0:00

This message comes from rinse. These

0:02

days you can do a lot

0:04

from your phone. Book a vacation,

0:06

buy and trade stocks, but you

0:08

can also make your dirty laundry

0:10

disappear. And then reappear, wash and

0:12

fold it with rinse. Schedule a

0:14

pickup with the rinse app and

0:16

before you know it, your clothes

0:18

are back folded and ready to

0:20

wear. They even do dry cleaning.

0:22

Sign up now and get $20

0:24

off your first order at rinse.com.

0:26

That's r-i-i-n-s-e.com. You

0:28

understand tech. You

0:30

already use a

0:32

VPN, a password

0:34

manager, and an

0:36

ad blocker. Surely

0:38

you are not

0:40

exposed. Right? Find

0:42

out for sure

0:44

by going to

0:46

cloaked.com/VOS. That's C-L-O-A-K-E-D-D-K-L-K-L-S-L-L-L-L-L-S-L-L-S-L-L-L-L-L-S-L-L-L-L-S-L-L-L-L-S-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-L-

0:49

Cloked can help

0:51

you get your

0:53

data removed from

0:55

hundreds of data

0:57

brokers, generate unlimited

0:59

working phone numbers to

1:01

protect your real one

1:03

and also offers one

1:05

million dollar insurance

1:07

against identity theft.

1:10

That's cloked.com slash

1:12

Vox, and check out with

1:14

Code Podcast to get 30%

1:16

off a Cloke subscription. Through

1:18

the ADT Plus app, easily

1:21

grant and automate event-based or

1:23

scheduled access for neighbors, friends,

1:25

and helpers. Notify trusted individuals

1:27

of events like alarms or

1:29

packages, and set access windows

1:31

for planned guests or even

1:34

the dog walker without interrupting

1:36

your day. Visit adt.com. When

1:38

every second counts, count on

1:40

ADT. Requires ADT Complete Pro

1:43

Monitoring Plan and Compatible

1:45

Devices. Copyright 2025 ADT

1:47

LLC. All rights reserved.

1:49

Hello and welcome to Decoder.

1:51

I'm Neelapetel, editor-in-chief

1:54

of The Virgin. Decoder is my

1:56

show about big ideas and other

1:58

problems. Today I'm talking... with Vimeo

2:00

CEO Philip Moyer. You probably know

2:02

Vimeo from its beginnings is an

2:04

artier, more creative competitor to YouTube.

2:06

But over the last few years,

2:08

and especially after Republic in 2021,

2:10

Vimeo has really turned itself into

2:12

an enterprise software company, selling video

2:14

hosting services to companies of all

2:16

sizes. I got to tell you

2:18

I was pretty excited to talk

2:21

to Philip because this episode is

2:23

a particularly fun full circle Dakota

2:25

moment. I interviewed Philip's predecessor, Anjali

2:27

sued, both when she was CEO

2:29

of Vimeo, and again more recently in

2:31

her new gig as CEO of Tube.

2:33

So I had a sense of how

2:35

Anjali ran Vimeo and what strategies she

2:37

took to Tube. And then it was

2:40

fascinating to close the loop and see

2:42

how Philip wanted to change Vimeo after

2:44

taking over himself. Especially because the entire

2:46

ecosystem of online video is itself changing

2:48

incredibly rapidly. You'll hear Philip be very

2:50

complementary of what Anjali accomplished at Phymio,

2:53

but now he's pushing to make Phymio

2:55

grow into a very different kind of

2:57

YouTube competitor, one that can support everything

2:59

from independent creators to huge corporations. It's a

3:01

shift from the strategy that Anjali used to

3:04

reset the company and take it public, and

3:06

there's a lot of interesting nuance to it.

3:08

It turns out everyone wants to put videos

3:10

on the internet. But only some of those

3:12

people want them to be ingested by YouTube's

3:14

advertising and recommendation systems. What's interesting about this

3:16

is that Philip has experience at Google. He

3:19

worked at Google Cloud. In fact, he has

3:21

experience at a lot of places. He's worked

3:23

at Amazon and Microsoft as well. And he's

3:25

deep in the weeds on both the tech

3:27

and the business. You'll hear us talk about

3:29

Google and YouTube in particular quite a

3:31

bit in this one, but we also

3:34

get into TikTok and what it means

3:36

that the incentives of algorithmic video platforms

3:38

have drastically influenced both creators worldwide and

3:40

the culture we all consume. Of course,

3:42

we also talked about AI, which is

3:44

upending every platform in different ways. Vimeo

3:46

markets itself now is an AI powered

3:48

video platform. So I wanted to know

3:50

how Philip is thinking about Vimeo's history

3:52

as a creator focus platform. It's present

3:54

as an alternative to the YouTube algorithms.

3:56

and how all of that collides with

3:58

the pitfalls of a... a lot of

4:00

creators and a lot of consumers do

4:02

not like AI generated content. But there's

4:04

still a place for these tools to

4:06

make some parts of the video process

4:08

easier. And Philip and I talked a lot

4:10

about where those lines might be. I

4:12

also asked Philip about a pretty simple

4:15

supply and demand bath problem I've been wrestling

4:17

with recently, one that seems like it

4:19

will change the creator economy drastically in

4:21

the years to come. If the total amount

4:23

of video on the internet explodes because

4:25

of AI... Well, the total amount of

4:28

time we can spend watching

4:30

that video remains relatively fixed.

4:32

Well, how is anyone going to make

4:34

any money at all? This is a

4:37

fun one. You can tell that Philip

4:39

and I could have kept going for

4:41

a long time. Okay, Vimeo, CEO Philip

4:43

Moir. Here we go. Philip

4:53

Moyer, you are the CEO of Vimeo.

4:55

Welcome to Decoder. Thank you. Hey, it's

4:57

great to be here. I am very

4:59

excited to talk to you. I gotta

5:02

tell you, just structurally, this is one

5:04

of the very first, like, truly full

5:06

circle Decoder episodes, because we had your

5:08

predecessor, Angelion, a CEO of Vimeo. And

5:10

then we had her on in her new

5:13

job as CEO of Tubi. And now we

5:15

have you on as her replacement. It's full

5:17

circle. It's full circle for a show that's

5:19

about structure and decision making and organizational culture.

5:21

This is as good as it gets for

5:23

me. So thank you so much for joining

5:25

us. That's fantastic. I have a lot of

5:27

respect for unduly and have to say a

5:29

big thank you to her for all she

5:31

built here. So it's wonderful. Yeah, I'm very

5:34

curious. You know, Angela, you know, when we

5:36

talked about Vimeo and when she was there,

5:38

she was in the middle of executing a

5:40

big pivot, right? People know Vimeo is what

5:42

it started as a consumer video service,

5:44

kind of the artier competitor to YouTube.

5:46

She pivoted it to a SAS business.

5:48

She was very open. This is now

5:50

a SAS company. We're doing a lot

5:52

of enterprise work. We're a video hosting

5:54

provider to a lot of people. We

5:56

are a video creation tool for small

5:58

businesses for small businesses. That was several

6:00

years ago. I talked to her three

6:02

years ago, Vimeo. You're the new CEO.

6:05

What do you think of Vimeo today?

6:07

You know, I think in a lot

6:09

of ways, I mean, a little bit

6:11

of what she started and a lot

6:13

of, you know, what we're continuing here

6:15

is going back to the roots of

6:17

what Vimeo is known for. You know,

6:19

people came to us because of the

6:21

time, you know, 20 years ago, it

6:23

was hard to upload video. I kind

6:26

of call that I'll say the second

6:28

or third epic of video was right

6:30

around that time period where video files

6:32

were getting bigger. It was hard to

6:34

stream video. They were no longer just

6:36

a single file. We were starting to

6:38

get multiple formats. And so people came

6:40

to us because of the quality of

6:42

our transcoding and then ultimately our ability

6:44

to be able to serve it directly

6:47

to the person, you know, that was

6:49

most important that they wanted to provide

6:51

that video to. And a lot of

6:53

what we're doing right now and what

6:55

we've kind of taken that Anjali had

6:57

started and extended pretty significantly is it

6:59

turns out that lots and lots of

7:01

organizations and individuals around the world want

7:03

to be able to keep a video

7:05

private or they want to be able

7:08

to serve it only to an intended

7:10

audience. They don't want to have someone

7:12

collect data. They don't want to have

7:14

an algorithm that tries to like send

7:16

them down a rabbit hole. Instead they

7:18

want to just literally be able to

7:20

provide that video. We're seeing video today

7:22

is about 82% of the world's internet

7:24

and we're seeing that same concept arrive

7:26

inside of with private individuals with doctors

7:29

with educators with large organizations. And so

7:31

yes, I would say that what we're

7:33

really becoming now is, you know, our

7:35

goal is to become the largest private

7:37

video distribution platform in the world because

7:39

we think that there's an increasing demand.

7:41

That video doesn't have to be public

7:43

or doesn't have to be algorithm during,

7:45

but instead it could be very personal

7:47

and deliberate at the right moment at

7:50

the right time, to the right individual.

7:52

Let me try to just understand that

7:54

in context. There was the beginning, right

7:56

you're saying to YouTube button. at the

7:58

operating system level because no one thought

8:00

that public videos were a big deal,

8:02

right? You just needed some place to

8:04

watch videos in YouTube. that. And then

8:06

YouTube grew into the social media juggernaut

8:08

it is now. Tic-talk exists, all that's

8:11

there. And the initial reaction was it's

8:13

very hard to compete with that, will

8:15

be for enterprises. Business, customers have video

8:17

needs, will service them. You're saying there's

8:19

some kind of middle ground in the

8:21

middle of the spectrum between big algorithmic,

8:23

consumer video platforms, and enterprise, where just

8:25

a lot of regular people want private

8:27

video sharing. We have hundreds of thousands

8:29

of school districts that use us. Teachers,

8:32

you know, want students to be able

8:34

to upload video assignments, especially in the

8:36

world of chat GPT and AI. They

8:38

want to see that the student is

8:40

actually doing the assignment. We've got tens

8:42

of thousands of medical professionals that want

8:44

to be able to send a video

8:46

to a patient without having some algorithm

8:48

capture the fact that their their disease

8:50

state or the question that they have.

8:53

We've got lots and lots of marketing

8:55

organizations that want to be able to

8:57

serve their video over to an individual

8:59

inside of a big company or to

9:01

their client and they don't want it

9:03

to be public. And then what we're

9:05

also finding is on the YouTube front,

9:07

I think organizations that we're hosting entire

9:09

YouTube sections of their website, they're finding

9:11

that YouTube is redirecting their customers over

9:14

to some other place. I was meeting

9:16

with a really large financial services company

9:18

CEO. And the chief economist, I said,

9:20

I took him to the website and

9:22

I said, let me show you your

9:24

chief economist speaking. And I clicked on

9:26

the video. And before the video played,

9:28

we had to watch some advertising on

9:30

some, you know, Bitcoin or something. And

9:32

then we had to go over and,

9:35

like on the right-hand side, it said,

9:37

here's how the top three credit rating

9:39

organizations are trying to control the world.

9:41

This is the video you should watch

9:43

next. And we're just trying to watch

9:45

the economist's video. And so organizations are

9:47

kind of getting tired of being redirected

9:49

or having their data captured. And so

9:51

a lot of organizations, you know, are

9:53

starting to. bypass what I call the

9:56

big walled gardens. And in a lot

9:58

of ways, we're kind of in the

10:00

back of the MSN and AOL error

10:02

for websites. That's what we are with

10:04

video right now. And people like want

10:06

to be able to go directly to

10:08

their consumer. They want to be able

10:10

to serve the message, you know, in

10:12

an unfederated way. And then ultimately, whether

10:14

they just want to be able to

10:17

ensure the highest quality and the most

10:19

personalized experience. And so we're seeing. tremendous

10:21

demand for those kind of scenarios. And

10:23

again, it's everything from the school teacher,

10:25

the fitness instructor, in some cases the

10:27

faith institution, all the way up to

10:29

some of the biggest companies in the

10:31

world. And I do mean literally the

10:33

biggest companies in the world are becoming customers.

