IO23: This Is The Way - Artificial Intelligence everywhere, Pixel Tablet/Fold/7a, Project Starline

IO23: This Is The Way - Artificial Intelligence everywhere, Pixel Tablet/Fold/7a, Project Starline

Released Wednesday, 17th May 2023
Good episode? Give it some love!
IO23: This Is The Way - Artificial Intelligence everywhere, Pixel Tablet/Fold/7a, Project Starline

IO23: This Is The Way - Artificial Intelligence everywhere, Pixel Tablet/Fold/7a, Project Starline

IO23: This Is The Way - Artificial Intelligence everywhere, Pixel Tablet/Fold/7a, Project Starline

IO23: This Is The Way - Artificial Intelligence everywhere, Pixel Tablet/Fold/7a, Project Starline

Wednesday, 17th May 2023
Good episode? Give it some love!
Rate Episode

Episode Transcript

Transcripts are displayed as originally observed. Some content, including advertisements may have changed.

Use Ctrl + F to search

0:00

Coming up next on All About Android, it's

0:02

me, Jason Howell. We've got Ron Richards and Wyn

0:04

Twitdown no longer in the studio sitting

0:06

with me, but they are, you know, zooming

0:09

in from their homes. We are also

0:11

joined by Mike Wolfson, who joins us every

0:14

year around I-O to either give a preview

0:17

or do a recap. This year, it's a

0:19

recap, so we're happy to have him. It is our

0:21

big Google I-O recap

0:24

episode. So we take a look at all the major

0:27

announcements, and let me just kind of spoil

0:29

the surprise a little bit. There's a ton of

0:31

AI at this year's Google

0:34

I-O, if you didn't already know. Hardware,

0:36

I mean, so much stuff coming from I-O. So we

0:38

have that. If you notice that this feed

0:40

is really long, that's because after

0:42

the credits, we've included our interview

0:45

from the Google campus with Dave Burke

0:47

and Samir Samat of the Google Android

0:50

team. We talk all about Android 14, the

0:53

keynote that they had just finished, literally

0:55

two hours prior to the interview. So

0:58

you don't want to miss it. Super-sized, all about Android coming

1:00

up next.

1:03

Podcasts you love. From people

1:05

you trust. This

1:08

is Twitch. This

1:12

is All About Android, episode 630, recorded

1:15

Tuesday, May 16th, 2023. I-O 23,

1:18

this is the way.

1:21

This episode of All About Android is

1:23

brought to you by CashFly. CashFly delivers

1:25

rich media content up to 10 times faster

1:28

than traditional delivery methods and 30%

1:31

faster than other major CDNs. Meet

1:33

customer expectations 100% of the time. Learn

1:37

how you can get your first month free at

1:39

cashfly.com. Also,

1:42

thank you for listening. As an ad-supported

1:44

network, we are always looking for new partners

1:47

with products and services that will benefit

1:49

our audience.

1:51

99% of our audience listens to most of

1:53

all of our episodes. So grow your

1:56

brand with authentic ad reads that

1:58

always resonate with our audience.

1:59

audience. Reach out to advertise

2:02

at twit.tv and launch your campaign

2:04

now. Hello,

2:07

welcome to All About Android. This is your

2:09

weekly source for the latest news, hardware, and

2:12

apps for the Android faithful. I'm Jason

2:14

Howell. And I'm Ron Richards.

2:17

And I'm Wyn Twedau. And we are

2:19

all in our places. We're not together.

2:22

We're not all together. Not sitting

2:24

physically at this desk

2:26

the way we were just a week ago. What

2:28

a wonderful week that was, or episode,

2:31

and week, because I was a lot of

2:33

fun. We're going to talk all about it. But before

2:36

we get into that, just want to make mention that

2:38

we have our esteemed

2:41

return guest Mike Wolfson

2:43

joining us. MikeWolfson.com, Android

2:46

Google Developer Expert, and

2:50

all-around awesome guy, especially pre or

2:52

post I.O. It depends on the year. But

2:54

Mike, it's good to have you back.

2:57

Thank

2:57

you. And thank you for calling me esteemed.

3:00

Esteemed. That's the nicest thing anybody said about me. I

3:03

mean, just look at all those badges. Who

3:07

could have all those badges and not be considered

3:09

esteemed? True.

3:11

I mean,

3:13

if you're wearing those, I don't know how you

3:15

walk around. It's probably very heavy. I

3:18

will say that like it's

3:21

Pavlovian in me now that whenever

3:23

we start talking about Google I.O., I get excited

3:25

to see Mike. And so while

3:28

Mike, you're on the post I.O. show, I'm

3:30

glad that the multi-year tradition

3:32

of Google I.O. and Mike being on the show

3:35

is still in play. That's right.

3:37

That's right. It's good to have you here, Mike. So

3:40

we should note

3:41

before we right off the

3:43

jump here, this episode

3:45

is very long. You probably saw it in your

3:47

podcast feed and you're like, holy

3:50

moly, what's going on here? The

3:52

reason for this is that and

3:55

we're going to talk all about Google I.O. coming up here.

3:58

After the credits of this episode.

3:59

you get a bonus episode tacked

4:02

on to the very end. So last

4:05

Wednesday from Google IO,

4:07

when Ron and myself, we

4:10

were on the Google campus, we had the opportunity

4:12

to speak with Sumer Samat and

4:14

Dave Burke from the Android, from Google's

4:16

Android team, all about

4:19

everything that they talked about during the keynote.

4:21

And this was like two hours after the keynote

4:24

ended. So we felt very privileged

4:26

to be able to sit down with them. They had very busy schedules

4:29

and they gave us like 45 minutes to

4:31

chat with them. It was a lot of fun. And

4:34

the interview went out on our twit news

4:36

feed, twit.tv slash news.

4:37

But I

4:39

mean, you all who subscribed to this

4:41

show, you're the core audience for this information.

4:44

And so we figured, you know, why not just

4:46

put this interview at the end of this episode,

4:49

if you've already checked it out on the twit news

4:51

feed, you don't have to listen to it, you don't have to watch

4:53

it. If you haven't, we're saving

4:56

you the need to jump over there. You can just check it

4:58

out here. And yeah,

5:00

so you get a little bonus. You get two episodes in

5:03

one. That's our gift to you. How about that? And

5:05

it was a great fun conversation that like

5:08

flew by. Yeah,

5:09

I think it ended up being like about 45

5:12

minutes recorded or so. And you

5:14

can hear both. We've had Dave on the show before.

5:17

It was great to meet him in person, but to meet Sumer

5:19

as well and like hear from both of them from their perspective

5:21

about everything, all the stuff we're gonna talk about here in tonight's

5:24

show is we break down IO. But

5:26

as you mentioned, Jason, just coming literally

5:28

like almost less than two hours

5:30

after the keynote, like you could still get

5:32

the adrenaline radiating off a day from

5:34

his, like so many recaps of

5:36

the keynote that I saw called out

5:39

Dave's performance of like, and he

5:41

did a great job with those live demos and stuff

5:43

like that. And he talked a little bit about that during the interview.

5:45

So seriously, like I know personally for me,

5:47

it was the absolute, I mean, on

5:50

top of being in person with you both,

5:52

Jason, and Wynn and Burke and

5:55

Anthony and Victor and the whole crew, it was the

5:57

highlight of my Google IO just being able to sit down with

5:59

them and have like

5:59

a great, you know, like exclusive conversation.

6:02

So I hope everybody enjoys listening to it. Yes, indeed.

6:05

That is after the credits roll on this

6:07

episode. And this episode could just by itself

6:10

be a little bit longer because there was so much to talk,

6:12

so much news from Google IO.

6:15

So, uh, let's not dance around

6:17

it anymore. It's time to turn off

6:19

the music. It's time to get serious. Let's,

6:21

let's start talking about Google IO.

6:24

It's a time for the news. So

6:28

thanks for bringing that up about last week.

6:29

Um, and, uh, thanks for the kind words,

6:32

Ron. Cause I was a little salty.

6:34

You guys didn't bring me last week. Yeah,

6:37

we tried.

6:38

We tried. I can't tell you how many hours

6:40

I spent on the phone with, with Google reps

6:43

saying, look, Victor wants to come. He

6:45

wants to be here. Why can't Victor

6:47

make it? And they were like, you know, next year. And

6:50

so maybe there's hope. Thanks. That

6:52

makes me feel a little

6:53

bit better. That

6:56

whole last part there I fed into

6:58

an AI system and it told me to tell you

7:00

that I

7:01

do so I don't know if it's true, but

7:03

anyways, well, you know, it's not

7:05

true because there is no way to call

7:08

Google. Certainly

7:10

hope to. That was the part that I knew.

7:13

That's how you do. It was an AI system. I'll

7:15

just say, I appreciate the thought. There

7:17

we go. Okay. All right. Even from AI

7:20

Jason.

7:22

Before we did, before we dig into the

7:24

shows, had some great

7:26

pictures from, from Google

7:29

IO. I mean, yeah, yeah. The

7:32

link, uh, Victor in the doc above the news section,

7:34

there's a little photo album

7:40

there, uh, and you

7:42

know, I mean, Rarely

7:45

if ever, actually never before

7:48

uh, last Tuesday and Wednesday were

7:50

all of us in the same place at the

7:52

same time. So, you

7:54

know, had to, had to capture

7:56

it and, uh, when took a number of

7:58

photos, we got a bunch of.

7:59

really awesome. There's Victor in there, Michelle.

8:02

Uh, we got a bunch of awesome, uh,

8:05

group pictures and everything. Soft flow.

8:08

Yeah. For our audio listeners, these are all photos of

8:11

us, uh, at Google IO right

8:13

now. But, uh, yeah,

8:13

this is the first time I got to sneak, I've tried

8:15

to sneak into the press boxes at Shoreline

8:17

amphitheater for years, uh, just because

8:20

they're nice to sit in. They're a little more roomy,

8:22

uh, but they're pretty good with the bouncers. But

8:24

this year I managed to sneak into the press box with

8:26

these lovely folks and sit from there and flow.

8:29

I miss flow. I miss flow.

8:31

And

8:34

I'm so sad. And she was very proud of her,

8:36

uh, cloud, uh, Zit

8:39

patch, which is a conversation starter. Um, but

8:44

yeah, it was just really, really great to see everybody

8:46

in person and, and to be at Shoreline

8:49

with everybody and kind of experiences together.

8:50

Yeah. So actually, actually,

8:53

so let's, let's, um, just real quick,

8:55

spend, spend like a minute, uh,

8:57

talking about that aspect, because, uh, Michael

9:00

or Mike, um, when both

9:02

of you were at IO last year, Mike, you were

9:05

not at IO this year, but when

9:07

you were, um,

9:09

I mean changes, I mean, it did, it

9:12

is different. They let in more people, uh,

9:14

arguably we as a team

9:17

had more going on, but was

9:19

it an upgrade in your mind when, what did you, what

9:21

do you think?

9:22

So I, as a

9:25

developer, I suppose

9:27

as a developer, no, there was less developers.

9:30

And I, I say not an upgrade

9:32

and I don't mean that to be insulting.

9:35

It's just that it's kind of the same where

9:37

there is, there's basically just the consumer

9:40

keynote and developer keynote. And that was it. And then the

9:42

rest of the content is online. So, you

9:44

know, if you like that, then that was more

9:46

of the same this year. There was not at least

9:49

as far as I could see, uh, any additional like

9:51

fun and festivals for us. I know it was different for press.

9:54

Um,

9:55

but yeah, I mean, I kind

9:58

of the same. So you can take that as you want. of

10:00

a press event then really. It

10:03

is. And that's kind of a shame

10:05

when it's a developer conference

10:07

or maybe what it is, is Google

10:09

IO the event is the press

10:12

event. Google IO the developer

10:14

conference is the 200 videos

10:16

that are released online, which we

10:19

had a conversation, Rana, I believe

10:21

you were there with this conversation. We were talking with

10:23

a Googler who helps coordinate

10:25

all this stuff. And he basically said, like,

10:28

look, you know, prior to the pandemic, we had no

10:30

reason to share all of our videos

10:32

because it was all about the event, bringing

10:35

people here. And if you were here, you got something special.

10:38

Now since the pandemic, they've

10:40

seen more engagement on the actual

10:42

developer con content online

10:45

than they feel like they ever saw with

10:47

the event itself. And so I kind of got

10:49

the impression that they were saying that

10:52

he was saying, you know, he didn't say

10:54

it outright, but I really got the impression that this

10:56

is kind of the way it's going to be. It's

10:58

going to be a single day probably and all

11:00

this stuff online. Yeah.

11:03

I think, and I think that it's kind of gives

11:05

them the best of both worlds because they get the gathering

11:07

event for people who want to go, but then people

11:09

who either they're there, they can't afford to

11:12

or employers don't want to send them or whatever it is,

11:14

can watch at home and get, and be in

11:16

and get it all. I mean, I, to me,

11:18

it seems like they found their footing and they found

11:20

the, what possibly works best for the

11:22

community. I don't know, Mike, what do you

11:24

think? What do you think?

11:26

So I do have a lot of thoughts about this, of course. First

11:29

thing I do think this is the way of the

11:32

future for big conferences. WWDC

11:34

this year is similar. It's

11:36

the same sort of thing where they're just having one day.

11:39

It's mostly online. I do think this is

11:41

the way this is the first Google

11:43

IO I have not been to live. And

11:47

it was actually the first one since last

11:49

year attending. And it was kind of not that,

11:52

you know, wonderful an event that I kind of just

11:54

okay with it. And

11:57

I'm okay with it in particular. Well, I'm okay with

11:59

it because a lot of the. And then online, the other thing I wanted

12:01

to mention is Google is trying

12:03

for the first time this year, a follow

12:05

on event called IO

12:08

Connect. And

12:10

they do these things a lot. But

12:12

this is kind of like the festival portion, perhaps

12:15

of Google IO. It's in a few weeks in

12:17

Miami, which is kind of not a great decision.

12:21

But it's going to be like the booths

12:23

and stuff that you would have set up at Google

12:25

IO, where you can ask DevRel questions

12:29

and the code labs and some things like that.

12:32

So they're

12:33

having that, but it's just

12:35

at a different event. Yeah. Yeah.

12:38

Definitely a different event all around. Interesting.

12:41

Okay. Well, I

12:44

missed seeing you at Google IO,

12:46

Mike.

12:47

Likewise. I will

12:49

say, and I won't name names, but I did

12:51

talk to some folks who

12:53

work at Google IO who lament the

12:55

lack of kind of in-person interaction

12:58

with developers. And that's always been a strong point of

13:00

the Android community that we have some access to the

13:02

Googlers. Yeah. And it goes both ways,

13:05

right? So we get access to the Google engineers

13:07

to ask them questions about either specific

13:09

or broad things. But the

13:11

people who work on the team also love hearing

13:13

from us as well. And not to say that that

13:16

can't happen now, like for sure

13:18

at the Connect events at well.

13:20

Mike, is the Connect event GDEs

13:23

only or is it like general? I'm sorry. It's

13:26

open to everybody. So I think that's where they

13:28

make it up. But especially with I

13:30

think Google like being a little more

13:33

like about local events, it's like you're not going to

13:35

see all your favorite Googlers or like all

13:37

of your favorite Googlers rather are not going to have as

13:39

much interaction with you. So I

13:42

think that is, we'll see how

13:44

it goes with a Connect and maybe like how both

13:46

us as developers, they're probably developers in the Google

13:49

folks feel about it. But it was

13:51

kind of a big, big deal to kind of have

13:53

this big event where all the Googlers

13:56

and as many devs as could make it could

13:58

all have access to each other. and

14:01

create that collaboration. So that's

14:03

gonna be the part that they need to work hard to make up for,

14:05

I think. And I mean, maybe Connect will be that,

14:07

but we'll see.

14:08

Keep

14:09

my hopes up. It won't be that. And

14:11

I think honestly, we won't have Google

14:14

I.O. like we had in the past, unfortunately.

14:16

Yeah, because I mean, how many developers are gonna

14:18

go out to fly out to the Bay

14:20

Area and stay at a hotel

14:23

and everything for a

14:25

single day that is taken up two thirds by

14:28

keynotes. Like, you know

14:30

what I mean? It's just fundamentally,

14:33

it's a very different event than it was

14:35

before. And the thing is, while we're on the top,

14:37

the keynote was very long. I

14:39

mean, that was a more than two hour keynote, right? And

14:44

then Jason, I feel like over the years, we've seen everything

14:46

from the tight hour,

14:48

hour and a half to like a crazy multi-hour

14:51

skydiving, kind of extravaganza,

14:53

right? Yeah. And I do think

14:56

it would be remiss without, and

14:58

we're gonna get into it in the details as we start breaking stuff down.

15:00

But like the going joke that

15:03

everyone made after the keynote for the rest of

15:05

the day, myself included was AI much,

15:08

right? Because after two hours of having

15:10

those two letters hammered into

15:12

your brain, like it did get taxing and

15:16

exhausting at some point where just like, it

15:18

was just very, it was very repetitive.

15:22

And I'll give my friend credit who was

15:24

texting me during the event and

15:26

he was describing what they were doing

15:29

as basically like, he said,

15:32

it feels a lot like a restaurant and they keep serving me vegetable

15:34

dishes instead of what I ordered. Right?

15:37

Right? So

15:41

then it's Google not reading the room, because I feel

15:43

like everybody's talking about AI, but Google,

15:45

I mean, and so just to

15:47

kind of give you

15:48

an idea of where this show is headed,

15:50

the entire first block here is filled

15:53

with AI news. I figured, well, just get the AI

15:55

done. And then the rest of the episode, we have

15:57

non AI stuff. And before.

15:59

Before we get into the AI stuff, I mean, I think that

16:02

the official kind of stuff,

16:04

the theme of the entire

16:07

event was AI. And

16:09

my takeaway from it was Google's

16:12

gotten beat up over the last

16:14

six months, maybe eight months

16:17

with chat GPT and open

16:19

AI kind of leading the pack of like

16:21

what AI can do and mid journey and all these

16:23

other, you know, all these other generative AI kind

16:26

of startups that are in the place and Google

16:28

needed to reassert themselves and say, no, actually

16:31

we're a leader in the space. They, you know,

16:33

Sundar opened it up by saying, you know, over seven

16:35

years of research into AI and development

16:38

and stuff like that. Like they kept on driving

16:40

home the fact that they've been working on this. They've been working

16:42

on this and it was almost, you know,

16:44

like Shakespearean and like the lady Doth

16:47

protests too much. But

16:50

it was, it was, it was definitely,

16:52

um, they said it a lot. They

16:55

said it a lot. Well, we were joking while we were there.

16:57

If we could take a transcript

16:58

of the keynote and run it through

17:00

Bard

17:01

and tell Bard, tell me

17:03

how many times the word that the

17:05

acronym AI was spoken during this keynote.

17:08

We figured it would probably be at least a thousand, if

17:10

not more. So let's

17:12

get into these things. And we already spent 10

17:15

minutes talking about the event itself, which I think is

17:17

really important because it is, like we

17:19

said, fundamentally different this year

17:21

and I think going forward. So let's focus

17:23

a little bit of time on AI. We don't have to go

17:25

into incredible detail on these

17:28

things, but I think you'll start to see the picture

17:30

that the theme wasn't just AI. It

17:32

was how Google is bringing AI

17:34

into the products and services that they already have.

17:36

So you don't have to go to a destination to find AI. AI

17:39

is where you already are. We kind of saw that a

17:41

little bit with assistant. We'll talk about that in a moment,

17:44

but this is in other ways. And one of the major

17:46

ways, um, is that

17:48

Google announced that it's bringing its generative

17:51

AI, uh, to search.

17:54

And

17:54

that in and of itself is a really big deal

17:56

because search, I mean, search has changed

17:58

a lot.

17:59

a lot over the years, but it's still really

18:02

fundamentally been the same thing. The 10

18:04

blue links ads,

18:07

you know, in ads throughout that experience

18:09

has changed a little bit, but how things

18:12

are structured, haven't changed a heck

18:15

of a lot, I would say, or either that, or

18:17

we've gotten used to it

18:18

over time. Well, this is a really big difference.

18:21

The key here though, is that you

18:23

have to opt into it. Once it's available, you'll

18:25

have to opt into it. There is a,

18:27

a wait list that you can get on

18:30

to get access to this, but you opt into

18:32

it via search labs. And

18:34

basically it's a, you know, it's a chat bot

18:36

integrated into search. It's basically

18:38

Google's answer to being

18:40

getting its own generative AI infusion

18:42

a few months back. So this would be,

18:44

and it's not barred. So it is, these

18:47

are different things. They're different efforts. But

18:51

the chat bot answers questions, you

18:53

know, that you ask it in real

18:55

language

18:57

style and kind of integrates

19:00

it above the

19:01

fold. So above where the blue links

19:04

are. So it really

19:05

entirely shifts how Google search

19:07

product operates

19:10

if you opt into it. And I can't help but

19:12

think that, okay, right now it's opt in, but

19:14

if this is the way,

19:16

this keeps being spoken on the show, it's obviously

19:18

a title. If this is the way, then will

19:22

there be a time somewhere down the line where this

19:24

is just kind of part of a part

19:27

of search is that it's no longer

19:29

this test thing that you have to opt

19:31

into. It's part of how, you know,

19:34

the fabric of how we search is

19:36

going forward. And I could totally see that.

