   - But the HomePod is obviously a computing device

01:18:57   and the iPod HiFi back in the day

01:18:59   was really just a consumer electronics device.

01:19:01   It was audio in and you get sound out and that's it.

01:19:05   Whereas now it's a computing device

01:19:08   and you can do all sorts of little things

01:19:10   and have people write software for it,

01:19:13   to make it work the way that it works.

01:19:16   - A world where more things are computers

01:19:18   is pretty good for us.

01:19:19   - That seems to me like the world that you're,

01:19:21   that's your world is more and more things

01:19:24   are becoming computers.

01:19:25   - Yeah, and in fact, Phil pointed it out

01:19:29   when he was introducing the product,

01:19:30   but there's actually iPhone silicon inside of the speaker

01:19:37   and to do the kind of signal processing that we have to do,

01:19:42   but all the networking, all the Siri.

01:19:45   The far field voice recognition,

01:19:47   which I know I've tested,

01:19:49   you haven't had the opportunity to test,

01:19:50   but is extraordinary, it's best in class.

01:19:54   The thing can be blasting music

01:19:57   and you can talk across the room to it

01:19:59   and in a way that a human can understand,

01:20:01   if you're sitting next to a speaker,

01:20:02   you wouldn't understand it, but the speaker can.

01:20:05   And that's great acoustic design,

01:20:08   but it's also a lot of powerful signal processing

01:20:11   that we can do on the silicon there.

01:20:13   And so it's great when more and more products

01:20:17   we're able to take advantage of this great

01:20:20   audio engineering team,

01:20:21   but the great silicon we built as well.

01:20:25   - Is it safe, I thought even the name alludes to it,

01:20:28   that HomePod sort of alludes to that it's AirPods,

01:20:33   AirPods are personal wireless listening devices,

01:20:38   and the HomePod is the shared one,

01:20:40   that it's sort of the speakers to the headphones of AirPod.

01:20:43   Okay.

01:20:45   Yes.

01:20:46   I think AirPod, I'll say it to you now,

01:20:48   because you're here,

01:20:49   but AirPods are my favorite Apple product in years.

01:20:52   (audience applauding)

01:20:56   I'm blown away by how much I like them.

01:20:59   I really am, and I see people here.

01:21:01   It's funny, you come to WWDC and everybody has the message.

01:21:04   It's like the one place where everybody has AirPods.

01:21:07   - The Apple Campus too.

01:21:08   - Right. (audience laughing)

01:21:12   Well, my favorite story along those lines was,

01:21:15   I had a meeting there, I forget why.

01:21:16   I don't know if it was the Antennagate thing or what,

01:21:20   but it was the time when the white iPhone 4

01:21:25   was delayed for a very long time.

01:21:28   But at Apple, people had them.

01:21:30   It was like, "Whoa, that's weird."

01:21:33   You see all sorts of stuff there,

01:21:34   and that's just the stuff people are walking around with.

01:21:36   (audience laughing)

01:21:38   - We got a test.

01:21:39   - That's pretty much the end of my list,

01:21:43   and you guys have been very generous with your time.

01:21:45   Is there anything else that you guys wanted to talk about?

01:21:47   Is there anything on your list that you wanted to say?

01:21:51   - There are a couple things, but before I do,

01:21:54   I don't wanna miss the opportunity

01:21:55   because the last two times, the two times we've been here,

01:21:59   which have been wonderful, about two or three weeks after

01:22:03   in a podcast you say, you know, I meant to ask that

01:22:06   and I didn't.

01:22:07   And then we got this cloud hangover,

01:22:09   like it was a great, great, great podcast,

01:22:12   but something was left unsaid.

01:22:14   And don't want anything left unsaid.

01:22:16   So if there's, here's your, you're on the spot now.

01:22:20   (audience laughing)

01:22:23   And again, I used to say things I regret,

01:22:28   But I want to at least make sure that there isn't something left that you want to ask

01:22:34   and it doesn't mean we'll answer.

01:22:37   But then you can't in three weeks from now say, you know what, I meant to ask that

01:22:42   and I totally didn't.

01:22:44   And while you're thinking I'll bring up something else just to give you some time

01:22:50   to think and I'm a masochist so I no longer regret this one.

01:22:54   But I'll talk about, I think one of the great things

01:22:59   that's been going on, of course, is the work with Swift

01:23:01   and Swift Playgrounds and education.

