6:00

We have lived through one

6:02

major platform transition already because we started

6:04

on web, right? Not on mobile. Mobile

6:07

phones and smartphones kind of got started

6:09

around the same time as Facebook

6:12

and kind of early social media was getting

6:14

started. So it didn't really get to play

6:16

any role in that platform transition, but going

6:19

through it, where we weren't born on mobile,

6:21

we kind of had this

6:23

awareness that, okay, web was a thing,

6:26

mobile is a thing, it is different. There

6:28

are strengths and weaknesses of it. There's

6:30

this continuum of computing where now you

6:33

have a mobile device that you can take with you

6:35

all the time and that's amazing, but

6:38

it's small, kind of

6:40

pulls you away from other interactions. Those things are not great.

6:43

But there was sort of this recognition that just

6:46

like there was the transition from computers to

6:48

mobile, mobile was not gonna be

6:50

the end of the line. So as soon as we

6:52

started becoming like a

6:55

more, I don't know, I

6:57

guess it's a stable company. Like once we found our

6:59

footing on mobile and we weren't like clearly gonna

7:01

go out of business or something like that, I was like, okay, let's

7:04

start planting some

7:06

seeds for what we think could be

7:09

the future, right? It's like mobile is already kind of

7:11

getting defined, you know, by 2012, 2014-ish, it was generally

7:13

too late to

7:17

really shape that platform in a meaningful way. I

7:19

mean, we had some experiments, I mean, they didn't

7:21

succeed or go anywhere. So pretty quickly, I was

7:23

like, okay, we should focus on the future because

7:25

just like there's a shift from desktop to mobile,

7:28

new things are gonna be possible in the future. So

7:30

what is that? I think the kind

7:33

of simplest version of it is

7:36

basically what you started seeing with Orion, right?

7:38

It's like the vision is a normal pair

7:40

of glasses that

7:42

can do two really fundamental things. One

7:45

is put holograms in the

7:47

world to deliver this realistic sense of presence,

7:49

like you were there with another person or

7:52

in another place, or maybe you're

7:54

physically with a person, just like we did, you can

7:56

pull up a virtual Pong game or, you know,

7:58

whatever. You can work on things. together,

8:00

you can sit at a coffee shop, pull up your

8:02

whole workstation of different monitors, you know, you can be

8:05

on a flight or in the backseat of

8:07

a car and pull up a full screen

8:09

movie theater and like, okay, all these things, great

8:11

computing, full sense of presence, like you're there

8:13

with people no matter where they are. The

8:15

thing too is

8:18

it's the ideal device

8:21

for AI. And the

8:23

reason for that is because glasses are

8:25

sort of uniquely positioned for

8:28

you to be able to let them see what

8:30

you see and hear what you hear and

8:32

give you very subtle feedback back to you

8:34

where they can speak in your ear or

8:36

they can have silent input that kind of

8:38

shows up on the glasses that other people

8:40

can't see and doesn't take you away from

8:42

the world around you. And I think that

8:44

that is all going to be really profound.

8:47

Now, when we got started with this, I

8:49

had thought that kind of the hologram part

8:51

of this was going to be possible before

8:53

AI. So it's sort of an

8:55

interesting twist of faith that the AI part

8:57

is actually possible before the holograms are really

8:59

able to be mass produced at kind of

9:01

an affordable price. But that was sort

9:03

of the vision. I think that it's pretty easy to wrap your head around.

9:06

And there's already a billion to 2 billion people

9:08

who wear glasses on a daily basis. Just

9:10

like everyone who didn't have smartphones

9:12

were kind of the first people to upgrade two

9:14

smartphones. I think everyone who has glasses is pretty

9:17

quickly going to upgrade to smart glasses over the

9:19

next decade. And then I think it's going to

9:21

start being really valuable. And a lot of other

9:23

people who aren't wearing glasses today are going to

9:25

end up wearing them too. That's kind of the

9:27

simple version. And then I think

9:29

as we've developed this out, there

9:32

are all these sort of

9:34

more nuanced directions that have emerged too. So

9:36

we've started, while that was kind of the

9:38

full version of what we wanted to build,

9:41

there are all these things that we

9:44

said, okay, well, maybe it's really hard to

9:46

build normal looking glasses that can do holograms

9:48

at an affordable price point. So what parts of that

9:50

can we take on? And that's where we did the

9:53

partnership with SLR Luxottica. So it's like, okay, before you

9:55

have a display, you can get

9:57

normal looking glasses that can have a camera, they can

9:59

have a microphone. great audio, can

10:01

capture content, you can stream video at this

10:03

point, but the most important feature at this

10:05

point is the ability to access

10:07

meta AI and just kind of have kind

10:09

of a full AI there, multimodal, because it

10:11

has camera. And I mean, that product is starting

10:14

at $300. And initially

10:16

I kind of thought, hey, this is sort

10:18

of on the technology path to building full

10:20

holographic glasses. At this

10:22

point, I actually just think both are gonna exist long-term. Right,

10:24

I think that there are gonna be people who

10:27

want the full holographic glasses. And I think that there are

10:29

gonna be people who prefer kind

10:31

of the superior form factor or lower

10:33

price of a device where they

10:35

are primarily optimizing for getting AI. I

10:38

also think there's gonna be a range of things in between,

10:40

right? So there's the full kind of field of view that

10:42

you just saw, right? 70 degrees, really wide

10:44

field of view for glasses. But I think

10:46

that there are other products in between that too. There's

10:49

like a heads up display version. For

10:51

that, you probably just need 20, 30 degrees. You

10:55

can't do kind of the full

10:58

kind of world holograms where

11:00

you're interacting with things. Like you're not gonna play

11:02

ping pong in 30 degree field of view, but

11:05

you can communicate with AI. You

11:07

can text your friends. You can get directions.

