0:00

Hi. Listeners, it's on you profumo or producer. Here

0:02

at Masters of Scale. We. Talk a lot

0:05

about the importance of a I, especially when it

0:07

comes to scaling your business. But we know that

0:09

many teams, including our own, are still wondering how

0:11

to better integrate Ai into their day to day

0:14

work. That's. Why We're inviting you

0:16

and your team to join us on

0:18

February twenty ninth for Masters of Ai

0:20

Day, the workshop that empowers teams to

0:22

accelerate and amplify their work with a

0:24

I. By. Registering for a I

0:26

Day, you'll get access to the Ai

0:28

Activity pack, a set of hands on

0:30

activities and experiments designed for teams to

0:33

explore ai and practical ways. Plus, you'll

0:35

also unlock access to the Ai Day

0:37

virtual event, the Ai deep dive collection,

0:39

and a robust set of free A

0:41

I resources. There's. No better moment

0:44

to leap into a I saw

0:46

register for Masters of Ai Day.

0:48

Now at Masters of scale.com Forward/masters

0:51

of Ai Again, that's Masters of

0:53

scale.com Forward/masters of Ai. Hey.

1:03

Everyone bob savvy in here. Every

1:06

few months I connect with Reid

1:08

Hoffman to get his perspective on

1:10

the key inflection points in the

1:12

business world are most recent conversation

1:14

which were sharing today starts with

1:17

open a eyes, new text, a

1:19

video tool, sorrow and why open

1:21

A I see Oh Sam Altman

1:23

is trying to raise seven trillion.

1:26

Dollars that's trillion with a

1:28

T. We. Also, talk about met

1:30

as dramatic resurgence. The Apple Vision

1:32

Pro Three things driving tech lay

1:35

offs, the Ai trap that many

1:37

businesses are falling into a more

1:39

as always reads insights both anchor

1:42

me and get my wheels turning.

1:44

I hope they do the same for you. Let's

1:46

get to it. We'll

2:09

start the show in a moment afterward

2:11

from our premier brand partner, Capital One

2:13

Business. Georgian

2:16

food is to Russia like Mexican food is

2:18

to the United States. It's comfort food. Georgian

2:21

wine was also the best wine, but the

2:23

Russians had embargoed it. And so I vowed

2:26

that if I ever did open a restaurant, I

2:28

would sell all the Georgian wine I could. That's

2:31

Rose Prevet, owner of the Michelin-starred

2:33

restaurant Maidan and Compass Rose. And

2:35

she was on a trip aboard the Trans-Siberian Railway in

2:37

December 2011 with her husband, former

2:40

NPR journalist David Green. Snowy

2:43

Siberia is the perfect place to have your

2:45

eat, pray, love moment. Only she was

2:48

able to like go around Italy and

2:50

eat pasta. I'm sitting on this

2:52

sweaty train eating stuffed pies. Back

2:55

in Washington, D.C., Rose had a career in

2:57

public policy, but she was contemplating a personal

2:59

pivot into the restaurant business. And

3:02

she had been particularly taken with Georgian food and

3:04

wine. The very first

3:06

smack in the face was trying to get my

3:08

liquor license. But first,

3:10

Rose had to navigate obtaining a liquor license.

3:12

We'll hear about that later in the show.

3:15

It's all part of the Refocus Playbook,

3:17

a special series where Capital One Business

3:19

highlights stories of business owners and leaders

3:21

using one of Reed's theories of entrepreneurship.