01:23:04   (audience applauding)

01:23:06   And there's some really great things

01:23:07   that have been going on in the last few weeks.

01:23:10   You pointed it out in your podcast that,

01:23:13   in fact, we put some stuff out last week

01:23:14   'cause we knew there wouldn't be time

01:23:16   to get them to the keynote this weekend.

01:23:17   We didn't want them to get lost in the shuffle.

01:23:20   So, great updates to Swift and Swift Playgrounds.

01:23:25   Swift Playgrounds supporting drones and robots,

01:23:29   and so I think kids are gonna love working with those,

01:23:32   doing great work with Lego and other great companies.

01:23:35   And we just rolled out a program with community colleges

01:23:38   so they can have a curriculum

01:23:40   to help students become developers.

01:23:43   (audience applauding)

01:23:46   We've also, as you know, over the last year or two,

01:23:49   open up education centers for SWIFT programming in Brazil,

01:23:53   in India, and in Italy.

01:23:57   And so this is an area, and a lot of credit to Tim

01:24:00   'cause he's really pushed that this is critically important.

01:24:03   There's no sales number tied to it,

01:24:05   there's no market share tied to it,

01:24:08   it's just the world's better off if we can help everyone

01:24:11   benefit from the opportunity to create software,

01:24:13   and that's what we're doing.

01:24:15   (audience applauding)

01:24:18   And it's all because of the software engineering team,

01:24:21   the developer tools team, the SWIFT team,

01:24:24   that make it possible to then create these curriculums.

01:24:26   And it's a great thing, and I think something

01:24:28   the whole community should be proud of.

01:24:31   - Well said.

01:24:31   All right, I found my card of questions

01:24:34   I've always been meaning to ask.

01:24:36   (audience laughing)

01:24:37   Never did, I have one more card.

01:24:39   I wanna ask you about something that I,

01:24:45   When I meet with you guys, you in particular,

01:24:49   but other people on your team, product marketing team,

01:24:52   when you bring up leaks to the press, you get fired up.

01:24:57   And it-- - You do.

01:24:59   - It angers you.

01:25:00   I think, it seems to me like very little software leaked

01:25:06   for this keynote.

01:25:07   This was a keynote full of things that we had no idea

01:25:10   what we were gonna see coming in.

01:25:10   And the stuff that did leak though,

01:25:13   seems to me either in the reports where it leaks,

01:25:16   they say that's where it came from,

01:25:17   or that must be where it came from,

01:25:20   that it came from the supply chain, hardware stuff.

01:25:23   Is that something that you feel like,

01:25:29   is it just luck that you guys,

01:25:31   that this keynote didn't really have as much stuff leak,

01:25:33   or is there a double down on secrecy,

01:25:36   as Tim said a couple years ago?

01:25:38   - Yes, there's absolutely a double down on secrecy.

01:25:42   there's a lot of work that's going on.

01:25:44   There are people and teams that work really hard at it

01:25:47   across every organization and how we manage secrecy

01:25:50   and with suppliers and partners.

01:25:54   It's a really, really hard challenge, no question,

01:25:57   and we may never be perfect at it,

01:25:59   but the teams work hard at it.

01:26:00   And it's really, there's lots of reasons

01:26:03   before we could go into it, it's a longer discussion.

01:26:08   But at the end of the day, it's really great

01:26:11   when you're a team who's worked so hard,

01:26:13   a couple years of your life on something,

01:26:17   and you get the opportunity to have Craig

01:26:19   or me or someone else present it,

01:26:21   and the whole world's hearing it for the first time,

01:26:24   they're hearing what you think about it

01:26:25   and why we think it matters,

01:26:27   and there's that excitement, enthusiasm,

01:26:29   and your heart goes out to the team

01:26:31   that's worked so hard on it

01:26:32   'cause they got their moment for

01:26:34   the love and energy they deserve, right?

01:26:36   And so that's one of the many reasons that it matters.

01:26:38   And so we had some of that.

01:26:40   There were many topics covered yesterday that weren't leaked,

01:26:45   weren't written up, weren't with screenshots.

01:26:48   And to me, the first thing I think about is I'm so happy for

01:26:52   those teams that they got their moment to go home to the kids

01:26:54   and say, this is what I worked on.

01:26:56   Now you can see, and it's fun.