11:09

You can see the content that

11:11

you're capturing. So I think that there's

11:13

a lot there that's gonna be compelling. And that's, I think, gonna

11:16

be at each step

11:18

along this continuum from display list to small

11:20

display to kind of full holographic, you're

11:22

packing more technology in. So each step up is gonna

11:25

be a little more expensive,

11:28

is gonna have a little more constraints on the form

11:30

factor, even though I think we'll get them all to

11:32

be attractive. You'll be able to do the kind of

11:34

simpler ones and much smaller form factors permanently. And

11:36

then of course, there's the mixed reality headsets, which kind

11:39

of took a different direction, which is going towards the

11:41

same vision. But on

11:43

that, we said, okay, well, we're not gonna try to

11:45

fit into a glasses form factor for that one. We're

11:47

going to say, okay, we're gonna really go for

11:50

all the compute that we want. And we're

11:52

gonna say, okay, this is gonna be more of like a headset

11:54

or goggles form factor. And my guess is that that's

11:56

gonna be a thing long-term too, because there are a bunch of

11:58

uses where people want the full. of

30:00

like R and D and talking about it, but like these,

30:03

the Ray-Bans are kind of a

30:05

signifier of that. And I'm wondering if you agree with that. I

30:07

agree. I mean, I still think it's early. I

30:09

think you really want to be able to not

30:12

just ask the AI

30:15

questions, but ask it

30:17

to do things. Yeah. And

30:19

know that it's going to reliably

30:21

go do it. And we're

30:24

starting with simple things, right? So voice control

30:26

of your glasses, although you can do that

30:28

on phones too, things like reminders, although you

30:30

can generally do that on phones too. But

30:34

I think as the model capabilities grow over the

30:36

next couple of generations and you get more of

30:39

what people call these agentic

30:41

capabilities, I think it's going

30:43

to start to get pretty exciting.

30:46

For what it's worth, I also think that all the AI

30:48

work is going to make phones a lot more exciting. It's

30:50

the most exciting thing I think that has

30:52

happened to our family of apps

30:54

roadmap in a long time, is all of the different

30:57

AI things that we're building. So if I were at

30:59

any of the other companies, I think that if I

31:01

were trying to design what the next few versions of

31:03

iPhone or Google's phone should be, I

31:05

mean, I think that there's a long and

31:07

interesting roadmap of things that they

31:10

can go do with AI that as

31:12

an app developer, we can't. So I think that that's

31:14

like a pretty exciting and interesting thing for them to

31:16

go do, which I assume that they will. On the

31:18

AI social media piece, one of the wilder

31:20

things that your team told me you guys are going to

31:22

start doing is showing people

31:24

AI generated imagery personalized to

31:26

them in feed. I think it's

31:29

starting as an experiment, but like

31:31

if you're a photographer, you would

31:33

see MetAI generating content that's maybe

31:35

personalized for you alongside content from

31:37

the people you follow. And

31:40

that's this idea that I've been thinking about of AI

31:42

and kind of invading social media, so to

31:44

speak. I mean, maybe you don't like the word invading, but

31:46

you know what I mean. And what does that

31:48

do to how we relate to

31:51

each other as humans? Like how much AI stuff

31:54

and AI generated stuff is going to be filling

31:56

feeds in the near future in your view. Well,

31:58

here's how I come at this. So

32:00

in the history of running the company where

32:02

we've been building these apps for 20 years

32:05

every cult three to five

32:07

years there's some new major

32:10

format that comes along that is

32:14

typically additive to the experience, right? So, you know

32:17

initially people kind of updated their profiles Then

32:19

they were able to post statuses that were

32:21

text and then links and you got photos

32:23

early on then you added videos then

32:26

with mobile Basically,

32:28

you know snap invented stories the first version

32:30

of that and that became a pretty

32:33

kind of widely used format the

32:36

whole version of short form videos

32:38

I think is sort of a still

32:40

ascendant format, but at

32:43

each step along the way you

32:45

want to Write

32:48

it's like you keep on making the system richer

32:50

by having more different types of content that people

32:52

can share and in different ways to Express themselves

32:54

and when you look out for the next 10

32:56

years of okay If this trend

32:59

seems to happen where every three five

33:01

years Whatever the pace is

33:03

that there are new formats. I

33:05

think given the pace of change in the tech industry

33:07

I think you'd bet that that continues or accelerates and

33:10

I think you'd bet that Probably

33:12

most of the new formats are going to be

33:14

kind of AI connected in some way given that

33:16

that's the kind of driving theme for the industry

33:19

at this point, so Given

33:21

that kind of set of assumptions. We're

33:23

sort of trying to understand What

33:26

are the things that are most useful to people within

33:28

that? There's one vein of this which

33:30

is Helping people and creators

33:32

make better content using AI that I

33:34

think is gonna be pretty clear right

33:36

just make it like super easy for

33:39

Like aspiring creators or advanced creators to make

33:41

much better stuff than they would be able

33:43

to otherwise That can take the form

33:45

out of like alright Like my

33:48

daughter is writing a book and She

33:51

wants it illustrated and like we sit down

33:53

together and work with meta AI and imagine

33:55

to help her come up with images to

33:57

illustrate It it's like okay. That's like a

33:59

thing That's like she didn't have the capability to

34:01

do that before. She's not a graphic designer, but now

34:03

she kind of has that ability. I

34:06

think that that's going to be pretty cool. Then I

34:08

think that there's a version where you have just

34:11

this great diversity of

34:13

AI agents that are as part of this

34:15

system. And this, I think is a big difference between

34:18

our vision of AI and most of the other companies

34:20

is, yeah, we're building MetAI as

34:22

kind of the main assistant that you can

34:24

build. That's sort of equivalent to, you know,

34:26

the singular assistant that may be like a

34:28

Google or an open AI or different folks

34:30

are building, but it's

34:32

not really the main thing that we're doing. You

34:35

know, our main vision is that we think that

34:37

there are going to be a lot of these,

34:39

right? It's every business, all the hundreds of millions

34:41

of small businesses, you know, just like they have

34:43

a website and an email

34:45

address and a social media account today, I

34:48

think that they're all going to have an AI that helps

34:50

them interact with their customers in the future that does some

34:52

combination of sales and customer support and all that. I

34:54

think all the creators are basically going to want some

34:57

version of this that basically helps them

34:59

interact with their community when they're just limited by their

35:01

own amount of hours in the day to interact with

35:03

all the messages that are coming in and they want

35:05

to make sure that they can show some love to

35:07

people in their, in their community. And

35:09

those I think are just the two most obvious

35:11

ones that even if we just did those, that's

35:14

many hundreds of millions, but then there's going to

35:16

be all this more creative stuff that's UGC that

35:18

people create for better kind of wilder use cases

35:20

that they want. And our view is, okay, these

35:22

are all going to like live across these social

35:24

networks and beyond. I don't think

35:26

that they should just be constrained

35:28

to waiting until someone messages them. Right.

35:31

I think that they're going to have

35:33

their own profiles. They're going to

35:35

be creating content. People will be able to follow them if they

35:37

want. You'll be able to comment on their stuff. They

35:40

may be able to comment on your stuff if you're connected with

35:42

them. I mean, there will obviously be different, different logic and rules,

35:44

but that's one way that there's

35:46

going to be just a lot more

35:48

kind of AI participants in the kind

35:50

of broader social construct that we have.