3:24

Always Playbook Insight, every successful

3:26

leader needs grit. All

3:33

right, we ready for this? Always

3:36

with you, Bob. This past

3:38

week, the new text-to-video tool that OpenAI

3:40

released, Sora, which I don't know if

3:42

you've played with it all yet. I

3:44

played with it a little bit, and

3:46

it does some amazing stuff. What

3:49

people have been working on pretty intensively for last

3:52

year or two is these multimodal, generative

3:54

models. And it

3:56

doesn't surprise me that OpenAI is the first

3:58

to release. the amazing

4:01

new jump in capability. They

4:04

just keep seeming to be ahead of everyone

4:06

else, even though so many

4:08

entities are racing to try to leapfrog

4:10

them. Well, I think one

4:13

of the key things that OpenAI

4:15

has stayed true to that

4:18

relatively few of the other players

4:22

are internalized to the degree to which

4:24

OpenAI does it, which is a scale

4:26

compute. So frequently, of course,

4:28

people want to talk about the narrative of what's

4:30

going on in AI as we've invented the new

4:32

algorithm. And there are new

4:34

discoveries and new algorithms, but really what

4:36

we are is applying the

4:39

scale compute lesson. We've

4:41

got these news stories of Sam Altman going

4:43

around for the trillion dollar chip

4:45

fund and all the rest, because he and

4:48

the organization has embraced the, we

4:50

are playing the scale. And by having played

4:53

the scale, and there's a lot of kind

4:56

of interesting academic results and other

4:58

things that say, hey, scale

5:00

is kind of trumping all. Yes,

5:03

this algorithm is a little better. Yes, this

5:05

algorithm is a little worse. Yes, this data

5:07

sets a little better. Yes, this data sets

5:09

a little worse. But it's the scale that

5:12

is primarily driving. It's what drove it in

5:14

the large language models, hence large out of

5:16

language models. And it's what's driving

5:18

it in the multimodal as well. Once

5:20

though you have that scale compute, it

5:23

allows you to create these executions in a

5:25

way that if you don't have that scale,

5:27

it's not that you can't, but it's not

5:29

going to be smooth the same way. It's

5:31

just going to be that much harder or

5:33

that much bumpier getting to those places. Yes,

5:36

exactly. The slower work is amazing. I

5:38

don't want to take anything away from

5:40

the fact that it was smart, high

5:42

quality people working intensely in a team,

5:45

but it's also creating these

5:47

super expensive computers. That's

5:49

a bold risk taking move from

5:52

both OpenAI and from Microsoft.

5:56

Now, you mentioned Sam Altman and his,

5:58

I think it's $7 trillion. trying

6:00

to raise for the chip

6:03

industry, which is a wild

6:05

number, right? 7 trillion dollars, especially

6:08

at the same time that he

6:10

is running this company. Is it

6:12

about that we really need that

6:15

number to get to the future

6:17

of, you know, this industry?