10:36

I think the thing that I'm keying on

10:38

there is there's the consumer facing video

10:40

platforms and then you have a bunch

10:42

of enterprise video needs. And you're kind

10:44

of in some way describing a bunch

10:46

of core enterprise customers, right? companies

10:48

big enough to have a chief

10:51

economist, I think those are classic

10:53

enterprise customers. And then somewhere in

10:56

the middle, those companies actually want

10:58

to go reach consumers without participating

11:00

in algorithmic media. And that consumer

11:02

surface, Vimeo has kind of gotten

11:04

away from. Are you saying you're

11:07

kind of, you're pushing back towards it? You know, we

11:09

get about 100 billion views of video a year on

11:11

Vimeo, and only 20% of those it on Vimeo. e-commerce

11:13

platforms. You know, I was meeting with a physician. He is

11:15

an independent practitioner or a fertility doctor and he records a

11:17

video before his patients come in. He sends that video over

11:19

to, to, you know, his patients in the fertility clinic.

11:21

And so we have individual proprietors, you know, we have

11:23

individuals that want to be able to share a family

11:25

video or, you know, something that they learned, you know,

11:27

if they had been into a doctor's visit or otherwise.

11:29

And so we have lots and lots and lots of

11:31

people that people that are using people that are using

11:33

individuals that are using individuals that are using individuals that are

11:36

using individuals that are using individuals that are using individuals

11:38

that are using individuals that are using individuals that

11:40

are using individuals that are using individuals that are

11:42

using individuals, what is most amazing to me right

11:44

now is the sheer number of filmmakers that I

11:47

have coming to me. And this has really started

11:49

kicking up since I've become CEO. I would tell

11:51

you it's been a trend about this past six

11:53

months. I have so many filmmakers that are coming

11:55

to me saying, you know, we don't like the

11:58

deal. We don't like the deal that we have. have

12:00

with the big studios. We don't like the

12:02

fact that if we go to YouTube, as

12:04

an example, they take 45 cents of every

12:06

advertising dollar. Or if we want to try

12:08

and go on to one of the big

12:10

platforms they'll take as much as 50 cents

12:12

of every dollar, is there a way for

12:14

me to be able to sell tickets to

12:16

my audience? In some cases, some of these

12:18

filmmakers that come to us, they have audiences

12:20

that are bigger that they might get on

12:22

one of those platforms. And so we're finding

12:24

people want to go direct. Our streaming business

12:26

is like that. When you post on some

12:28

of these big platforms, I really encourage people

12:31

to look at the terms of service of

12:33

the big major consumer -based video platforms. It

12:35

says in their terms of service, they're able

12:37

to monetize your content any way they want

12:39

to. They can reuse your content.

12:41

They can serve the content. And

12:43

quite frankly, they're capturing 45 cents

12:45

of every dollar in that process. And

12:48

so a lot of these organizations want to

12:50

be able to bypass those kind of economics

12:52

and those kind of IP. So is that

12:54

you building a consumer interface? You're saying it's

12:56

only a small percentage that's coming to vimeo.com.

12:58

Where are they finding audiences? Is it all

13:00

on their own websites? Is it in other

13:02

people's platforms? Where is that audience actually going?

13:05

You know, it's been interesting because there

13:07

was just this trend among streamers in

13:09

particular, well, they'll go to the large

13:11

platforms, they'll get some following, and

13:13

then when they want to be able to serve

13:15

premium content, that's where they'll come

13:17

over and say, you know, we want to put

13:19

SBOT. We want to be able to put

13:22

a gate in front of this content. Or they

13:24

may want to go live or and they

13:26

may want to go asynchronous. And so they'll come

13:28

to us and say, look, you know, we

13:30

want to be able to have a common library

13:32

so that people can see past live events.

13:34

Plus they want to be able to serve new

13:36

content. In some cases, we're getting some organizations

13:38

that want to do more interactive. So like clickable

13:40

videos as an example. And so it's a

13:42

whole variety of creators that are kind of saying,

13:44

look, we might not want to adhere to

13:46

the requirements or the, let's call it

13:48

the compliance requirements or the economic

13:50

requirements or the IP requirements of

13:52

the big platforms. Give us an

13:54

environment that we can control ourselves. And

13:57

that environment lives on their websites. Exactly.

14:00

Like I'll give you an example. Sideman is

14:02

a great example of a group. They're one

14:05

of the biggest followerships over in the UK.

14:07

They sold out Wembley Stadium in like two

14:09

hours and they famously played a soccer match

14:11

where when a yellow card was shown one

14:14

of their one of the Sideman held up

14:16

an Uno Reversi card to the ref and

14:18

it blew up the internet over the UK.

14:20

They serve on Vimeo. They have both content

14:23

they put on a YouTube or they put

14:25

on Instagram, but then some of their more

14:27

extended content, they actually put on Vimeo. And

14:30

then that library lives on us as

14:32

well, you know, for a lot of

14:34

things. Dropouts is a great example. That

14:36

Try Guys is a great example. That

14:38

Zeus Networks, Martha Stewart, where they want

14:40

a little bit more control over the

14:42

monetization, more so than what what the

14:44

traditional platforms give you. I'm going to ask

14:46

you a question and you're just going to

14:49

have to bear with me on sort of

14:51

the mathematical nature of this question. Hopefully it

14:53

makes sense. I have a lot of

14:55

CEOs of web hosting companies on the

14:57

show because I'm very curious about the

15:00

web in the age of platforms. How

15:02

will it grow? Where will the audience

15:04

come from? The last great refer of

15:06

web traffic as everyone knows is Google

15:08

Search. Google Search is undergoing some sort

15:10

of gigantic AI-powered identity crisis. Who

15:13

knows what's going on over there? And

15:15

so I have the CEO of Squarespace

15:17

or a CEO of Wicks or whatever

15:19

other hosting platform on. I say, why

15:21

does anyone build a website? Why would

15:23

you do that instead of starting a

15:25

Tiktok channel now if you're a small

15:28

business or an individual creator? And

15:30

they all kind of say, well, it's

15:32

to do e-commerce, right? And they all

15:34

kind of say, well, it's to do

15:36

e-commerce, right? And they all kind of

15:39

say, well, it's to do your audience,

15:41

and you're going to your audience. You're

15:43

describing the content itself being valuable,

15:45

right? Being more valuable when it's hosted on

15:47

Vimeo. Maybe you're selling it. Maybe you're doing

15:49

subscriptions, whatever you're doing. But that's happening on

15:52

a website because you can't transact that way

15:54

on YouTube or TikTok. You can't make the

15:56

content valuable. But you're still stuck with how

15:58

do you get all. to come to the

16:01

website, right? That's still just some fraction

16:03

of a search audience or some fraction

16:05

of conversions from one of the social

16:07

video platforms. And that, this is the

16:09

math, that seems like the upper bound

16:11

of your growth, right? Because the people

16:13

have, some number of people have to

16:16

come to the website, some number of

16:18

people have to choose to transact on

16:20

a video from the tri-guies, and that

16:22

can only grow insofar as all of

16:24

those individual customers can get people to

16:26

websites. Do you see that the web

16:28

is the limiter in that way? No, not

16:30

at all. I've worked for Google,

16:32

Amazon, and Microsoft in my life.

16:35

You know, most recently, you know,

16:37

Google, I've worked in all manners

16:39

of businesses and data problems. And

16:41

I have this foundational philosophy that

16:43

there's actually more data behind firewalls

16:45

and paywalls and there is in front.

16:48

There's more information behind those firewalls and

16:50

paywalls than in front. And you know,

16:52

when I take a look at the

16:54

enterprise market for video, or you might

16:56

have like a couple product videos like

16:58

for e-commerce, or you might have, you

17:00

know, for example, the CEO's message. Video

17:02

is coming to literally every single element

17:04

of business. In the same way, it's

17:06

82% of the internet. It's coming in.

17:08

And so whether or not it's that

17:10

e-newsletter, whether or not it's for sales,

17:12

like if you look at an organization

17:14

like seismic or gong, it records sales

17:16

calls, and then it helps to

17:18

coach individuals. If you watch a

17:21

video, you're 67% more likely to

17:23

buy a product. And so we've

17:25

got very large e-commerce customers where

17:27

they now have millions of videos

17:29

on us that are serving to

17:31

every single product on their website.

17:33

And so what I'm seeing is

17:35

quite frankly that there is an

17:37

explosion of video. It's such an engaging

17:39

media. Like when you watch a video,

17:41

you have 91% better retention than when

17:43

you read something. So what I'm seeing

17:45

is that a lot of the stuff

17:47

that's behind the firewall and behind the

17:49

paywalls now getting video enabled and that

17:51

it's going across every single division inside

17:53

of an organization. And it actually dwarfs

17:55

what I'll call kind of a lot

17:57

of the content. We're going to see

17:59

video. show up in so many different ways

18:01

and in so many different businesses. Like people

18:04

are starting to use video to determine efficiencies

18:06

like inside a quick service restaurants. They're starting

18:08

to use video to be able to evaluate

18:10

what's on a shelf and whether or not

18:13

there's a stockout in the shelf. And so

18:15

when I think about this, I don't think

18:17

about it just in terms of like one

18:19

segment. of our organization, I actually, you know,

18:22

the beauty of Vimeo is that we're able

18:24

to live inside and outside the firewall. And

18:26

there's very, YouTube does not live inside of

18:28

the firewall. Like we're able to hook in

18:31

and sign a BAA, you know, like an

18:33

agreement to do HIPAA for a doctor. YouTube's

18:35

not going to do that. And then you

18:37

think about all the interactions in the healthcare

18:40

industry, you know, that actually can be video

18:42

enabled. And so. Our upper bound of growth

18:44

is, you know, I actually kind of feel

18:46

like it's a larger opportunity than what YouTube

18:49

is focused on right now. I see, I

18:51

mean, YouTube right now is focused on video

18:53

podcasts. It's like very, they've picked their shot

18:55

and they're going to take it. By the

18:58

way, every time anyone says that stat about

19:00

video retention, I feel like a dinosaur because

19:02

I need to read for as much video

19:04

podcasting as I literally do. I'm a reader

19:07

to the core. I'm an underliner. I'm an

19:09

underliner, right? I learned to highlight in five

19:11

colors in law school. It's still where I'm

19:13

at in maybe the future is video and

19:16

that's why we do a video podcast, but

19:18

still a reader to my core. There's like

19:20

three parts of just sort of the video

19:22

business. We've talked a lot about distribution where

19:25

you might distribute that video. It sounds like

19:27

that's where you think there's a lot of

19:29

growth across organizations even to consumer in some

19:31

new way. Then there's monetization, which I don't

19:34

I get to. But the first part is

19:36

the hardest, and I think, undergoing the most

19:38

change in terms of what we expect videos

19:40

to be, and that's obviously creation, right? You

19:43

need to make a video, you need to

19:45

distribute it, you need to monetize it. Creation

19:47

of video right now is super interesting because

19:49

you have in particular, not just the young

19:52

generation, but everybody learning to speak the language

19:54

of TikTok, I think is most importantly expressed

19:56

to people as a video editor, not just

19:58

a scrolling video editor. Then

20:01

there's AI, which is making it a

20:03

lot easier to make all kinds of

20:05

videos and all kinds of ways. And

20:07

then there's like the sort of like

20:09

smaller AI components, right? That it's going

20:11

to write a script for you that

20:13

you can read. And like, maybe that's

20:15

good, maybe that's bad. But it's all

20:17

just in the mix, right? Like everyone

20:20

is expecting the tools to guide them.