35:53

And then I think you get to the test that you mentioned, which

35:55

is maybe the most

35:58

abstract, which is just having. kind

36:00

of the central meta

36:02

AI system directly

36:05

generate content for you

36:08

based on what we think is going to be

36:10

interesting to you and putting that in your feed.

36:12

On that, there's been this trend over time where

36:14

the feed started off as primarily

36:17

and exclusively content for

36:19

people you followed, friends. I

36:21

guess it was friends early on, then it kind

36:23

of broadened out to, okay, you followed a set

36:26

of friends and creators. And

36:28

then it got to a

36:30

point where the algorithm was good enough where we're actually

36:32

showing you a lot of stuff that you're not following

36:34

directly because in some ways that's a better

36:36

way to show you more interesting stuff than only constraining it to

36:38

things that you've chosen to follow. And I

36:42

think the next logical jump on that is like, okay,

36:44

we're showing you content from your friends and

36:46

creators that you're following and creators that you're

36:48

not following that are generating interesting things. And

36:51

you just add onto that a layer of, okay, and we're also

36:53

going to show you content from that's generated

36:55

by an AI system that might be something

36:57

that you're interested in. Now, how big do

36:59

any of these segments get? I

37:01

think it's really hard to know until you build them out over

37:03

time, but it feels like it is

37:06

a category in the world that's going to exist and

37:08

how big it gets is kind of dependent on

37:11

the execution and how good it is. Why do

37:13

you think it needs to exist as a new

37:15

category? I'm still wrestling with why people

37:18

wants this. I get the companionship stuff that

37:20

Character AI and some startups have already shown

37:22

there's like a market for, and you've talked

37:25

about how meta is already being used for

37:27

role-playing. But the

37:29

big idea is that AI has been used to intermediate

37:32

and feed how humans reach

37:34

each other. And now all of a

37:36

sudden, AIs are going to be in feeds with us.

37:39

Well, I think the main difference. And that feels big.

37:42

Yeah, but in a lot of ways, the

37:44

big change already happened, which is people getting

37:46

content that they weren't following. The

37:49

definition of feeds in social interaction has

37:52

changed very fundamentally in the last 10 years. Now,

37:55

in social systems, most

37:58

of the direct interaction is happening

38:00

in more private forums in messaging our

38:02

groups. This is one of the

38:04

reasons I think why we were late with Reels initially

38:07

to competing with TikTok is because we hadn't made this

38:09

mental shift where we kind of felt like, no, feed

38:11

is where you interact with people. Actually

38:14

increasingly feed is becoming a

38:16

place where you discover content that you then take

38:18

to your private forums and interact with people there.

38:20

So a lot of the way that I interact

38:22

with people, it's like, yeah, I'll still have the

38:24

thing where a friend

38:26

will post something and I'll comment on it

38:29

and engage directly in feed. Again, this is

38:31

additive, you're adding more over time. But

38:33

the main way that you engage with Reels isn't

38:36

necessarily that you go into the Reels comments and

38:38

comment and talk to people you don't know. It's

38:40

like, you see something funny and you send it

38:43

to friends in a group chat. I

38:45

think that that paradigm will absolutely continue

38:47

with AI and all kinds of interesting

38:49

content. It is

38:51

facilitating connections with people, but

38:55

I think already we're in this mode where

38:58

our connections through social media are shifting to

39:00

more private places and

39:02

the role of feed in the ecosystem is

39:04

more as a, I call it a discovery

39:06

engine of content to kind of, of icebreakers

39:09

or interesting kind of topic starters

39:11

for the conversations that you're having

39:13

across this broader spectrum of places

39:16

where you're interacting. Do you worry

39:18

about people interacting with AIs like

39:20

this, making people less

39:22

likely to talk to other people, like it

39:24

reducing the engagement that we have with humans?

39:28

I mean, the sociology that I've

39:30

seen on this is that most

39:32

people have way fewer friends physically than they

39:34

would like to have. I think people cherish

39:36

the kind of human connection that they have

39:38

and the more we can do to make

39:41

that feel more real and give you more

39:43

reasons to connect, whether it's through something funny

39:45

that shows up so you can message someone

39:47

or a pair of glasses that lets your

39:50

sister show up as a hologram in your living room when

39:52

she lives across the country and you wouldn't be able to

39:54

see her otherwise. That's always kind of

39:56

our main bread and butter in the thing that we're doing. In

40:00

addition to that, I mean,

40:03

if the average person, I think, maybe

40:05

they'd like to have 10 friends and I

40:08

mean, there's the stat that it's sort of

40:10

sad, but I think the average

40:13

American feels like they have fewer than three

40:15

real kind of close friends. So does

40:18

this take away from that? My guess is

40:20

no. I think that what's

40:22

gonna happen is it's going to help

40:25

give people more of the support that

40:27

they need and give people

40:29

more kind of reasons and ability to connect

40:31

with either a broader range of people or

40:33

more deeply with the people that they care

40:35

about. We

40:38

need to take another quick break. We'll be right back. What

40:49

is AI actually for? Everybody is

40:52

talking about AI. Wall Street is

40:54

obsessed with AI. AI will save

40:56

us. AI will kill us. It's just

40:58

AI, AI, AI, everywhere you turn. But

41:01

what does any of that actually mean, like

41:03

in your life right now? That's

41:05

what we're currently exploring on the VergeCast. We're

41:08

looking for places and products where AI might

41:10

actually matter in your day-to-day life. And I

41:12

don't just mean for making your emails less

41:14

mean, though I guess that's good too. Lots

41:17

of big new ideas about AI this month on

41:19

the VergeCast, wherever you get podcasts. I'm

41:23

Anu Subramanian from Vox Media. While

41:25

I see them all around the city, I've

41:27

never ridden in an autonomous vehicle myself. I

41:30

do have some questions about the tech. You may

41:32

as well. Hello,

41:34

from Waymo. This experience may feel

41:36

futuristic. This is so cool. Vox

41:41

and Waymo teamed up for an in-depth

41:43

study about AV perception. And what they

41:45

found was that as people learned more

41:47

about Waymo, their interest in choosing one

41:49

over a human-driven vehicle almost doubled. Person

41:52

approaching. Waymo can see 360 degrees and

41:54

up to 300 meters away, which

41:57

helps it obey traffic laws and get you where you're

41:59

at. going safely. Swiss

42:01

Re found that compared to human drivers, Waymo

42:04

reported 100% fewer injury claims and 76%

42:08

fewer property damage claims. And

42:10

speaking of safety, folks identifying as

42:12

LGBTQIA and non-binary showed the highest

42:15

interest in AVs, and women showed

42:17

the greatest increase in interest after

42:19

learning more. Arriving

42:22

shortly at your destination. So

42:25

that actually felt totally normal. AVs

42:28

are here, and the more you know, the more exciting

42:30

this tech becomes. You

42:32

can learn more about Waymo, the world's

42:34

most experienced driver, by heading to waymo.com.