6:20

Or that's a target that pushes

6:22

us in the right direction? What

6:25

I would say is they've probably done

6:27

some internal calculations about what scale gets

6:29

them potentially to artificial

6:31

general intelligence. There's

6:33

several different definitions of artificial intelligence,

6:35

but obviously the rough thought is

6:38

something intelligent like we are, capable

6:41

of having metacognition and strategic

6:43

planning and one-shot learning and a

6:45

bunch of other stuff, and

6:47

that this scale approach of this

6:50

will get there. So my

6:52

guess is that's how

6:54

the number got calculated just

6:56

from knowing the people. Now

6:58

that being said, part of

7:00

the thing I think is amazing is

7:03

whether or not you think AGI

7:05

is high probability, medium probability, low

7:07

probability, wherever you are

7:09

in that spectrum, you will be

7:12

creating really interesting increases in

7:14

cognitive capability of these systems. Say

7:17

they can't raise seven trillion or not

7:19

in one bold swoop or

7:22

whatever else, you're still gonna have

7:24

amazing progress. So kind of doesn't

7:26

matter if it was set out as

7:28

a we need this to get

7:30

to AGI or it's

7:32

a target number with

7:35

a aspirational focus. That

7:37

continued investment to scale

7:40

is gonna be one of the things

7:43

that's gonna net a lot of the

7:45

steam engine of the mind revolution and

7:47

this amazing cognitive amplifiers

7:49

for human work. That

7:51

will come out of some portion of that whatever

7:54

portion of the goal plays out. Sometimes

7:57

I feel like the story in technology

8:00

development is like it's software

8:05

which feels like it's been more recently. But now

8:08

with AI of course it's software but it sounds

8:10

like it's the hardware also. These two things are

8:12

being married or relying on each other to

8:15

advance in a way that maybe

8:17

we haven't had for a little while. Well

8:19

I think it's always been a combination

8:21

of software and hardware. We were operating

8:23

for decades on so-called Moore's law. Really

8:25

kind of Moore's hypothesis or

8:27

principle or something. Which

8:29

by doubling the number of transistors and getting

8:32

the kind of compute better, we could

8:34

continue that software progression. And there's some

8:36

really compelling graphics that kind of show

8:38

that the progress in this modern way

8:40

of AI came about

8:43

from when as Moore's law flattened,

8:45

it just changed its shape

8:47

of what the hardware loop was. Was opposed

8:50

to the hardware loop being we're also doing

8:52

Moore's law. It's the we're doing

8:55

massive parallel configurations. By the

8:57

way what that means is that

8:59

tends to go to certain kinds

9:01

of algorithms. It requires a learning

9:04

versus a programming artifact

9:06

tends to be probabilistic computing. There's a

9:08

stack of things that kind of go

9:10

into how that

9:12

scale plays but

9:15

the hardware software combination continues. Sometimes

9:19

sort of the valuations rise

9:21

for one group and then the other

9:23

group. They sort of seem to

9:25

sometimes go in sequence although right now with

9:28

Microsoft on the one hand and open AI

9:30

and with Nvidia on the other end they're

9:32

both going. Valuations

9:35

are actually in fact kind

9:37

of a market prediction about where future

9:40

value will lie. And

9:43

you go okay the hardware like Nvidia

9:45

has got these great business with 80%

9:47

margins. We have

9:50

this fairly strong hardware edge that

9:52

so far we have not

9:54

had anyone catch us. And

9:56

so people are okay. There's a lot of value there. We'll bet on

9:58

that. But on the other hand people also. to know there's going to

10:00

be a ton of value on the software side. And

10:02

so the short answer is the

10:04

market saying, we're betting on both the software

10:06

and hardware on the AI side. And

10:09

so place your bets. And

10:11

who wins in the long run, or who

10:14

is the bigger winner? It's kind of hard

10:16

to tell right now. I mean, right now,

10:18

we see open AI obviously winning. But I'm

10:20

thinking of the earlier waves of social media.

10:23

Like Facebook was not necessarily considered to be

10:25

the winner in the beginning, and

10:27

yet they ended up being the winner. This

10:30

happens in business and in technology in particular.

10:33

And also, it's a little bit

10:36

reminiscent also of the question is, is this

10:38

going to be, on a technological change, things

10:40

that large companies win from or startups win

10:42

from? And the answer

10:44

that I've been given fairly consistently along this

10:46

whole path is both. It

10:49

isn't going to be the David and Goliath where

10:51

David invests his new thing AI, and it kind

10:53

of resets the apple carts of

10:56

the Goliath that are appropriately

10:58

bought into AI, which is

11:01

intensely Microsoft and Google and

11:03

some Amazon. Definitely Facebook. The

11:06

AI will be a massive increase for the

11:08

large companies, but will also be very

11:11

valuable across a whole set of startups.

11:13

And so I think it's going to

11:15

be a realization of the software revolution,

11:17

that transformation of industries by software across

11:20

the entire stack of size. And I

11:22

think that's also true hardware and software.

11:25

The concentration of wealth right now

11:27

in the top tech names, Microsoft,

11:30

Nvidia, Apple, Amazon,

11:33

it's like 25% of the market value from

11:35

the S&P 500. I mean, it's

11:37

not been that way before. Is that troubling?

11:41

Well, there's definitely places where scale can

11:43

be not good. That's an odd statement

11:45

to be making on the Masters of

11:47

Scale podcast. But it's not

11:49

what most people think. Most people

11:51

think, oh, big is bad. Oh, look, it's

11:54

continuing market dominance of the top

11:56

tech companies. And you're like, well, actually, in fact,

11:59

if we were five. big US

12:01

tech companies or seven big

12:03

US tech companies heading

12:05

to three, I'd be actually quite

12:08

concerned. We're actually five to seven

12:10

heading to 10 to 12.

12:13

And the competition between these organizations is

12:15

fierce. And it creates a lot of

12:18

opportunities for startup. It creates a lot

12:20

of services for consumers. It

12:22

creates a lot of value within the American

12:24

tech industry, which benefits America.

12:27

Scale is concerning when it distorts it

12:29

because it crushes competition. If

12:32

tech company X were to

12:34

get to scale and then lock

12:36

out other players to the detriment

12:38

of society, detriment of industries, detriment

12:40

of consumers, but that's not

12:42

an absolute scale number. That's a relative

12:44

scale number to your competitors and other

12:47

players. That's the mistake that

12:49

lots of press and everyone else just

12:52

mistakes. That's the reason why if it was five

12:54

to seven going to three, well, if

12:56

it's five to seven going to 10 to 12, then it's

12:59

good. Nvidia's inclusion

13:02

in the trillion dollar club is

13:04

part of the going from five to seven to 10 to 12. Right?