20:22

Are you thinking of that component of

20:24

it? Like we need to build an

20:26

enterprise tick-talk editor for people in order

20:28

to just bring them in the pipeline?

20:30

You know, I think there's a couple

20:32

dynamics that are happening right now, and

20:34

this is what gets me so excited

20:36

about this one. It's one of the

20:38

biggest things that brought me here, is

20:41

the barriers to video creation are dropping

20:43

so dramatically, which leads to that mass

20:45

proliferation of video, and then the difficulty

20:47

in being able to manage at that

20:49

scale. Like that's just foundation like the

20:51

market forces that are behind us. I

20:53

always pause for a second. I always

20:55

tell people, you know, I can, I'll

20:57

be able to talk to you for

21:00

a long time about artificial intelligence in

21:02

a couple seconds, but let me talk

21:04

to you about like what's happening in

21:06

video formats. You're 100% right. Right now,

21:08

mobile video, we're kind of in that

21:10

error where mobile video is becoming much

21:12

easier. People were becoming much easier. People

21:14

were becoming much easier. People were becoming

21:16

more comfortable. Culturally people are getting much

21:18

more comfortable shooting video. The proliferation of

21:21

tools has been extraordinary. Now we made

21:23

some acquisitions of the past, you know,

21:25

like Magista was an example of this

21:27

because we really felt that we had

21:29

to make it easier and easier to

21:31

be able to create video. Well, I

21:33

have been thrilled. with proliferation of tools.

21:35

We shot a video for our reframe

21:37

conference, I don't should say shot a

21:40

video, we actually like created a video

21:42

using 16 AI tools that didn't exist

21:44

18 months ago and it was over

21:46

$15 billion of venture capital had gone

21:48

into. creating those tools. That's just one

21:50

small set of tools. But you're absolutely

21:52

right is that you've got all these

21:54

tools that are being created. And so

21:56

we're thrilled about that. But simultaneously, the

21:58

format of video is also proliferating. And

22:01

so you've got traditional like 1068, you've

22:03

got 4K that's starting to become more

22:05

commonplace. 8k is arriving and when you

22:07

do 8k it's roughly about six times

22:09

the size of a 4k well 16k

22:11

and 32k televisions are on the horizon

22:13

right now you've got wide stream formats

22:15

you got square formats you know for

22:17

podcasting you got rectangular formats and then

22:20

you know we just recently released Apple

22:22

Vision Pro support to be able to

22:24

stream on Apple Vision Pro which is

22:26

8k per I 36 frames per second

22:28

and so While simultaneously, well, the tools

22:30

are proliferating, the format types are also

22:32

proliferating. And so your ability to be

22:34

able to both accept video from any

22:36

format, you know, in some cases you

22:38

accept something that's an old old format,

22:41

you know, that you have to have

22:43

improved or whether or not it's a

22:45

super high quality, giant wide screen format

22:47

that needs to be cut up for

22:49

all the different areas that you're going

22:51

to serve that video. What I would

22:53

tell you is that it's becoming more

22:55

complex on the creator about which tool

22:57

to use when and then how do

23:00

I ensure that the right format gets

23:02

served at the right moment. And so

23:04

the two simultaneously simultaneous things that are

23:06

happening in our in our business are

23:08

creative tools and formats are just are

23:10

exponentially growing right now. You're exponentially growing

23:12

the amount and a video that a

23:14

creator has to deal with. What's going

23:16

the fastest? You know, it's really interesting.

23:18

As you can imagine, I think square

23:21

format is popping up a lot. We

23:23

are seeing a lot of demand for

23:25

4K, you know, in live formats and

23:27

in in serving formats. I think people

23:29

are, you know, people are starting to

23:31

use, demand that format more, which is

23:33

obviously for us, we have to move

23:35

more bits, we have to store more

23:37

bits, we've got to, you know, transcode

23:40

more bits. And so I would tell

23:42

you, that's probably the thing that we're

23:44

seeing that we're seeing spiked the most

23:46

the most in terms. of like consumption.

23:48

Like the traditional mobile stuff is going

23:50

to be there and it's going to

23:52

be constant. I think it's kind of

23:54

almost like growing at the speed of,

23:56

you know, just so-called mobile phones basically,

23:58

but like I'm actually surprised about how

24:01

many people are coming to us asking

24:03

for 4K. What do you think that

24:05

is? I look at the broader industry

24:07

and you see the big streamers are

24:09

pushing everybody to basically 1080P with ads.

24:11

Like that's sort of the default for

24:13

Max or Netflix and then you got

24:15

to pay extra for 4K. Are you

24:17

seeing that demand in the same way

24:19

they do, which is people will pay

24:22

extra for it? Or are you seeing

24:24

that demand as this is now the

24:26

understood industry norm? It may be the

24:28

place that we sit in the industry.

24:30

You know, as I mentioned at the

24:32

top of our talk was that people

24:34

have always come to us for quality.

24:36

You know, and so it may just

24:38

be that because we actually are, you

24:41

know, we've been known for quality, we've

24:43

been known for the quality of our

24:45

stream, the quality of the serve that

24:47

we do. It may just be that

24:49

people are not finding that kind of

24:51

support elsewhere and coming to us for

24:53

it. But I would just say I

24:55

think that people are experimenting with those

24:57

formats. I've been pleasantly surprised with the

24:59

sheer number of people that have come

25:02

to us since we've launched the Apple

25:04

Vision Pro and are coming to us

25:06

with really interesting film projects to do

25:08

8K per eye, you know, stitching all

25:10

the camera work together. So there's also

25:12

some like really interesting stuff and you'll

25:14

see us talk a lot about this

25:16

at South by Southwest, about like what

25:18

we think, you know, I'm seeing some

25:21

good excitement in those 8K formats as

25:23

well, as all I'll say. But it

25:25

might just be the position that we

25:27

sit in the position that we sit

25:29

in the industry. It's so interesting to

25:31

see the rest of the industry basically

25:33

insist that consumers don't care about 4K.

25:35

You and I are talking two days

25:37

before the Super Bowl, and for all

25:39

of Fox's talk about 4K, they're still

25:42

producing that in 1080P and then upscaling

25:44

it. And it's kind of fascinating to

25:46

see the consumer side of the market

25:48

land at one sort of standard quality

25:50

level while you're saying the enterprise of

25:52

the market, the more discerning part of

25:54

the market, is now not just assuming

25:56

that 4K it will exist, but that

25:58

you will support. 8K per I, 36

26:01

frames per second in the vision prep.

26:03

When there's 16K for TVs out, I

26:05

think people will be buying them. Is

26:07

that, does that, that's a lot of

26:09

cost, right? I mean, you're talking about

26:11

moving an enormous amount of data. Are

26:13

you just getting ahead of it? Because

26:15

that's what the customers expect? You have

26:17

a background in cloud services and big

26:19

data. Like, is that something where you

26:22

say, okay, this is just scalable, like

26:24

we can just solve this problem with

26:26

the tools we have? Or is it

26:28

we have to re-archate the systems? Or

26:30

is it we have to re-archate the

26:32

systems? I'll use the corollary of what's

26:34

going on with token. is going to

26:36

come down over time. The cost of

26:38

storage comes down, the cost of bandwidth

26:41

that comes down, and then even the

26:43

innovations that are in the televisions, those

26:45

costs are going to be coming down.

26:47

When you think about the quality of

26:49

the TVs we have now versus even

26:51

just 10 years ago, it's like so

26:53

discernably different. And I think that as

26:55

those costs come down, you know, somebody

26:57

has to serve that content. and you

26:59

know our infrastructure we have the infrastructure

27:02

to be able to do it. And

27:04

so some of us we have we

27:06

have to stay slightly ahead so that

27:08

we are that place that's always viewed

27:10

as quality. And so yeah I guess

27:12

it's just it's part of our it

27:14

has to be part of our DNA

27:16

that we're always going to support those

27:18

cutting edge formats. We need to take

27:21

a quick break. We'll be right back.

27:26

Support for this show comes from

27:28

Attentive. Millions of marketing messages are

27:31

sent out every day. So if

27:33

you're trying to make sure your

27:35

business's message actually gets delivered and

27:38

sticks in customers' minds, you might

27:40

want to try Attentive. Attentive is

27:43

the SMS and email marketing platform

27:45

designed to help brands build and

27:47

connect with their ideal audience. It

27:50

helps marketers create unique messages for

27:52

every subscriber transforming the consumer shopping

27:55

experience and maximizing marketing performance. Here's

27:57

how it works. Attentives AI learns

27:59

what subscribers actually want based on

28:02

their real-time interactions with your brand.

28:04

That means it customizes the content,

28:07

tone, and timing of every message

28:09

so they always resonate. You can

28:11

send out truly personalized messages at

28:14

the perfect times, and you'll be

28:16

able to keep supporting customers throughout

28:19

the buying process with conversational AI

28:21

that's actually helpful. All in all,

28:23

when you use attentive, you can

28:26

connect with the right customers because

28:28

they'll feel a connection with the

28:31

messages you send. For messages that

28:33

perform and results that transform, check

28:35

out attentive. Visit attentive.com/decoder to get

28:38

started. Support for the show comes

28:40

from Vanta. Trust isn't just earned.

28:42

It's demanded. Whether your startup founder

28:45

navigating your first audit or season

28:47

security professional scaling your GRC program,

28:50

proving your commitment to security has

28:52

never been more critical or more

28:54

complex. That's where Vanta comes in.

28:57

Vanta says they can help businesses

28:59

establish trust by automating compliance needs

29:02

across over 35 frameworks like SOC

29:04

and ISO 27001. They also centralized

29:06

security workflows, complete questioners up to

29:09

five times faster, and proactively manage

29:11

vendor risk. In short, VANA wants

29:14

to save you time and money,

29:16

and a new IDEC white paper

29:18

found that VANA customers achieved $535,000

29:21

per year in benefits, and the

29:23

platform pays for itself in just

29:26

three months. You can join over

29:28

9,000 global companies like Atlacian, Cora,

29:30

and Factory who use Vanta to

29:33

manage risk and prove security in

29:35

real-time. For a limited time, our

29:38

audience gets $1,000 off, vanta.com/Dakota, for

29:40

$1,000 off. It's been reported that

29:42

one in four people experienced sensory

29:45

sensitivities. making everyday experiences like a

29:47

trip to the dentist is especially

29:50

difficult. In fact, 26% of sensory

29:52

sensitive individuals avoid dental visits entirely.

29:54

In sensory overload, a new documentary

29:57

produced as part of Sensodyne's Sensory

29:59

Inclusion Initiative, we follow individuals navigating.

30:01

I've only had two full-time jobs

30:04

where I felt safe, loud sounds,

30:06

and unexpected touches can turn routine

30:09

moments into overwhelming challenges. Burnett Grant,

30:11

for example, has spent their life

30:13

masking discomfort in workplaces that don't

30:16

accommodate neuro divergence. I've only had

30:18

two full-time jobs where I felt safe,

30:20

they share. This is why they're advocating

30:22

for change. Through deeply personal

30:25

stories like Burnett's, sensory overload

30:27

highlights the urgent need for

30:30

spaces, dental offices and beyond,

30:32

that embrace sensory inclusion. Because

30:34

true inclusion requires action with

30:37

environments where everyone feels safe.