42:40

When it comes to business, you know this

42:42

podcast has you covered. But who do you

42:44

turn to when you need smart financial decisions?

42:47

If your answer is our sponsor, NerdWallet, then

42:49

you're absolutely right. And if it's not,

42:51

allow us to change your mind. Not

42:54

only have the nerds over at NerdWallet spent thousands

42:56

of hours researching and reviewing over 1,300 financial

42:59

products, but they have the tools you need

43:01

to make smarter decisions. Looking

43:03

for a credit card? Go beyond the basic

43:05

comparisons. At NerdWallet, you can filter for the

43:08

features that matter to you and read in-depth

43:10

reviews. Ready to choose

43:12

a high-yield savings account? Get access to

43:14

exclusive deals and compare rates, bonuses, and

43:16

more. House hunting? View

43:18

today's top mortgage rates for your home sweet

43:20

home. Make the nerds

43:23

your go-to resource for smart financial decisions.

43:25

Head to nerdwallet.com/learn more.

43:29

NerdWallet. Finance smarter. NerdWallet Compare

43:31

Incorporated. And MLS 1617539. We're

43:42

back with Metaseo Mark Zuckerberg, talking about the

43:44

current state of threads and why the company

43:46

is trying to back out of politics. How

43:52

are you feeling about how threads is doing

43:54

these days? I mean, threads on fire. It's

43:56

great. I mean, it's I mean, these things, it's

43:58

like there's only so quickly that something can get to a billion. people. So

44:00

it's going to, you know, we'll kind of keep

44:02

on pushing on it over. I've heard it's still

44:04

using Instagram a lot for growth. Like I'm, I

44:07

guess I'm wondering when do you see it getting

44:09

to like a standalone growth driver on its

44:12

own? I mean, I think that these things

44:14

all connect to each other. I mean, I think threads

44:16

helps Instagram. I think Instagram helps threads. I don't know

44:18

that we have some strategic goal, which is like make

44:21

it so that threads is completely disconnected from Instagram or Facebook.

44:23

I actually think we're going the other direction. It started off

44:25

just connected to Instagram and now we also connected it so

44:27

that content can show up. You know, taking a step back,

44:30

I mean we just talked about how for

44:32

most people they're interacting in more private forums. If

44:34

you're a creator, what you want to do is

44:36

have your content show up everywhere, right? Because you're

44:39

trying to build the biggest community that you can

44:41

in these different places. So it's this huge value

44:44

for people. If they can generate a real

44:46

or a video or some text based content,

44:48

and now you can post it in threads,

44:50

Instagram, Facebook, um, and more places

44:52

over time. So I think the direction there is

44:55

generally kind of more flow, not

44:57

less and kind of more interoperability. And that's,

44:59

that's why I've been kind of pushing on

45:01

that as a theme over time. I'm not

45:03

even sure what X is anymore, but I

45:05

think what it used to be and what

45:07

Twitter used to be was a place where

45:09

you went when news was happening. I know

45:11

you and the company seem to be distancing

45:13

yourself from recommending news, but with

45:15

threads, it feels like that's

45:18

what people want and people thought threads might

45:20

be, but it seems like you all are

45:22

intentionally saying we don't want threads to actually

45:25

be that. Yeah. I mean, it's

45:27

interesting. There's there are different ways to look at

45:29

this. I always looked at Twitter, not as

45:31

primarily about real time news, but

45:33

as a kind of short form,

45:36

primarily text discussion oriented app to

45:38

me, the fundamental defining aspect of

45:40

that format is that you

45:42

don't, when you make a post, the

45:45

comments aren't subordinate to the post. The

45:47

comments are kind of at a peer level, and

45:50

that is a very different

45:52

architecture than every other type

45:54

of social network that's out there. It's

45:56

a subtle difference, but within these

45:58

systems, these subtle differences. is lead to very

46:01

different emerging behaviors. So because of that, people

46:03

can take, they can fork discussions, and

46:06

it makes it a very good discussion-oriented platform. Now, news

46:08

is one thing that people like discussing, but it's not

46:10

the only thing. And I always looked at Twitter and

46:12

I was like, hey, this

46:14

is such a wasted opportunity. Like this is

46:17

clearly a billion person app. You

46:19

know, maybe in the modern day when you have

46:21

multiple, like many billions of people using social apps,

46:23

it should be multiple billions of people. For whatever

46:25

reason, I mean, there are a lot of things

46:27

that have been complicated about Twitter and the corporate

46:30

structure and all that, but they just weren't quite

46:32

getting there. And eventually, I kind

46:36

of thought, hey, I think we can

46:38

do this, right? I think we can get this,

46:40

like build out the discussion platform in

46:43

a way that can get to a billion people and

46:45

be more of a ubiquitous social platform

46:47

that I think achieves the, like

46:50

its full potential. But our

46:52

version of this is we want it to be a kinder place.

46:55

We don't want it to start with kind of the

46:57

like direct kind of head

46:59

to head combat of news and especially

47:01

politics. And they feel

47:05

like that constrains the growth of the product at all. I mean,

47:07

I think we'll see. Or that that needs to exist in the

47:09

world. Cause I feel like with X's

47:11

seeming implosion, it's not really existing anymore.

47:13

Maybe I'm biased to someone in the

47:15

media, but I do think people want,

47:18

when something big happens in the world, they want an

47:20

app that they can go to and see everyone that

47:22

they follow talking about it immediately. Yeah,

47:24

well, we're not the only company. And

47:27

there's like a ton of different competitors

47:29

and different companies doing things. And I

47:32

mean, I think that there's a talented team over at Twitter

47:34

and X and I wouldn't write them off.