19:39

What do you all think? Do you think this is, do

19:41

you think this is the way? So

19:43

we talked about this a little bit in the pre-show and

19:46

the only way

19:46

that this becomes our normal

19:48

pattern is if these

19:51

chat bots return good results. Because

19:53

right now, I mean, when we search Google,

19:56

we expect the correct answer. And

19:58

like

19:59

what we you know, when we were playing around with this before

20:01

the show, we learned that none of these answers are

20:03

correct. Like

20:05

not any of them, it's 90% incorrect. Well,

20:09

a great example of this is that Jason,

20:11

while you were just talking about it, I went to Google Bard

20:14

and I said, how many times did they say AI

20:16

during the Google IO keynote? And

20:18

the first response I

20:20

got was the word AI was mentioned 125 times

20:23

during the Google IO keynote. It was most

20:25

often used in reference of new AI featured in

20:27

products like pixel seven, a lambda and Bard,

20:29

blah, blah, blah. And you know how in Bard,

20:32

you can say view other drafts, right?

20:34

So I click that. And the second

20:37

draft says, according to the YouTube transcript

20:39

of the Google IO 2023 keynote, the word AI

20:41

was said 43 times. That

20:44

is so wrong. So

20:47

what is right? Like, like, and here's the thing is like,

20:49

we're, we're, we're assuming that these,

20:51

you type stuff,

20:53

a prompt in and you get a response back and there's

20:56

a, there's a commitment to the answer

20:58

you're given is truthful,

21:00

right? Like, and this is a whole

21:03

another layer of like science fiction here

21:05

where it's like, what if you're not getting

21:08

correct answers back, you know, like it,

21:10

so it's definitely the thing that you're using

21:12

to, to find facts, which largely

21:14

I'd say people use search

21:17

to find facts about things, things that are

21:19

posted somewhere that are factual. Maybe I'm wrong.

21:21

You know, I guess if you're searching for fiction,

21:24

that would be wrong. But largely

21:26

when I go to a search engine, it's because I have

21:28

this thing that I want to know about.

21:30

I want to know the details that I want it to be factual.

21:33

And if the thing that you're using can't give you factual

21:35

information or rather it gives you fact, it

21:38

gives you information that appears to

21:40

be factual, but you only

21:42

find out that it's not factual by actually

21:44

researching it and spending the time

21:46

to look into it. Then like, is anybody

21:48

going to use it?

21:50

That's a good question. There's,

21:53

there's also the element of like, uh, that like, so just like we've

21:55

had comps, like so many conversations about generative

21:58

AI, especially when it comes to

21:59

and kind of like the aggregation

22:02

and training on other people's art. And then you create something

22:04

from other people's art. And whether that

22:06

is theft

22:08

of art,

22:09

theft of ideas. It

22:13

was really funny. We were watching the keynote. My husband

22:15

was on chat with me. And I think there

22:17

was one thing in the search results where it was like aggregating. I

22:20

think he was planning for a vacation, right? And

22:22

it had aggregated information from three different

22:24

blog posts. And he made a comment like, so

22:26

now that means that your content will just be sucked

22:29

up by Google and spit out. And

22:31

to be fair, they did have annotations there. But

22:34

I do think there is kind of a feeling, oh,

22:36

it's now it's not just something kind of more

22:39

visual and necessarily inherently

22:41

individualistic like art. But imagine

22:43

your content just kind of gets sucked up and

22:46

then kind of rolled into this generated content.

22:49

And it's not as easy to attribute you. I

22:51

mean, it's there, but that's not the first thing a person

22:53

sees. And that feels less good to me. I

22:57

think it kind of goes with, okay, what is it

22:59

pulling together? What is it synthesizing?

23:01

Is it synthesizing fact? Is it synthesizing opinion?

23:04

Is it synthesizing someone's creative work? Whether

23:06

that is a guide to hiking in the Sierras or something or an image.

23:09

So

23:14

that kind of stuff also kind of makes me wonder

23:16

as well. I know that it

23:18

was interesting because they seem to be kind of trying to address some

23:22

of these like attribution slash like how can

23:24

we tell if a thing is real, like it was made by a person or

23:26

made by generation. And that also

23:28

seemed to pop up as a theme of privacy, of

23:32

attribution and of like disclosure

23:35

and that as well. But I don't know, I feel

23:37

like there's a lot of kinks to work out,

23:39

but I

23:40

can't imagine that conversational

23:43

and kind of helping us try to do like the

23:45

nitty gritty stuff and like the kind of like

23:48

low hanging fruit type tasks a day is going to go away.

23:50

But I think there's a lot of work to be done.

23:52

There's no question.

23:54

A lot of work to be done. 100%. Indeed.

24:01

But

24:02

a lot of the work has been done already. And

24:05

what, at least during the keynote,

24:08

Sundar Pichai kind of, you know,

24:10

he set it all up and he positioned, you know,

24:12

the kind of AI conversation and it was

24:15

all kind of framed around introducing

24:18

Palm II

24:19

or Palam, as I like to say, because

24:21

capital P, lowercase A, capital

24:23

L. Palam, Palam,

24:26

Palam, Palam II. Palam

24:28

II, basically this is Google's new language

24:31

model and it's smaller.

24:33

So the second version of Palm, that was the first one,

24:35

smaller in size than the previous version but more efficient

24:38

and with multilingual understanding, which is powerful.

24:41

And it's different model scale for different devices,

24:43

you know, so whether you're using a

24:45

smartphone or a browser or whatnot. And

24:48

really like it's the underlying tech

24:50

that all this stuff is kind of built on. And

24:53

you know, actually if you

24:53

dig into the Google blog, there's

24:55

actually a link to

24:58

the

25:00

scientific paper

25:03

explaining Palm II, the Palm II tech

25:05

report, which is like a

25:08

hundred page document that goes

25:10

through and gives the background and all the methodology

25:13

and all the stuff that came before. It is really, really

25:15

kind of fascinating and way out of my

25:17

realm of understanding. But

25:20

yeah, so it was like, as

25:22

they're telling the story of the keynote, it's

25:24

like, hey, look at the underlying tech underneath all

25:27

this, under the hood, we're developing

25:29

it, we're advancing it, we're adapting it for

25:31

the needs that we have for our various devices or different

25:33

kind of interfaces. And this is what it's all built on.

25:36

You know, they did push that multilingual support.

25:39

You know, they did announce like when they're, I don't know, I don't want

25:41

to steal some thunder later on, but they're talking

25:43

about Bard, how Bard is now going to

25:45

be available in Korean and Japanese. And

25:48

then I think they're planning on rolling out like 80 languages soon.

25:50

So it's going to be multi, you know,

25:52

multi-language, which has got to be even more

25:54

common, you know, training and English. That's a strength.

25:56

That's a strength that Google coverage. I don't know, honestly,

25:59

off the top of my head. chat, GPT

26:01

and the other systems, how they are with multilingual

26:03

support. But I mean, Google's been doing multilingual stuff

26:06

for quite a while when it comes to its training.

26:09

And that was, and that was the second, and that

26:12

was the, and we're going to chase and you're going to talk about later on, this

26:14

is a little bit of tease, but like for

26:16

me, the theme of the keynote was AI, right?

26:18

And what Google's doing in AI, but then

26:20

what, what really kind of pushed it across the

26:22

finish line for me is like, yeah, chat, GPT and open AI

26:25

are doing great stuff and they're building, you

26:27

know, API is underlying tech for other people to go build

26:29

stuff, but Google's got a whole bunch of stuff

26:31

that they can implement indirectly and get

26:33

this stuff to, you know, in

26:36

front of users much faster than anyone else. And

26:38

that's their strategic advantage, you know, so the multilingual

26:40

support and just like the applications that

26:42

are ready to use us. We're going to talk about that in a little bit.

26:45

But like that's really, I think going to be the game changer, whether

26:48

or not

26:49

Palm two is better than open

26:51

AI is model or not. It's

26:53

open to debate, but you can't argue what

26:56

the power that Google has in front

26:58

of the AI now. Yeah.

26:59

Yeah. certain

27:03

languages that are not, you know,

27:06

like, you know, romance

27:08

languages, Western languages like Korean and Japanese

27:11

already are hard enough to translate without

27:14

sounding a little bit like robotic and unnatural

27:16

and not idiomatic. But I

27:19

know they talked last year about improving the models for

27:21

that, where they're not doing like a direct translation model. So

27:23

they compare that, you know, like language model with,

27:26

you know, Bard and with like the accessibility and

27:28

locality of like, Hey, we're putting

27:30

Bard in our stuff and we have all these features

27:32

in our stuff. Like maybe that, that could be the secret

27:34

sauce that can help them kind of

27:36

not feel like they're chasing chat

27:39

GPT and Bing and all those other things. And

27:41

which is like the strength that we've heard about all

27:43

these different years at IO 23, just kind of synthesize,

27:45

bring it all together, mix it all up. I

27:48

keep saying that word. That was been my

27:50

last, that's like the new

27:51

drinking game. Shake

27:55

up a drink or something. I don't know. Yeah. Well,

27:58

speaking of Palm two.

28:00

actually win this leads really

28:02

well into yours because

28:04

this was something that you brought up that definitely

28:07

wasn't on my radar because I'm not a developer but

28:09

this is actually very cool and and I think

28:11

based on the Palm II technology.

28:14

Yeah,

28:15

yeah so there is a descendant of Palm

28:17

II, Cody,

28:18

code E, code Y, Cody,

28:20

that is now powering an

28:24

AI conversation based helper

28:26

inside of Android Studio. So during developer

28:29

keynote which have an app as a consumer keynote, this

28:31

actually took me by surprise but yeah,

28:34

Jay Eason came out and announced that there's a studio

28:37

bot and I think there was about 30 seconds

28:39

of a little bit of skepticism like oh no why

28:41

do I have a chat bot in my integrated

28:44

development environment but and

28:46

I think this goes into kind of what

28:49

we're talking about earlier it's actually really cool

28:51

y'all so basically you talk

28:54

to studio bot you ask it natural language

28:56

questions and it can return

28:59

to you all kinds of things it can return to you information

29:01

from documentation it can return

29:03

to you information about material guidelines and

29:06

it can return to you templates like code templates

29:08

it can help you refactor things from say Java

29:11

to Kotlin it can even tell you about

29:13

when you have a crash or an error it can tell you what

29:15

that error is what it means and

29:17

for kind of low hanging fruit things

29:20

that are just like little oh you forgot to turn on

29:22

the internet permissions

29:23

it can do some of those fixes for you

29:26

and I

29:27

actually really love this and the thing

29:29

here which makes me super excited about this where

29:31

I'm a little bit you know kind of like Debbie downer

29:33

on other things is that all of these

29:35

things like refactoring templates even

29:37

documentation and even like things like static static

29:40

analysis where we have someone telling us

29:42

hey you shouldn't do that in your code exists

29:44

already it's just that now we have a

29:46

conversational interface for it so you can

29:48

just ask rather than you googling

29:50

and figuring out what to Google and

29:53

again studio studios in beta so it's not quite

29:55

as like fluent as a bar to something but

29:58

imagine like not having to Google or

30:00

open up like 10 or 15,000 different websites,

30:02

you just ask the studio. And

30:05

it is pulling on all of these information sources

30:07

that we already have. It's factual or,

30:10

I mean, not quite a factual, but kind of, these are all

30:12

kind of resources that we use today. It's just in one place. So,

30:14

um, yeah, based on Palm II and

30:16

Cody. And I mean, I thought

30:18

that we were done with AI and chatbots.

30:21

But no, it's only

30:23

just begun. No, no, no, developers,

30:27

you get a chatbot and you get a chatbot

30:29

and you get a chat. I'm curious

30:31

to know, uh,

30:32

Mike, I want to throw this over to

30:35

you. Uh, I mean, you know,

30:37

obviously when I'm curious to know what you think too,

30:39

but Mike, as, as one

30:41

of the developers on this panel is

30:43

an AI informed chatbot.

30:46

Like, is that an answer to

30:48

something that you actually want? Or

30:51

is that, are you getting clippy vibes?

30:54

Like I'm curious to know where you stand

30:56

on this. So I kind of am and you kind of, it's

30:58

a good segue because that's kind of what I was going to mention. Um,

31:01

I will start this by mentioning

31:03

that I currently use GitHub Co-Pilot,

31:06

which is AI assisted development.

31:09

It's different though, in that it

31:11

basically provides like advanced, um,

31:13

IntelliSense like code completion.

31:16

Um, and it's really magic for a lot of things.

31:19

Like when we're just mentioning like the template

31:21

stuff that is just like boiler plate, um,

31:24

code comments, it kind of like writes

31:26

out my comments for me. I don't know how it kind of knows,

31:28

but

31:29

those things are magic. Awesome.

31:31

Yeah. Um, I don't

31:34

like having a chat interface to that

31:36

is of

31:37

value, but it's kind of different.

31:40

So I kind of have to get used to it. There

31:42

are times when Co-Pilot doesn't tell me what I need

31:45

and I still have to go search for it and maybe

31:47

this would be that solution. Like Co-Pilot's

31:49

not giving me what I want. I asked studio

31:51

about, Hey, but how do I, um,

31:54

make this text bold? You know, um,

31:58

and I welcome.

31:59

all the assistance I can get.

32:02

I welcome my AI robot overlords,

32:06

and if they make my job easier,

32:08

bring it.

32:09

Well, and that's what, and we talked

32:11

about that in the interview with Dave and Samir, and

32:14

the context of this is that like, there's so much

32:16

stern and drang over AI and

32:19

how AI is taking over, and I've

32:21

dealt with it in my day job, you know,

32:23

like where I'm trying to be

32:25

innovative and trying to cultivate my team to

32:27

be aware of technology and how can we use it as tools,

32:30

and like a lot of folks' first reaction, and not, and

32:32

I wouldn't say Luddite, but it's very like, I don't

32:34

wanna lose my job, do a robot, and robots can't

32:36

do my job, you know, but like the great

32:39

example or the great kind of, you know, the

32:42

point that Mike makes in terms of saving time, and

32:44

we talked about it on the revelation I had, which,

32:47

you know, like 150 years ago, it took a week

32:49

to do the laundry, and then they invented the

32:51

washing machine, and now it takes two hours,

32:53

right? And so what do you do with that time back? And

32:56

from a developer standpoint, when

32:58

Mike, how much time will you get back with these

33:00

tools?

33:02

Yeah, I mean, it's not too, I mean,

33:04

it's a bit of, it's not quite the same

33:06

evolutionary step, but if you imagine like how many years

33:08

ago, we had punch cards, right? And so you spent most of your time

33:10

punching out the dang punch cards. And

33:13

part of, I mean, this is all for like

33:15

modern tooling, even things like Kotlin and things

33:17

like that. It takes away the drudgery

33:20

so that you can focus on high-level things, and that's,

33:22

time is kind of one of those resources that developers

33:25

never have enough of.

33:25

And so

33:28

anything that saves me time helps. And I think

33:30

what is interesting about the chat is that, like

33:32

Google has

33:34

touted some of these, the part I think

33:36

that's interesting about conversational is that the conversational

33:38

bot keeps context. So it's like that shopping

33:41

thing that they always mention where you're saying, hey,

33:44

show me some pants, right? And

33:46

then out of the results, you can say, okay,

33:49

show me this one, but an orange. It keeps the

33:51

context of what you're talking about. So it

33:53

saves you from having to figure out, okay,

33:55

how do I filter these results and not

33:58

lose them? You know what I'm saying? Like it's building.

33:59

on the previous thing that

34:02

you asked just like any other conversation that you have with

34:04

a human being. I think that's what's interesting about Studio

34:06

Bot versus, as Mike mentioned, Copa

34:09

and stuff like that too, is that maybe by

34:11

using a bot that can keep that context,

34:13

you can say something like, hey, what is this error

34:16

and how do I fix it rather than having

34:18

to

34:19

copy paste a stack trace, go out, search

34:21

for Google, you get back some answers and

34:23

then you have to kind of individually copy paste

34:26

the kind of things. There's a lot more

34:28

of a manual process there and it's just

34:30

like, as you said, washing the clothes rather than me taking

34:34

the clothes out the basket

34:37

and smacking them on the rock. I

34:39

don't know what that might be analogous

34:41

to in coding, but the

34:44

spin cycle does that for me. And so similarly,

34:46

these things like copy paste, put

34:48

it in Stack Overflow, look over here, that

34:50

gets taken care of for me so I can actually worry about architecture

34:53

and UI and all that kind of jazz. So I'm

34:55

not talking now. I do think that's the power of it.

34:58

Interesting. We have, I'm

35:01

looking at a rundown. I'm like, oh, we have so much

35:03

more to talk about with AI.

35:05

I think I'm going to do an on the

35:07

fly, make an on the fly

35:10

decision and kind of change things around. So

35:12

that might be a little bit inside

35:14

baseball. Ron, why don't you throw

35:17

to the ad and I'm going to do some restructuring

35:20

because we have a lot more AI to talk about. And

35:23

we do have to go to an ad and we're

35:25

going to get to all the really important stuff. So Ron,

35:28

take it away. And yes, we

35:31

will take a break in all about AI to

35:33

tell you how this episode of all about Android

35:35

is brought to you by cash fly. And

35:38

listen, we've all been making internet content

35:40

for so many years and we know for a fact that viewers

35:42

don't hang around for videos that buffer

35:45

shoppers, abandoned carts and e-commerce sites

35:47

that are slow and gamers leave bad

35:49

reviews when the latency is high. Be

35:51

ready for those fluctuations with cash fly.

35:54

Customers expect a faultless experience when engaging

35:57

with content on any device, anytime,

35:59

anywhere in the world.

35:59

Building trusted CDN relationships

36:02

since 1999, Cashfly has held

36:04

a track record for high performing, ultra

36:06

reliable content delivery for two decades. Cashfly

36:09

pioneered the use of TCP Anycast

36:12

in 2002, an innovation that CDNs

36:14

continue to build upon.

36:16

And quality of experience is the single

36:18

most critical metric when serving content simultaneously

36:21

to a large and distributed audience on a global

36:23

scale. And your delivery stack can be

36:25

your secret weapon. With

36:27

Cashfly, you'll get ultra low latency

36:29

video streaming that will deliver video to more than

36:31

a million concurrent users. You'll

36:33

also get lightning fast gaming, which will deliver downloads

36:36

faster with zero lag, glitches or outages,

36:39

mobile content optimization that offers automatic

36:41

and simple image optimization so your site loads

36:43

faster on any device. And

36:46

Cashfly is the only CDN built for throughput,

36:48

delivering rich media content up to 10 times

36:50

faster than traditional delivery methods

36:53

and 30% faster than other major CDNs.

36:56

Plus, you never have to pay for service

36:58

overlap again. You can get flexible month

37:00

to month billing for as long as you need it and discounts

37:02

for fixed terms once you're happy. Design

37:04

your own contract when you switch to Cashfly.

37:07

With more than 3,500 clients in over 80 countries, organizations

37:11

consistently choose Cashfly for scalability,

37:13

reliability and unrivaled performance. And

37:16

you might be wondering, how do we know all this? Well,

37:18

we've been using Cashfly for over a decade and we

37:20

wouldn't have it any other way.

37:22

Learn how you can get your first month free at cashfly.com.

37:25

That's cashfly.com and we thank them

37:27

for keeping the content flowing fast

37:30

and

37:31

right on getting to our viewers.

37:33

Thank you, Cashfly, we appreciate it.

37:35

Cashfly is awesome. Been

37:38

with them for so very long. They're

37:40

responsible for so much of what we do. So

37:42

thank you, Cashfly. Okay,

37:45

I, through the magic of on-the-fly

37:48

producing, have reordered the doc. So

37:51

we're

37:51

going in, the

37:53

show's not gonna be any longer. It's just, we have more stuff

37:55

to talk about with AI. I thought we could do it all

37:57

in one block, but it was clearly

37:59

it was.

37:59

is too much. So let's keep this show

38:02

rolling here and continue with

38:04

the AI discussion because I think this is all

38:06

of these topics are actually really important. I didn't want

38:08

to do it short, you know, give a short service. So

38:11

first of all, you alluded Ron to

38:13

the ways in which Google is taking

38:16

these many different AI efforts

38:18

and integrating them into services. And when I

38:21

have a couple of examples

38:23

of this and there's some really interesting discussion

38:25

that we can get into around this, but just real

38:27

quick to set the scene, Gmail

38:29

getting generative AI features

38:32

called help me write. This

38:35

will be a feature integrated into

38:37

Gmail. I think you have to opt into it

38:39

that you could say like,

38:41

I need to write it in email to this person

38:44

that says, I'm not going to make it on time. You

38:47

know, I got caught up in traffic, but I want to reschedule

38:49

for next week. I checked my calendar

38:51

for three open time slots or whatever.

38:54

And then you generate it, generate it, help

38:56

me write it. And boom, instead of me having

38:58

to write this entire email, it sends

39:00

that person all the information you

39:02

can expand it, contract it, that sort of stuff.

39:05

Really weird messages, magic

39:07

compose, which we alluded to last week on

39:09

the episode where you can rewrite

39:11

text messages. So you put in a short text

39:13

message and you tell it the style that

39:15

you want it to rewrite in and it creates a

39:19

new text message for what you've already put there.