13:08

That's the instance of it. Like I've been saying

13:10

this for years and it's like, look, here's an

13:12

example proof. And you said, well, you didn't say

13:14

Nvidia before it's like, you don't know which ones,

13:16

but you know that the competitive ground swell is

13:18

coming. And by the way, one of

13:20

the benefits that you have when

13:22

you have this is all of

13:25

these companies are investing massively in

13:27

technological R and D. It's not

13:29

what people frequently are saying scale

13:31

is bad. It's saying you have to watch for

13:33

certain elements of scale, which I don't think we

13:36

are actually triggering yet. So

13:38

this is like qualitatively different than

13:40

say regulators looking at the airline

13:43

industry and saying, JetBlue and Spirit

13:45

shouldn't get together or the grocery

13:47

industry and Kroger's and Albertson shouldn't

13:49

get together. The tech industry

13:51

is sort of qualitatively different. Well,

13:54

if you were saying Microsoft

13:56

and Google should combine into one company,

13:58

I would definitely go. that

14:00

would not be a good idea, right?

14:03

But if you said, should large

14:05

company X be able to buy small

14:07

company Y, the answer is, sure, they

14:10

should be able to. The competition between

14:12

these organizations is ferocious. And

14:14

by the way, as a venture capitalist, my primary identity, there's

14:17

a lot of opportunity on the startup side.

14:19

Back in the day, decades

14:21

ago, when Microsoft was the

14:23

one big tech company, that was a

14:25

bigger challenge. It was good to limit

14:27

it. But I think

14:29

everyone was surprised by how quickly,

14:31

through the browser and through other

14:33

things, that new tech companies

14:36

could emerge and challenge it. Now, the

14:38

last point to your earlier question, these

14:40

top tech companies are a quarter of

14:42

the S&P. I think this is beginning

14:44

to really show the realization that all

14:46

companies are heading towards becoming

14:49

tech companies, that you

14:51

need that tech amplification in

14:53

all of it. And that's what we need

14:55

to be figuring out across all industries. And

14:58

it's the thing that we need to be embracing for

15:01

the future of our economies, the

15:03

future of our prosperity as societies,

15:05

individuals. The classic dialogue thing is,

15:08

we got to limit the big tech. And you're

15:10

like, well, that's if you want to

15:12

limit your economic future, right? It's

15:14

like, no, no, what you want to do

15:16

is say, how do we leverage the fact

15:18

that we have some technological advantages to benefit

15:21

society broadly? And that's where the intellectual work

15:23

needs to be. Reed is

15:26

such an impassioned advocate of technology, but

15:28

that doesn't make him wrong. The future

15:30

is coming faster than ever. And you

15:32

want to be on the right side

15:34

of it. After the

15:36

break, we talk about Mark

15:38

Zuckerberg and Meta, the Apple

15:40

Vision Pro, tech layoffs, and

15:42

more. Stick around. We'll

15:44

be back in a moment after a word

15:46

from our premier brand partner, Capital One Business.

15:52

They saw a very young woman who had never owned

15:54

a business before and thought the fight

15:56

they were going to fight was going to be the

15:58

one against me because the bunch of other

16:00

restaurants opening, but no one went after them

16:02

as hard as they would after me. These guys wouldn't

16:05

give up. We're back

16:07

with Washington, D.C. restaurant owner Rose Preffitt.

16:10

A decade ago, she was opening her

16:12

first restaurant, Compass Rose, in the rapidly

16:14

gentrifying 14th Street Corridor. But

16:16

some disapproving members of the community were determined

16:19

to stop her, so Rose pounded the pavement

16:21

in search of support. We

16:23

knocked on doors and got hundreds of

16:25

petitions signed by our neighbors. And

16:27

eventually, we did win. And what ended

16:30

up happening is those people felt very invested

16:32

in our cause. So when Compass Rose opened,

16:34

all those people came. Rose

16:37

was dedicated to her vision, and that grit

16:39

would prove essential as she faced the challenges

16:41

of being a first-time business owner, says Lauren

16:44

Trusco, head of brand partnerships and insights at

16:46

Capital One Business. When

16:48

you're starting out as a business owner, it's

16:50

really easy to get discouraged. The challenges you

16:53

have to face are daunting and constant. In

16:55

order to succeed, you really have to be

16:57

your own biggest advocate. But

16:59

getting a liquor license would only be the first

17:02

hurdle Rose faced, because in the middle of D.C.,

17:04

she wanted to serve food, cook outdoors,

17:06

over an open fire. We'll

17:09

hear about that later in the show.