30:39

Watch Sensory Overload now, streaming

30:42

on Hulu. We're

30:47

back with Vimeo CEO Philip Moir. Before the

30:49

break, we were discussing the kinds of

30:51

opportunities that still exist for online video

30:53

distribution and the types of tools that

30:56

are dramatically changing video creation. But before

30:58

we really got into AI, I wanted

31:00

to ask the big decoder questions about

31:02

structure and decision making and how Philip

31:04

is changing Vimeo now that he's in

31:06

charge. We've opened the door to AI and I

31:08

definitely want to talk about that and... in

31:10

particular using AI as a creative tool and

31:13

how your customers might be thinking about that

31:15

and their relationship to it. But first of

31:17

all, I just want to get to the

31:19

decoder questions. We've talked a lot about how

31:21

you're thinking about growing Vimeo's business. What

31:23

would you say is the most tangibly

31:25

different thing you're doing than your predecessor,

31:27

Anjali? There's a couple things. Anjali, she

31:29

was supporting a lot of different businesses. I'll

31:32

say as you went through COVID. And as

31:34

you went through kind of when when video

31:36

was hard and I would say it's it

31:38

wasn't as culturally ingrained as it is right

31:40

now I think you know she had to

31:42

make a lot of decisions around the business

31:44

when I got here a lot of people

31:46

asked me this question like well do we

31:49

serve the consumer or do we serve the

31:51

enterprise do we serve the filmmaker or do

31:53

we serve the physician when I really spent

31:55

time down with you know who our customer

31:57

was like I really had to get deep

31:59

deep deep. inside and go, you know, who

32:01

really uses this? Show me the type of

32:04

company, show me the names of the companies,

32:06

the industries that we're in. You know, it

32:08

was very clear to me that we serve

32:10

the creator that is professional, somebody that is

32:12

using video for their business, like professionally,

32:15

it's not a hobby, it's actually like

32:17

to get a job done. You know,

32:19

so being able to consensus around that

32:21

creative pro. not trying to go and

32:24

create a YouTube competitor or trying to

32:26

create, you know, create a low-cost, you

32:28

know, tool for the hobbyist, but truly

32:30

that we serve that professional creator. And

32:33

then being able to describe very succinctly,

32:35

you know, the fact that all of

32:37

that comes together, that sometimes a professional

32:39

creator, and then being able to describe

32:42

very succinctly, you know, the fact that

32:44

all of that comes together, that sometimes

32:46

a professional creator, or square formats. I

32:48

think one of the core things I

32:50

did when I got here was really

32:53

obsessed around the customer. Every meeting we

32:55

start, we start by telling a customer

32:57

story. I learned a lot of this,

32:59

I would tell you between Google and

33:01

Amazon, you know, Amazon is, I would

33:03

say, renowned for this, but we really

33:06

tell the stories of who our creators are.

33:08

And then really building the ability to be

33:10

able to move quickly and listen to those

33:12

enterprise requirements. You know, when you start a

33:14

company, and I... I started at Microsoft in

33:17

the very early days of Microsoft. It wasn't

33:19

an enterprise company then, believe it or not.

33:21

Amazon, you know, when I got there, there

33:23

weren't a lot of financial services companies using

33:25

the cloud. I went through that transition and

33:28

Google, I kind of, you know, it wasn't

33:30

really known as like an enterprise-grade cloud when

33:32

I first got there. And so I've been

33:34

through this transition and when you start a,

33:36

you know, when you start a product that

33:39

is going to serve the enterprise, you know,

33:41

what I always tell people is easier to

33:43

get more complex, but it's really hard to

33:45

be complex and get easier. And so we

33:47

were starting from these routes as a kind

33:49

of a consumer and filmmakers product, and a

33:52

lot of what I've focused on is really

33:54

listening quickly, you know, to not just our

33:56

individual customers, but like the entire spectrum and

33:58

being able to say yes. to the requirements.

34:00

I also come with a tremendous amount of

34:03

experience. As you can imagine, you could look

34:05

at my background. It's years of enterprise experience.

34:07

And so I also, I know, you know,

34:09

what's required to be able to do that.

34:12

I understand what it is to be able

34:14

to GDPR. I understand compliance requirements. And so

34:16

when we go into things like artificial intelligence

34:19

or we go into storage and distribution, like

34:21

I have, you know, a lot of instincts

34:23

around that. And so spending the time to

34:25

explain. where we're going as a company to

34:28

be able to serve both inside and outside

34:30

the firewall and then the requirements that

34:32

we're going to have architecturally and just

34:34

explaining that to the organization and weaving

34:37

together. Hey, that filmmaker once their content

34:39

protected the same way that may be

34:41

the largest, you know, retailer in the

34:43

world or that CEO wants their content

34:46

protected and weaving those two messages together

34:48

and building a product roadmap is going

34:50

to serve both. I would tell you.

34:52

I feel like probably the biggest thing

34:55

that I've probably done since getting here

34:57

is unify the vision into a single

34:59

cohesive vision. And then the second thing is

35:01

that really making sure that all of us

35:03

are telling the stories, hey, did you know

35:05

that this fertility clinic is using us in

35:07

this way? Did you know that this schoolteacher

35:09

is using us in this way? Did you

35:11

know that this faith organization uses us in

35:13

this way? Do you know that the largest

35:15

retailer uses us this way? Getting people to

35:17

tell the stories internally, I think was an

35:19

important. The last thing I'll tell you is

35:21

that, you know, the company kind of like

35:23

there was a little bit of a shine

35:25

that came off the company after the IPO.

35:28

And I think, um, given the company

35:30

confidence. and saying, let me tell you

35:32

who's actually using us. I don't think

35:34

that a lot of people really realized

35:36

internally even the broad array of customers.

35:38

And I put this in our shareholder

35:41

letter where we have, you know, eight

35:43

out of the top eight big box

35:45

retailers, you know, eight of the top

35:47

eight big box retailers, you know, we've

35:49

got air on the top eight media

35:51

companies that are using us internally, we've

35:54

got, you know, huge numbers of insurance,

35:56

and financial tool in a world

35:58

of increasing complexity. I think

36:00

it's given the company a lot more confidence

36:02

to take to be even bolder as we

36:04

go forward. You're talking a lot

36:06

about culture change, renewed focus. Kind of

36:08

the thesis of this show is that

36:10

comes out of structure. I'm just looking

36:12

at our notes here from the producers.

36:15

In the past few months, you've named

36:17

a new chief marketing officer, a new

36:19

chief information security officer, a new people

36:21

officer, a new revenue officer, you're obviously

36:23

making some changes in the organization. How

36:25

was Vimeo structured and how are you

36:27

restructuring it? You know we had in some

36:29

cases our technology organizations were incredibly siloed.

36:32

How we did trust and safety as

36:34

an example you know in some cases

36:36

how we did data in some cases

36:39

who owned what parts of engineering they

36:41

were actually broken up in a

36:43

lot of ways. I think oftentimes our

36:45

marketing organization had a mandate while the

36:47

sales organization might have had another mandate.

36:49

A lot of what I would tell

36:51

you what was really important to me,

36:53

like one of my first hires was

36:56

the chief technology and product officer, you

36:58

know, Bob Petroselli, bringing an individual in

37:00

that unifies product engineering and, you know,

37:02

is able to have like the single-threaded

37:04

leadership over top of parts of the

37:06

business, but are that important. They have

37:08

to be working together well. some of

37:10

the things that we're doing in trust

37:12

and safety, we actually think we can turn

37:15

around and expose that to our enterprise customers

37:17

because it turns out they probably don't want

37:19

things on their platform the same way that

37:21

we don't want certain things on our platform.

37:23

And so getting that kind of like any

37:25

internal function can become an external function and

37:27

getting that kind of view that we all

37:29

serve the customer in some way shape of

37:31

form, super important inside of our product and

37:33

technology organizations. One of the most important things

37:36

I had to do as well, you know,

37:38

our chief marketing officer that came in, you

37:40

know, he has a really extensive experience in

37:42

Charlie Ungersick in being able to both do

37:44

marketing to individuals as well as to enterprise.

37:46

We are going out and talking about how

37:48

we can protect videos as well as serve

37:51

videos as well as, you know, as well

37:53

as provide AI to that entire audience. And

37:55

so we had to get an individual that

37:57

was able to oversee both parts of the

37:59

business. recent restructuring where we put an

38:01

individual completely in charge what we call

38:03

our self-service business and to be able

38:05

to move even faster in that part

38:08

of the business and obsess on everything

38:10

from the top of the funnel all

38:12

the way through you know when a

38:14

subscriber comes in right down to what

38:16

are we actually using in the product

38:18

and this is a big element as

38:20

well. We will tell you one of

38:22

the other change they didn't mention is

38:24

that we're obsessing on use like right

38:26

down to the feature level. where I

38:28

look at those kind of reports on

38:30

a weekly basis, like how many people

38:32

are using our edit feature, how many

38:34

people are using our live feature, how

38:36

many people applied permissions to a video,

38:38

you know, did that increase week over

38:40

week? What did we do? And so

38:42

that self-service leader is really now single-threaded

38:44

leader. And we just, you know, we

38:46

had as well a single-threaded leader around

38:48

our streaming business and we really started

38:50

seeing, you know, some of the results

38:52

from that internally inside of the company.

38:55

So giving single-threaded leadership, I will tell

38:57

you, it is talked about a lot,

38:59

but oh my God, it's beautiful to

39:01

actually be able to call up somebody

39:03

that owns the number that owns the

39:05

resources that worries about it as much

39:07

as you do, like every single-threaded leadership

39:09

is an Amazon concept. You've worked at

39:11

all the companies, so I can pick

39:13

out where the concepts come from. It's

39:15

pretty fun for me. That's a classic

39:17

Amazon concept. You need to have a

39:19

pretty small team that owns a thing,

39:21

and there's a leader who's responsible for

39:23

the whole stack. That is how you

39:25

get silos. You can look at Amazon's

39:27

product and say, oh, there's a much

39:29

single-threaded leaders here. This is not necessarily

39:31

cohesive. Everything is running as fast as

39:33

it can, but the holistic as it.

39:35

You started out talking about having too

39:37

many silos and we're talking on single-threaded

39:39

leaders. How are you managing that tension?

39:42

I was asked one time to give

39:44

a talk on what it was like

39:46

to work because you know at Microsoft

39:48

Amazon and Google, you know, I got

39:50

an opportunity to work directly with Gates

39:52

and Balmer, directly with Jesse and then

39:54

certainly with Thomas Curry and Sundar. One

39:56

of the most important lessons I learned

39:58

very early on, you know, at Microsoft

40:00

was around having like really establishing a

40:02

strong single vision for the entire company

40:04

about what we're trying to do in

40:06

the future and you know you wake

40:08

up every single morning and I mean

40:10

I was there in the early days

40:12

when the vision was a PC on

40:14

every desktop in every home. Now that

40:16

was like extraordinary the time now we

40:18

have a PC in every you know

40:20

pocket at this point but We all

40:22

knew that what we were trying to

40:24

do was unlock information for the world

40:26

by putting this powerful computing device in

40:29

someone's hands. And so regardless of the

40:31

the divisions or otherwise, it all feathered

40:33

into a common vision. And it was

40:35

a lot of what I had to

40:37

do when I was when I got

40:39

here was I owed, you know, the

40:41

company a strong agreement among everybody in

40:43

the company of what we're trying to

40:45

build. You know, are we trying to

40:47

build the best live stream product? Are

40:49

we trying to build the best marketing

40:51

platform? Are we trying to build just

40:53

a product for filmmakers? You know, we

40:55

settled on this common vision. And then,

40:57

you know, being able to say, okay,

40:59

this is the individual, you know, that

41:01

owns this part of the business. There's

41:03

a huge portion from our individual business

41:05

where people swipe a credit card and

41:07

start using us or that register for

41:09

free. Huge number of those customers actually

41:11

end up as enterprise customers. You know,

41:13

You know, I started talking to him

41:16

about, you know, Vimeo. He said, well,

41:18

first of all, he goes, you don't

41:20

have to tell me who you are.