39:21

Maps, getting an immersive mode

39:24

that flies through, this looked really cool, that

39:26

flies through

39:28

the route that you're navigating to give

39:30

you kind of a virtual tour of what that route

39:32

is going to look like. It's

39:35

almost like you get to experience your

39:37

routes

39:38

in, you know,

39:41

they use AI and computer

39:43

vision to kind of merge together this

39:45

experience. So you can see what

39:47

that route is going to look like. Photos,

39:50

getting a new magic editor, which is kind

39:52

of similar to some of the AI editing

39:54

features that we've seen in Google Photos

39:57

and on Google Camera before, but

39:59

this is almost

39:59

like that taken to the next level. It's like

40:02

the example they have is this kid sitting on a

40:04

bench and it's like, we want the bench to be more

40:06

centered in the photo. We want the clouds

40:09

to be bluer. We want to remove,

40:11

you know, this, that, the other thing. I mean, it's just like

40:13

all these edits being made to the

40:15

image that normally, you know, again,

40:18

it's an example of how AI is doing things

40:20

that normally we would have needed to be skilled in the

40:24

in the feature set of an app

40:26

like Photoshop in order to do all this stuff.

40:28

And it would take a while to make

40:29

it all happen. And now it's like

40:32

you click it and you tell it what you

40:34

want it to do. And it does the thing. So

40:37

I think the question there that could be interesting

40:40

perspective that I've been talking a little bit

40:42

with Micah Sargent here at Twit. He brought

40:44

this up and ever since then I've been thinking about it

40:47

is the the authenticity

40:50

or lack thereof when we

40:52

have AI systems writing

40:55

all of our communications for us. He

40:57

mentioned to me that using an AI

40:59

to write a text message, he's

41:02

like, I hate it. I was like, you

41:04

know, I want to know when I'm talking that when I'm when

41:06

I'm having a coming,

41:08

having contact with someone, communicating with

41:10

someone that we're actually communicating

41:12

and it isn't a robot or a computer

41:15

that's kind of standing in for me.

41:18

And I don't know, there's some there's something

41:20

there. Does that

41:22

does that impact you guys at all? Like,

41:25

how do you feel about that? I mean,

41:27

personally, I think that texting

41:30

is such a like I'm going to spend more time

41:32

telling the AI what to say in response.

41:34

So I can just text to myself. Exactly.

41:37

That's my thought.

41:38

Yeah. What do you think, when? I

41:41

do think there is

41:43

that there might be some room

41:45

for this, but this is kind of like by case basis

41:48

or your mileage might vary. It was

41:50

really funny because not in

41:52

text. So maybe this invalidates the

41:54

argument. But I know I think sometimes we do talk

41:56

to different people, like maybe

41:58

whether it's introducing our ourselves to a

42:01

new person that we want to engage in with business

42:03

or something. I don't know, sometimes I

42:06

know my husband and I often have arguments about, okay,

42:08

I want to, I don't know, engage

42:10

this landscaper. And some people do

42:12

actually still do a lot of work and

42:15

contracting through tax, through word

42:17

of mouth, through recommendations. So I don't know.

42:19

I feel like I'm just very focused on, well,

42:22

sometimes people are just not sure how to approach people or

42:24

not sure how to sound like, like

42:27

not weird when you just text

42:29

someone out the blue. So I could see there's

42:32

probably places where this might just help someone

42:34

kind of like write that first into your text or be able

42:36

to kind of analyze things. I don't know. I

42:39

think

42:39

it almost feels like a prerunner to

42:42

something else, like an experiment to let people,

42:44

hey, here's some, here's magic compose,

42:46

play around with it and see if

42:49

it can be of use to you or even

42:51

get like, you know, human use feedback

42:53

on whether you ask it

42:56

to jush up your text and it accidentally insults

42:59

the person that you're talking to. I don't

43:01

know. Like, I just feel like it's like

43:03

a starter, you know, it's like a little test bed that they're

43:05

sticking their toe into the water

43:06

to kind of more broadly do that. And

43:08

this isn't the starter because that was

43:10

one thing that I mentioned to, I mean, isn't

43:13

necessarily the starter I

43:15

mentioned to Micah. Like I already

43:18

on my text messages that come in, you

43:20

know, it gives me those

43:22

little one word answers. Yes, not

43:25

really. Blah, blah, blah, you know, and I

43:27

do use those. So is that

43:29

me using an inauthentic communication

43:32

because the computer presented that or is that just

43:34

me going quicker to the

43:36

thing I was already going to say, Mike, how

43:38

do these features land for you? Okay,

43:42

so this is a great all about

43:44

Android moment because I want to call out

43:46

Ron for

43:47

the example he gave us before the meeting

43:50

about he wrote a love letter to

43:52

his wife using chat

43:55

GBT. So I just did

43:57

it actually myself. I did it in Bard

43:59

just to know. I did it in Bard as well, actually,

44:02

just to be perfectly clear and on the record.

44:04

So I actually just did it while I'm kind of sitting

44:06

here. And it is remarkably

44:08

good and feels very authentic,

44:11

talking about, you know, how she makes me a better man.

44:14

You know, it's very generic,

44:16

but,

44:17

you know, I think I'm gonna get, you know, some

44:20

good, you know, kisses tonight

44:22

after she reads this. I did it. Yeah,

44:24

I did it. So it's part of the announcement, they

44:26

said they were going to embed the little

44:28

like the little magic wand or the sparkly icon

44:31

or whatever in Gmail. I signed

44:33

up for it. It got enabled to my Gmail account. And

44:35

so the first time I opened up the prompt and I said, you know,

44:38

help me write an email to my wife about how much

44:40

I love her or something like that, whatever I said. And

44:43

it wrote a very, like to Mike's

44:45

point, a very believable kind

44:47

of thing, you

44:47

know, and I sent it to her and she came in and she was like, oh

44:50

my God, that was so nice. And I was

44:52

grinning and she was like, you didn't write it, did you? It was like,

44:54

no,

44:54

Bard did it. But and

44:57

like the examples they gave in the demo, like

44:59

Dave, Dave Burke of Google, who we talked to

45:01

in the interview that you can listen to at the end of the show, did

45:04

the live, you know, did the live demo of it. And it was

45:06

like, you know, I want to send an email to someone congratulating

45:08

him doing a good job. And like, it's

45:10

this idea of like,

45:12

short attention span approach. Like I want

45:14

to, I want to email Jason and thank him for doing all

45:17

about Android and being so awesome. And it will write

45:19

this whole thing and insert it. It's almost like greeting

45:21

cardification of our emails. That's exactly where

45:23

my mind was at.

45:24

Exactly. You pulled

45:27

the words out of my brain, Ron, because as

45:29

I was hearing this, I started

45:32

to kind of really clued into the fact that

45:34

like we already kind of do this. We

45:36

already put our control

45:40

over what we say about how

45:42

we feel about someone just using this as

45:44

an example in the hands of someone else.

45:47

We go into a store, we pull a card

45:49

off the shelf and we read it and we go, Oh,

45:52

those words match what I want to say.

45:54

So

45:54

I'm going to buy this and I'm going to take

45:56

these words and offer them as if they

45:59

are mine. And I'll see you next time. Bye. I mean, obviously they're

46:01

not. It's a greeting card, you know? Are

46:03

we losing the skill or the training to

46:08

write those words yourself sincerely?

46:10

Like because there was a time, there

46:13

was a time in our civilization in 1835 when

46:17

someone growing up said, I wanna be a poet.

46:21

I wanna be a writer. I wanna learn how to do

46:23

all this sort of stuff. And like, when was the last time you met somebody?

46:25

Well, actually I know somebody who's a poet, but still. But

46:29

I was just saying like, going

46:31

back, and I know this is counter to what we were just saying

46:34

about programming, about the chatbot in Dev

46:36

Studio and how that making easier, does

46:38

it make my life any more efficient to

46:40

have a AI bot

46:43

write a love email to my wife? I don't

46:45

know. Like I'm looking for ways to

46:47

introduce this into my life

46:51

and make it realistic.

46:53

I will say another example of how I used Bard recently

46:55

was I was at, it wasn't in email,

46:58

but it was just like, and the idea of them integrating it into email,

47:01

into workspace, into docs, into

47:03

spreadsheets and stuff like that, I think it

47:05

was really compelling because like I was on a work call

47:07

the other day and we were brainstorming

47:09

to come up with a new product

47:11

kind of naming type thing. And I was in Bard and

47:13

I was just like, give me suggestions for a product

47:16

and with these attributes or whatever. And it gave me a list

47:18

and none of them were viable.

47:20

But of like the 10 things that suggested,

47:23

at least three triggered sparks

47:25

amongst the people I was talking to that led to

47:27

possibly viable options. There you go.

47:30

Yeah.

47:30

So I don't

47:31

know. Launchpad for something else.

47:34

I don't know. So we're all people that talk

47:36

on podcasts. So there's a level of

47:38

comfort with discourse, with words.

47:41

And I know maybe this is just me

47:43

kind of trying to optimistically

47:46

apply this, but I can't imagine like

47:48

so many times in my life where I was less

47:50

confident enough and had some social

47:52

anxiety where, you know, like maybe if

47:54

I wanted to like, let someone know something,

47:57

whether that was something serious or important or

47:59

something where I just. I just couldn't find the words or

48:01

was worried that it sounded aggressive

48:04

or not, whether it sounded sensitive

48:07

or not to

48:08

whatever the situation was. And

48:11

I'm a person who very often when I send someone

48:13

something, even if it's innocuous, I'd be like, go to my husband,

48:15

go to my sister. Hey, does this sound

48:18

mean or whatever?

48:21

And to some of the points that you all said,

48:23

at some point you realize

48:26

that there's a level

48:27

of facsimile

48:29

and artificiality. So

48:34

I can't help but think that if someone

48:36

really cares about crafting their own words that

48:39

they hopefully would still do it. But I don't know, maybe that's

48:41

just me being optimistic. But I do think

48:43

there is probably a place for this. It's

48:45

just not maybe what

48:47

we expect or maybe what,

48:50

especially us as people that are

48:52

people creating content who

48:54

show up and put our faces and our voices

48:57

on the interwebs and that who write for a

48:59

living might,

49:00

maybe like we're not

49:02

the audience for this per se.

49:04

I don't know. Yeah. I mean,

49:06

when I think about it, like that feature, I'm super torn because

49:09

as they were showing it off, the Gmail is the one

49:11

that I'm primarily thinking about, the write

49:13

for me Gmail feature. I

49:16

was like, yeah, you know, there are emails that I

49:18

write that are very, you

49:20

know, very much the same between

49:23

every person that I send it to. I don't copy

49:26

top to bottom, but I write basically the same

49:28

things with slightly different words, you know, invites

49:30

for shows and everything. Like there's only so many ways

49:33

you can invite, you know, 20 people

49:35

a week on to our shows and

49:38

have it be differentiated and why differentiate

49:41

when the facts need to be there and all

49:43

these kinds of things. But at the

49:45

same time, like I don't, I don't feel

49:48

right

49:49

handing the keys over to an AI

49:51

to write all of my communication for me. But

49:53

I bet you there's somebody out there that's

49:55

willing to go that far and willing to say, you know

49:58

what, I'm not writing another email.

49:59

ever. I'm just going to tell, help me

50:02

write

50:03

what email I want to write and

50:05

do some light editing every single time.

50:07

I guarantee you that people are going to use that.

50:10

Honestly, it's not that like, like it reminds

50:12

me of when I would ask a professor to write a recommendation

50:15

letter for me or a teacher and they say, write it for

50:17

me and give it to me and I'll, I'll edit it. And then

50:19

send it like, I do it at work all the time. I tell

50:21

people I'm going to write this up, send it to me, I'll make the edits and

50:23

then send it and put it, you know, kind of like, I don't

50:26

know. It's like, I'm, I'm vacillating between it because

50:28

I can see that like Jason, you're right. Like there's a lot of

50:30

redundant emails that I write often

50:32

that had to do with the same thing. And I can, and it's easier to

50:35

do the heavy lifting of the work and

50:37

then just edit it to make it sound in my voice than

50:39

do it

50:40

all from scratch, you know? So I don't know. I think,

50:42

I think there is value to it, but I don't think it's replacing

50:45

the thought I put into what I

50:47

do. That was something that the big reaction

50:49

that I saw online during the keynote,

50:51

you know, they were doing live coverage

50:53

here, Leo and Jeff Jarvis. And so I was logged

50:56

into the discord and kind of live, live blogging

50:58

my own thoughts and everything. There were a lot of people

51:00

really commenting on kind of the blandness

51:03

and the saminess of the AI

51:05

responses and how it's, oh, really

51:07

it couldn't get more creative. And I was, and when I was thinking about

51:10

it, I was like, yeah, but how much of our communication

51:12

that we do on a regular daily basis is

51:14

very bland and is very, you

51:16

know what I mean? Like there's so, there's so

51:18

much of our communication. We like to think that we're really creative

51:21

and everything like that, but so much of business

51:23

is derived around these very set

51:25

in, you know, in stone processes

51:28

and things. It's just a reflection of

51:31

how we write. So if it's,

51:33

thank you for

51:33

your time. Best wishes. Yeah.

51:35

So anyways, so it's interesting.

51:40

I think this is, you know, when we're talking about AI and

51:43

ethics and, and, uh, best

51:45

practices, you know, these are things that I

51:47

think businesses are going to have to start making decisions

51:49

around. Do we allow our employees

51:52

to allow, you know, to use these

51:55

AI tools to write emails or do we say,

51:57

Hey, no, that's not okay because of secrets

51:59

or because

51:59

Uh, it doesn't impart a human and a

52:02

human quality to the conversation. And we're

52:04

a business. We want people to feel like we're connecting

52:06

with them and these systems remove the

52:08

human, you know, these things are going to have to be discussed

52:11

and figured out on a company-wide basis. I

52:13

think, uh, I think that's in our future.

52:15

Anyways, we could talk about this forever. Uh,

52:18

we have a few more things before we get to the next

52:20

break. So when, uh, I

52:22

guess this kind of falls into similar territory.

52:28

Oh, sorry, are we talking about Android 14? Now we're talking

52:31

about the other thing. No, we're talking about the, the wallpapers

52:33

we're in. Okay. Yeah. Sorry. I

52:36

know I shifted things around.

52:37

Yeah. So, you

52:40

know, none that we don't have enough generative content

52:42

and things to discuss already, but

52:45

Google is kind of aiming its generative

52:47

AI at your wallpaper. And

52:49

so there's a couple of like interesting elements to this, right?

52:52

So there is the ability to create

52:54

like a wallpaper from, you know, whatever emojis

52:56

that you would like, but in more Googly

52:58

flavor, there's also the ability to kind of leverage

53:01

AI to create cinematic

53:03

wallpapers, um, that kind of like,

53:05

you know, like, Ooh.

53:06

Ah. Um,

53:09

and so yeah, it's, it's kind of, and as well as,

53:11

let's see, what else, what else do we have? Oh yeah. And as well as

53:13

kind of more, if you think about more like

53:15

mid journey, simple diffusion, you can prompt

53:18

Google to create you, you know, this, whatever,

53:20

whatever your mind can, can,

53:23

whatever your mind can prompt, uh, it will

53:25

also simulate you with generative art, uh,

53:27

wallpaper. So emoji wallpapers, cinematic

53:30

wallpapers from your photos, where it kind of analyzes it

53:32

and gives it, gives a static image motion and

53:34

generative AI wallpaper for if

53:37

you want to have a wallpaper of yourself,

53:40

uh, as a D and D character.

53:42

There you go. Um, that's kind

53:45

of neat, I guess. I mean, but again,

53:47

it's like, it's wallpaper. It's

53:49

wallpaper.

53:49

So I think that was the one

53:51

thing that I feel like, so,

53:54

uh, like a bunch of us, Android does, we're in a chat and I think

53:56

we got a little bit snarky about this point to

53:58

be fair, like all of these are really cool.

53:59

isolation, but isolation is kind of

54:02

thing. It's wallpaper. Yeah. Um, so

54:04

I think it's like a small example

54:06

of like a big technology. Um,

54:09

but yeah, are they pushing this

54:11

because they want to, uh, push the dynamic

54:14

theming part of material three that's

54:17

driven by your wallpapers.

54:19

It took me. That's a really good point. Like giving,

54:22

giving, giving people more. We

54:25

need to switch your wallpaper. Yeah.

54:28

I mean, at the end of the day, it's, it's material

54:31

use core promise.

54:33

If, if I, you

54:35

know, based on what I feel like I

54:38

know about material, you is making

54:40

your device as customized to you

54:42

as possible within the framework of what

54:45

they've set out and the generated AI

54:47

system really allows

54:49

you to get as unique as you want, because

54:52

literally that image doesn't exist before

54:55

you put in your prompt and it comes back with something

54:57

that's about as unique of a wallpaper as

54:59

you can come up with, uh, Mike, unique,

55:02

you. Yeah. Mike,

55:05

you've been spending a lot of time with G with

55:07

generative AI and prompts and everything

55:09

like that. Like are, are people

55:12

like, how do you think this is going to go? Cause I mean, this

55:14

is largely, this is going to be much

55:16

more consumer facing. There's going to be a lot of people that have

55:18

never used something like a chat GPT

55:21

or a mid journey to use

55:22

the system. I have to imagine Google is going

55:25

to make it friendly, right? Um,

55:30

well, I don't know. So

55:32

I don't know if this is Google integrations with Firefly,

55:35

but what I will say is using

55:38

these generative, uh, text

55:40

to image tools is

55:42

so much fun. It

55:45

kind of gives you weird stuff, but it's, it reminds

55:47

me of the early days of Google search where you just

55:49

were kind of searching for stuff just to see

55:51

what it would give you. Um,

55:53

it's really delightful and really fun to play

55:56

with. Yeah. I mean,

55:58

it looked like a lot of fun and

55:59

having it kind of baked into Android 14, which

56:02

we're going to talk about after the next

56:04

break in a little bit

56:06

more detail. But at the same

56:08

time, can help, but just kind of

56:11

chuckle a little bit because I

56:14

think the wallpaper as feature ship

56:17

has kind of sailed. You

56:20

know what I mean? It kind of felt like, oh

56:22

no, we're back to wallpapers. There's

56:24

really not a whole lot going on here, is there? If

56:26

like the wallpaper is- Gateway

56:28

drug to more generative

56:30

AI because it's easy, it's simple, it

56:32

won't disrupt their life. It's not disrupting

56:34

their lives so far. Well, I think

56:36

this came up in our interview with Dave and Samir

56:38

as well also. So go back and listen to Dan and

56:42

you can hear Dave in his own words saying it, but they wanted

56:44

to give it a practical, like

56:46

this is something that actually people can do

56:49

on their phone and immediately see the power of it. So

56:51

when I think you're right in terms of like a gateway drug, it's

56:54

like it hurts no one. It gives you a chance

56:56

to express yourself. And here's the thing,

57:00

in terms

57:02

of this, you could

57:04

do this and not

57:07

thank the AI for it. Right?

57:11

Like you don't need to position this as AI.

57:13

You could position this as material, you, and

57:16

all that sort of stuff. And then let the under

57:18

the hood people be like, oh my God, is there AI thing

57:20

doing it? Like you could just be like, type this in, you can

57:22

make your own wallpaper type, you know, like- Yeah,

57:24

that's true. But they chose to position, you know, like, yeah. Yeah,

57:27

it's interesting. But of course they're going to make

57:29

it an AI thing because AI

57:32

is very now and Google's mission

57:35

was to prove to you, hey, we've been doing this a long

57:38

time. We deserve to be in the conversation when

57:41

it comes to AI. Exactly. Which is a perfect

57:43

segue to you, Ron, because

57:45

you ranked an article about the

57:48

assistant and how

57:50

it was missing. Yeah,

57:52

so yeah, thank you, Jason. We

57:55

talked a couple of months ago about how

57:57

Google was shifting a lot of key staff members from-

57:59

the Google Assistant team to the

58:02

Bard and AI team. And

58:04

then as we went through the keynote, I forget

58:06

who said it, but at some point I was talking to somebody

58:09

and someone's like, I don't think they said assistant

58:11

once in the presentation at all.

58:13

And they're like, Jason, it might've been you actually. I was

58:15

like, yeah, you're right. They didn't say it at all. And sure enough,

58:18

you know, I'm referring to this article in wire that I saw

58:20

the hit today that was, you know, the curious case

58:22

of the missing Google Assistant. And

58:24

it's just basically saying how, you know,

58:26

the focus is on AI

58:28

and assistant

58:29

has been the thing that

58:32

they built their AI interface around.

58:34

And it was gone.

58:36

It was just absent from the keynote. And

58:39

I would love to know why or how or whatever

58:42

and why that choice

58:44

was and what it means to the future of assistant.

58:46

So we'll see. I

58:48

mean, maybe they're going to get rid of it, right?

58:51

And then it just becomes how you interact with

58:53

Google stuff. And it's just a native part of it. And

58:56

it could be, I've heard some people speculate that, you know,

58:59

people don't like it anymore. They're down on Siri,

59:01

like they're down on assistance. So like, don't focus

59:04

on it. But yeah, what the, what

59:06

the reasoning was. Yeah.