17:11

It's all part of Capital One Business'

17:13

Spotlight on Entrepreneurs following Reed's Refocus Playbook

17:15

at all levels of scale. Before

17:22

the break, Reed Hoffman shared his perspective

17:24

on how AI is altering the game

17:27

of scale. Now we talk

17:29

about Mark Zuckerberg and Meta, the Apple

17:31

Vision Pro, tech layoffs, and the trap

17:33

that many business leaders are falling into

17:35

in 2024. Let's

17:38

jump back in. So I want

17:40

to ask you about Mark Zuckerberg and Meta.

17:43

Facebook was the quintessential social media

17:45

company. And then it sort of

17:48

seemed like they were seeding ground,

17:51

particularly the TikTok, which

17:53

sort of captured the cultural heat. Zuckerberg's

17:56

talking about the metaverse and spending

17:58

a lot of money and maybe not getting

18:00

a lot for it, the Applevision Pro is

18:02

getting all the buzz more than Quest

18:04

VR did, and Sheryl Sandberg steps aside

18:07

as COO and then she's leaving the

18:09

board. And then on one day, Meta

18:12

stock, bam, goes up like 20%. What

18:15

happened? What did people

18:18

sort of miss appreciate or

18:20

misunderstand? And I don't know whether

18:22

it's about Meta or about the

18:24

industry or about Mark. It

18:27

didn't surprise me, although I'm not a

18:29

public market trader, so I tend to

18:31

be structuralist over years. What

18:33

is the position over one year, three years, five

18:36

years, 10 years? It's part of why

18:38

doing early stage venture. Now one,

18:40

Mark is a bold innovator.

18:43

We'll take bold bets. Some

18:45

of the bold bets are pretty amazing, like when

18:47

he bought Instagram when it was really small, bought

18:49

WhatsApp. Some of the bold

18:51

bets I personally don't think work out as well.

18:54

Oculus and Meta as they focus,

18:56

but he's a constant infinite learner.

18:59

And so I saw

19:01

when the light bulbs came on of, oh my God,

19:03

this AI thing is going to be really important

19:05

and we should be doubling down on that. I

19:08

don't think they're agents and so forth

19:10

of that good jet. But the

19:12

notion of how do you tie the AI

19:14

stuff into the advertising system and have an

19:16

alternative, really compelling advertising

19:18

system. How do you

19:20

continue to do engagement with the feed and

19:22

other kinds of things, which I think they

19:24

do now. I think it didn't surprise me

19:26

at all that they would have a vigorous

19:28

recovery given a focus in

19:31

AI, given a natural set

19:33

of assets in aspects that are

19:35

important in human life. I

19:37

mean, he made the business more

19:40

efficient, I think, than people expected.

19:43

And I think the stickiness

19:45

of the different parts of

19:47

Meta proved to be stronger

19:49

than certainly the marketplace was

19:51

anticipating. I mentioned the

19:54

Quest and Apple Vision Pro. Have

19:57

you tried these things out? I know Microsoft

19:59

has its HoloLens. Is this an

20:01

area that you play with, that you dabble

20:03

in, or is it not really your jam?

20:06

Well, my very first product management

20:08

job was in virtual worlds. Having

20:11

gone into the promise of it and realized

20:13

it was so under-delivered that I'm a little

20:15

bit overly skeptical, I'm

20:18

part of the reason why Greylock didn't

20:20

invest in Magic Leap and other

20:22

things because I go, it is really

20:24

amazing technology, but I just don't see

20:26

it coming together as the new tech

20:28

platform yet. Now, I've had

20:30

a couple of my trusted friends, David Z., other people

20:33

say, I really need to play with Vision Pro, I

20:35

need to see what it is. I

20:37

have played with the Oculus, I have played

20:39

with HoloLens, I have yet

20:42

to think that any of these things are

20:44

a new platform. I definitely think that the

20:47

economic cost of

20:49

the Vision Pro means

20:51

it certainly won't be a platform yet.

20:54

Now, will it be, as some

20:56

people are speculating, is it a limited n

20:58

number of iterations from here to there to

21:01

make it work? Is it two or

21:03

three? I think that's

21:05

part of the reason why David has been like,

21:07

okay, you really need to go play

21:09

with this? And so I've ordered one to

21:12

get a sense of it. I caution everything about how

21:14

you get to a platform. Here we

21:16

are, you and I, talking, wearing glasses.