41:22

He goes, my son is on your

41:24

on Vimeo every weekend. He says, an

41:26

independent filmmaker. He goes, so I know

41:28

who you are. Why are you calling

41:30

me? And I said, we have 2,600

41:32

accounts, self-service accounts that are on us.

41:34

We should do an enterprise agreement. And

41:36

so being able to explain to the

41:38

organization to the organization. between where we're

41:40

applying more features, maybe in one part

41:42

of the business or another, and how

41:44

those features actually feather, how we might

41:46

start them for an individual, but they

41:48

have to grow to be, you know,

41:50

for an entire enterprise, really, really, really,

41:52

really important. And so I would tell

41:54

you these, starting with a strong vision

41:56

that everybody buys into, that they understand

41:58

their piece of it, really critical. And

42:00

then each one of those, those leaders,

42:03

I expect them to have a, another

42:05

important that you can't let their vision

42:07

exist in the absence of the rest

42:09

of the company's vision. So you have

42:11

to actively stitch those visions together. You

42:13

brought up decisions, that's the other classic

42:15

decoder question. I will warn you, this

42:17

is a honeypot for former Amazon executives,

42:19

because when you ask Amazon executives how

42:21

they make decisions, everybody sings chapter and

42:23

verse. But you've been a bunch of

42:25

places you are now the CEO of

42:27

this company. How do you make decisions?

42:29

What's your framework? The other company I

42:31

didn't talk about that I learned a

42:33

lot from was Google. And one of

42:35

the things that I would tell you

42:37

that Google gave me was I think

42:39

they've managed something like 10 out of

42:41

the top 11 billion user products in

42:43

the world and really like thinking big

42:45

actually like giving an organization like incredibly

42:47

lofty goals that sometimes you only make

42:50

like 80 to 90% of them. One

42:52

of the things that I do, you

42:54

know, first and foremost is that I,

42:56

you know, I really am a believer

42:58

that you've got to set these very

43:00

high goals. You've got to have this

43:02

vision and you've got to be willing

43:04

to like put yourself out there to

43:06

set extremely high goals. And then backing

43:08

into that from a decision-making process, I

43:10

would tell you that we've had to

43:12

make some a number of decisions around

43:14

what products we focus on, what areas

43:16

we deprecate. I come back to the

43:18

customer. And one of the things I

43:20

really try to hold people accountable to,

43:22

and I think it's really important, I

43:24

think I learned a lot of this

43:26

both at Google and at Amazon, but

43:28

actually explaining the customer problem that we're

43:30

trying to solve. And there's all kinds

43:32

of studies that you can do. There's

43:34

user studies, there's data and so forth,

43:37

but truly being able to assess on

43:39

what that workflow looks like, you know,

43:41

what are we trying to solve? What

43:43

is the most challenging thing for the

43:45

customer for the customer most? And really

43:47

having a strong sense for your customer

43:49

and the customer anecdotes as well as

43:51

what we call, you know, Google, we

43:53

call it a customer empathy, like actually

43:55

putting yourself in the shoes of the

43:57

customer. One of the things that we

43:59

ask all of our everyone inside of

44:01

being able to do is be a

44:03

user of the product. And so the

44:05

things that are frustrating us, we actually

44:07

are elevating those things into our decision-making

44:09

process. And so we both... bring the

44:11

voice of the customer in, we bring

44:13

our own voice in, and then we

44:15

also are saying, okay, well, what's going

44:17

to help us grow, you know, what's

44:19

going to grow the next million users

44:21

for us, or what's going to grow

44:24

us to 10X. And so I can't

44:26

just kind of tell you it's like

44:28

one thing. It's a little bit of

44:30

a framework of the customer, making sure

44:32

we're tethered to big ideas and really

44:34

make sure that we're being innovative enough

44:36

in how we push the team. Well,

44:38

I applaud you for being the only

44:40

former Amazon executive to not talk about

44:42

one way and two way doors when

44:44

I ask that question. You've done it.

44:46

You've done it. You've achieved escape velocity.

44:48

Well done. I mean, I appreciate the

44:50

one way and two way doors. I'm

44:52

just saying. I don't know that I

44:54

buy that. I mean, like, you know,

44:56

Google, I think they've proven that there's

44:58

not a lot of one way doors.

45:00

Fair enough. Google is. Come out against

45:02

YouTube, but you have entire blog posts

45:04

about how your search capabilities are better

45:06

than YouTube search capabilities or you're a

45:08

better platform You've talked even in this

45:11

episode about wanting things to be private

45:13

not being part of the algorithmic ecosystem

45:15

or the advertising ecosystem on YouTube Google

45:17

is big. They think so big that

45:19

sometimes they let opportunities just slide away

45:21

because they think it's such massive scale.

45:23

How are you thinking about competing with

45:25

a YouTube at that scale when they

45:27

seem to own so much of the

45:29

attention space and video? I'll say it

45:31

again. I don't think that a lot

45:33

of product companies love the fact that

45:35

you got to go to YouTube to

45:37

get some customer support for one of

45:39

their products and meanwhile one of their

45:41

competitors could be rolling right next to

45:43

you. When I look at YouTube, I

45:45

was at Google and I spent a

45:47

lot of time with customers and I

45:49

really foundationally believe, like I love YouTube.

45:51

I'll watch, you know, YouTube as much

45:53

as the next person. And I think

45:55

that what they're doing, you know, for

45:58

all call it kind of the attention

46:00

economy, for what they're doing around content

46:02

for democratizing access to more and more

46:04

content, I think it's absolutely wonderful. a

46:06

lot of our customers are great YouTube

46:08

customers as well. You know, people will

46:10

house their videos on video and post

46:12

on YouTube in a lot of ways.

46:14

But I really do think, you know,

46:16

there is in the same way that,

46:18

you know, you don't do a lot

46:20

of your business on Facebook, or you

46:22

don't do it on LinkedIn, you kind

46:24

of do it behind closed walls. I

46:26

think that a lot of the economy

46:28

runs behind fire walls and pay walls.

46:30

And so I think that we can,

46:32

we can go directly at that. Think

46:34

about what happened in content and why

46:36

some of these platforms rose. You know,

46:38

think about, and I, you know, again,

46:40

I'm old enough to have remembered MSN

46:42

and AOL. The reason why we went

46:45

there was because news had to be

46:47

consolidated. It was hard to create websites.

46:49

It was difficult to find information. It

46:51

had to be curated. Well, you know,

46:53

Netflix and YouTube were born in an

46:55

error where it was really hard to

46:57

categorize content. to say, hey, this video

46:59

is about a cat. This video is

47:01

about how to, I don't know, plug

47:03

a HDMI cord in the back of

47:05

your LGTV. And so there was categorization

47:07

that had to take place. There was

47:09

standardization of the data and the metadata.

47:11

And then recommendations, I don't know if

47:13

you remember them, but Netflix famously paid

47:15

a million dollars to be able to

47:17

write their recommendation engine that they went

47:19

out and said, whoever builds the best

47:21

recommendation engine will win a million dollars.

47:23

you know with an AI model I

47:25

can like categorize content in seconds now

47:27

with a recommendation engine I can buy

47:29

recommendation engines off the shelf and quite

47:32

frankly the metadata that we can produce

47:34

now out of a video is extraordinarily

47:36

more detailed than a human being can

47:38

even write like I can tell you

47:40

precisely when the purse like went left

47:42

the beach and like you know who

47:44

was carrying the person what the brand

47:46

was of shoes that the individual was

47:48

wearing you know all stuff that maybe

47:50

missed in a traditional human being after

47:52

entering all that metadata. And so what

47:54

I see is that, you know, there

47:56

is going to be kind of a

47:58

democratization of content classification and content recommendation

48:00

and context discoverability. You know, searches, there's

48:02

been a single search place that you

48:04

go to be able to get your

48:06

content and your videos and your maps

48:08

and, you know, pick your favorite things.

48:10

But I think that there's billions of

48:12

dollars going into another part of, you

48:14

know, in other ways to be able

48:16

to find and interact with information. And

48:19

so. I think that Vimeo can serve that kind

48:21

of information outside of like a traditional Google

48:23

search in a lot of ways. Whether or

48:25

not you're inside an internet, inside of a

48:27

company, whether or not you're inside an AI

48:30

model, you don't want to leave the ecosystem

48:32

of the AI model, you know, we can

48:34

provide an answer to a question as well.

48:36

And so I guess what I'd say to

48:38

you is that I think YouTube is fantastic,

48:41

I love it. But I think that

48:43

the discoverability, accessibility, indexing, and recommendations

48:45

is there's a whole new error

48:47

coming and we intend to be

48:49

part of it. There's another piece of

48:51

that dynamic that I'm wondering if you are

48:53

considering, which is a Google is a huge

48:56

company that is under an enormous

48:58

amount of pressure right now. And maybe

49:00

so much pressure that it will be

49:02

hard for the company to execute, right?

49:04

There's an antitrust trial in this country

49:06

that has resulted in the government suggesting

49:08

that Google break itself up and sell

49:10

chrome. There's a Donald Trump in the

49:12

mix who makes some sort of deal.

49:14

There's a Donald Trump in the mix

49:16

who's done a tariffs regime with China

49:18

that has resulted in a Chinese antitrust

49:20

investigation in Google, which is amazing because

49:23

Google doesn't really operate in that country.

49:25

there's Europe exists much to the chagrin

49:27

of many of our tech companies. There's

49:29

just a lot going on, right? There's

49:31

a lot of pressure on Google to

49:33

not flex that dominance. And then there's

49:35

competitive pressure from the AI products, right?

49:38

From chat GPT, from search GPT, from

49:40

Bing to whatever extent that Microsoft believes that

49:42

Bing is a real competitor to Google. Is

49:44

that create an opening for you? Do you

49:46

see that as a real opening? Or is

49:48

that just, well, if those doors open, you'll

49:50

be ready? You'll be ready. It absolutely does.

49:52

I mean, I would tell you, I have a

49:54

lot of friends at Google. I really enjoy my

49:57

time there. You know, I don't wish them ill

49:59

in any way. I really hope that they,

50:01

you know, sail through this, you know, this

50:03

error right now of challenge for them, like

50:05

in a really great way. But it does,

50:08

we, I clearly, you know, I came here

50:10

from Google because I saw the opportunity, you

50:12

know, I really, I really did see the

50:14

opportunity that we're about to go through a

50:17

seismic shift. in accessibility of information, in

50:19

new ways to access, you know, to go

50:21

and access it. Whether or not you're using

50:23

chat GPT, you're using anthropic, you know, whether

50:25

or not you're using, you know, mistral, there

50:27

are so many different ways to be able

50:29

to discover information. And, you know, the notion

50:31

of, like, the common crawl on the web,

50:33

the ability to be able to crawl the

50:35

whole web index it and then be able

50:37

to ingest it into these models, you know,

50:39

it's just kind of showing that it's democratizing

50:42

that it's democratizing that it's democratizing access to

50:44

that information access to that information, Video is

50:46

a very important element of video and you

50:48

can't imagine I mean you I

50:50

think you'd agree You can't just

50:52

imagine only one platform is going

50:54

to serve all the video answers

50:56

You know in the world, and

50:58

so that's where I just see it's

51:01

such a super opportunity for us

51:03

at Bimeon We need to take

51:05

another quick break. We'll be back in

51:07

just a minute Support

51:13

for this show comes from Oracle.