59:08

I mean, it is interesting. I mean, especially

59:11

when you consider, you know, as

59:13

we have talked about on this show in recent

59:16

months with this whole AI

59:19

thing that's happening right now, Google

59:21

was there. They were there, you know, many

59:23

years ago, they fired off the, the

59:27

voice assistant trend,

59:29

I would say

59:30

largely, but they were very accessible

59:33

when it came to, to that. And they

59:35

proved, I think, you know, in that

59:37

first, that first go, Google

59:40

assistant was the winner, you know, certainly

59:42

it wasn't Siri assistant

59:45

for the longest time,

59:47

gave you what you were looking for. And

59:49

I would say it's broken down in recent years, but

59:52

Google was there. It had the pedigree to

59:54

really dominate with AI. So it is

59:57

telling that there was no real assistant

59:59

news.

59:59

I went to just now the Google

1:00:02

blog, the keyword blog where,

1:00:04

and the, uh, articles, a hundred

1:00:06

things we announced at IO 2023, a hundred things. And

1:00:10

I did a search on the page for assistant.

1:00:13

It appears twice. Number 14

1:00:15

is thanks to machine learning technology, assistant

1:00:18

voice typing on pixel tablet is

1:00:20

almost three times faster than regular typing.

1:00:23

Okay. And then the last

1:00:25

one is number 99 out of a hundred.

1:00:29

Um, with the new app actions test

1:00:31

library and the Google assistant plugin

1:00:33

for Android studio, uh, it's

1:00:36

now easier to code, easier to emulate

1:00:39

the user experience to forecast, uh,

1:00:41

user expectations and blah, blah, blah. So

1:00:44

not a whole lot going on with assistant at Google

1:00:46

IO this year, even according to Google

1:00:49

and a hundred things that were announced only to very,

1:00:51

you know, very small things.

1:00:53

It really is like a Sherlock murder mystery

1:00:55

with the curious case of the missing Google assistant. And

1:00:57

then we had like, I mean, I don't know if we talked

1:00:59

about it specifically, but there was a story months ago

1:01:01

about losing $10 billion for

1:01:04

Amazon. So it's just all nefarious,

1:01:06

right? Like one assistant loses $10 billion. Another

1:01:08

one is like missing from the scene. What's happening

1:01:10

to all the assistants? Where are they going? What's happening? But

1:01:13

yeah, I mean, like,

1:01:13

they're being promoted. Assistance are no longer

1:01:15

assistance to others. They're, uh, commanding

1:01:18

their own ship. I don't know. Yeah. This

1:01:22

pivot, pivot. Yeah. The assistants

1:01:24

are pivoting. That's what they're doing. They're

1:01:28

entering the witness protection program and becoming chat

1:01:30

bots. That's, that's

1:01:33

it. Um, okay.

1:01:36

Let's take a break

1:01:37

and, uh, we will

1:01:40

come back and talk about even more IO

1:01:42

news that doesn't really have a whole

1:01:44

lot to do with AI, at least not

1:01:46

directly. Watch it prove us

1:01:48

wrong when we get there. Um, but

1:01:50

I did want to take a moment and tell

1:01:53

you a little bit about us. We here at

1:01:55

twit do these shows because

1:01:57

we love technology. We also know.

1:01:59

you, well at

1:02:01

least we hope you love technology as much

1:02:03

as we do. We are dedicated to bringing

1:02:06

you awesome content about technology.

1:02:09

It's what we've been doing. I mean, Leo's been doing

1:02:11

it since 2005 with the This Week in Tech

1:02:13

podcast. That was when it started 2005, like

1:02:16

almost 20 years ago. Isn't that crazy? We've

1:02:18

been in the business for a very, very long time

1:02:21

and we create partnerships

1:02:24

with brands that we

1:02:26

trust. So any of the brands that we talk

1:02:28

about on the shows that

1:02:31

sponsors,

1:02:32

we have worked with

1:02:34

them to select the brands that we actually

1:02:38

like, that we agree with, that we

1:02:40

approve, that we use. A lot

1:02:42

of thought goes into that.

1:02:45

And it's kind of our job

1:02:47

to introduce you to the audience to those

1:02:49

brands as a result. Half

1:02:51

of our listeners are in management positions

1:02:54

or above. 65% are involved in

1:02:56

their company's decision making. So we

1:02:58

have a lot of high level folks watching

1:03:01

our shows and that just

1:03:04

means that our audience is really raising the bar

1:03:06

for us and our partners, explaining

1:03:09

why our network continues to stay at the

1:03:11

top of delivering this tech news

1:03:13

that's reliable and engaging. Partnering

1:03:16

with Twitch, if you happen to be in that

1:03:18

position, means that you're going to get the gold standard

1:03:21

in podcast advertising with valuable services.

1:03:24

We have a whole suite of valuable services,

1:03:26

a full service continuity

1:03:29

team that supports everything from copywriting to graphic

1:03:31

design. We've got embedded ads. Those

1:03:34

are unique ads. They're unique every single

1:03:36

time and we always over deliver on

1:03:38

our impressions. Onboarding services,

1:03:41

ad tech with pod sites, which

1:03:43

is free for direct clients and detailed

1:03:45

reporting along with that.

1:03:47

You get courtesy commercials that are shareable

1:03:50

across social media and landing pages. And

1:03:52

we have a bunch of other free goodies. Things

1:03:54

like mentions in our weekly newsletter sent

1:03:56

to thousands of fans, bonus ads,

1:03:58

social media promotion.

1:03:59

We really

1:04:02

want to make partnering

1:04:05

with twit appealing and worth

1:04:07

it, worth the time, worth the money. And

1:04:11

we want to work with you. We want long-term

1:04:13

partners that want to grow with us. And

1:04:15

we have partners who have shared

1:04:18

with us on their own, kind

1:04:20

of how they feel about it. Tim Broome is the

1:04:22

founder of ITProTV. And

1:04:25

if you've been listening or watching

1:04:27

our shows for the last 10 years, you

1:04:29

know they've been with our network for 10 years. He

1:04:31

said we would not be where we

1:04:33

are today without the Twit network. Mark McCreary,

1:04:36

the CEO of Authentic, has partnered with

1:04:38

Twit for 16 years, said

1:04:40

the feedback

1:04:41

from many advertisers over 16 years

1:04:43

across a range of product categories is that

1:04:45

if ads and podcasts are going to work

1:04:47

for a brand, they're going to work on

1:04:49

Twitch shows. That's all to say. We know

1:04:52

what we're doing and we want

1:04:54

to work with you. So if you're ready to elevate your business,

1:04:56

reach out to

1:04:57

us. You just have to email us, advertise

1:04:59

at twit.tv. And you can work with us,

1:05:02

work with Twit.tv's world-class

1:05:04

audience. You can reach them. That's what it's all

1:05:06

about. That's advertise at twit.tv.

1:05:10

All right. We've got more news, but

1:05:12

Ron, you have something. I was going

1:05:15

to say, before we moved on, I did ask

1:05:17

Bard why didn't Google talk about Assistant

1:05:20

at I.O. And I just wanted to

1:05:22

share what it said. It said there are

1:05:24

a few possible reasons why Google didn't talk about Assistant

1:05:26

at I.O. It said the Assistant is already

1:05:28

mature. It's been around for several

1:05:30

years now. It's already a very powerful and capable product.

1:05:33

As a result, may not be seen as a priority. Second

1:05:37

option was Google's focusing on other things. They're

1:05:39

always working on new products and features. It's possible

1:05:41

they're prioritizing other things at I.O. For example,

1:05:44

they talked about new product products like the Pixel

1:05:46

7a and new AI

1:05:48

powered features like Bard. And the last

1:05:50

one is Google's facing competition. Another number

1:05:53

of other companies are developing voice assistants, such

1:05:55

as Amazon's Alexa and Apple Siri. As

1:05:57

a result, Google may be feeling pressure to focus

1:05:59

on. developing new features. And they

1:06:01

said, ultimately, only Google knows for sure why

1:06:03

they didn't talk about assistant at IO this year. However,

1:06:06

it's likely a combination of factors played a role in

1:06:08

that decision. So,

1:06:10

okay. I'm,

1:06:12

Bard is like a part of my life now. I love it.

1:06:17

Oh, Bard. Okay.

1:06:21

So let's see

1:06:23

here. I'm trying to make sense of all the, just the

1:06:26

reorganizing

1:06:27

that I did. I think we

1:06:29

are at Android 14

1:06:32

and when this falls in your

1:06:34

lap, this was kind of an interesting thing. And

1:06:36

again, we will tease forward to the

1:06:38

interview that you will catch after

1:06:41

the credits with Dave Burke and Samir Samat because

1:06:43

they even said it, I mean, a must, much

1:06:46

of that interview is about Android 14,

1:06:48

but they even said like Android 14 is

1:06:50

kind of an under the hood update.

1:06:52

What are your thoughts? Like what are we up to?

1:06:55

Yeah. So I like to call, like, I feel

1:06:57

like a lot of, uh, late lately,

1:06:59

uh, Android has been what we would call a dot release. So

1:07:01

when we version things in developer world, very often

1:07:04

we save like whole numbers, like 1.0 for

1:07:06

like big flashy things that you would get marketing about.

1:07:08

And then kind of like the nice like work,

1:07:10

horsey, fixing bugs, minor like

1:07:13

updates, things that aren't really like flashy and

1:07:15

awesome that you want to hold like keynote for. They're

1:07:18

still very important. Like tend to be like that releases.

1:07:20

So 1.1,

1:07:21

1.2. So it kind of feels

1:07:23

a little bit like kind of like that, right? That

1:07:26

maybe the reason

1:07:28

that Android 14 was not so mentioned

1:07:30

is that it's just doing the hard work under the hood,

1:07:33

but I mean, like each individual thing in

1:07:35

isolation might not seem so crazy and awesome, but

1:07:37

it's an aggregate. It's a lot of

1:07:40

much better experiences, uh, in

1:07:42

small ways because the small things add up, right? So

1:07:45

just to kind of go down, uh, this

1:07:47

list and again, I, I thought it was great.

1:07:50

Uh, and to give Dave Burke a lot of credit when

1:07:52

we asked him about Android 14 or Android 14

1:07:54

and not to spoil anything,

1:07:56

the ban was on, he was on message and he had a lot of

1:07:58

great things to say about Android.

1:07:59

But yeah, just to

1:08:02

give the folks here an understanding

1:08:05

of what

1:08:06

you would see in

1:08:08

Android 14, once it

1:08:10

comes out, would be things like clock customization

1:08:12

for the lock screen. Again, a small

1:08:16

quality of life, fun little enhancement, maybe

1:08:18

not necessarily worth the keynote though. I

1:08:20

don't know, I'll

1:08:20

kind of be kind of fun. But

1:08:22

little things like data privacy. So

1:08:25

they actually, one thing that was really important

1:08:27

is that when a dev changes the way they share data,

1:08:29

like if they share data with a third party, you

1:08:31

get to know about it. Very important

1:08:34

stuff. Again, not the most flashy thing that you'd want to make a Super Bowl commercial about,

1:08:36

but super important. Camera

1:08:39

flash notifications. So if you're the kind of person who goes

1:08:41

heads down a lot and you really, really,

1:08:43

really need some extra, like

1:08:46

a nudge when you get a notification,

1:08:48

you're going to ask Android 14 to

1:08:51

flash the camera flash.

1:08:54

Health Connect is actually getting integrated into the

1:08:56

OS and you're

1:08:58

going to get easier access for hearing and controls

1:09:00

as well as warnings when you listen to high volume

1:09:03

for an extended period of time. So kind of

1:09:05

wellness and wellness and health of users.

1:09:10

There's something really cool called Smart File Transfer

1:09:13

that allows developers like us

1:09:16

to basically create a... Like if you're downloading

1:09:18

your favorite podcast or whatever,

1:09:21

like a YouTube video, if you start that

1:09:23

transfer,

1:09:24

the OS will do the work of making

1:09:26

sure to reschedule or pause it if you lose connectivity.

1:09:28

So again, that's a really

1:09:30

awesome feature. Just kind of

1:09:33

hard to make a keynote around. But it

1:09:35

will be there for developers as well

1:09:37

as third party support for Paskies. Not

1:09:40

that we've ever, ever talked about Paskies on this show

1:09:42

or

1:09:42

Paskies in general. Definitely not.

1:09:45

But if we did, we would have to

1:09:47

definitely mention that Android 14 is going to give

1:09:49

Paskie support to third party developers. And

1:09:52

then there's...

1:09:53

Oh, this is very developer-y, but

1:09:55

art is going to be upgraded to

1:09:58

R14, which gives support for Java 17, which is...

1:09:59

is really awesome, Kotlin,

1:10:02

jargon, jargon, jargon, but this is really cool stuff for

1:10:05

us to have. It kind of allows us to write things a little

1:10:07

bit better and with more modern technology

1:10:09

under hood, more paraphasing. And

1:10:11

something we talked about last week with Michelle is again

1:10:13

that forced letterboxing,

1:10:16

vertical boxing, on larger screens,

1:10:19

partial screen share for portion screen and ultra

1:10:22

HDR support. So, I mean, there

1:10:24

is so much in Android 14

1:10:27

that unfortunately, by design

1:10:30

is getting overshadowed by AI. But

1:10:32

I mean, there's a very, I'm getting

1:10:34

letterboxed. I'm getting vertical box. I'm getting vertical.

1:10:37

Pillar boxing is what it's called. Pillar boxing.

1:10:39

Pillar boxing. Yeah, we got any help

1:10:41

to that effect. Oh, we did. Thank you,

1:10:43

Matt.

1:10:43

I think it was Matt, but yeah. Thank

1:10:46

you, Matt, for letting us know that I'm being

1:10:48

pillar boxed right now. But yeah, all

1:10:50

of these things are just

1:10:52

honestly really great. Oh, I forgot to mention share sheets

1:10:55

because again, we've never ever talked about share

1:10:57

sheets on a show or complained about them at

1:10:59

length. But I mean, the AI

1:11:02

just a lot of great stuff. Anything

1:11:04

that I just ran down really quickly that sounds super

1:11:07

exciting to anyone that they wish was an Aikino?

1:11:09

Well,

1:11:09

yeah, Mike, I'm curious to kind

1:11:11

of hear your thoughts because I think IO, you

1:11:15

know, at its core, I don't have to

1:11:17

remind you of this, it's a developer conference.

1:11:20

And so many people are

1:11:22

there because of the Android,

1:11:24

the Android developments.

1:11:28

But again, this was kind of a light. I mean,

1:11:30

the keynote hardly talked about Android.

1:11:33

I mean, it was hardly even a topic. In two

1:11:35

hours, you barely heard anything

1:11:37

about Android. How do you feel about

1:11:39

how Android 14 is shaping up from

1:11:43

your perspective and

1:11:45

how Google is choosing to or

1:11:48

not to emphasize it?

1:11:50

What do you

1:11:51

think? So I think that Android

1:11:54

is fairly mature. So we're not ever

1:11:56

going to be looking at these drastic changes.

1:11:58

It's all going to be small. small things, but

1:12:01

the small things that I think are trending really positive

1:12:04

for Android and that are really empowering

1:12:06

for us as Android users are all the permissions

1:12:09

and the control over data that

1:12:11

there were some examples of that in that list, but

1:12:16

us being in control of how our information

1:12:18

is shared and understanding that is

1:12:20

really powerful and I feel like there's a lot of moves

1:12:23

for the OS to kind of control that.

1:12:27

That's really useful. Yeah, yeah.

1:12:31

Interesting. I mean we're gonna blink and Android 14

1:12:34

is gonna be releasing to our

1:12:36

Pixel devices, those of us who have a

1:12:38

Pixel, so you're not gonna have to wait too

1:12:40

long. I still have yet to actually

1:12:43

get on the beta train and this could be the longest

1:12:45

that I've gone in many years of not being

1:12:48

on the beta train, but life has been busy,

1:12:50

but I do feel the tug in that direction.

1:12:54

I feel like I need to kind of hop on the train. Tonight's

1:12:57

tonight.

1:12:57

Go ahead. Tonight. Do it

1:12:59

tonight. I like it. I like it. The

1:13:02

community wants you to, Jason, just to be clear. I

1:13:05

will do it and then I'll be like, it

1:13:07

looks the same, you know. It's

1:13:11

so under the hood, but anyways, so

1:13:13

yeah, I will probably do that soon. But

1:13:17

that's and then what is it? Is

1:13:19

it upside down cake? That's where we're at

1:13:21

now.

1:13:22

Yes, yeah. Yes. I said down cake. Okay,

1:13:25

good to know. I don't know if there's even an Easter egg

1:13:27

that we can look forward to. Oh, there's the upside down

1:13:30

cake. Does not look delicious to

1:13:32

me.

1:13:33

By the way, that is Alex and

1:13:35

Chris on the on the Android team

1:13:37

and they actually made this upside down

1:13:39

cake as part of like an Android developers video.

1:13:42

So if you on the, on the audio stream, you're going to see two

1:13:44

lovely gentlemen that work at Android holding up an upside down

1:13:46

cake. And if you actually are interested in watching a couple

1:13:48

of people on the Android team make an upside down cake, you

1:13:51

should go to Android developers YouTube channel. But yeah, they

1:13:53

actually made that so good for them.

1:13:55

I don't know that I've ever had an upside

1:13:57

down cake. It does not look delicious to me. It

1:14:00

kind of looks like a fruit

1:14:02

of fruit. Well, much

1:14:04

like the much like the Android 14 beta, don't

1:14:06

judge it until you try it. OK, all right. Fair

1:14:08

enough. Tonight. So this is what you're going to do tonight

1:14:11

after the show. You're going to start up. You're going to

1:14:13

you're going to install the beta on your phone and then

1:14:15

you can go have an upside down cake and then it'll be great.

1:14:17

So where do I get an upside down cake?

1:14:19

Victor, get me an upside down

1:14:22

cake. Oh, go to

1:14:24

one every week. I

1:14:26

was thinking like when it was Oreo.

1:14:28

General, I don't know. I

1:14:31

can never again, never again. Anyways,

1:14:35

so I saw something really cool

1:14:37

while we were at Google I.O. And

1:14:40

I don't know. I don't know if you remember

1:14:42

this couple of years ago when Google

1:14:45

part of Google's announcement at I.O.

1:14:48

was something called Project Starline

1:14:51

and Starline. And this was, you know, peak

1:14:55

of the pandemic. Right. So we're all at home.

1:14:57

The Google I.O. presentation was all,

1:14:59

you know, it was virtual. And I

1:15:01

think much of it was most of it was prerecorded,

1:15:04

that sort of thing. Maybe this this must have been the year

1:15:06

where they had the like outdoor kind of

1:15:08

small stage with like little

1:15:10

groupings of people sitting in random places

1:15:13

in the in the yard and stuff. It was very strange.

1:15:15

But anyways, they announced Project

1:15:17

Starline then, which was not necessarily

1:15:20

a product that you could get, but

1:15:22

a product that they were working with and testing.

1:15:25

And it was essentially take

1:15:26

take a video conference

1:15:29

like a meet or a Zoom video conference,

1:15:31

but put it in a room with

1:15:34

a screen that is very, very

1:15:36

large and three dimensional. So

1:15:38

it's a three dimensional view. You

1:15:40

had to be in a fixed position looking

1:15:42

at the screen, talking to someone else who

1:15:45

has that technology on the other side of the Internet,

1:15:47

wherever they're calling from. And these cameras

1:15:50

that would essentially turn you and

1:15:52

your presence into a three dimensional object.

1:15:56

And I say object

1:15:56

because it's not just straight video. There

1:15:59

is, you know,

1:15:59

course, um, a computer that's

1:16:02

processing those images and making it so

1:16:04

that when you're looking in the screen, you're actually making

1:16:06

eye contact with the other person it's

1:16:09

three dimensional it's human size. And

1:16:11

so the idea was

1:16:13

then, you know, to make something that's

1:16:16

not just you looking on your laptop and

1:16:18

chatting with someone like you're actually kind

1:16:20

of in the same room, even though you're not,

1:16:23

I was very intrigued by this. I

1:16:25

know a lot of people were like, it, it seemed

1:16:27

like one of those like big promises, can

1:16:29

it deliver sort of things? And they

1:16:32

had the current version, the current

1:16:34

iteration of this technology at Google

1:16:37

IO

1:16:38

and, uh, somehow I was

1:16:40

able to talk myself into sitting down

1:16:43

in it, even though they were totally booked up

1:16:45

for most of the day, but I kept being persistent.

1:16:47

It was like, I have to check this out. And

1:16:50

I ended up speaking with a couple of the founders of the technology

1:16:52

and, uh, was able to sit

1:16:55

in the room, uh, on one end. And

1:16:58

then in the other room was, you know,

1:17:02

I wish I remembered his name, but, uh, the

1:17:04

guy who I was talking with and, but,

1:17:07

but I should clarify

1:17:08

all of that video was being sent

1:17:11

out into the vast internet. It wasn't

1:17:13

like they were connected directly. So I was getting

1:17:15

the actual experience. It was all happening

1:17:18

in real time online and

1:17:20

you know, it was updated technology. So

1:17:22

the cameras, there were less of them. Um,

1:17:26

I had to say it was really impressive.

1:17:28

Like not that I think that everybody's

1:17:30

going to have these, you know, the specialized

1:17:33

video conferencing system in their

1:17:35

homes. Yeah, that's exactly what it looked like. Um,

1:17:38

but it was really, really

1:17:40

cool. Like

1:17:41

it, like the, the, um, the.