47:37

And then obviously there's all these other folks. There's a

47:39

lot of startups that are doing stuff. So I don't

47:41

feel like we have

47:44

to go at that first. I think that

47:46

like maybe we get there over time or maybe

47:48

we decide that it's enough

47:50

of a zero sum trade or maybe even a

47:52

negative sum trade where like that use case should

47:54

exist somewhere. But maybe that use

47:56

case prevents a lot more

47:59

usage. kind of a lot more

48:01

value in other places because it makes it a somewhat

48:03

less friendly place. I don't think we

48:05

know the answer to that yet, but I do think the

48:07

last 10 years, eight years

48:10

of our experience has been that

48:12

the political discourse, it's

48:14

tricky, right? It's on the one hand, it's

48:16

obviously a very important thing in society. And

48:19

on the other hand, I don't think it

48:21

leaves people feeling good. So I'm torn between

48:23

these two values. On the one

48:25

hand, I think like people should be able to

48:27

have this kind of open discourse and that's good.

48:29

On the other hand, I don't want to design

48:32

a product that makes people angry. Right?

48:34

It's like, I mean, there's an informational lens

48:36

for looking at this and there's

48:38

kind of a, you're designing a product and like, what's

48:40

the feel of the product, right? It's like, I think

48:43

anyone who's designing a product cares a lot about how

48:45

the thing feels and- But

48:48

you recognize the importance of that discussion happening

48:50

in the world. I think it's useful. And

48:52

look, we don't block it. We just

48:54

make it so that, for the content

48:56

where you're following people, if you want to talk to your

48:58

friends about it, if you want to talk to them about

49:00

it in messaging, there can be groups about it. If

49:03

you follow people, it can show up in your feed. But we

49:05

don't go out of our way to recommend

49:08

that content when you're not following it. And

49:10

I think that that

49:12

has been a healthy balance for us and

49:14

for getting our products to generally feel the

49:17

way that we want. And

49:19

culture changes over time. Maybe this stuff will

49:22

be like a little bit less polarized and

49:24

anger inducing at some point. And maybe it'll

49:27

be possible to have more of that while

49:29

also at the same time having a product

49:31

where we're proud of how it feels. But

49:34

until then, I think we want to design a product

49:36

that, yeah, people can get the things that they want,

49:39

but fundamentally I care a lot about

49:41

how people feel coming away from the

49:44

products. Do you see this decision to

49:46

downrank political content for people who aren't

49:48

being followed in feed as a political

49:50

decision, I guess? Because I don't

49:53

know. You're also at the same time, not

49:56

really saying much about the election this year. You're not

49:58

donating. You've said you... kind of want to stay out

50:00

of it now. And I see

50:02

the way the company is acting and it

50:04

reflects your personal kind of way you're operating

50:07

right now. And I'm wondering like how

50:09

much more of it is also about what you and the

50:11

company have gone through and the

50:14

political environment and not necessarily just what users

50:16

are telling you. Like, is there

50:18

a through line there? I mean, I'm sure it's all

50:20

connected. I think in this case, it wasn't a trade-off

50:22

between those two things because this actually was what our

50:24

community was telling us. And people were saying generally, we

50:27

don't want so much politics. Like

50:30

this isn't, you know, like we don't feel good. Like we want

50:32

content, we

50:35

want more stuff from our friends and family. We want more stuff

50:37

from our interests. That was kind

50:39

of the primary driver. But I think

50:41

it's definitely the case that our corporate

50:43

experience on this shaped this. And

50:45

I think there's something, there's a

50:48

big difference between something being political and being partisan.

50:51

And the main thing that I care

50:53

about is making sure that we

50:55

can be seen as a

50:57

nonpartisan and as much as something

50:59

can in the world in 2024 be

51:02

sort of like a trusted institution by

51:04

as many people as possible. And I

51:07

just think that the partisan politics is so

51:09

tough in the world right now that

51:12

I've made the decision that I kind of feel like for

51:14

me and for the company, the best thing

51:16

to do is to try to be as nonpartisan as

51:18

possible in all of this and kind

51:20

of be as neutral and distance ourselves as much

51:23

as possible. And it's not just the substance. I

51:25

also think the perception matters. So

51:27

that's why maybe it doesn't matter

51:29

on our platforms, whether I endorse a candidate or

51:31

not, but I don't wanna go anywhere near that.

51:35

And yeah, sure. I mean, you

51:37

could say that's a political strategy, but

51:39

I think for where we are

51:41

in the world today, it's very

51:43

hard. Almost every institution has

51:46

become partisan in some way. And

51:48

we are just trying

51:51

to resist that. And maybe I'm too

51:53

naive and maybe that's impossible, but we're

51:55

gonna try to do that. On the

51:57

acquired podcast recently, you said that the

51:59

political. miscalculation was a 20 year mistake.

52:01

Yeah, from a brand. From a brand perspective.

52:04

And that it was gonna take another 10

52:06

or so for you to fully work through

52:08

that cycle. Yeah, yeah. What makes you think

52:10

it's such a lasting thing? Because you look

52:13

at like how you personally have kind

52:15

of evolved over the last couple of years and I think

52:17

perception of the company has evolved and I'm wondering

52:19

like what you meant by saying it's gonna take another

52:21

10 years. I'm just talking about

52:23

where our brand is and

52:25

our reputation are compared to where I

52:28

think they would have been. There's no

52:30

doubt that even now here and okay,

52:32

yeah, sure. Maybe things have improved somewhat over the

52:34

last few years. You can feel the trend, but

52:37

it's still significantly worse than it was in 2016. You

52:40

know, it's I mean the internet industry overall

52:43

and I think our company in particular

52:46

just we're seeing way more positively

52:49

and now look, there

52:51

were real issues, right? So I

52:54

think that it's always very difficult to talk about this stuff in

52:56

a nuanced way because I think

52:58

to some degree before 2016, everyone was sort of too

53:00

rosy about the internet overall and didn't talk enough about

53:02

the issues and then the pendulum sort of swung and

53:04

people only talked about the issues and didn't talk about

53:06

the stuff that was positive and it was all both

53:09

there the whole time. So when I talk about this,

53:11

I don't mean to come

53:13

across as simplistic or like

53:15

you guys didn't do anything wrong or anything like that.

53:18

There weren't issues with the internet or things like that.

53:20

I mean, obviously every year, whether it's politics or other

53:22

things, there are always things that you look back at

53:24

and you're like, Hey, yeah, like if I were playing

53:26

this perfectly, I would have done these things differently. And

53:29

but I do think it's the case that I

53:32

didn't really know how to react to something as big

53:34

of sort of a shift in the world as what

53:36

happened. And it took me a while to

53:38

find my footing. And I do

53:40

think that it's tricky when

53:42

you're caught up in these kind of big debates

53:44

and you're not kind of

53:46

experienced or sophisticated in engaging with that. I

53:49

think you can make some big missteps.