51:15

Even if you think it's a

51:17

bit overhyped, AI is suddenly everywhere,

51:19

from self-driving cars to molecular medicine

51:22

to business efficiency. If it's not

51:24

in your industry yet, it's coming.

51:26

Fast. But AI needs a lot

51:28

of speed and computing power. So

51:31

how do you compete without cost

51:33

spiraling out of control? Time to

51:35

upgrade to the next generation of

51:37

the cloud. Oracle cloud infrastructure, or

51:40

OCI. OSI is a blazing fast

51:42

and secure platform for your

51:44

infrastructure, database, application development, plus

51:46

all your AI and machine

51:48

learning workloads. OSI cost 50%

51:50

less for compute and 80%

51:53

less for networking, so you're

51:55

saving a pile of money.

51:57

Thousands of businesses have already

51:59

upgraded to OCI, including Vodafone,

52:01

Thompson Reuters, and Suno

52:03

AI. Right now, Oracle

52:06

is offering to cut

52:08

your current cloud bill

52:10

in half if you

52:12

move to OCI. For

52:14

new US customers with

52:16

minimum financial commitment, offer

52:18

ends March 31st. See

52:20

if your company qualifies

52:22

for this special offer

52:24

at oracle.com/Vox. Support

52:29

for this show comes from

52:31

ADT. From ADT comes trusted

52:33

neighbor, the new standard in

52:35

home access. Through the ADT

52:37

Plus app, easily grant and

52:40

automate event-based or scheduled access

52:42

for neighbors, friends, and helpers.

52:44

Notify trusted individuals of events

52:46

like alarms or packages and

52:48

set access windows for planned

52:50

guests or even the dog

52:52

walker without interrupting your day.

52:54

Visit adt.com. When every second

52:56

counts, count on ADT. requires

52:58

ADT complete pro-monitoring plan

53:01

and compatible devices. Copyright

53:03

2025 ADT LLC. All

53:06

rights reserved. Okay, business

53:08

leaders, are Okay, business

53:10

leaders, are you playing defense or are

53:12

you on the offense? Are you just...

53:15

Excuse me. Hey, I'm trying to talk

53:17

business here. As I was saying, are you

53:19

here just to play or are you

53:21

playing to win? If you're in it

53:24

to win, meet your next MVP. Next

53:26

week by Oracle. Net Suite is your

53:28

full business management system in one suite.

53:30

With Net Suite, you're running your accounting,

53:33

your financials, HR, e-commerce, and more, all

53:35

from your online dashboard. One source of

53:37

truth means every department's working from the

53:39

same numbers with no data delays. And

53:41

with AI embedded throughout, you're automating manual

53:44

tasks, plus getting fast insights for your

53:46

next move. Whether you're competing on your

53:48

home turf or looking to conquer international

53:50

markets, Net Suite helps you get the

53:53

W. Over 40,000 businesses businesses of already...

53:55

I made the move to Netsweet,

53:57

the number one cloud ERP. Right

53:59

now, get the CFO's guide to

54:02

AI and machine learning at netsweet.com

54:04

slash stereo. Get this

54:06

free guide at netsweet.com/stereo.

54:08

Okay guys. CEO Philip

54:10

Moyer, discussing what generative

54:13

AI might do to video

54:15

creation and how the attention

54:17

economy can possibly sustain an

54:19

even larger amount of AI

54:22

generated content. Let's talk about AI,

54:24

I want to start by again asking

54:26

sort of a mathematical question. One of

54:28

my the thesis for the year is

54:31

that the creator economy is under an

54:33

enormous amount of pressure, not just from

54:35

AI, but just from what you're describing,

54:37

this huge shift to video. You can

54:40

see that there's just an exponential

54:42

increase in supply of video in all

54:44

these platforms, right? More and more kids

54:46

are making videos, more and more people

54:48

are choosing to communicate in video first

54:50

ways, more enterprises are doing it, and

54:52

then you have AI, which is just

54:55

making it easier and easier to produce.

54:57

a massive amount of video. So the platforms

54:59

are getting flooded with supply. There's not as

55:01

much ad revenue as there is increasing video

55:03

supply. So just, you do the division,

55:05

you're like, well, the ad rates are going to

55:07

go down. And then attention is sort of fixed,

55:10

right? There's only so many people with so many

55:12

hours in a day, and presumably they do have

55:14

to eat and do productive work. So attention

55:16

is kind of fixed, right? It's just like

55:18

a fixed number that you can capture, That

55:20

all just seems like it's a

55:22

bubble that's going to pop, right?

55:24

You flood an ecosystem that has

55:26

been pretty stable for a few

55:28

years with an enormous amount of

55:30

supply, the ad rates go down, attention

55:32

stays fixed, something happens in there.

55:34

And it seems like to me,

55:37

AI is the most important component

55:39

of that, because it's the thing

55:41

that can change the economics of

55:43

the supply the fastest, right? You just

55:45

say, make 50 videos about my product

55:47

on whatever platform. Is that an opportunity

55:49

for you? That this whole sort of

55:51

the creator economy or the video creator

55:53

economy as we know it seems like

55:55

it's going to go just a pretty

55:57

basic shift in its economics? You

56:00

know, it's a huge question. To me, it

56:02

feels like the question of 2025. Like

56:04

if you ask me, what is the

56:06

version of the number in 2025? There's,

56:08

there's Elon and Toehan and there's, what

56:10

happens to the creator economy? Yeah, you

56:13

know, I mean, I think the creator

56:15

economy, we are, I think we are

56:17

reaching saturation. I mean, I don't know

56:19

what we're going to get to the

56:21

post mobile phone error, like, but this

56:23

is not a way we're going to

56:25

live for the rest of the rest

56:27

of eternity. Yes, you know, there is a

56:29

saturation point, but I also think that people are

56:32

looking for a little bit of a higher quality

56:34

experience. I think people are getting tired of like

56:36

the doom scrolling. I think the mere fact that

56:38

we name it, you know, the fact that we

56:40

are now, you know, acknowledging that we get sent

56:43

down rabbit holes. I do think that people will

56:45

like storytelling. I do think there's going to be,

56:47

you know, really different opportunities. I get asked all

56:49

the time, when will AI be able to take

56:51

my favorite book and turn it into a movie

56:54

and turn it into a movie? Think about

56:56

how wonderful that would be. Think about

56:58

being able to take your child's favorite

57:00

book and be able to turn it

57:02

into a video as an example for

57:04

them that has like an extended storyline.

57:06

I think storytelling is as old as

57:08

humanity and it's going to continue forward.

57:11

Can I just stop you there for one

57:13

second? I know the Dakota audience fairly

57:15

well. A lot of people just started

57:17

screaming at you in their car because

57:19

they think that's a bad outcome. Right. Like

57:21

I'll just, I have a young child,

57:23

right? The idea that we're going to

57:25

read the Wild Robot and then some

57:27

AI tool is going to make the

57:29

movie The Wild Robot instead of the

57:31

beautiful actual movie made by people, the

57:33

Wild Robot. I would argue that that's a

57:36

bad outcome. I think it is for a,

57:38

for a class. Like here's what I, here's

57:40

what bugs me the most right now. You

57:42

know, last year the Big Six Studios

57:44

only put out 88 movies. 88. Right,

57:46

because the economics of video have collapsed on

57:48

them, right? They don't have the distribution monopoly.

57:50

And I think there's like so many stories

57:52

to be told, you know, if the creator

57:54

of, you know, the giant robot gets paid,

57:56

you know, for having a now movie and

57:58

is able to have... monetized, you know, in

58:01

some ways should perform in a really

58:03

beautiful way. I actually think we're supporting

58:05

storytellers in a foundation away. I think

58:07

that's a decade away. I would say

58:09

maybe five to seven years away. And

58:11

so, you know, first and foremost, I

58:13

do think that AI is going to

58:15

help people create more stories. I think

58:18

they are. I think they're going to

58:20

be able to illustrate more stories. Let's

58:22

put it that way. I talked to

58:24

a lot of creative types that tell

58:26

me that like, look, you know, AI

58:28

is fairly disjoint it right now. It's

58:30

indeterministic. I don't know what I'm going

58:32

to get out of it. Human curation

58:34

of AI creation is going to be

58:37

a necessity. And, you know, the same

58:39

way that shooting a green screen and

58:41

then being able to... I'm not, you

58:43

know, when I say this, what I'm

58:45

saying is that like I do think

58:47

that longer form stories are going to

58:49

be more compelling. I think people are

58:51

going to want to stay inside of

58:53

a story a little bit longer. That

58:56

doesn't mean that the creator is going

58:58

away. It doesn't mean that the creator

59:00

is going away. It doesn't mean the

59:02

human curation is going away. I just

59:04

think that we're going to be able

59:06

to uplift them and make them faster.

59:08

So I think this is a real

59:10

tension. And I see it expressed all

59:12

the time. I've heard it from your

59:15

peers on the show when they tell

59:17

me about the tools that people use

59:19

in Photoshop, right? Generative fill in Photoshop,

59:21

according to Sean Thine Neu Ryan and

59:23

Adobe, is like 100% usage rate. But

59:25

then everyone yells about Jennifer, fill existing.

59:27

And there's a real mismatch between consumer

59:29

expectations and how people feel about AI

59:31

and then the creatives actually using the

59:34

tools at high rates. I get all

59:36

that. But I also think there's a

59:38

mismatch between you saying you're for filmmakers

59:40

and how marketers want to use AI.

59:42

I think we're careening towards a world

59:44

of basically custom creative being shown to

59:46

individual users, right? Where some brand uploads

59:48

their assets to a video platform and

59:50

literally ads get assembled for you in

59:53

AI. for you as specific user targeted

59:55

to your interests. We're headed there. The

59:57

big platform is already talking about it.

59:59

But those really commercial uses of AI,

1:00:01

we're going to make a whole bunch

1:00:03

of ads, and we're going to do

1:00:05

some of the most creative filmmaking that

1:00:07

exists. They don't seem like they're happening

1:00:09

at the same rate or with the

1:00:12

same level of acceptance, or even like

1:00:14

they should happen with the same tools.

1:00:16

And you have every piece of the

1:00:18

puzzle in front of you. Where do

1:00:20

you see the biggest pushback? One of

1:00:22

our creators, Jake Olson, recently shot Currents,

1:00:24

you know, for Applevision Pro. And when

1:00:26

you shoot for an Applevision Pro, you

1:00:28

know, you have to maintain perfect stillness

1:00:31

in the camera, you'll shoot four to

1:00:33

six cameras, and by hand, you have

1:00:35

to stitch all these things together. If

1:00:37

you get an opportunity to watch Currents,

1:00:39

it's absolutely stunning. You kind of look

1:00:41

at and go, oh my God, I've

1:00:43

seen the future filmmaking. I truly have.

1:00:45

This is where I kind of saying

1:00:47

that I'm. for watching content and creativity.

1:00:50

And I think that we'll experience film

1:00:52

in new ways. We'll experience stories in

1:00:54

new ways. I would tell you that

1:00:56

I think that the best creators are

1:00:58

going, I'm seeing them blend together, you

1:01:00

know, the content, blend together AI usage,

1:01:02

you know, with traditional techniques to make

1:01:04

something incredible. Most filmmakers that I talk

1:01:06

to start with something they've shot and

1:01:09

then enhance it with AI. One of

1:01:11

the things that is most interesting to

1:01:13

me as well is in the marketing

1:01:15

world, the thing that's taking off the

1:01:17

most right now are not avatars. And

1:01:19

avatars I like to call and like

1:01:21

nobody wants to talk to a robot.