1:17:44

Kind of the impact of it was very convincing.

1:17:47

Um, I'm looking at a screen,

1:17:50

but it's three dimensional. The person that's

1:17:52

in front of me is exactly, you

1:17:54

know, the right scale. At one point

1:17:56

he held up an apple and

1:17:58

I feel like I could.

1:17:59

could have just reached out and grabbed the apple and the scale

1:18:02

to the apple on the screen to my

1:18:04

hand right in front of me was perfect. He

1:18:06

ended up putting his fist out and I gave him a fist

1:18:09

bump. Even though I couldn't feel

1:18:11

his fist,

1:18:12

like my brain was really messed up. I was

1:18:14

like, I,

1:18:15

I want to feel like I expect to feel

1:18:18

the fist. It looked that real. And the

1:18:20

eye contact of the experience

1:18:22

was really, it was really something

1:18:24

I like. I was really impressed by it.

1:18:27

Yeah. So I don't remember the name of the guy you did the demo

1:18:29

with, but we, you and I both had the opportunity to

1:18:31

talk with Andrew, who was, um, uh,

1:18:34

one of the, uh, one of the, uh, he was chasing

1:18:36

the G. Yeah. He chased. Okay. You

1:18:38

talked to Jason, but, uh, Andrew is the GM of project

1:18:40

star line. Uh, we talked to him both at the star line

1:18:42

demo. Uh, also, uh, uh, just

1:18:44

kind of chatting, uh, later on, uh,

1:18:47

totally awesome, really engaging,

1:18:49

really believer of the product, really excited.

1:18:52

Like when you get that, like that's what I did like about

1:18:54

the IO experiences, like talking to the people

1:18:56

who are building

1:18:57

this stuff and getting a sense of that excitement,

1:18:59

it just, uh, so it was like, it was somewhat

1:19:01

infectious and like star line looked really, really

1:19:03

impressive. Andrew Nartker is

1:19:06

the general manager of project star line. And

1:19:08

then Jason Lawrence is

1:19:10

the director and they both have been

1:19:12

working on it for like six years. So I'm

1:19:14

happy. We finally got their name on, on these things

1:19:16

cause I talked about it on TNW and I could not

1:19:19

remember the name and I didn't have the time to search it. Anyways,

1:19:21

it was a really interesting and,

1:19:23

um, and powerful experience. Like

1:19:26

I really enjoyed

1:19:27

it. And I'm just like, Hey man,

1:19:29

if video conferencing was that it's,

1:19:31

it's almost, it's not quite the same

1:19:33

as like stepping on a teleporter and taking

1:19:36

me to wherever that person is, but

1:19:38

it's a close, I mean, but it's as close as

1:19:40

I'm probably going to get. You know what I mean? It was

1:19:42

very, very real. Like, and the

1:19:44

eye contact thing, I'll tell you the eye contact

1:19:47

thing was a little disconcerting because I'm

1:19:49

looking at this screen the way we're used

1:19:51

to looking in screens, right? But I'm

1:19:54

making eye contact with this person that I'm talking

1:19:56

to and he's making eye contact with

1:19:58

me and like.

1:19:59

it felt kind of uncomfortable. It was

1:20:02

like, at some point I wanted to like look away

1:20:04

because like, I don't know.

1:20:07

It was, it was, it was real, but it, but

1:20:09

at the same time it was a TV. And,

1:20:12

um, anyways, it was a pretty

1:20:14

remarkable experience. So

1:20:16

if you ever get the opportunity,

1:20:19

if you happen to walk into a room and there's a project star

1:20:21

line, just try it. Just give it a

1:20:23

shot. Yeah. You're upside

1:20:26

down cake and show the person you're talking to. And

1:20:28

yeah, so there you go. But

1:20:30

they, they, they, they didn't just talk

1:20:33

about AI and apps

1:20:35

and stuff at, uh, we're seeing, we're

1:20:37

seeing the best for last coming

1:20:40

up. It's some hardware from

1:20:42

Google IO.

1:20:46

All right. So yes, we

1:20:48

knew all about the hardware in

1:20:50

advance, but we might as well

1:20:53

tell

1:21:01

you, you know, give it 10 minutes or so to kind

1:21:03

of talk about a little bit of what was announced

1:21:05

and when we'll start with you.

1:21:07

Yeah. And so the very

1:21:10

equally awaited pixel seven eight was

1:21:12

announced and our very own Florence

1:21:15

has written up a little review of it. So check that

1:21:17

out on gizmodo.com. But I

1:21:19

mean, um, I, I think

1:21:22

unabashedly, uh, flow really liked it.

1:21:24

I mean, the title of her article is

1:21:26

Google's pixel seven. A is just

1:21:28

as good as the regular pixel seven.

1:21:31

And yeah, I yes,

1:21:33

for a hundred dollars less. And, and flow makes a

1:21:36

lot of really good kind of, she did a lot

1:21:38

of really awesome testing and comparisons, but basically

1:21:40

there's not, there's not

1:21:42

a huge difference in a lot

1:21:44

of ways between the two phones. And

1:21:47

there's a lot that the seven A seems to do as

1:21:49

well, or even better than the seven.

1:21:52

So for example, flow the

1:21:54

camera, the camera action, the seven A seems just

1:21:56

a tiny, tiny, tiny bit better. Um, and

1:21:58

if you want to go like blow for

1:21:59

or blow spec for spec. The

1:22:02

Pixel 7a, for example, has a 64 megapixel

1:22:05

main camera compared to the 50 megapixel

1:22:07

camera of the Pixel 7, a 13 megapixel

1:22:11

ultra wide versus 12 megapixel ultra

1:22:13

wide. And even the front camera is 13

1:22:16

compared to the 10.8. I think Flo even said in her

1:22:18

article that while she found the Google

1:22:20

Pixel 7's front camera to be a little muddy,

1:22:23

does not feel the same way about the Pixel

1:22:25

7a. Not only are the camera units themselves

1:22:27

good, but the 7a is also

1:22:29

rocking

1:22:29

the tensor 2 chip, the

1:22:34

Tensor G2 chip, which is bringing you that super awesome

1:22:37

computational photography

1:22:39

and

1:22:45

taking your picture to the next level as Google does.

1:22:48

And just in general, Flo found this phone

1:22:50

to be as snappy and performant generally as

1:22:52

the Pixel 7. And

1:22:55

there's a lot of things that are just a little bit different,

1:22:57

like it's a little bit smaller. The storage

1:23:00

is a problem. The storage max that you get for

1:23:02

7a is 128 gigabytes. So

1:23:05

I mean, if you do like using that slightly

1:23:08

better camera, you'll have less space

1:23:10

to take

1:23:11

pictures and video with. But

1:23:13

apps can offload to the cloud, right?

1:23:16

And stuff like that. But I

1:23:18

mean, the 7 does go up to 256. I

1:23:21

mean, just generally, that

1:23:24

is kind of one of the big things. I think the battery

1:23:26

life is slightly, yeah, trade off. Thank you.

1:23:29

It's slightly less. And there

1:23:32

is an MmWave version that costs $50 more. But

1:23:35

I mean, just generally, for $500 for a mid range

1:23:38

phone, you're getting the same chip as

1:23:43

the flagships. You're getting roughly the same

1:23:45

size

1:23:46

screen. You're

1:23:48

getting the same amount of RAM, which is not great at 8, but not

1:23:50

bad at

1:23:50

all. Comparable

1:23:53

capacity, fast charging, wireless

1:23:55

charging, a better

1:23:58

camera.

1:23:59

IP67 versus IP68. And

1:24:03

interestingly enough, while the 7

1:24:06

has in-screen fingerprint in terms of biometrics,

1:24:08

the 7A has fingerprint and face

1:24:11

unlock. So it's

1:24:15

really kind of confusing, right? Because

1:24:17

obviously you have, you know, it

1:24:19

seems like it's starting to intrude on the space that

1:24:22

the Pixel 7 is holding in terms

1:24:24

of like, you know, the Pixel system. So

1:24:26

I know, Ron, you're my mid-range man.

1:24:29

Thoughts? Like, I

1:24:31

loved it. I got to hold it. I

1:24:34

got to get to use it. I got to play with

1:24:36

it. It does not feel like a mid-range phone.

1:24:39

Like the A series phones

1:24:41

have felt plasticky

1:24:43

in comparison to their non-A counterparts,

1:24:46

which should feel very medley and glassy. If

1:24:49

you handed me this Pixel 7A and told me it was the 7,

1:24:51

I would believe you. Like if you erase everything

1:24:53

I knew about everything we knew, we'd be like, here's the Pixel 7. It's

1:24:56

like, oh, wow, cool. Like it felt great. Like this

1:24:58

is a, like, I would confidently

1:25:00

recommend this to anybody who's looking for, you know,

1:25:02

for an Android phone in the mid-range and

1:25:05

be satisfied with what you get for 500 bucks. I

1:25:07

mean, I think it's, I think it's a home run personally. Yeah. Very,

1:25:10

very surface, like just playing with it in the demo pit

1:25:12

area. I'm hoping knock on wood, we

1:25:14

get a, you know, I get a review in it and can, you

1:25:17

know, kind of take it through spaces a little more. But at first

1:25:19

glance, I was super impressed. Jason, what did you think? Yeah.

1:25:22

I mean, and unsurprisingly,

1:25:24

so at this point, I feel like Google

1:25:27

has kind of a good formula for the A series

1:25:30

devices that they're releasing. It

1:25:32

is interesting that, you know, it's, it's a hundred

1:25:34

dollar difference from the Pixel It's

1:25:37

making me wonder like, why have the

1:25:39

A series be away

1:25:41

from the regular

1:25:42

device? Because the family

1:25:45

is, you know what I mean? Like Samsung, well,

1:25:48

I guess they kind of do that with their, um,

1:25:51

do they even do the fan edition anymore? I don't think

1:25:53

that they do. I can't remember the

1:25:55

FE, but anyways, my point being

1:25:58

Samsung, you know, releases like three

1:25:59

three galaxies at the same time, small,

1:26:02

medium and large, you know, the ultra

1:26:04

and whatever they call the rest of them. The

1:26:07

seven a is really the small, you know, it's 6.1 inches,

1:26:09

pixel seven is 6.3, the pixel seven

1:26:13

pro is 6.7 and they

1:26:15

kind of, it's, it's just like, you

1:26:17

know, a hundred dollar difference from the seven. So

1:26:20

I don't, I don't know. It's kind of like, why do

1:26:22

we even really need to wait for this? It's almost

1:26:24

like they should all release at the same time and

1:26:27

you just have three options depending on the size,

1:26:29

depending on the materials. And

1:26:32

those are really the only two things that are kind

1:26:34

of different. I mean aside, you know, the battery,

1:26:36

like you said, is a little bit larger actually.

1:26:38

The other thing that you don't get is battery share.

1:26:41

So the ability to share the battery of the device

1:26:43

with another device, you don't get that on the,

1:26:46

on the seven a, but that's not really a feature that I use.

1:26:49

So and you get the lower storage. So there

1:26:51

is that, that is pretty significant for a lot

1:26:53

of people.

1:26:54

But is that enough for the seven

1:26:56

a not to cannibalize like the

1:26:58

audience and the buyers for the seven?

1:27:01

And is that,

1:27:01

and you know what, that's probably the reason why they stagger

1:27:04

because if that did come out, then you know what I

1:27:06

mean? Maybe that's a part of the reason why they stagger. But

1:27:10

then are people just pissed that they, sorry,

1:27:12

are they just mad? My bad that they

1:27:14

spent a hundred bucks extra for a slightly

1:27:18

to somewhat better phone. Like is that, I don't know.

1:27:20

It's just

1:27:22

a little muddled, but I'm happy that

1:27:25

the a is awesome. Let's the positive

1:27:27

note. I'm happy that they is awesome.

1:27:28

You know, once again, the a, the a stands

1:27:30

for awesome. Apparently it does. It

1:27:32

really does. Sure does. Um,

1:27:35

also in the hardware parade was

1:27:37

the, uh, a year in

1:27:40

the making the pixel tablet finally was unveiled.

1:27:43

Um, I feel like we've seen it.

1:27:45

We knew it existed. We've been waiting

1:27:47

for it. The question was, uh, what

1:27:50

was it going to cost? Uh,

1:27:52

and, and would it come with the doc

1:27:54

or would they charge for the doc or whatever it would be?

1:27:57

And happy to report that.

1:27:59

It was, it's, it's, it's

1:28:02

priced at a dang good price. For some

1:28:04

reason the Google store is giving me UK

1:28:06

prices, which I don't really understand why. Um, so

1:28:09

I can't tell you the exact price. It was, yeah,

1:28:11

there we go. I got to fix it. It fixed it for some reason.

1:28:15

Yeah, I think it's 4.99. Yeah. Sorry. I

1:28:17

wanted to get prices too. Did I link to it? I

1:28:19

think the link in the doc goes to the, it goes to the, oh,

1:28:21

G B. Oh, I see. Yeah, yeah,

1:28:24

exactly. But I was super

1:28:26

excited to see the tablet get announced at $499. Starting

1:28:30

that's with 128 gig of memory. You

1:28:35

can get 256 gig of memory and get, and that knocks

1:28:37

the price out to 599 comes in three colors,

1:28:39

porcelain,

1:28:42

hazel and Rose. And

1:28:44

it comes with the doc. And

1:28:47

here, if you're watching the video version, you can see my dumb face

1:28:49

holding the tablet itself. But

1:28:53

the big question for me is that tablet

1:28:55

to doc,

1:28:56

a snap in moment, and

1:28:58

it has a satisfying

1:29:00

snap. It has a satisfying click. The

1:29:03

magnets click in and it grabs it and

1:29:05

holds it to that doc. Um, I saw some

1:29:07

articles and some coverage that were touting and saying,

1:29:09

you know, not only is this an upgrade to

1:29:11

the smart speaker kind of, you know, kind

1:29:13

of segment, but it's going to destroy it because

1:29:15

it now has created a new product, which is the

1:29:18

smart speaker with the tablet that detaches.

1:29:21

And it's like Google has innovated in it

1:29:23

by innovating in the space. They're going to destroy a

1:29:25

product line, which is kind of what Google does.

1:29:28

Um, but, uh, I will say

1:29:30

that I ordered this

1:29:32

and when can, uh, you can, you can attest

1:29:34

this cause you're right next to me. I ordered it as

1:29:37

they announced it. I was sold immediately

1:29:39

upon seeing it. So, um, the

1:29:41

one, my one little bit of criticism is they did show

1:29:44

off a very cool case that had

1:29:46

a little metal ring that also worked

1:29:48

as a stand. And you can see if you're

1:29:50

watching the video, you can see me playing with that, playing

1:29:52

with that. Um, uh,

1:29:55

and the, the metal ring snaps into the dock

1:29:57

really nicely and it all fits very organically. That

1:29:59

case.

1:29:59

like 80 bucks, which is like, that's a

1:30:02

little, I did not buy that, but, um,

1:30:04

that was a little expensive for

1:30:05

my, for, for people on the, who

1:30:07

are only on the audio stream, the look of joy, uh, kind

1:30:10

of like enjoyment on, on Ron's face,

1:30:13

uh, as he plays with the pixel tablet is very, I don't

1:30:15

know. It's

1:30:15

been a long time coming for, yeah, it's

1:30:18

awesome. I love a good tablet. Mike,

1:30:20

I want to know what you think about the tablet revolution that's coming.

1:30:22

Do you think that this is, is this it or,

1:30:25

or are we still have a long uphill battle

1:30:27

to go? Will this, will this device change the

1:30:29

fate of tablet apps?

1:30:31

I don't know about that. And I

1:30:33

am a tablet user. I actually already have two

1:30:35

perfectly good Android tablets in my house. I

1:30:37

have a S eight plus, which is a Samsung

1:30:40

S eight plus, which is I think a 12 inch tablet.

1:30:43

Um, and then a six, which is a nine

1:30:45

inch. I'm still going to buy this. Um,

1:30:48

but I actually think I'm going to buy it more as

1:30:50

a home, a replacement for my original

1:30:52

home seven inch home with

1:30:55

screen, which I love. It's like the best,

1:30:57

uh, photo album in the world. Um, you

1:31:00

know, great location for just like central

1:31:02

hub. So I

1:31:04

don't know if I'm going to purchase it as a tablet,

1:31:07

um, as much as a home device,

1:31:09

which is what you just said a moment ago. It's kind of like a

1:31:12

killer

1:31:13

feature for a different product product segment.

1:31:15

If you're buying it as a home device though,

1:31:18

you're spending a lot more on this than you were spending

1:31:20

on other home devices, like two to

1:31:22

three times as much. You're buying it as

1:31:25

a tablet that also has that functionality.

1:31:27

It makes a lot more sense. And the fact that that doc

1:31:29

is included at the $500 price point. I think

1:31:32

that for me, that was once I

1:31:34

heard that I was like, okay, this, this

1:31:36

has a, this is better

1:31:39

than I had hoped. Like I really thought

1:31:41

that doc was going to be like another hundred hundred

1:31:43

fifty dollars. And then it

1:31:46

would be self control. I buy

1:31:48

pretty much every device that comes

1:31:50

out. So certainly

1:31:53

not every device, but if it

1:31:55

has any interest, even a little bit,

1:31:57

I'm like, Oh yeah, I'll take my credit

1:31:59

card.

1:31:59

Okay. Well, that's, that's interesting.

1:32:02

Cause I want to know, did

1:32:04

you pull out your wallet for the pixel fold?

1:32:08

My wall, it's not thick

1:32:10

enough. Ironically, maybe

1:32:13

my company will get me one of those. Yeah.

1:32:16

I mean, $1,799 for the pixel fold. Yes. Announced

1:32:21

at Google IO. Uh, they

1:32:23

showed it off. Dave Burke,

1:32:26

as we said earlier in the show, did a bunch

1:32:28

of demos from the stage and they

1:32:30

went really well. And, uh,

1:32:32

you know, they, they had this device

1:32:35

in the press area as well. So I spent a bunch of time

1:32:38

pawn at the pixel fold. And

1:32:40

yes, it is, it has the nice

1:32:43

kind of thin quality to it.

1:32:45

When it's unfolded, it feels thinner than

1:32:47

any of the other fold devices

1:32:50

that I've used. So it feels

1:32:52

really nice. I didn't notice this, but

1:32:54

a bunch of other people noticed that it

1:32:56

doesn't fold 100% flat. Like

1:32:59

you kind of have to bend it a little

1:33:01

further than it feels like it's supposed to go in order

1:33:03

to get it to be completely

1:33:05

flat. And the only, I'd say the only way you'd

1:33:07

probably notice

1:33:08

that is if you had it laying on a table, either

1:33:10

that, or if you're very OCD, you

1:33:12

would notice it in the palm of your hand. I didn't notice

1:33:15

it, but I mean, materials were excellent.

1:33:17

It felt really nice in the

1:33:19

hand as far as a foldable

1:33:22

is concerned. This style of foldable

1:33:24

though, for me, I just,

1:33:26

I don't trust myself with a

1:33:28

device that folds like that. Because

1:33:31

I feel like I'm going to want to hold it like

1:33:33

I hold a phone, which is with one hand. And

1:33:36

you know, once it's unfolded, it's really

1:33:38

a two hand device. And if you drop that thing,

1:33:41

that is a pricey mistake to make

1:33:43

at $1,800. I mean, if you drop a tablet, you're

1:33:46

spending $500 to replace it. If you drop

1:33:48

this thing, that's $1,800. That's

1:33:51

a lot, but

1:33:53

they did announce that if you order it now, I

1:33:55

don't think it's forever, but if you order it now, you

1:33:57

get a pixel watch thrown in and a couple of other

1:33:59

free.

1:33:59

So they're trying to, you know,

1:34:02

goose the pixel watch numbers a little bit. But,

1:34:05

uh, yeah, it's a pretty, pretty sharp

1:34:08

looking, uh, device as far as foldables are concerned. I

1:34:10

really, I really thought it was nice. The bezels

1:34:12

on the inside didn't bug me. That didn't really

1:34:14

bug me at all. So,

1:34:17

yeah. What did you think, Ron? Cause you got to play

1:34:19

around with it a little bit, right? Yeah, I

1:34:21

got to, I definitely wanted to feel it and

1:34:24

feel the full, feel the full, then get a sense for

1:34:26

it. I really like

1:34:28

what they did from a software

1:34:31

standpoint, you know, like the, the, the camera

1:34:33

modes, you know, like being able to, you know, being

1:34:35

able to, you know, the, the, the, the Uber selfie, you

1:34:37

know, in the example. And, and,

1:34:40

and some of the viewing modes and stuff like that, when

1:34:42

you're like watching YouTube and all that sort of stuff. Like

1:34:45

those are the things that like Google's going to bring

1:34:47

to the, to the platform that Samsung

1:34:49

did, that, that differentiation that, you know, like, like

1:34:51

the Google approach to phones versus Samsung

1:34:54

or, or Oppo or anybody else like that. I like the way it

1:34:56

looks. I like the little Star Trek, you

1:34:58

know, camera bar to match the, the pixel

1:35:00

line. The folding felt good.

1:35:03

It felt sturdy. It felt snap. You

1:35:05

know, went in place. I didn't, I didn't care about the folding

1:35:07

flat type thing. I think that was fine. Bezel.