53:51

And I do think that some of the things that

53:53

we were accused of over time, it just, you know,

53:55

I think it's just been pretty clear at this point,

53:57

you know, now that all the investigations have been done

53:59

that. they

54:01

weren't true. And- You're

54:03

talking about Cambridge Analytica. I think Cambridge Analytica

54:06

is a good example of something that it's

54:08

like people thought that all this data had

54:10

been taken and that it had been used

54:12

in this campaign. And I think it turns

54:14

out it wasn't. It turns out it wasn't.

54:16

Yeah, so it's all the stuff. Okay, and

54:18

the data wasn't even accessible to

54:20

the developer. And we'd fixed the issue five years

54:22

ago. So in the moment, it was

54:24

really hard for us to have

54:26

a rational discussion about that. And

54:28

I think part of the

54:31

challenge is that for the general

54:33

population, I think a lot of people, they read the

54:35

initial headlines and they don't necessarily

54:37

read the, and

54:41

frankly, a lot of the media, I don't think

54:43

was like as loud to write about when all

54:45

the investigations concluded that said that like a lot

54:47

of the initial allegations were just completely wrong. So

54:49

I think that's like a, that's a real thing.

54:52

So, okay, so you take these hits. I

54:54

didn't really know how to kind of

54:56

push back on that. And maybe some of

54:58

it you can't, but I like

55:01

to think that I think we could have played some of

55:03

the stuff differently. And I do think

55:05

it was certainly the case that when you

55:07

take responsibility for things that are not your

55:09

fault, you become sort

55:11

of a weak target for people

55:14

who are looking to blame other things and

55:16

find a target for them. It's sort of

55:18

like, this is a different part of, it's

55:21

somewhat related to this, but when you think about like, like

55:24

litigation strategy for the company. One of the

55:26

reason why I hate settling lawsuits is

55:28

that it basically sends a signal to

55:30

people that, hey, this is a company that

55:33

settles lawsuits. So maybe like we can sue

55:35

them and they'll settle lawsuits. So- So

55:38

you wouldn't write a blank check to the government

55:40

like Google did for its antitrust case. No, I

55:42

think the right kind of way

55:44

to approach this is when you believe in something,

55:47

you fight really hard for it. And I think

55:49

this is a repeat game. This isn't like, it's

55:52

not like there's a single issue and we're

55:54

going to be around for a long time. And I

55:56

think it's really important that people know that

55:59

we're a company that has conviction and that

56:01

we believe in what we're doing and we're

56:03

gonna back that up and defend ourselves. And

56:05

I think that that kind of sets the

56:07

right tone. Now, I

56:10

think over the next 10 years, I think we're

56:12

sort of digging ourselves back to neutral on this,

56:14

but I like to think that if we hadn't had a lot

56:16

of these issues, we would have made progress over the last 10

56:18

years too. So I sort of give it this timeframe, maybe 20

56:20

years is too long, maybe it's 15, but it's

56:23

hard to know with politics. It feels like mental health

56:25

and youth mental health may be the next wave of

56:27

this. That I think is the next big fight. And

56:29

on that, I think a lot of the data on

56:31

this, I think is just not

56:34

where the narrative is. The narrative, yeah, I

56:36

think the narrative is a lot of people

56:38

sort of take it as if it's like

56:40

an assumed thing that there's some link. And

56:42

like, I think the majority of the

56:45

high quality research that's out there suggests that there's

56:47

no causal connection like a kind

56:49

of a broad scale between these things. So, no,

56:51

look, I mean, I think that that's different from

56:53

saying like in any given issue, like is someone

56:55

bullied? Should we try to stop bullying? Yeah, of

56:57

course. But yeah, overall,

56:59

I think that this is one

57:01

where there

57:04

are a bunch of these cases. I think that there will be a

57:06

lot of litigation around them. And it's

57:08

one where we're trying to make sure

57:10

that the academic research that shows something

57:13

that I think is, to me, it sort of

57:15

fits more with

57:19

what I've seen of how the platforms operate, but

57:21

it's counter to what a lot of people think.

57:23

And I think that that's gonna be a reckoning

57:25

that we'll have to have is basically as

57:28

the kind of the majority of the high

57:30

quality academic research gets shown is

57:32

like, okay, can people accept this? And I think that's

57:35

gonna be a really important set of debates over the

57:37

next few years. At the same time, you guys have

57:39

acknowledged there's affordances in the product, like the teen rollout

57:41

with Instagram recently that you can make to make

57:43

the product a better experience for young

57:45

people. Yeah, and I think this is an interesting

57:47

part of the balance is you

57:51

can play a role in trying to make something

57:53

better, even if the thing wasn't

57:55

caused by you in the first place. There's

57:57

no doubt that being a parent. is

58:00

really hard. And there's a big

58:02

question of in

58:05

this internet age where we have

58:07

phones, what are the right

58:09

tools that parents need in order

58:11

to be able to raise their

58:13

kids? And like, I

58:16

think that we can play a role in

58:18

giving people controls over parental controls over the

58:20

apps. I think that parental controls are

58:22

also really important because parents have different ways that they wanna

58:24

raise their kids. Or just like schooling

58:27

and education, people have very significantly

58:29

different local preferences for how they wanna raise

58:31

their kids. I don't think that

58:33

most people want some internet

58:36

company setting all the rules for this

58:38

either. So obviously, when there

58:40

are laws passed, we'll kind of

58:42

follow the government's direction and

58:44

the laws on that. But I actually think

58:47

the right approach for us is

58:49

to primarily kind of

58:52

align with parents to give them the tools that they want

58:54

to be able to raise their kids in the way that

58:56

they want. And some

58:58

people are gonna think that more technology use is good.

59:00

That's sort of how my parents raised

59:03

me growing up. I think it worked pretty well. Some

59:05

people are gonna be kind of more, you

59:08

know, wanna limit it more. And we wanna give them the tools to be able

59:10

to do that. But I don't think that this is primarily

59:12

or only a social media thing. Even

59:16

the parts of this that are technology. Age

59:18

verification. I think phones, like the phone platforms have

59:21

a huge part of this. I mean, it's, yeah,

59:23

there's this big question of how do you do

59:25

age verification? I mean, I

59:27

can tell you what the easiest way is, which is

59:29

like, all right, like every time you go do a

59:31

payment on your phone, I mean, there already is child,

59:33

you know, basically like child age verification. So I

59:36

don't really understand, well, I guess I

59:38

understand, but I think it's not very,

59:40

you know, excusable from my perspective why

59:43

Apple and I guess to

59:45

some extent, Google don't wanna just

59:47

extend the age verification that they already have

59:49

on their phones to be a parental control

59:51

for parents to basically be able to say,

59:53

you know, what apps can my kids use?