1:01:23

It's actually people that are kind of

1:01:25

sitting there going, hey I just bought

1:01:28

this piece of furniture, this looks really

1:01:30

cool, let me show you what it

1:01:32

looks like inside of my house. And

1:01:34

it's like actually authenticity in a world

1:01:36

of robots I think is actually like,

1:01:38

like I'm already seeing some avatars. And

1:01:40

then we've also offered to them, hey

1:01:42

we have this like super like down

1:01:44

and dirty create tool where you can

1:01:46

record, we can put a teller prompt

1:01:49

or up in front of you so

1:01:51

you can like do your own script

1:01:53

and we can either make the avatar

1:01:55

look. perfect or you can kind of

1:01:57

like you know be sitting in your

1:01:59

living room and do this quick thing

1:02:01

about your product or about your service

1:02:03

and inevitably all of them go to

1:02:05

the the real human being doing this

1:02:08

and it's just I'm just going to

1:02:10

tell you point blank I'm not seeing

1:02:12

the robots take off I'm not seeing

1:02:14

it and and we've we've tried to

1:02:16

serve both and so humans have always

1:02:18

kind of risen above like they've always

1:02:20

kind of brought through the authenticity they've

1:02:22

always brought through you kind of know

1:02:24

when you're getting something and when you're

1:02:27

not like even in the chatbot world

1:02:29

you're like how many people get frustrated

1:02:31

when they're like talking you know in

1:02:33

a chatbot online they quickly want to

1:02:35

talk to a human being and so

1:02:37

I don't know how to say it

1:02:39

to you but like there's we sense

1:02:41

that there's no ghost in the machine

1:02:43

so I don't know how to say

1:02:46

it anyway. I studied artificial intelligence for

1:02:48

a long time and I'm very confident

1:02:50

in the beauty of the human soul

1:02:52

in the context of creativity. I feel

1:02:54

like I'm more cynical than you, but

1:02:56

I spend more time on the social

1:02:58

platforms, it feels like. And the problem,

1:03:00

generally, is you can sense it, some

1:03:02

people can feel it, and a lot

1:03:05

of people cannot, right? Or they just

1:03:07

let it wash over them. And then

1:03:09

you end up in sort of interminable

1:03:11

fights about metadata or labeling, or Google

1:03:13

just wrote about synth ID for images

1:03:15

that you edit and Google photos. And

1:03:17

none of that stuff seems to have

1:03:19

landed, sort of landed, right? Vimeo has

1:03:21

some labeling features, you have some ideas

1:03:24

about how you might show people AI

1:03:26

generated content or expose that meditated to

1:03:28

people. Do you think that's working? Is

1:03:30

that something you're going to continue? Is

1:03:32

that something that you think needs to

1:03:34

be expanded? You know, when I was

1:03:36

at Google, there were about 42 different

1:03:38

regulatory bodies that were working on AI

1:03:40

legislation. The last week checked, there's over

1:03:43

a thousand. One world by basis. And

1:03:45

I'm raising that kind of say that

1:03:47

like when you do translations in certain

1:03:49

states inside of the United States like

1:03:51

Illinois or Texas, you can't actually modify

1:03:53

people's lips and put words in their

1:03:55

mouth. And so that's like one of

1:03:57

the just. one of the regulations. You

1:03:59

know, over in Europe, you actually do

1:04:02

have to identify that something's been AI

1:04:04

enhanced or modified. I think we as

1:04:06

humanity are wrestling with, you know, kind

1:04:08

of like, when do we want to

1:04:10

know, you know, that something's not real?

1:04:12

The mere fact that that's coming from

1:04:14

all over the world, that you're seeing

1:04:16

that kind of like desire to know

1:04:18

when something's not real, you know, I

1:04:21

can't say whether or not that's good

1:04:23

or bad, but I can tell you

1:04:25

that it's actually a human desire. Getting

1:04:27

something done like changing the credit card

1:04:29

on my telephone bill, I'll deal with a

1:04:31

bot to do that. But if I'm like

1:04:33

really having a problem with my elderly father,

1:04:35

you know, father-in-law is having a problem, I

1:04:38

actually do want a human being to pick

1:04:40

up the phone and just talk him through

1:04:42

it. I guess what I'd say to you

1:04:44

is so I think it's filmmaking as well.

1:04:46

We've always used tools to tell our

1:04:48

story. You know, it's been the invention

1:04:51

of so many tools to be able

1:04:53

to be able to tell stories. I

1:04:55

would certainly like to know when characters

1:04:57

aren't real. This is

1:04:59

the hardest question. This is

1:05:01

like an existential philosophical question,

1:05:04

but where do you draw the line?

1:05:06

Where do you personally think the line

1:05:08

should be for when you have to

1:05:10

put the label on? I'll give you

1:05:12

two examples. I think I

1:05:15

stabilized this video with AI does

1:05:17

not require that. I think I'm

1:05:19

talking to a deep fake. avatar of

1:05:21

a creator that is photo realistic is

1:05:23

just lying to me probably requires that,

1:05:25

right? I've replaced products in this

1:05:27

movie with other products, maybe requires

1:05:30

it, right? Where do you think the line

1:05:32

is? You know, I haven't been asked this question

1:05:34

for, but as I reflect on it, you

1:05:36

know, for me, I would like to know

1:05:39

that a particular character or a particular animal

1:05:41

or something that's in that in that film

1:05:43

is actually not real. that's completely made up.

1:05:45

Now you can tell that in innovation, but

1:05:47

like in a real film, if I know

1:05:50

that a certain character or a certain scene

1:05:52

is actually like completely fabricated with a single

1:05:54

individual in it, I probably would like to

1:05:56

know that. Or when there's dialogue involved

1:05:59

or something... talking back to me, that's

1:06:01

not for real. Like, you know, I probably

1:06:03

want to know about it. You know, when

1:06:05

I look at some of the Marvel movies,

1:06:07

clearly, then you start crossing the line, well,

1:06:09

you know, does the Fox, you know, Guardians

1:06:11

of the Galaxy do it? We all know.

1:06:14

I mean, the latest Marvel movie, they're, they

1:06:16

just sounds fantastic for, they made the poster

1:06:18

with AI, and there's fan backlash to it.

1:06:20

And so I'm just wondering, do you see

1:06:22

the norms moving faster than the norms moving

1:06:24

faster than the psychology, Right, I mean you

1:06:26

were shipping these products and you have such

1:06:28

a direct line to creators. It feels like

1:06:31

you're caught up right in the middle of,

1:06:33

oh, where do we put the labels?

1:06:35

Yeah, I mean, I got to tell

1:06:37

you, like the thing that really does

1:06:39

bother me right now is like these,

1:06:41

the influencers that don't even exist. If

1:06:43

I'm looking at something that's clearly animated,

1:06:45

I'm okay with it. You know, I

1:06:47

would love to know that somebody's voice

1:06:49

was actually used for real in that

1:06:51

by that by that individual. you know, will

1:06:53

we become comfortable that basically we're being,

1:06:56

the whole thing is simply animated. Like,

1:06:58

because that's really what we're talking about. We're

1:07:00

just creating like animation that's higher and higher

1:07:02

fidelity in a lot of ways. But I,

1:07:04

you know, I think it should probably be

1:07:07

noted that at some point the human doesn't

1:07:09

exist. I kind of feel like that's probably

1:07:11

where I'd cross a line or that that

1:07:13

dog doesn't actually exist. especially the dog's a

1:07:15

main character. So you might end up doing

1:07:17

it based on the classification of the importance

1:07:20

of the character and you know whether or

1:07:22

not there's actual true existence there and like

1:07:24

how much modification was done to the individual

1:07:26

based on the classic character in the story.

1:07:28

I asked that question three different ways

1:07:31

and pushed on it again because it feels

1:07:33

like the pressure on the creator economy and

1:07:35

the social platforms is just going to

1:07:37

go up. Metah, Mark Suckerberg is out there

1:07:39

openly saying we will have AI content

1:07:41

in people's social feeds on Facebook and Instagram.

1:07:44

How they choose to label it, whether

1:07:46

or not the mean dogs have labels.

1:07:48

I don't know, I don't know what Mark Sucker work

1:07:50

is going to do. No one can see into

1:07:52

his soul. But it's pretty obvious that if he

1:07:54

could get a bunch more cute cat videos on

1:07:56

Facebook using AI, he's going to do it. Like

1:07:59

it's obvious why. he would do that. Does

1:08:01

that create opportunity for you to say

1:08:03

Vimeo is for real content or real

1:08:05

people? Is that something you would lean

1:08:08

into? Because you are, you do have

1:08:10

the AI tools. When you open the

1:08:12

website today, it says you're an AI

1:08:14

powered video platform. There's a lot of

1:08:17

conflation between we can do better classification

1:08:19

and better recommendations and we can do

1:08:21

better marketing videos and we're gonna steal

1:08:23

everyone's data and make cat videos for

1:08:26

Facebook attention spans. And appealing that

1:08:28

apart seems hard. I think I kind of

1:08:30

like as I sit here in protecting the Vimeo

1:08:32

brand I aspire to actually like be that place

1:08:34

that people trust. When I first got here we

1:08:37

were approached to basically crawl our content you

1:08:39

know as many companies were approached and we

1:08:41

talked to the creators and creators said listen

1:08:43

don't replace us and protect us make sure

1:08:45

that we're always the Vimeo is always a

1:08:48

place where we're protected and that's why we

1:08:50

come to you and that's why we want

1:08:52

to continue to come to you and so

1:08:54

we made a decision not to allow that

1:08:56

crawling. and then we very quickly like shortly

1:08:59

thereafter had to like say any time you

1:09:01

use AI on bimio that we're gonna actually

1:09:03

guarantee that none of the AI models will

1:09:05

actually improve based on your usage of it

1:09:07

unless you say improve based on my usage

1:09:10

like you know understand my storyline or understand

1:09:12

my you know my filters or understand you

1:09:14

know my dialogue my style of dialogue we'll

1:09:16

do that for you as an individual creator

1:09:18

will create your own private AI Then we

1:09:21

were actually approached as well to say, listen,

1:09:23

we need you, a number of the creators

1:09:25

said we want you to help us identify

1:09:27

like when content has been generated by AI.

1:09:30

We do a lot of business over the

1:09:32

EU and so we said, yes, we're going

1:09:34

to do that as well. You know, the

1:09:36

thing that I describe about AI was,

1:09:38

you know, back when in the era

1:09:40

of the production line, humans stopped being

1:09:42

able to keep up in a lot

1:09:45

of ways. And so we started creating

1:09:47

creating machines that turned on screws and

1:09:49

so forth. And next year at this

1:09:51

time, there's going to be more information

1:09:53

created in the next year than there

1:09:55

has in the history of mankind coming

1:09:57

up into this point. Humans are starting.

1:09:59

not to be able to keep up with

1:10:01

the production line of information. And so

1:10:04

we've invented these machines. I also think AI

1:10:06

is going to help us identify these things

1:10:08

and help us be able to filter and

1:10:10

help us be able to say this is

1:10:12

AI generated or this content cannot be verified

1:10:14

from the source. We have to do some

1:10:17

work around what we call KYC like Know

1:10:19

Your Creator. I mean, some jurisdictions that we

1:10:21

operate in, we have to actually say, yes,

1:10:23

this is a human being that created this

1:10:25

is the company that created this. I actually

1:10:27

think we have an opportunity to serve as

1:10:29

that as like being a, yes, this was

1:10:31

created by a real human, which actually stands

1:10:34

out in a world of robots. And

1:10:36

so I think there's a lot of

1:10:38

opportunities to protect the viewer and protect

1:10:40

the creator as well as, you know,

1:10:42

serve them in helping them produce stories

1:10:44

faster. There's a lot of talk

1:10:46

about costs right now in the AI

1:10:48

world, right? There's Deep Seek, which might

1:10:50

have brought down the cost of training.