1:35:10

I got a dinging on the bezel bezel was noticeable

1:35:13

and distracting to me personally. But

1:35:17

it's the first phone, you know, they'll get there. Yeah.

1:35:20

It was definitely, it was a good first start,

1:35:22

I thought for sure. Yeah. Interesting

1:35:24

stuff. So that's hardware that

1:35:27

there were no like, you know, Google IO

1:35:29

gimmes or freebies or anything like that. But there

1:35:31

were a lot of hardware announcements. Obviously

1:35:34

we all, we all saw that coming, right? So

1:35:37

that's Google IO in a nutshell. We're

1:35:39

going to give you a little Amoose Boosh here before

1:35:42

we actually, that's at the

1:35:44

beginning of the meal, isn't it? We're going

1:35:46

to give you a little Android dessert. It's not

1:35:48

an upside down cake. It is J.R. Ray

1:35:50

feel he's going to round out the show before

1:35:53

we say goodbye with an Android event. I'm Jay with an Android intelligence

1:35:56

tip. J.R.

1:35:58

What you got? Good day. So,

1:36:00

IO is officially behind us, and while

1:36:02

we're waiting for all the newly announced

1:36:05

googly goodies to actually reach

1:36:08

us, we've got a fun little improvement you

1:36:10

can bring onto any Android

1:36:12

device this minute. It's

1:36:14

a really cool new feature that just showed up

1:36:16

in my favorite Android launcher,

1:36:19

Niagara. And if you haven't used Niagara

1:36:21

before, it's a teensy bit different

1:36:24

from your standard Android home screen

1:36:26

setup.

1:36:27

Niagara replaces your phone's

1:36:29

default home screen with a simple list

1:36:31

of your most commonly accessed apps.

1:36:35

And everything else lives within a scrolling alphabetical

1:36:37

menu that you pull up by swiping

1:36:39

along either side of the screen.

1:36:42

Now, Niagara's all about simplicity,

1:36:45

and eliminating all of the efficiency-harming

1:36:48

clutter most of our home screens are weighed down

1:36:50

by.

1:36:51

So to that end, the setup supports

1:36:53

just a single on-screen widget

1:36:56

at its uppermost edge. It's part of

1:36:58

its mission to encourage thoughtful configuration,

1:37:01

keep only the stuff you actually interact

1:37:04

with regularly, front and

1:37:06

center. And now, Niagara's

1:37:08

offering up a new way to put even more

1:37:10

pertinent info at your fingertips

1:37:13

without abandoning its minimalist focus-centric

1:37:16

philosophy. It's an option to stack

1:37:18

multiple widgets on top of each

1:37:20

other in that one single space

1:37:23

that gives you on-demand access

1:37:26

to more useful info while

1:37:28

still maintaining the launcher's trademark

1:37:30

clutter-free environment. So

1:37:32

in my setup right now, I've got a color-changing

1:37:35

MaterialUEnabled clock

1:37:37

in the default widget position.

1:37:40

Then, with a single swipe to the left

1:37:42

on that area, I can see a list of my

1:37:45

latest tasks from Todoist.

1:37:48

And with a swipe to the right, I can see

1:37:50

the native Android 13 battery

1:37:52

status widget. The new widget

1:37:54

stacking feature is included in the latest

1:37:57

stable version of Niagara Launcher.

1:37:59

So just go download it from the Play Store

1:38:02

if you don't already have it. Then once you've got

1:38:04

it up and running, press and hold your finger

1:38:06

onto whatever widget you've got

1:38:08

present at the top of the screen

1:38:10

and then select the Add Custom

1:38:12

Widget option. Repeat that same

1:38:14

process for each additional widget

1:38:17

you wanna add into the mix. And then all

1:38:19

you've gotta do is swipe your finger horizontally

1:38:22

on that widget area of the screen to

1:38:25

explore everything you've added.

1:38:27

Niagara Launcher is free to use,

1:38:30

though it does require a $10 a year or $30

1:38:34

lifetime upgrade to its pro version

1:38:37

if you wanna maintain access to the stackable

1:38:39

widgets and some other advanced options.

1:38:42

If all that stuff enhances your efficiency as

1:38:44

much as it does mine, well, it's pretty easy

1:38:46

to justify as money

1:38:48

well spent. And hey, if you

1:38:50

want even more advanced efficiency

1:38:52

in your life, come check out my Android

1:38:55

Shortcut Super Course. It's

1:38:57

a completely free week-long e-course

1:38:59

that'll show you all sorts of awesome tricks

1:39:02

for flying around your phone, typing

1:39:04

out text faster than ever, managing

1:39:06

your inbox more efficiently, and oh,

1:39:08

so much more. Just head over to

1:39:11

androidintel.net slash twit

1:39:14

and scroll to the bottom of the screen to get started.

1:39:16

That site again is androidintel.net

1:39:20

slash twit. That's

1:39:22

all for now. We'll pick up with even more

1:39:24

Android enhancing enchantment

1:39:27

next week. Back to you, gang.

1:39:30

Enchantment, enchanting.

1:39:33

Niagara Launcher, that's

1:39:35

a classic. JR Raphael,

1:39:38

also a classic at

1:39:40

androidintel.net slash twit. Thank

1:39:44

you, JR, for rounding out this episode.

1:39:46

It all came together in the

1:39:48

end. We had to kind of shuffle things

1:39:51

around, but we made room for it. And

1:39:53

yes, the supersized episode is not

1:39:56

over because you have an interview after

1:39:58

the credits. So definitely.

1:39:59

stick around for that. But

1:40:02

this was a heck of a lot of fun. And Mike, I'm really

1:40:04

happy that we could continue the tradition

1:40:07

of having you on around I.O. Thank you for

1:40:09

carving out some time. I've watched as your room has

1:40:11

gone from light to dark.

1:40:14

And you're muted. You're muted. It

1:40:18

is all good. It's my pleasure, of

1:40:20

course. Thank you so much for

1:40:22

having me. Yeah, of course. Mike Wolfson

1:40:24

dot com. Anything you want to leave people with?

1:40:28

Nothing important. I want to promote. But I

1:40:30

will say that if you are in the San Francisco

1:40:32

area and are a developer,

1:40:35

Droidcon SF is coming up January

1:40:38

9th and I would love to see you there. Ah, right

1:40:41

on. So SF dot droidcon dot

1:40:44

com is

1:40:46

the place to go to check out information for that.

1:40:48

Mike, always a pleasure. We'll have you back

1:40:50

soon. Appreciate you.

1:40:52

Thank you again. Thank you again. And Wyn,

1:40:55

appreciate you and appreciate

1:40:57

the fact that I got to hang out with you last

1:40:59

week. It was just so great. What do you

1:41:02

want to leave people with?

1:41:03

Yeah, also an Android developer.

1:41:05

You can find my talks on accompanying

1:41:08

code and videos at my website, randomly

1:41:10

typing dot com. And, you know, find

1:41:12

me on the interwebs at Queen Code Monkey. If you did

1:41:15

actually someone in Code Monkey,

1:41:17

probably, hopefully that's me. What's the story

1:41:19

with the with the purple?

1:41:22

Everybody always asks that one. OK, so my

1:41:25

local gym is awesome. They're called Endorphin and they have

1:41:27

what are called stink and drinks where you

1:41:30

advisable or inadvisably drink a little bit and you

1:41:32

work out a lot of it. And that

1:41:34

was I forgot what the stink and

1:41:37

drink. Yeah. They actually provided

1:41:39

us a Coors Light on the like

1:41:42

exercise bikes on the cycle. So in the cycling

1:41:44

section, there was actually like a Coors

1:41:47

Light or something. I say a light beer

1:41:49

and that was the workout equipment. We literally

1:41:51

were like doing balance things with it.

1:41:53

Anyway,

1:41:54

they had they had face paint. I put

1:41:56

it on. I liked it.

1:41:57

I've never heard of a gym.

1:42:00

Hey, come work out and drink. It'll be

1:42:02

great. I mean,

1:42:04

I guess that's why they had a light beer. Yeah.

1:42:07

I Definitely

1:42:10

did not drink the whole thing. Yeah, there was no

1:42:13

tequila rower station, right?

1:42:15

I mean there should have been The

1:42:18

actually but not for me. Yeah Now

1:42:21

we actually they made us do handstand holds there and to

1:42:23

so anyway, so yeah Yeah,

1:42:27

yeah, okay. Oh gosh See see

1:42:29

we could go many directions with this I'm gonna have to

1:42:31

tell my gym that they need to have a drink and stink

1:42:34

I'll tell him door for then and we'll have to credit you

1:42:36

if they use that Thank

1:42:39

you in always a pleasure Ron

1:42:42

Finally, what do you want? Yeah, finally,

1:42:44

just go follow me on Twitter

1:42:47

and on Instagram at Ron XO I'm

1:42:50

on blue sky. Thanks Jason on Ron

1:42:52

XO dot blue sky, whatever Mashed

1:42:55

it on all that all that nonsense But

1:42:58

yeah No, I posted all my photos

1:43:00

from Google IO on my Instagram account and

1:43:02

got more likes that I've gotten in ages Everybody

1:43:05

likes to see our smiling faces. So

1:43:07

that was great. Thanks everybody. Cool.

1:43:09

Thank you, Ron. Thank you win. Thank you, Mike Thank

1:43:12

you Victor here in studio

1:43:14

in-person Every once

1:43:16

in a while pulling over that microphone and speaking. Thank

1:43:19

you, man

1:43:19

Thanks, and thank you Burke who

1:43:22

was in here a little while ago helping out Thanks

1:43:24

to J. R. Ray feel Android Intel net

1:43:26

slash twit. You can find

1:43:29

me just here on to it I'll just leave you that

1:43:32

You know tech news weekly every Thursday twit

1:43:34

TV slash T&W. Also, don't forget we have

1:43:36

club twit It's really important to us because

1:43:38

it brings you closer to the creation

1:43:41

of our shows that integrates you into

1:43:43

keeping us Keeping

1:43:46

us going and we appreciate that ad

1:43:48

free subscription tier. So all of our shows with no

1:43:50

ads exclusive twit plus

1:43:52

podcast feed with tons of extra

1:43:54

content shows you can't find outside

1:43:57

of the club members only discord $7 a

1:43:59

month

1:43:59

month. Um, and you can

1:44:02

find all the information you need to know there at twit.tv

1:44:04

slash club twits. Uh, and I'll

1:44:06

just leave you with that. This show can

1:44:08

be found at twit.tv slash AAA. We

1:44:11

record every Tuesday evening. So just

1:44:13

go there and subscribe. Seriously. Subscribe

1:44:16

to the podcast and it will be delivered to you.

1:44:19

We're probably going to get that hero image replaced

1:44:21

here pretty soon with some of the images that we

1:44:23

shot last week.

1:44:24

As well together. Yeah, we were all together.

1:44:26

Yeah. And we need to replace that. Twit.tv

1:44:28

slash AAA. And you can see what the

1:44:31

new hero images once it hits the

1:44:33

web. Thank you everybody once

1:44:35

again for hanging out with us this evening.

1:44:37

And again, there's an interview

1:44:40

coming up after the credits. So check that out

1:44:42

too. And, uh, yeah, that's

1:44:44

it. We've reached the end of this super size episode

1:44:47

of all about Android. We'll see y'all next week. Bye

1:44:49

everybody.

1:45:07

Hello everyone. And welcome to Google

1:45:10

IO 2023. We are, well, we're outside

1:45:13

of the Shoreline amphitheater where Google

1:45:15

IO has happened this morning. I'm Jason

1:45:17

Howell,

1:45:18

went to a Dow Ron Richards

1:45:20

to be here for many years. This is my first

1:45:23

IO since like 2015 to be in person. So

1:45:25

I'm very excited. Yeah, that's awesome. That's a great

1:45:27

reason for everybody to meet up at the same place.

1:45:29

We had a lot of fun at all about Android last night and

1:45:32

we're having a lot of fun today because we're sitting down with Dave

1:45:34

Burke, VP of engineering or

1:45:36

Android and Samir Samat, VP

1:45:38

of product management. And

1:45:41

we

1:45:41

just always look forward to this opportunity.

1:45:43

So first and foremost, thank you for giving us

1:45:45

some, some of your time. And I have to say

1:45:47

like you finished the keynote like

1:45:50

an hour, hour and a half ago maybe.

1:45:52

And here you are sitting down at a table with

1:45:54

us. So I feel incredibly fortunate for the adrenaline

1:45:57

still pumping. We come down a little.

1:45:59

Yeah, I might like full. So how did you

1:46:02

feel about how did you feel the show from

1:46:04

the audience standpoint, it was great. But yeah, it's nice

1:46:06

to be back. It's nice to be back. I mean, that's that's

1:46:08

the summary. I like, you know, I feel like if there's

1:46:10

one thing we've learned the last couple of years is that you know,

1:46:12

these in person event in person events where

1:46:15

we can all be together again, or it's just so unique and

1:46:17

special. And so for me this

1:46:19

year, I was like, well, if I'm gonna do IO, I want to go all

1:46:21

in on live demos, because that's what people come for,

1:46:23

right? And they, you know, they kind of like get excited

1:46:25

with you when it works. And they maybe

1:46:28

commiserate when it doesn't and they feel the tension.

1:46:29

And so, so I don't I had

1:46:32

a lot of fun today. We just demoed a lot. Did you do

1:46:34

live demo? I missed. I'm just

1:46:36

kidding. Yeah, Dave had quite a few

1:46:38

live demos. I think he has the dubious

1:46:41

distinction of coming

1:46:43

back

1:46:44

in two IO segments to

1:46:46

do demos. I don't think we've ever had anyone

1:46:49

else do that. So but it all went well.

1:46:51

So demo gods. I was gonna say

1:46:53

it all worked out all right. I mean, I'm sure

1:46:55

you have it. You both have enough experience that you

1:46:58

go into those, no matter how prepared

1:47:00

you feel with trepidation, because

1:47:02

anything could possibly have a plan a plan C

1:47:05

plan, plan ABC, and then plan

1:47:07

D was my personal device, my back pocket. No,

1:47:10

that works. And planning is to run

1:47:12

screaming out of the room,

1:47:14

hands over your head. It all went really

1:47:16

well. It all worked out.

1:47:19

Well, I think if it

1:47:21

wasn't abundantly clear, the overarching

1:47:24

theme and you know, we talked about this last night

1:47:26

on the shows, like if we had to guess

1:47:28

what we're gonna see tomorrow, I mean, there was no guessing.

1:47:30

There was there was no question. It was going to be artificial

1:47:33

intelligence. And I think

1:47:35

there's so many announcements around

1:47:37

AI that maybe just from a broad

1:47:40

kind of view of the

1:47:43

of what you guys are bringing to the table here,

1:47:45

it really felt like the story

1:47:47

and correct me if I'm wrong, is that generative

1:47:50

AI is here.

1:47:50

And here's how

1:47:52

it plays into our products. Because when

1:47:54

I think of something like chat GPT, I

1:47:57

think of this destination I have to go to

1:47:59

in order to use it.

1:48:00

But Google is basically bringing that

1:48:03

AI into the products that we're all already

1:48:05

using. Would you say that's a fair assessment? That's

1:48:08

kind of like the story of AI with

1:48:10

Google, right? I think that's right. I think that's right. I

1:48:12

think like, you know, you just step back for a minute, you know, AI is something

1:48:14

that Google has invested a lot in over the years.

1:48:17

And you know, a friend of mine asked me recently, he was like,

1:48:19

why is it all suddenly like, why did it suddenly take off

1:48:21

in the last six months? And what happened? And,

1:48:24

you know, I think of it as like a trifecta. So

1:48:27

three things. So it's, you know, more

1:48:29

data. And we all have an intuition for that. Oh, yeah, more data,

1:48:31

more compute. We all have an intuition for that because like

1:48:33

we see our phones are faster and our laptops are faster.

1:48:36

But the real difference is the models got better.

1:48:38

That's what caused that sort of nonlinear shift.

1:48:42

And the model, the key, there were really two models

1:48:44

that really made the difference. One is transformer

1:48:46

models. And the other one is our

1:48:48

diffusion models. And actually, these are things that we worked

1:48:50

on at Google, you know, quite early on. And

1:48:52

actually, you know, Google researchers published the original

1:48:54

transformer paper. And we can learn it out at all

1:48:57

on if you want. But I won't go too deep on it. But

1:48:59

it's really kind of ushered in this whole

1:49:01

like generative AI sort

1:49:04

of era,

1:49:06

if you like. And we've been using it a

1:49:08

lot. So if you use search and you type

1:49:10

in a very esoteric query, you'll actually

1:49:12

sort of bypass all the caches and you'll

1:49:15

hit what we call BERT, which is it's

1:49:17

actually an encoder. It's like a transformer

1:49:19

encoder. And so we've been

1:49:21

deploying these technologies for quite a long time.

1:49:24

But I think, you know, deploying them in a generative

1:49:26

form is something that we haven't done as much

1:49:29

of, right? And so I think that's what you see at today's

1:49:32

keynote. And

1:49:32

I mean, there's a range of stuff from workspace

1:49:35

to the search generative experiences to the things

1:49:38

we showed on Android. But

1:49:40

yeah, it was a pretty, pretty exciting, you know. Yeah.

1:49:43

And I think it's exactly what you said, though. The opportunity I think

1:49:45

that we've always seen with this kind of technology

1:49:47

is the technology super cool. We want

1:49:50

to make it available for developers to use. They'll build amazing

1:49:52

things with it. But you

1:49:53

know, there's so many developers inside Google

1:49:56

who build products that we all depend on

1:49:58

every day. How can those products be available?

1:49:59

products get better with this technology. And

1:50:02

so I think what we wanted to do today was to showcase

1:50:05

not just the technology, but also

1:50:07

how the technology comes into each and every

1:50:09

one of those products that we all use and

1:50:11

what can it do for me? Because that is, I

1:50:13

think, a thing that people are curious about is,

1:50:16

okay, I've heard a lot about it, I've played with

1:50:18

these chatbot things and they're cool, I

1:50:20

can definitely see things I can do, but what's

1:50:23

the big deal? How's it really gonna change? I

1:50:26

love the Gmail example that Aparna

1:50:29

and Dave showed where you can write

1:50:31

a very short terse prompt

1:50:33

and it'll generate the email for you. I mean,

1:50:36

who doesn't think that they're gonna use that like every

1:50:38

day? And so

1:50:40

there's so many possibilities.

1:50:43

So I was really excited about how we showcased some of that.

1:50:45

And then I think with Android, we wanted

1:50:47

to kind of keep it a little lighter. And we

1:50:49

were working on a whole roadmap of stuff, some

1:50:52

we can probably hint at today, but I think the things that

1:50:54

we showed, and we wanted

1:50:56

to keep it a little bit lighter and just also

1:50:58

show that AI can be fun and

1:51:01

help you

1:51:02

just express yourself in creative

1:51:04

and interesting ways. So obviously

1:51:06

Dave showed what you can do with messages

1:51:08

by Google and just rewriting what you're about to

1:51:10

say to somebody and then also the generative wallpapers,

1:51:13

which are just neat. And I think really

1:51:15

relatable. So that was the goal of today.

1:51:17

So in my experience, so much of the conversation with AI has

1:51:19

been like, oh, the robots are coming, they're gonna take my

1:51:21

job and all this sort of stuff. And it's interesting

1:51:24

to see, and like we were talking about is the application

1:51:27

of how you take this technology and like the

1:51:29

wallpaper is a great example. It's like, I love San Francisco,

1:51:31

let me press

1:51:32

a button and get a illustrative way

1:51:34

of it. And I feel like that kind of helps bridge

1:51:36

the gap between this kind of like fear

1:51:39

of AI versus where it's no

1:51:41

different than the code, it's just

1:51:43

another way to do the stuff we do on our phones.

1:51:45

Yeah, I like that. Yeah, I like how you

1:51:48

said that because for us, so

1:51:50

I was nerding out a little bit earlier about diffusion models

1:51:52

and transform models and all that stuff. But actually the hard

1:51:54

part in my view is like how do you make

1:51:56

this approachable, right? And

1:51:59

relatable.

1:51:59

fit for purpose and also safe.

1:52:02

You got to be careful because like, so like think about the generative

1:52:04

AI wallpaper, the way it works, there's structure prompts, you

1:52:06

pick the different options you generate. We could have had an open

1:52:08

text box, but an open text box is not

1:52:10

approachable because it's like, here's a blank canvas, draw

1:52:13

something. You're like, ah, get creative. They're creative.

1:52:16

Create. And then also like there's

1:52:18

a safety aspect to it. Well, okay, what the hell kind of create

1:52:20

is not even going to be safe for work kind of thing. So we're

1:52:22

safe for my kids. And so, you

1:52:24

know, I think the solution we came up with, it seems kind

1:52:26

of obvious in retrospect, but like when you were

1:52:28

thinking of all the different types

1:52:29

of ways, internally, we call it Mad Libs,

1:52:32

by the way, where you have these like drop towns.

1:52:35

But, but I think the idea is like very simple.

1:52:37

And it's what Samir was getting at. It's like, it's applying

1:52:39

this stuff and figuring out how to apply it.