59:55

It's hard for me to not see the

59:57

logic in it either. I don't, I don't.

1:00:00

I think they don't wanna take responsibility. But

1:00:03

maybe that's on Congress then to pass who

1:00:05

has to take responsibility. Yeah. Yeah.

1:00:07

And we're gonna do our part and we're

1:00:09

gonna build the tools that we can for

1:00:11

parents and for teens. But

1:00:14

at the end of the day, and

1:00:16

look, I'm not saying it's all the phone's fault

1:00:18

either. Although I would say that like the ability

1:00:20

to get push notifications and get kind of distracted

1:00:23

is from my perspective seems like

1:00:25

a much greater contributor to mental health issues than

1:00:28

kind of a lot of the specific apps. But

1:00:30

there are things that I think everyone should kind

1:00:32

of try to improve and work on. But

1:00:35

yeah, I mean, that's sort of my view on all of that.

1:00:38

I guess on the regulation piece, as it

1:00:41

relates to AI, you've been very

1:00:43

vocal about what's happening in the EU. And

1:00:45

you recently signed an open letter and

1:00:47

I believe it was saying basically that

1:00:49

you guys just don't have clarity on

1:00:52

consent for training, how it's supposed to

1:00:54

work. And I'm wondering

1:00:57

what you think needs to happen there for things to

1:00:59

move forward because like med AI is not available in

1:01:01

Europe, new llama models are not. Is

1:01:04

that something you see getting resolved

1:01:06

at all, I guess, and

1:01:09

what would it take? Yeah, I

1:01:11

don't know. It's a little hard for

1:01:13

me to parse the European politics. I

1:01:15

have a hard time enough with American

1:01:18

politics. I mean, it's, and I'm American,

1:01:20

but in theory, my

1:01:22

understanding of the way this is supposed to work is

1:01:24

they kind of passed this

1:01:26

GDPR regulation. And

1:01:30

you're supposed to have this

1:01:32

idea of sort of a one-stop

1:01:35

shop, like home regulator,

1:01:37

who can basically on behalf of the whole EU interpret

1:01:41

and enforce the rules. We

1:01:43

have our European headquarters and

1:01:46

we work with that regulator and I

1:01:48

think they're like, okay, they're pretty tough on

1:01:50

us and pretty firm. But at least

1:01:52

when you're working with one regulator, you can kind of understand

1:01:55

how are they thinking about things and you

1:01:57

can make progress. And the thing that

1:01:59

I think has been... tricky is

1:02:01

there have been, from my perspective,

1:02:03

a little bit of a backslide where

1:02:06

now you get all these other DPAs

1:02:08

across the continent, sort of

1:02:10

also intervening and trying to do things and

1:02:12

it just seems like more of an kind

1:02:14

of internal EU political thing, which is like,

1:02:16

okay, do they want to have this one

1:02:19

stop shop and have clarity for companies so

1:02:21

companies can kind of like can execute or

1:02:24

do they just want it to be this kind

1:02:26

of very complicated regulatory system? And

1:02:28

look, I think that's for them to sort out, but I think

1:02:30

that there's no doubt that when you have like dozens

1:02:33

of different regulators that can ask you kind

1:02:35

of the same questions about different things, it

1:02:37

makes it a much more difficult environment to

1:02:39

build things. Do you understand

1:02:41

that that's just us? I think that that's all

1:02:43

the companies. But do you understand the concern people

1:02:45

have about training data and how

1:02:47

it's used and this idea

1:02:49

that their data is being used for

1:02:52

these models, they're not getting compensated and

1:02:54

the models are creating a lot of value. And

1:02:56

I know you're giving away llama, but you're

1:02:58

still, you've got med AI and I

1:03:00

understand the frustration that people have about that. I

1:03:03

think it's a naturally bad feeling to be like,

1:03:05

oh, my data is now being used in a

1:03:07

new way that I have no control or compensation

1:03:09

over. Do you sympathize with that? Yeah. I

1:03:12

mean, I think that there in

1:03:14

any new medium in technology, there's

1:03:16

like the concepts around fair use

1:03:18

and like where the boundary is between

1:03:21

what you have control over, but when you put

1:03:23

something out in the world, to what degree do

1:03:25

you still get to control it and kind of

1:03:27

own it and license it? I think

1:03:29

that all these things are basically going to need to

1:03:31

get relitigated and

1:03:34

rediscussed in the AI era.

1:03:36

So I'm going to get it. I think that

1:03:38

these are important questions. I think this is not

1:03:40

like a completely novel thing to AI in the

1:03:43

grand scheme of things. I think it was a

1:03:45

lot of them, there were questions about it with

1:03:47

the internet overall too and with different technologies over

1:03:49

time. But I think getting to

1:03:51

clarity on that is going to be important

1:03:53

so that way the things that society

1:03:56

wants people to build, they can go

1:03:58

build. look like to you there.

1:04:01

I mean, I think it starts with having some framework

1:04:03

of like, okay, what's the process going to be if

1:04:06

we're working through that? But

1:04:08

you don't see a scenario where creators get

1:04:10

like directly compensated for the use of their

1:04:12

content models. I think that there's a lot

1:04:14

of different possibilities for how stuff goes in

1:04:17

the future. Now I do think

1:04:19

that there's this issue, which is a lot of,

1:04:21

like, while psychologically I understand what you're saying, I

1:04:25

think individual creators or

1:04:27

publishers tend to

1:04:29

overestimate the value of

1:04:32

their specific content. Right, so it's

1:04:34

like, okay, maybe like, in

1:04:36

like in the grand scheme of this, right? So

1:04:39

we have this set of challenges with, you know, news

1:04:41

publishers around the world, which is like, okay, they're

1:04:43

like a lot of folks are constantly kind

1:04:45

of asking to be paid for the

1:04:47

content. And on the other hand, we have our community, which

1:04:50

is asking us to show less news because it makes them

1:04:52

feel bad, right? And I mean, we talked about that. So

1:04:54

it's like, there's this issue, which is, okay,

1:04:56

it's like, actually, we're showing some amount of

1:04:58

the news that we're showing because we think

1:05:00

it's socially important against what our

1:05:03

community wants. Like if we were actually just following what

1:05:05

our community wants, we'd show even less than we're showing.