1:10:52

There's all this argument about whether the

1:10:54

cost of inference will drop. That's at

1:10:57

the same time Sam Altman is saying

1:10:59

he's going to build 500 billion

1:11:01

dollars for the data centers all around

1:11:03

the world, Soft Bank. You'd use these

1:11:05

tools, right? You're deploying these tools against

1:11:07

large data sets in video, which is

1:11:09

where the costs tend to go up

1:11:11

the fastest, the fastest. Sure answer is

1:11:14

yes, I would tell you that I

1:11:16

expect the cost of inference that dropped dramatically.

1:11:18

You know, we were experimenting, you know, with

1:11:20

some of the exact same things that Deep

1:11:22

Seat claimed to do to be able to

1:11:24

use like really low-cost chips to be able

1:11:26

to do inference. And I do expect inference,

1:11:29

the cost of inference is going to... go through

1:11:31

the floor. I used the joke to

1:11:33

say, you know, if I need to

1:11:35

order a frosty and a double burger

1:11:37

at Wendy's, I don't need to wait

1:11:40

through all of Taylor Swift's boyfriends and

1:11:42

songs to be able to get through

1:11:44

that. So distillation of models to be

1:11:46

able to serve like at the exact

1:11:48

moment of time that's necessary, like whatever

1:11:51

the language is or whatever the function

1:11:53

is. I think that you're going to

1:11:55

see distillation will help with this, you

1:11:57

know, the most recent Blackwell, black-well chips,

1:11:59

to us. And so we're going to solve

1:12:01

a lot of the inference problems and the

1:12:04

cost associated with inference. And I'm seeing it

1:12:06

drop dramatically for what we do. And so

1:12:08

it'll be very manageable over time. I think

1:12:10

that the real big cost for a lot

1:12:12

of these companies is the training of some

1:12:14

of this stuff. And that is going to

1:12:16

come in line as well. like we'll get

1:12:18

to a point of like kind of diminishing

1:12:20

returns that like do you really need to

1:12:22

go to 10 trillion parameter models or do

1:12:24

you need something that's just a lighter way

1:12:27

to be able to do chemistry to be

1:12:29

able to do biology or to be able to

1:12:31

do you know security you know coloration as an

1:12:33

example so i just i think right now we're

1:12:35

in the air of big models yeah i don't

1:12:37

think that'll last do you think we need to

1:12:39

spend 500 billion dollars on data centers all

1:12:41

around the world i think Here's more Google

1:12:43

Cloud executive. Do we need to spend 500

1:12:46

million dollars on data centers all around? You

1:12:48

know, we need to get, we need to

1:12:50

lay down a new infrastructure of silicon. Like

1:12:52

the silicon that's around the world right now

1:12:54

is highly optimized for general compute. And this

1:12:56

is a new mathematical model that has to

1:12:58

be supported with silicon. Otherwise, we'd actually consume

1:13:00

more power if we didn't have specialized chips

1:13:02

that run this this math equation. And so

1:13:04

all that's happening right now is that yes,

1:13:06

we got to run our current compute. And

1:13:08

then we have a new algorithm, we have

1:13:10

to run, and we're going to have an

1:13:12

optimized silicon to be able to run that extra

1:13:14

algorithm. So short answer is that we have to,

1:13:16

we actually do have to duplicate this silicon around

1:13:19

the world. Is that something that you can drive,

1:13:21

VEMIA, right? We talked about needing to serve 8K

1:13:23

and the price of storage and compute for that,

1:13:25

you know, falling on a pretty predictable basis. We

1:13:28

need to invent new silicon to support AI workloads.

1:13:30

That's like a whole of industry

1:13:32

effort, right? And the pressure is

1:13:34

all in maybe a handful of

1:13:36

companies and one foundry to pull

1:13:38

that off. Like, how do you make those

1:13:41

bets? You know, I would tell

1:13:43

you that we're doing a lot

1:13:45

in evaluating quality across a whole

1:13:47

spectrum. Like, one of the things

1:13:49

that as we're doing a lot

1:13:51

in evaluating quality across a whole

1:13:53

spectrum, like, one of the things

1:13:55

that, as I said, you at

1:13:57

the very start, like, how are

1:13:59

creators? or you know, this is what we

1:14:01

need to do to be able to support like

1:14:03

understanding what changed frame to frame. So a lot

1:14:06

of what we're doing is that we're saying, okay,

1:14:08

what are the, what's the best model for the

1:14:10

job that is going to get, that our creators

1:14:12

are going to need to get done. And then

1:14:15

under the surface, we're stitching all that together so

1:14:17

the creator doesn't know there might be multiple models

1:14:19

that are supporting them. And we've established quality and

1:14:21

then also handoffs, you know for that creatoroffs, you

1:14:23

know for that creator for that creator, you know

1:14:26

for that creator, And so we're going to remember

1:14:28

whether or not it's over here in the translation

1:14:30

world or whether or not it's over and

1:14:32

ask a question or indexing or otherwise. We

1:14:34

kind of stitch it all together for the

1:14:36

creator so they don't even know that we

1:14:38

might be using multiple AI. But it's establishing

1:14:41

quality bars for each one of those things.

1:14:43

And then also, you know, economics, like making

1:14:45

sure we get the best economics and also

1:14:47

the best performance like queries per second from

1:14:49

the model providers, you know, for that area

1:14:51

that they can serve our massive. of minutes

1:14:53

of video and our number of creators. So

1:14:55

we're managing performance, cost, and quality on behalf

1:14:58

of the creator across multiple models. Yeah. Well,

1:15:00

as you can probably tell, I could talk

1:15:02

to you forever, but we are out of time.

1:15:04

What's next for me? Let people give people a

1:15:06

preview of what's coming up next and watch it

1:15:08

out here. I'm probably most excited right now, as

1:15:11

I mentioned to you, about the massive formats

1:15:13

that are coming at the creator. I'm super

1:15:15

excited about what can be done with immersive

1:15:17

formats. You know, I'm also starting to see

1:15:19

a lot of people that want to kind

1:15:21

of go back into these like sphere-like experiences

1:15:23

and I do think that that's going to

1:15:25

be exciting and you'll see us continue to

1:15:27

push the edge there. you're going to see

1:15:29

us invest more in the filmmaking community like

1:15:31

on Monday I'm headed over to the Berlin

1:15:33

Film Festival after hopefully my Philadelphia Eagles do

1:15:35

well in the Super Bowl and so you're

1:15:37

going to see us even do more around

1:15:40

staff picks and celebrating filmmakers in every geography

1:15:42

we serve around the geography we serve around

1:15:44

the world I mean it's been a fairly

1:15:46

US-centric and you're going to see us get

1:15:49

a lot more global and supporting filmmakers and

1:15:51

then also just I would tell you as

1:15:53

we look over at our enterprise customers we

1:15:55

think we can support customers in their customer

1:15:58

organization in the service of customers. We're going

1:16:00

to do really well at just in time

1:16:02

videos serving just the right video to just

1:16:04

the right person at the right moment of

1:16:07

the customer interaction. And so you'll see us,

1:16:09

like, really come out with some exciting things

1:16:11

about that between the format and the AI

1:16:13

that we can do to transform storytelling. Amazing.

1:16:16

Well, we'll have to have you back when we

1:16:18

do just in time immerse the video, 8K, to

1:16:20

both eyes at the same time. So thanks so

1:16:22

much for being on a decoderoder. Okay. Thank you.

1:16:24

Thank you. I'd

1:16:27

like to thank Philip Moyer for taking the time to talk to

1:16:29

me and thank you for listening. I hope you enjoyed it. If

1:16:31

you'd like to know what you thought about this episode or really

1:16:33

anything else, drop us a line. You can email us at decoder

1:16:35

at the birdish.com. We really do read all the emails. You can

1:16:37

also hit me up directly on threads or blue sky. And we

1:16:39

have all of the emails. You can also hit me up directly

1:16:41

on threads or blue sky, and we have a Tik. You can

1:16:44

also hit me directly on threads, on threads, on threads, on threads,

1:16:46

on threads, on threads, on threads, on threads, on threads, on, on

1:16:48

threads, on, on, on threads, on, on, on, on, on, on, on,

1:16:50

on, on, on, on, on, on, on, on, on, on, on, on,

1:16:52

on, on, on, on, on, on, on, on, on, on, on, on,

1:16:54

on, on, on Decoder is a production of urgent part of the

1:16:56

boxing new podcast network. Our producers are

1:16:58

Kate Cox and Nick Sta, our editor

1:17:01

is Ursa Wright. Decoder Music is

1:17:03

my breakmaster cylinder. We'll see

1:17:05

you next time. Support for the show comes

1:17:07

from Alex Partners. If you're in

1:17:10

the tech industry, wondering how

1:17:12

artificial intelligence is going to

1:17:14

affect your business, might seem

1:17:16

like the new normal by

1:17:19

now. Alex Partners is a

1:17:21

global consulting firm dedicated to

1:17:23

helping you navigate the changing

1:17:25

headwinds of AI. without getting

1:17:27

lost in the noise. Learn

1:17:30

how your business can navigate

1:17:32

AI while making sure your

1:17:34

strategic initiatives are aligned by

1:17:36

reading Alex Partners' latest technology

1:17:39

industry insights available at

1:17:42

www. Alex partners.com/VOX. In

1:17:44

the face of disruption,

1:17:47

businesses trust Alex partners

1:17:49

to get straight to

1:17:52

the point and deliver

1:17:54

results, when it really matters.

1:17:57

Support for this show comes from

1:17:59

ADT. From ADT comes trusted neighbor,

1:18:02

the new standard in home access.

1:18:04

Through the ADT Plus app, easily

1:18:06

grant and automate event-based or scheduled

1:18:09

access for neighbors, friends, and helpers.

1:18:11

Notify trusted individuals of events like

1:18:13

alarms or packages, and set access

1:18:16

windows for planned guests or even

1:18:18

the Dog Walker without interrupting your

1:18:20

day. Visit adt.com. When every second

1:18:23

counts, count on ADT. Requires ADT

1:18:25

Complete Pro Monitoring Plan and Compatible

1:18:27

Devices. Copyright 2025 ADT LLC. All

1:18:30

rights reserved. Sometimes a single performance

1:18:32

can define an artist's legacy. Think

1:18:34

about Hendrix's fiery Woodstock National Anthem

1:18:37

or Biance's Homecoming at Coachella. Coming

1:18:39

up on Switched On Pop, we're

1:18:41

exploring artists who've had recent transformative

1:18:44

live shows. First is Missy Elliott,

1:18:46

who recently put on her first

1:18:48

world tour where she taught everybody

1:18:51

to get the freak on. And

1:18:53

then there's her collaborator Timbalin, who

1:18:55

recently evolved from Beatmaker to orchestra

1:18:58

conductor at the Songwriter Hall of

1:19:00

Fame. And then Lady Gaga, whose

1:19:02

chromatic ball featured a theatrical museum

1:19:05

of brutality revealing the darker side

1:19:07

of Gagga's mayhem. Listen to these

1:19:10

live moments on Switched on Pop

1:19:12

wherever you get podcast. Where you

1:19:14

get podcast. brought to you by

1:19:17

defender.