01:26:57   [APPLAUSE]

01:27:05   Yeah, that's exactly what always seems to get fired up,

01:27:07   is it's not that the surprise was ruined for the keynote

01:27:11   that you're helping to put together,

01:27:13   it's the people, the teams,

01:27:16   it's the people who made that product

01:27:18   who had their hard work spoiled

01:27:20   with a crummy screenshot or something.

01:27:22   - They get, yeah, they get really angry

01:27:25   when one of these happens.

01:27:27   It's just a huge disservice to the amount of work

01:27:31   they put into it when it does,

01:27:34   and so glad we had a--

01:27:37   And not to be misunderstood, we get that there's,

01:27:41   the reason this happens and more with us

01:27:42   than any other company is the interest is so high, right?

01:27:46   And that's great, that's a good thing.

01:27:48   Heaven forbid there's a day where nobody,

01:27:49   there's leaks and nobody cares to beat it, right?

01:27:52   (audience laughing)

01:27:53   And so we get that there's a passion out there

01:27:56   and there's a voracious appetite to understand

01:27:58   what's going on and to gain insight

01:28:00   and there are sites that love Apple that post stuff

01:28:03   just out of a misplaced love of us, honestly.

01:28:06   And that's good.

01:28:09   We can't ever be mad at that or upset about that.

01:28:12   It's more the lost opportunity to make a lot of people

01:28:16   really happy with the news.

01:28:19   That about does it.

01:28:21   I thank you gentlemen for your time.

01:28:22   I certainly thank all of you for coming.

01:28:25   Thank you.

01:28:26   [APPLAUSE]

01:28:29   (cheering and applause)

01:28:32   I'll just run through, I'll thank MailChimp for the bar again.

01:28:43   (cheering and applause)

01:28:46   Our sponsors for the event, Jamf Mobile Device Management

01:28:50   with a total Apple focus on Apple products.

01:28:53   Mac Stadium, totally professional hosting

01:28:56   for Mac OS X server and distributed doing builds

01:28:59   with Xcode and stuff which might be of interest

01:29:01   to people at WWDC and Setapp,

01:29:04   a really cool subscription service for indie Mac apps.

01:29:07   So my thanks to them.

01:29:09   I would like to thank here today at the event,

01:29:11   I would like to thank Caleb Sexton,

01:29:12   the audio editor of the show.

01:29:15   (audience applauding)

01:29:16   He's been here working for two days

01:29:19   to make sure everything sounds as good as we can have it

01:29:22   and I really appreciate working with him.

01:29:24   Caleb, thank you for everything you do for the show.

01:29:26   Marco Arment is up there somewhere.

01:29:29   I see him waving, and he knows how to live stream audio.

01:29:33   So if there's, Marco, did it work?

01:29:36   Thumbs up. - Thumbs up.

01:29:38   So everybody out there who's on the web

01:29:40   listening to me tell you this right now,

01:29:42   thanks to Marco Arment for that.

01:29:44   (audience applauding)

01:29:46   And we're going to have video.

01:29:48   It probably won't be up 'til tomorrow.

01:29:51   Jake Schumacher, director of App the Human Story,

01:29:54   which screened just the other night a fabulous movie,

01:29:57   which is coming out in final form

01:30:00   probably in about five or six years.

01:30:02   But it's in really good shape.

01:30:05   But my thanks to Jake Schumacher,

01:30:07   he's here shooting this somewhere,

01:30:09   and his colleague Charles Davis.

01:30:12   And then I heard somebody call him Charlie,

01:30:14   and I said, well, I wanna thank you.

01:30:16   Are you a Charlie or a Charles?

01:30:18   And Jake told me his nickname is Clutch.

01:30:20   So if your nickname is Clutch, I'm just going

01:30:22   to call you Clutch Davis.

01:30:23   Thanks for your help shooting the video.

01:30:25   And last but not least, the staff here

01:30:29   at the California Theater, the entire staff,

01:30:31   back of the house, front of the house.

01:30:33   Everybody here is a total professional

01:30:35   and are just really nice people.

01:30:38   It's been great being here.

01:30:40   So thank you, Phil, Craig.

01:30:43   Thank you.

01:30:43   >> Thank you, John.

01:30:44   >> Good night.

01:30:45   [APPLAUSE]

01:30:49   Thank you.

01:30:50   Perfect.

01:30:50   [MUSIC PLAYING]

01:31:08   [music]

01:31:23   [ Silence ]