1:05:08

And you see that in the data that people just

1:05:10

don't like to engage with this stuff. We've had these

1:05:12

issues where sometimes we, like, publishers

1:05:14

say, okay, if you're not gonna pay us, then

1:05:16

pull our content down. And it's just like, yeah,

1:05:18

sure, fine, pull your content down. I mean, that

1:05:21

sucks. I'd rather people be able to share it,

1:05:23

but to some degree, some

1:05:25

of these questions have to

1:05:27

get tested by their

1:05:30

negotiations and they have to get tested

1:05:32

by people walking. And then

1:05:34

at the end, once people walk, you

1:05:36

figure out where the value

1:05:39

really is. If it really is

1:05:41

the case that news was a big thing

1:05:43

that the community wanted, then, I mean, look,

1:05:45

we're a big company. We could probably,

1:05:48

we pay for content when it's valuable to

1:05:51

people. We're just not gonna pay for content when it's not valuable

1:05:53

to people. So I think

1:05:55

that you'll probably see a

1:05:57

similar dynamic with AI, which is...

1:05:59

is my guess is that there are

1:06:01

going to be certain partnerships that get made when content

1:06:03

is really important and valuable. And

1:06:05

I guess that there's probably a lot of people who

1:06:08

kind of have a concern about

1:06:10

like the feel of it, like you're saying, but

1:06:13

then when push comes to shove, if

1:06:15

they demanded that we don't use their content,

1:06:18

then we just wouldn't use their content. And

1:06:20

it's not like, you know, that's going to

1:06:22

change the outcome of the stuff that much.

1:06:24

To bring this full circle where we started

1:06:26

as you're building augmented reality glasses and

1:06:28

what you've learned over just the societal implications of

1:06:30

the stuff you've built over the last decade.

1:06:34

How are you thinking about this as

1:06:36

it relates to glasses at scale, because you're

1:06:38

literally going to be augmenting reality, which is

1:06:40

it's a responsibility. I think that's going to

1:06:42

be another platform too. And you're going to

1:06:44

have a lot of these questions as well.

1:06:46

I mean, I think the interesting thing about

1:06:49

holograms and augmented reality is it's going to

1:06:51

be this intermingling of the physical and digital

1:06:53

much more than we've had in other platforms

1:06:56

or on your phone. It's like, okay, yeah, we

1:06:58

live a primarily physical world, but then you have

1:07:00

the small window into this digital world. And I

1:07:02

think we're going to basically have this world in

1:07:04

the future that is increasingly,

1:07:07

you know, call it half physical, half digital,

1:07:10

or I don't know, 60% physical, 40% digital.

1:07:13

And it's like going to be blended together. And I think that

1:07:15

there are going to be a lot

1:07:17

of interesting governance questions around that, right? In

1:07:19

terms of is kind of

1:07:21

all of the digital stuff that's overlaid

1:07:24

physically going to fit within sort

1:07:26

of a physical kind of national

1:07:29

regulation perspective? Or

1:07:32

is it sort of, is it actually coming from

1:07:34

a different world or something? You know, and I

1:07:36

think that these will all be very interesting questions

1:07:38

that we will have a perspective. I'm sure we're

1:07:40

not going to be right about every single thing.

1:07:43

I think like the world will kind of need to sort out

1:07:45

where it wants to be. Different countries will have different values and

1:07:47

take somewhat different approaches on things. And I think that that's, it's

1:07:50

part of the interesting process of this, that

1:07:52

the tapestry of how it all gets built

1:07:54

is like, you need to work through so

1:07:56

that it ends up being positive for as

1:07:59

many of the possible. stakeholders as possible. More

1:08:02

to come. A lot to come. Thanks, Mark. I'd

1:08:08

like to thank Mark Zuckerberg for joining the Coder and

1:08:10

thank Alex Heath for guest hosting. I'd also like to

1:08:12

thank you for listening. I hope you enjoyed it. You

1:08:14

should subscribe to Alex's newsletter command line, which comes out

1:08:16

every week. It is absolutely jam-packed with industry insight, scoops,

1:08:18

and smart analysis. It's a must-read. If you'd like to

1:08:20

let us know what you thought about this episode or

1:08:23

really anything else, drop us a line. You can email

1:08:25

us at decoder at the verge.com. We really do read

1:08:27

all the emails. Or you can hand me up directly

1:08:29

on Threads, a meta product. I'm at Reckless 1280. You

1:08:31

also have a TikTok, or as long as TikTok lasts,

1:08:34

it's at Decoder Pod. It's a lot of fun. If

1:08:36

you'd like to be Coder, please share with your friends

1:08:38

and subscribe wherever you get your podcasts. Decoder is a

1:08:40

production of The Verge and part of the Vox Media

1:08:42

Podcast Network. Our producers are Kate Cox and Nick Stat.

1:08:44

Our editor is Callie Wright. This episode was additionally produced

1:08:46

by Brett Putman and Viren Pavik. Our supervising producer is

1:08:48

Liam James. Decoder music is by Breakmaster Cylinder. We'll see

1:08:50

you next time. What

1:08:57

does impactful marketing look like in today's day and

1:08:59

age? If you're a marketing professional, what decisions can

1:09:01

you make today that will have a lasting impact

1:09:04

for your business tomorrow? We've

1:09:06

got answers to all of the above and

1:09:08

more on the PropG Podcast. Right now, we're

1:09:10

running a two-part special series all about the

1:09:13

future of marketing. It's time

1:09:15

to cut through the noise and make a

1:09:17

real impact. Tune into the future of marketing,

1:09:19

a special series on the PropG Podcast sponsored

1:09:21

by Canva. You can find it

1:09:23

on the PropG feed wherever you get your podcasts.

1:09:28

This episode is brought to you

1:09:30

by Choiceology, an original podcast from

1:09:32

Charles Schwab. Hosted by

1:09:34

Katie Milkman, an award-winning behavioral scientist

1:09:37

and author of the best-selling book

1:09:39

How to Change, Choiceology is a

1:09:41

show about the psychology and economics

1:09:43

behind our decisions. Hear true

1:09:45

stories from Nobel laureates, authors, athletes,

1:09:47

and everyday people about why we

1:09:50

do the things we do. Tune

1:09:53

into Choiceology at schwab.com/podcast

1:09:55

or wherever you listen.