1:52:42

And actually, you know, honestly, this is like such

1:52:44

an exciting time, like this is the perfect thing. Being

1:52:47

on Android at this time in

1:52:49

computing history is just like perfect. Because we have all

1:52:51

this amazing set of like algorithms,

1:52:54

technologies, infrastructure, and then we

1:52:56

have all these surfaces and user problems. And

1:52:58

so, honestly, it's

1:52:59

just a lot of fun. Yeah, I know, like

1:53:02

the operative word a lot in like the zeitgeist is generative,

1:53:04

but watching the keynotes, both the kind

1:53:06

of main keynote and the developer keynote, I couldn't help

1:53:08

but and this is corporate speak, maybe in my

1:53:10

bad, but synthesis, like a lot of it is just

1:53:13

bringing together different tools. And I think I know

1:53:15

you talked about maybe,

1:53:16

like letting people like myself developers

1:53:18

kind of see the possibilities of it. And for I guess

1:53:21

most of our viewers aren't necessarily watching the developer

1:53:23

keynote, but something I was surprised about was that

1:53:25

even AI and like there's going to be

1:53:27

an Android Studio chat bot. And that feels

1:53:29

like the like epitome of a synthesis where,

1:53:31

hey, like here are these different resources,

1:53:34

here's things that usually look up. Yeah, let's let

1:53:36

the bot kind of like do things for you. And I thought that was really

1:53:38

cool.

1:53:38

Yeah, I am super excited

1:53:40

about Android Studio bot. We had it in the developer

1:53:42

keynote. And I mean, it really could have like held

1:53:45

its own in the in the consumer, even because

1:53:47

it's it's sort of a tops into your imagination. And

1:53:49

it makes programming, I think more accessible.

1:53:53

But if you think about what you're what is actually

1:53:55

doing, so you know, today or yesterday,

1:53:57

everything's changing so fast, but yesterday.

1:53:59

You would be in Android Studio and you'd be

1:54:02

coding along and you're like, ah, it's gonna

1:54:04

curse, I'm not allowed to curse,

1:54:07

dough. Like, how do I do this

1:54:09

thing? And so you would go to developer.android.com

1:54:11

or frankly, you'd go to Stack Overflow and

1:54:13

you would sort of go to these resources and you'd search and

1:54:16

you sort of look at the results, is that what I want? And maybe it

1:54:18

is, copy and paste, take some pieces, clean it up

1:54:20

or whatever. Now you just go to your point,

1:54:22

you go to StudioBot and you're like, hey, I got this,

1:54:25

there's an error in the compiler, in the Kotlin compiler,

1:54:27

click, what is this? And it will tell you what it is and how to fix

1:54:30

it. Or you could say, hey, I'm

1:54:32

stuck in this point, tell me how to do this. And it will just give

1:54:34

you the code and you press a button, it's syntax, beautifully

1:54:36

syntax highlighted and then you press a button

1:54:39

and it inserts the code for you. And so it's

1:54:41

just, it's sort of, it basically just makes

1:54:43

that whole process so much easier because the information is

1:54:45

encoded in the foundational model. And

1:54:48

so we've done a ton of work in like, in

1:54:50

fine tuning that and like figuring out how to make

1:54:52

it seamless. And then, you know, coming back to

1:54:54

this, you know,

1:54:55

this technology, like I was talking

1:54:57

about safety earlier, I

1:54:59

mean, you also have to think about like, okay, so I work in an enterprise,

1:55:02

I don't want that code, leave it, you know, that's my company's

1:55:04

code, I don't know what my, what the rights are for that

1:55:06

to leave the IDE. And so we had to think through all that

1:55:09

about not sharing the code. And so I

1:55:11

think what's interesting, like from a product management

1:55:14

point of view today is like how do you have to think through all

1:55:16

these different angles that are new,

1:55:18

but yeah, but yeah, StudioBot.

1:55:20

And I think StudioBot's kind of like that, as you said,

1:55:22

synthesis, it's in that genre of like, you

1:55:25

know, sort of a co-pilot

1:55:27

or whatever you want to call it and an assistant

1:55:29

or, you know, that's helping you and it works really

1:55:31

well, yeah.

1:55:34

I mean, there were a few examples given

1:55:36

today and we've certainly heard this about generative AI,

1:55:39

you know, earlier than today, about

1:55:41

the ability of these generative AI systems

1:55:44

to write code. And I guess

1:55:47

where my mind ends up, and we joked about this

1:55:49

a little bit, it's like, oh, I'm a developer

1:55:51

too, I can, you know, Bard and

1:55:53

plug in what I want and I get code, I'm a developer,

1:55:56

yay. But what

1:55:58

do you think about how tools?

1:55:59

like this actually transform

1:56:02

the role of a developer. Because I mean,

1:56:04

it does kind of change, it does kind of move the goalposts

1:56:06

as far as what it takes to be a developer.

1:56:09

Sure, you need to have developer chops. And I'm not

1:56:11

saying the system like this is gonna completely replace developers,

1:56:14

but it does alter the needs,

1:56:17

the requirement of that developer and

1:56:19

their knowledge to a certain degree. Maybe it propels

1:56:22

them to another level. But what are your thoughts

1:56:24

on that? Like, how does it impact that? I mean, one of the things

1:56:26

I've always found is that,

1:56:29

as an engineer, if you have time

1:56:32

to actually dig into

1:56:34

the whole business problem that you're solving,

1:56:37

you're just so much more motivated by what you're doing, as

1:56:40

opposed to like, here's a spec, like, let me go,

1:56:42

nobody really wants to do just that. And

1:56:45

so I think, but sometimes it's pretty hard because it's like,

1:56:47

you have a lot of minutia to get through, right? So

1:56:50

I think the first step is just like, can we use

1:56:52

this technology to get as little

1:56:54

focus on the minutia as possible? Not to the detrimental

1:56:57

products, but maybe we can increase the quality while

1:56:59

we do that too. And I

1:57:01

think it'll just let everybody pick

1:57:03

their head up just a little bit and just say like, okay,

1:57:05

like, how can I actually, like, what are we doing

1:57:08

here? Like, what are we solving? Is this

1:57:10

the best way to solve it? I think the

1:57:12

line at Google, honestly, historically

1:57:15

between like software engineer and product manager,

1:57:17

I don't think product manager's right. It's on a go to Google,

1:57:19

but I think software engineers play a lot of

1:57:21

the product manager role as well. And I think that's really

1:57:24

good. And I think over time, as we've just seen

1:57:26

in our company, there's just like a lot of code and

1:57:28

a lot of maintenance and a lot of things to do. And

1:57:30

they've started doing like more and more and more

1:57:32

software engineering and a little bit less of the product.

1:57:35

And so I'm excited about kind of rebalancing

1:57:37

that because I think just people are so much more motivated when

1:57:39

they get that time back. The other reality

1:57:41

is like, if you're a full-time coder, and

1:57:44

especially

1:57:44

if you're a rusty coder, that's me, like

1:57:46

there's just a lot of TDM. Like there's a lot of like, oh,

1:57:48

how do I do this boilerplate thing again? And

1:57:51

like, I just want to focus on the innovative idea of this product

1:57:53

that I'm building. And so this, you know, that's what this does.

1:57:58

You know, I think in the limit, you know, if you think about coders or magicians.

1:57:59

they can make things appear

1:58:02

and do things. We're bringing

1:58:04

in more magicians in the limit, but

1:58:06

really we're just automating the kind of TDM in

1:58:08

a lot of ways. Maybe that's not ambitious

1:58:11

enough to say, but that's why I think when you see

1:58:13

the developers in Shoreline today, when they saw

1:58:15

Studiobot, they're like, oh yes, hell

1:58:17

yes. They're not worried about their job. They're like, this

1:58:20

is going to make me so much more productive. It becomes how do you

1:58:22

use the tools to be creative? Exactly.

1:58:24

I can be more creative, more innovative,

1:58:26

more productive. It's interesting that

1:58:28

you say the TDM of this

1:58:29

because it's easy to hear that and be like, because

1:58:32

I will take a step back. I do

1:58:34

know, I was participating in the ClubTwit

1:58:37

discord for some of our members

1:58:39

during the live event and

1:58:42

some of the demonstrations, right? Like writing,

1:58:44

composing an email, that you write

1:58:47

in a couple of words and then you can expand it and everything

1:58:50

like that. Some of the people were joking, well, this

1:58:52

is all like baseline TDM

1:58:54

stuff. This is so not

1:58:57

creative. Why are we kind of writing

1:58:59

to this level?

1:58:59

Why doesn't it elevate? Why doesn't this or that?

1:59:02

And when I really thought about it, I was like, but wait a minute, like

1:59:05

a lot of what we actually do in our lives

1:59:07

is very tedious. The

1:59:09

forms that we create, the standardized emails

1:59:12

that we send

1:59:14

out to people to ask, hey, can you blah, blah, blah,

1:59:16

blah. A lot of that is incredibly tedious.

1:59:18

And these systems are just mirroring

1:59:21

what we're already doing. So if we don't actually

1:59:23

have to do it, it's interesting.

1:59:25

Because if you think about like the, now we're getting kind of more

1:59:27

technology, look at the evolution of technology. Like

1:59:29

people spent days doing the laundry and then

1:59:32

we got a washing machine. Exactly. I was

1:59:34

literally going to make that a knowledge. Well,

1:59:38

if you think about, you

1:59:39

know, for any of the product managers that are listening,

1:59:42

I mean, how many times have you, you know,

1:59:45

late at night created some presentation deck

1:59:47

and

1:59:47

then you're like, oh, I got to add notes,

1:59:49

you know, like, because it's someone else is going to

1:59:51

give it or you're going to give anyone. And like, I just love

1:59:53

the feature that Parna showed today, where it's like automatically

1:59:56

generate the notes from the slides. And it's like,

1:59:59

Now, now you're really starting

2:00:02

to talk to me about what can help me. And

2:00:04

so I think this is where AI can

2:00:06

really be made a little bit more understandable.

2:00:10

And frankly, I think

2:00:13

I've heard so many people today talk about just when

2:00:15

can I use it? Oh yeah, I was

2:00:17

there signing up for every wait list. I'm like,

2:00:19

sign up, sign up. Well,

2:00:21

not to end the beauty of it. When can I use

2:00:24

this? How can I use this? And how can

2:00:26

I use this is, well, just keep using the apps you're already

2:00:28

using. Eventually, this button's going to appear.

2:00:30

That's right. Or however that is presented. That's

2:00:32

right. I think it's that pervasive. And I think it has that broad

2:00:34

an impact on our products. So that's exactly what is going

2:00:37

to happen. It's just going to permeate and appear everywhere.

2:00:41

So switching gears, I mean, like staying

2:00:43

on the AI topic. But like Google I-O keynote,

2:00:46

you mentioned it's a consumer keynote,

2:00:48

right? Very focused on what it does. Historically,

2:00:50

a lot of the conversation that I has been around

2:00:52

the next version of Android. I don't

2:00:54

think you guys said Android 14 once in the keynote.

2:00:57

Was that by that design this year? Or how

2:00:59

does that? Yes and no. I

2:01:01

mean, Dave can elaborate a lot on

2:01:04

sort of how we've been evolving our thinking

2:01:06

around that. But just like taking a quick step

2:01:08

back, one of the realizations that we had is,

2:01:10

as the consumer keynote

2:01:13

broadens its audience, it used

2:01:15

to be long ago that I-O, like most

2:01:18

everybody that watched it was a developer actually.

2:01:20

And over time, what's happened is it's become

2:01:22

more than just Android. It's become more than just

2:01:24

developers. It's kind of like the whole Google

2:01:26

narrative. And what's interesting

2:01:29

actually, just as a quick sidebar, is that

2:01:32

a lot of the folks who are Android

2:01:35

developers would tell us, could

2:01:37

you all just take a quick step back? Before we get into

2:01:39

the APIs, could you just tell us, what are you trying

2:01:41

to do? Because developers

2:01:45

are users too. And so they're consumers

2:01:47

as well. So they're interested. So I think what we

2:01:49

wanted to do when

2:01:51

we talk about Android or really anything

2:01:53

is just talk about it from a consumer experience

2:01:56

standpoint. What will you get this

2:01:58

year? That's cool. leave

2:02:00

the vehicles,

2:02:01

uh, to the details, um, because I,

2:02:03

and people, a lot of folks, I mean, I think

2:02:05

you're the audience that listens to this. And I think

2:02:07

all of us here were like pretty interested

2:02:10

in those vehicles, you know, like what's in Android 14, what's

2:02:12

in this, what's in that? Um, and so that

2:02:14

is important and we should, we can go through that. But I think

2:02:16

for the general audience, what we realized

2:02:18

is that they just want to know what's going to be new

2:02:21

and that they're going to get it all. You know? And

2:02:23

so we took, we just sort of take it from that perspective

2:02:26

instead of chopping it up, we have more vehicles.

2:02:28

I mean, we've been, you probably noticed we had these like quarterly

2:02:30

spotlights and then we lean into,

2:02:32

we have vehicle, you know, we can update apps independently

2:02:34

of the OS we've been able

2:02:35

to do that for a long time. That was innovation

2:02:38

for some people recently. Um, uh,

2:02:41

and it still feels pretty innovative. I

2:02:43

know what it was like before all the

2:02:45

way it is now. It's miles different. I love

2:02:47

it. And then we have, and then we have like play system updates where

2:02:49

you can actually, like we can literally update the

2:02:51

virtual machine on an Android, Android

2:02:54

S, you know, version and increase performance

2:02:56

by just changing the VM with an updatable module. Like

2:02:58

we have like very sophisticated update mechanisms

2:03:01

in Android. We probably don't talk about it enough. I

2:03:03

mean, we like enroll these AB updates. We can

2:03:05

test them, roll them back. I mean, it's pretty amazing.

2:03:08

Um, and then we, and then we have like Google play services

2:03:10

and we have like jetpack libraries and so we have all

2:03:13

these, these vehicles. And so exactly what some

2:03:15

are said. And actually like maybe a little bit of inside

2:03:17

baseball, we used to like, when we did our product

2:03:19

planning, we used to sort of have like, we'd write an at a glance

2:03:21

or a summary docs. We'd be like Android 11 at a

2:03:23

glance. Now it's like Android 24

2:03:26

at a glance, right? Like we just think about the year and

2:03:28

then we talk about the different, you know, we talk about what we want to do

2:03:30

to Samira's point and why we want to deliver. And then we have the vehicles

2:03:33

as sort of a subtext. So, so there's multiple vehicles

2:03:35

as well as the core OS. When you said 24,

2:03:38

I immediately thought, wow, you guys are planning like 10

2:03:40

versions ahead. I

2:03:42

was blown away. But

2:03:44

it's funny you mentioned it because I, the one during

2:03:46

your demo, I think when you, you giggled when you mentioned

2:03:48

jetpack with the, with the Dragon

2:03:51

drop jetpack composed Dragon drop library.

2:03:53

I'm sorry. I was like really excited because I was

2:03:55

getting ready to like write my own system, but I

2:03:57

was great. It was like almost wins.

2:03:59

I was like, uh, like just in time

2:04:02

before you do that. No, no,

2:04:04

we play the compose a drinking game on the show

2:04:06

where every time I get to talk about compose, uh,

2:04:08

we'll win takes a drink, but then it's a

2:04:11

little more confusing now with magic composed, but that's okay.

2:04:12

Yeah. Yeah. We,

2:04:15

um, we may or may not have changed the name of that 24 hours ago. Um,

2:04:19

yes, it's, this is how we do product design.

2:04:22

Uh, yeah. Compose is mate. We should have like

2:04:24

composed another time. Compose is amazing. I love Kotlin

2:04:26

and compose. It's like so elegant. It's so nice.

2:04:29

I keep telling folks internally, you don't

2:04:31

go to Google like, this is so good. You realize it like,

2:04:33

it's really nice.

2:04:36

I bet it, but I mean, it connects to other things that we've talked about. As

2:04:38

you've mentioned, we were trying to take the minutia, the

2:04:40

boilerplate out of daily life, whether it's

2:04:42

a developer or a consumer. And so I feel

2:04:44

like the, it feels like the,

2:04:46

the, the roadmap, the game plan, the last

2:04:48

few years, at least from a developer's perspective has been, Hey,

2:04:51

do you actually want to have time to think about the product level

2:04:53

or the higher level engineering? So it all kind of fits.

2:04:55

Yep.

2:04:55

Yep. And to be clear,

2:04:58

when I say TDM, it's like, there's, it's sort of like, what's

2:05:00

not TDS is when you're innovating and using the code to

2:05:03

create this, here's my idea. And I bring my idea to

2:05:05

life. What's tedious is like, how do I close the

2:05:07

camera handler again? Or like the camera

2:05:09

object? Like where's the camera? I don't

2:05:11

know the camera one API. That's the thing where you're like, that's

2:05:13

the TDM part where you want just the bot to be like, Oh, here's the

2:05:15

idea. Oh yeah. Paste the code. Boom,

2:05:18

done.

2:05:18

Let's go. Oh, there are about like

2:05:20

dozens of developers sharing when I think one of the examples

2:05:23

that Jamal gave for the, sort of

2:05:25

for the, Android

2:05:27

Studio bot was remembering to put internet position,

2:05:30

internet permissions. So just

2:05:33

for the audience, everyone has

2:05:36

forgotten to give the app internet permission.

2:05:38

So that's, that's kind of one of those things that this

2:05:40

will

2:05:40

help us. I think I do it every time. Every single time. I

2:05:43

create something I forget. Like, why is it, oh yeah. Internet

2:05:45

permissions. Yeah. Yeah. I

2:05:48

can't believe I said camera handle that dates me is you can guess what I'm

2:05:50

saying. I started with handler. I mean, we can get switch operating

2:05:52

system. I started with anyway, not for that. Well,

2:05:55

so just to go back just real quick

2:05:57

on as far as Android 14 is concerned.

2:05:59

Yeah, considering everything that we've just talked about the

2:06:02

past five, 10 minutes, what would you say is

2:06:04

the story of Android 14? I

2:06:07

would say it's more, this

2:06:10

release is a little, sometimes we do very like user

2:06:13

facing sort of visual changes. Like

2:06:16

when we brought in material to you, that was a really big

2:06:18

one, right? I think this one

2:06:20

is more, what's the right word? I mean,

2:06:22

plumbing just sounds too negative, but

2:06:24

it's

2:06:25

under the hood sort of architecture,

2:06:29

making this up on the fly. Which is maybe not

2:06:31

like a sexy, like someone

2:06:34

sees that and they're like, oh, I've got to have Android 14

2:06:36

because it's got under the hood or architecture changes.

2:06:39

But it's important. It's faster. It's

2:06:42

got higher performance, right? And so we did a lot

2:06:44

of work on, we have two projects, the code

2:06:46

names internally, I don't mind sharing them, is

2:06:49

Snowbird and Falcon. The

2:06:51

way I think of how I remember Snow is the

2:06:53

S sound is close to Z, which is freeze. Okay,

2:06:55

I'm

2:06:55

losing myself here. And what we're

2:06:57

doing is we're trying to reduce the number

2:07:00

of the impact of broadcasts. And

2:07:02

so what we often find is, you know, Android

2:07:04

has this idea of a broadcast and an app can listen to it

2:07:06

and the app can wake up or come out of a cached state

2:07:08

and you can get like thundering herd issues

2:07:11

and you can get a lot of churning on IO. And

2:07:13

so what we're doing now is we're not actually taking, we're

2:07:16

actually allowing apps to basically stay cached for

2:07:18

a longer time. So we sort of queue up all

2:07:21

the broadcasts for an app and then wake the app up and go, here's

2:07:23

your broadcast rather than one at a time doing it and

2:07:25

it's coming in and out. And then

2:07:27

another one we're doing is like, if there's a broadcast

2:07:29

like battery change that happens all the time, we sort

2:07:31

of batch them up because it doesn't, you don't care that

2:07:34

which you just need to get caught up. And so you say

2:07:36

you collapse it to once. That's an example of, and

2:07:39

that makes a huge difference because it's sort of fundamental to

2:07:41

operating system. So that's that's that's

2:07:43

no word. And then Falcon is about really

2:07:45

refining our foreground services. They tend to get,

2:07:47

these are the notifications that pop up for an app and

2:07:50

you can't swipe the ones that won't swipe away. But

2:07:55

they're,

2:07:55

they're an important in a way cause it's, it's kind of

2:07:57

conferring to the user that this app is running. but

2:08:00

it's pesky. And so what we realized

2:08:02

was like, OK, well, first of all, there are some legitimate

2:08:05

cases where you want that. And then there's some cases that

2:08:07

are probably not ideal. And

2:08:10

so one case that was they were kind of, I would say,

2:08:12

I misused the wrong word, but they were used for that because

2:08:15

there was no other choice. It was like when an app needs to download

2:08:17

some data. Maybe it's a weather app and needs to grab

2:08:19

a snapshot of the data or it's a news app

2:08:21

or something. And so what we've done is we've basically

2:08:24

created a new set of foreground services. There's

2:08:26

not one for data. Instead, there's now a new API

2:08:28

job, basically, for data that

2:08:30

doesn't require a foreground service. And then we've

2:08:32

categorized the rest of them into like, is it a camera? Is it

2:08:34

a health app? Is it a location

2:08:37

or something like that?

2:08:38

And so what you should hopefully see over time is less

2:08:40

nuisance foreground services. And

2:08:44

that will also help our performance. So that's an example

2:08:46

of one of the things we're doing internally. HealthConnect

2:08:49

is another one. I don't know if you want to mention that. Yeah, HealthConnect

2:08:52

is another one. So HealthConnect, I

2:08:54

think for a long time, we have

2:08:57

been

2:08:58

focused on helping

2:09:00

people connect different health

2:09:02

services to each other. And that's

2:09:04

important because if you have a Peloton

2:09:06

device and you have Peloton at home

2:09:09

and you have a WearWatch and you have

2:09:11

an Oor Ring and a Withing scale, you're

2:09:13

like, oh, I got to, where do I go to dock

2:09:15

all this information? It's actually important to

2:09:18

realize

2:09:19

that you're building

2:09:21

your own

2:09:22

personal health record. There's

2:09:25

what happens at the doctor's office and then there's what happens

2:09:27

with all this information. And so in the ideal

2:09:29

future world, you'd be able to dock all

2:09:31

that together, right? And then share it

2:09:33

with only who you want. And I

2:09:36

think there's gonna be a lot of new insights that come from that

2:09:39

connectivity because your doctor today doesn't really have

2:09:41

full access to that information. They

2:09:44

don't track that. And you don't necessarily

2:09:46

have all your records either. So

2:09:48

being able to bring all that together. So HealthConnect

2:09:51

is a system that sits

2:09:53

on your Android phone to essentially

2:09:56

manage the data flows between those things and do it

2:09:58

in a safe and private way. way where

2:10:01

you provide access for data

2:10:03

to move from one app to another

2:10:05

and it's all in your control, it's all on

2:10:07

the device, it's encrypted. We

2:10:10

made that part of the platform. We've been working on

2:10:12

it independent of the platform and we decided

2:10:14

it reached APIs and so forth and

2:10:17

reached a maturity level where it's not part of the platform.

2:10:19

That's really exciting. Is that a mainline

2:10:21

module as well? Yes. Yeah,

2:10:24

exactly. Exactly. One of the quick mention

2:10:26

is we've been working on security

2:10:29

and privacy every Android release so

2:10:31

it would be a miss if I didn't mention

2:10:33

a couple of things. One of the things that I'm pretty excited

2:10:36

about that we're doing and it seems small

2:10:38

but I just wanted to mention it because

2:10:40

I think the impact will be outsized

2:10:43

for folks who really care about this which

2:10:45

is when you're giving an app permission,

2:10:47

for example, to location, we

2:10:49

of course added a while back

2:10:52

while in use permissions. But

2:10:54

now from the safety labels, the

2:10:56

data that you, when

2:10:59

you upload your app to play, you fill out a thing called

2:11:01

the safety label which is

2:11:03

what do you do with this data that your app

2:11:05

collects? Well, that information from the safety

2:11:08

label is now piped through to the

2:11:10

notification, to the prompt, the permission

2:11:12

prompt. When it asks you, do you

2:11:15

want to give this app

2:11:17

location permission? It's actually a

2:11:19

little snippet there from their safety label which will

2:11:21

say this app does share

2:11:24

data with third parties for location. Which

2:11:27

is like, yeah, that's information because when we

2:11:29

realized that we're like the safety label and then we looked

2:11:31

at the permission and a lot of times

2:11:33

you answer the permission and then you're like, wait, but the

2:11:35

safety labels over there, information

2:11:38

I could have used so we pull that in

2:11:40

now. Which I think is a really, it's a

2:11:42

small insight but I actually

2:11:44

think it's kind of a big deal.

2:11:46

A couple of other things that, so on privacy

2:11:48

that reminded me, we've improved

2:11:51

foreground services. We've intense, so one of

2:11:53

the issues we've seen, not so much in the US but

2:11:55

I mean Android is such a huge platform globally.

2:11:58

We've seen some abuse where apps.

2:11:59

sort of will launch a full screen intent

2:12:02

takeover of the phone. And so

2:12:04

now we've changed the policy so that really it can

2:12:06

only be like a clock, like an alarm clock or a dialer,

2:12:08

which are kind of what you would expect to

2:12:10

do it and nothing else. And so that's, it might seem

2:12:13

like a small thing, but it's important. Couple other

2:12:15

things to mention. This

2:12:17

one is not, this is not so obvious. That's why I want to mention

2:12:19

it is share sheet. So one of the challenges

2:12:21

we have a share sheet is, We went deeper.

2:12:23

Everybody. It

2:12:27

is not our, it's not our proudest moment. It's

2:12:29

okay.

2:12:29

It's okay. Even, even on, even some of the Google

2:12:32

apps have their own custom share sheets. And so what we've been doing

2:12:34

over time is improving the system one. And

2:12:37

so for example, we've improved some of the

2:12:39

custom targets and share sheet in

2:12:41

Android 14. But what's not really easy

2:12:43

to understand from blogs and stuff is that

2:12:45

we now have a plan for all

2:12:47

the Google apps to adopt the system share sheet. So

2:12:50

you'll see, because now the share sheets brought up to a level where they,

2:12:52

because they had valid reason, each team at Google

2:12:54

had a valid reason to have a custom thing, just like anyone

2:12:57

building an app. And so now I think we've got, I think

2:12:59

we finally got to the

2:12:59

system share sheet to a level that

2:13:02

is where the apps

2:13:04

need them. And the Google apps are pretty sophisticated

2:13:06

set of apps. So, so you'll see over

2:13:08

time, you'll see sort of, you'll

2:13:11

get, you'll see more coherence coming. So

2:13:14

that's one thing that's not clear. And then something

2:13:17

that's hard to pick up from blogs. The other one is ultra

2:13:19

HDR. That's our fancy name for, so,

2:13:22

so HDR is like high dynamic range. So

2:13:25

you've got like, brighter range of colors, effectively.

2:13:29

And one of the challenges is like, you

2:13:31

can't just introduce a new format, because then if I share

2:13:33

HDR photo to an app that doesn't understand

2:13:36

it, it gets de-saturated. So we

2:13:38

have a new approach where you have a JPEG

2:13:40

and has the standard dynamic range in the sort of main

2:13:43

body of the JPEG. And then we have a gain map that goes

2:13:45

alongside it in the, in the,

2:13:47

it's in the file format container. But

2:13:50

it means that if an app is savvy and understands HDR,

2:13:52

it will take the gain map and then expand the SDR

2:13:55

to the high dynamic range. It doesn't change the color of space, the color

2:13:57

of space is the same. And so that's a

2:13:59

way.

2:13:59

of us introducing HDR

2:14:02

in the platform, and why is that important? Well, I think

2:14:04

one of the things is we have these amazing cameras on many

2:14:07

devices like a Samsung or Pixel, but the

2:14:09

photos tend to be the best taken from the camera. And

2:14:11

when you then use the social networking app

2:14:13

to take the photo, it's not so good. This will help those

2:14:15

cases. So that's another new file format. No,

2:14:18

yeah, exactly. Just JPEG. Yeah.

2:14:21

Yeah. So actually a lot of what we do is like, how

2:14:23

do you introduce these things with a sort

2:14:26

of a back compatible approach that

2:14:28

doesn't make things worse as you're, as

2:14:30

everybody's catching up. And so that was one of those things

2:14:32

that took a little while to get right. Yeah, but

2:14:35

that's something I was just looking at my little notes here

2:14:37

to try and remember it's somebody, this is somebody other

2:14:40

things. I mean, system UI, we talked about the lock screen clock,

2:14:42

which was just something that a lot of people asked us for it to

2:14:44

customize. And it's like, I mean, it feels like that's been

2:14:46

a long time. It's a long time. I mean,

2:14:49

I mean, you know, within the realm of material,

2:14:51

you know, this idea

2:14:52

of a device that you

2:14:55

can make fully customizable. And then the clock

2:14:57

is kind of like, well, what about that thing? So we got, we

2:15:00

got around to that and no AI was harmed

2:15:02

in the process. And

2:15:05

then custom, custom shortcuts, which is what was,

2:15:07

was what we requested. A lot of people requested as

2:15:10

well. The other thing that we're doing, it's still

2:15:12

not GA or whatever

2:15:14

you call it, generally available. It's, it's predictive

2:15:16

back, but I'm really excited about it. It's this thing, you know, back

2:15:18

is such a core concept in Android. And

2:15:21

so the challenge is like, it has a history

2:15:22

and the challenges like, you know, you're

2:15:24

deep in N activities and you come back to the final

2:15:27

one and now it's going to fall back to home, but you don't know.

2:15:29

And so predictive app app shows you

2:15:31

a peak and animation peak. We were planning

2:15:33

to launch it in 14, but it just wasn't quite

2:15:36

the quality bar wasn't quite right. So we decided we

2:15:38

still, we still did a bunch of work. You still turn

2:15:40

it on and developer options, but it wasn't quite to the

2:15:43

bar we wanted. So hopefully

2:15:45

we'll land it for next cycle. But, but

2:15:48

yeah, but that's just some, it's a cool feature

2:15:50

that I like. You mentioned clocks. I

2:15:52

wanted to add one more clock thing,

2:15:55

but it's not for phones. It's for, for

2:15:57

watches. So we have this watch face

2:15:59

for.

2:15:59

that we introduced, which

2:16:02

is pretty cool. It's

2:16:04

declarative XML file, basically,

2:16:07

XML format that works with

2:16:09

this watch studio. So you can, with

2:16:12

this IDE interface, define a watch face. And

2:16:14

the reason that's cool is it's just

2:16:17

fun to design these watch faces. I

2:16:19

expect a lot of developers to do it, and

2:16:21

there'll be a lot more watch faces, which I think

2:16:23

for all of us who have wear watches is exciting.

2:16:26

But the other thing is what we found

2:16:28

is that people

2:16:29

were kind of building watch faces in

2:16:32

a lot of different ways. And you can

2:16:34

get yourself in trouble battery-wise pretty

2:16:37

quickly when you're designing a watch face. And

2:16:39

there are some safeties in the OS that obviously help

2:16:41

with that. But you can

2:16:43

go ahead and get yourself kind of run up against

2:16:46

the rocks. And so

2:16:48

what's nice about the watch face studio

2:16:51

is that all that code is generated for you. And

2:16:54

as the platform

2:16:57

adds new capabilities for

2:16:59

improving the efficiency of watch faces, including

2:17:02

more offload to the little processor

2:17:04

that's on the device, et cetera, the watch face

2:17:07

studio will just recompile, and you'll

2:17:09

get all the benefits of that stuff. So you don't have to

2:17:11

sort of recode your watch face over and over

2:17:13

again, which is pretty neat. Yeah,

2:17:15

I imagine developers love when

2:17:18

those kinds of things are made. Yes,

2:17:20

please. We have a thing for you. It's less reliant

2:17:22

upon your absolute attention

2:17:24

to it. But of course, it's always there. I

2:17:27

know that we're running a little

2:17:29

bit short on time, and we haven't even talked about

2:17:31

devices. So why don't we spend just a couple of minutes

2:17:33

just kind of focused on how

2:17:36

Android is kind of relating some

2:17:40

of the changes to Android over the last couple of years,

2:17:42

right? We're talking about every 12L. Yeah,

2:17:44

I feel like

2:17:45

today was the culmination of what a couple of

2:17:47

the 12L being like, we want to embrace larger

2:17:50

screen devices. And we're like, hmm, wonder why. Then we find

2:17:52

out the tablet's coming. And now we unfold.

2:17:54

Yeah, I mean, actually, when you were asking,

2:17:56

what's Android 14 about, that's kind

2:17:58

of where I was going to go.

2:17:59

which is, you know, there is a ton of, as Dave said,

2:18:02

you know, behind the scenes, you know, work that's been

2:18:04

done. But

2:18:04

to be honest, it's a culmination

2:18:07

of multiple releases, you know, 12L, which

2:18:09

is kind of a bit of an off cycle release that

2:18:11

we did. But because we really wanted to start getting

2:18:14

this in the hands of device

2:18:16

makers and developers, because we, when

2:18:18

we saw foldables initially,

2:18:20

you know, start to hit the market

2:18:22

and take off, we were just incredibly fascinated

2:18:25

by this form factor. And, you know, Android

2:18:27

usually leads in these form factors.

2:18:29

But I think one thing that we could do a

2:18:31

better job on is, you know, making

2:18:35

sure that we hit that polished

2:18:37

level more quickly. And so

2:18:39

with foldables, we just saw an opportunity

2:18:41

to say, Hey, wait a minute, large

2:18:44

screens are a thing, you know,

2:18:46

because of this new innovation,

2:18:49

it's an opportunity to reinvigorate that whole

2:18:51

category of devices all the way from tablets

2:18:54

to foldables and kind of like the different

2:18:56

size factors in between. And,

2:18:59

and so what you see now with with this release

2:19:02

is and you saw Dave demo a whole bunch of this,

2:19:04

like it's very powerful. And,

2:19:06

and the developer capabilities are there.

2:19:09

So it's, it's easy, it's not complicated

2:19:12

to build apps that make take advantage of drag

2:19:14

and drop and so forth and so on, you know, we have set

2:19:17

ways of doing it. So it's not a choose your own adventure,

2:19:20

you can always do that. But at the

2:19:22

same time, it's nice to have defined clear paths

2:19:24

for golden paths for people to walk through

2:19:27

to make this happen. And, and

2:19:29

I think that what you're, we're

2:19:32

now on, you know, nth

2:19:35

generation of foldables, right? And what I

2:19:37

really want to give a lot of credit to the Samsung team

2:19:39

here, you know, because I think they pioneered

2:19:41

a lot of this. Yeah, they paved the way they

2:19:43

super paved the way and we not only do they do

2:19:45

that with their hardware, but we've worked with them really

2:19:48

closely. Like this is the first time that a pixel

2:19:51

is launching a fold, but you know, Samsung's had

2:19:53

a number of foldables up until now

2:19:56

and we've worked with them on every single one of them.

2:19:58

And a lot of the features that you saw today. actually

2:20:00

a lot of the innovation was done in tandem

2:20:02

with them. And a lot of those components are

2:20:05

actually shipping on their current devices. And

2:20:07

so what's really nice is we've been able to kind

2:20:09

of build this category together.

2:20:12

And now people

2:20:14

are jumping in and making

2:20:16

it even more interesting category of,

2:20:19

you have OPPO and Xiaomi

2:20:21

and

2:20:22

Google and a whole bunch of other companies jumping

2:20:24

in. So we're super excited about it. And I think that does come

2:20:26

all together in 14. Yeah, for sure. And

2:20:29

then with the Pixel tablet, of course, like we've

2:20:31

talked a lot about tablets, because we feel as if the

2:20:34

tablet space in Android is very murky

2:20:36

because iPad dominates

2:20:39

in terms of people think of tablets. I

2:20:41

have a proud owner of a Lenovo tablet. I use

2:20:43

it all the time, but they tend to fall under a media device

2:20:45

or use it as instead of a Kindle. But

2:20:48

now with the Pixel tablet, it seems like you guys kind of

2:20:50

returning back to that tablet spec, but

2:20:52

over 50 Google apps being

2:20:55

developed to optimize for tech, take advantage of it. Other

2:20:57

apps like

2:20:58

Calm and Disney Plus optimizing it. It's

2:21:00

a kind of a chicken and the egg kind of thing, because

2:21:02

in order to get developers to put the resources

2:21:05

into adapting their apps, there need to be users,

2:21:07

but in order to be, you know, so how do you guys approach

2:21:09

that from a developer relations standpoint? I mean, it's always

2:21:11

been that way for even when we started Android

2:21:13

at the beginning for mobile. And I, you know, I was really

2:21:15

close to the start, you know, helping start Android TV. It's

2:21:18

always, it's chicken and eggs precisely what it is. And

2:21:20

you just sort of got an ooch and scooch and making a little bit

2:21:22

of progress. And then the apps come in, then they make the platform and you

2:21:24

just have to iterate. I think,

2:21:26

you know, I think it's important

2:21:28

for Google to lead by example here. And that

2:21:30

means that you mentioned that, you know, over 50 apps,

2:21:32

so we need to show that we are committed. I think

2:21:35

it's important for Google to have hardware in the tablet

2:21:37

space and show we're serious about

2:21:39

it. And that, you know, that will

2:21:41

help push forward. And then, you know, more developers

2:21:43

come, they see us do it. So you get more developers on

2:21:46

and then the products get better, and then more developers

2:21:48

come on. And that's how you do it. One of the things you'll

2:21:50

see, we had a blog post today about some of the third party

2:21:52

apps, you know, some of the big names, and you mentioned

2:21:54

Peloton, I think that was in the article. And

2:21:57

you'll probably, you know,

2:21:58

we're planning every, I don't know what the exact,

2:21:59

Zach Hadens has got to be, but we'll have a series of blog posts

2:22:02

because we're working with so many third party developers and we'll show

2:22:04

the progress. So this isn't all sort of a one

2:22:06

and done for Google I.O. This is a path. And

2:22:09

the momentum is really good. And

2:22:12

then the Pixel tablet, I think, is

2:22:14

a really nice product. I think you saw the

2:22:17

foldable work that I demo

2:22:19

today, and all that translates to tablet

2:22:21

two, which is really nice.

2:22:23

And then I think one of the things I really

2:22:25

like about the Pixel tablet, just to talk about that product for a

2:22:27

second, it's very simple.

2:22:29

It has a dock, so it's always charged. So

2:22:32

I've had this thing for a year in my

2:22:34

home, and it's awesome. It's like, oh, my tablet's ready.

2:22:37

Well, it's a satisfying click. Yeah,

2:22:39

it's very satisfying. It

2:22:41

just pops right on there. But I did find that

2:22:43

actually a pretty unexpected

2:22:46

compelling moment from the keynote, where the presenter,

2:22:50

I'm blanking on her name, but she had the tablet in the

2:22:52

drawer. Rose, that's right. And

2:22:54

the drawer opens, the tablet's in the drawer, and it's dead.

2:22:57

And I've been there a million times. It's like, just

2:22:59

get the tablet

2:22:59

a home. Give it a home where it can

2:23:02

get the juice that it needs so that it could

2:23:04

be your companion. Yeah, it's

2:23:07

often the simplest idea, the most powerful. And

2:23:09

then it has this nice speaker. When you dock it, the audio

2:23:12

instantly translates. And then you sort of extrapolate

2:23:15

this, and you're like, OK, so it's on the dock, and

2:23:17

it's a nice angle. OK, so let's make it proactive.

2:23:19

Let's have information that's useful for me. Or I

2:23:21

turn off the light, and it will dim its screen, and it will

2:23:23

turn into a clock mode. And now it can be a photo frame, and

2:23:26

now it can be basically

2:23:28

a Google Assistant. So it's basically a home hub now,

2:23:29

and now it can be a controller. And so it's sort

2:23:32

of just all things. And hey, we're Google. We've

2:23:34

done a lot of art displays. So

2:23:36

we can do that over here, too. Exactly, yeah. That

2:23:39

was kind of the question that popped up in my

2:23:41

head. Is this a tablet first, or

2:23:44

is it a smart display first with tablets? You don't want

2:23:46

to be in which direction? I mean, all of the above. But I

2:23:48

mean, as an Android person, I'm most excited

2:23:50

about it's a tablet first. And

2:23:53

I think if you talk to the team in Google,

2:23:55

they'll tell you it's a tablet first, too. Because

2:23:57

it has to be. Actually, it's one of the

2:23:59

best.

2:23:59

the premises at the beginning. It's like it has to be a great

2:24:02

tablet first. Yes. And so that was a focus.

2:24:05

Yeah. Yeah. Interesting. Well,

2:24:07

I know we got a, we actually am looking at the time.

2:24:09

We got to round things out. Do you like,

2:24:11

have we missed anything that we need to throw out there before

2:24:13

suddenly we don't have our opportunity anymore? I got to say satisfying clicks.

2:24:15

So I'm like, check that

2:24:18

part. I did get a chuckle from the, we

2:24:20

were in the, in the product

2:24:22

demo pit for the press and the

2:24:24

woman from Google was over my shoulder and I did it. And she's like,

2:24:27

it's nice, isn't it? Yes, it really

2:24:29

is. I actually knew

2:24:29

exactly what I was doing. Well

2:24:33

we are delighted every time we get a chance

2:24:35

to come to Google IO, even more delighted

2:24:37

when we get the chance to speak with you both. Just

2:24:40

really want to thank you for taking time out

2:24:42

of what is always a crazy day, Google IO

2:24:44

keynote day to give us like 45

2:24:47

minutes of your time is we're just

2:24:49

over the moon. So thank you very much for carving out

2:24:51

the time. Thank you. And we, we,

2:24:53

we love what you're doing and the fans

2:24:55

of, of, of, of this podcast

2:24:57

are, are really important to us. So thank you so

2:24:59

much. We're fans too. So thank you. Right

2:25:02

on. Excellent. Well, Dave Burke, Samir Samat,

2:25:05

we will talk to you next time hopefully. And

2:25:08

yeah, everybody we have

2:25:10

our live coverage. If you missed it of

2:25:12

the keynote, if you want to hear what Leo

2:25:14

and the crew were saying while everything was

2:25:16

happening, you can just check it out. It's with TV

2:25:19

slash news where you find this podcast

2:25:22

as well. So thank you so much here

2:25:25

from the Google campus. We'll see you next

2:25:27

time. Bye everybody. Thank you.

Unlock more with Podchaser Pro

  • Audience Insights
  • Contact Information
  • Demographics
  • Charts
  • Sponsor History
  • and More!
Pro Features