21:20

Even these things, which are an amazing part

21:22

of early technology, a

21:24

number of human beings pay $5,000 to

21:27

have a laser applied to their eyes so

21:29

they don't have to wear these things. So

21:32

that's what you have to clear in

21:34

terms of the value proposition to

21:36

get to a general platform. And

21:39

if it isn't a double-dick platform,

21:42

maybe like a doctor's, nurses, police

21:44

people, fire people, other kinds of

21:46

things, is part of why I

21:49

think the HoloLens and Microsoft

21:51

kind of realize better than the consumer

21:53

folks for these things. But

21:55

as a general platform, you

21:57

have a very high hurdle.

22:00

The benefits really have

22:03

to be powerful enough to make

22:05

it worth the inconvenience of, much as

22:07

we love that we have these

22:09

spectacles, they allow us to see, they're

22:11

pain. They're pain. Yes.

22:14

Exactly. In optimistic,

22:18

good times that the tech industry is

22:20

having right now, there have

22:22

also been a bunch of low-offs talked about

22:25

at places like Amazon

22:27

and Ineta and Alphabet. I've

22:29

heard three different theories about

22:31

why these layoffs happen. One

22:34

being overhiring from the pandemic, which

22:36

is something that Zuckerberg has said

22:38

at Meta, they overhired some. Another

22:41

argument is copycat layoffs, basically doing

22:43

it for performative reasons to appease

22:45

public market investors to make it

22:48

look like you're being sharper. The

22:51

third being that AI is making things

22:53

more efficient and that these

22:56

businesses just don't need people the same

22:58

way. Now, obviously, there are many more

23:00

reasons why layoffs could happen. I'm curious

23:02

whether you have a theory

23:04

about why this is happening. The

23:08

first one is certainly the case, the overhiring

23:10

from the pandemic and reconfiguration and so forth

23:12

because people overly generalize from, oh,

23:15

during the pandemic, this is the new normal.

23:18

To some degree, it was almost like a collective

23:20

mistake where everyone's saying everyone's hiring, so we should

23:22

be hiring too. Now,

23:24

the copycat layoffs is actually a touch

23:27

more sophisticated as a question, which is

23:29

I don't think that any of the leaderships of

23:31

any of these companies is so banal as to

23:36

simply be, well, because the investors are demanding

23:38

it for margins, other people doing it, I

23:40

have to do it too. And

23:43

so it doesn't say there isn't something there. But

23:45

I think what actually, in fact, goes into this

23:47

is when there is

23:49

a zeitgeist, a time of layoffs,

23:52

you can then kind of go, okay, I

23:55

now have a permission to do

23:57

some of the resetting in the business. I

24:00

should be doing less of X, more of

24:02

Y, things I should be doing

24:04

anyway, but suddenly I can do it and it's

24:06

sort of part of the zeitgeist and I'm not

24:08

going to pay as much of a price for

24:10

it. Exactly. Because the price you

24:12

use, you're the only person doing these kind of

24:15

layouts. Is it because you suck? You've

24:17

made bad decisions. And so there's

24:20

a pressure to not do it in

24:22

bold stroke. If we do that and

24:24

we're standing out there by ourselves, it

24:27

causes a bunch of negative speculation. And

24:29

then we go, oh shoot, we can

24:31

now take opportunity, like the

24:33

industry itself is doing layoffs, and

24:35

we can do some reconfiguration. And I

24:37

think that's much more of what you're

24:39

happening. It's all kind of

24:42

simplistic press stories. And what you should really

24:44

do is look underneath, okay, what

24:46

are they reshaping their businesses for? Now,

24:48

they may be reshaping their businesses for an

24:51

anticipated AI productivity. I think

24:53

there will be a bunch of AI productivity. I

24:55

don't think that much AI productivity is currently baked

24:57

into the numbers or baked into the way that

24:59

people are operating. So it's certainly not a post-fact

25:02

thing. I think the other cards

25:04

to consideration are this gives us a way

25:07

to reshuffle. And that

25:09

reshuffling, that reconfiguration is what

25:11

you should be looking at in each of

25:13

these tech companies. And I do feel like

25:15

after over the last four

25:17

years between pandemic and supply chain

25:19

and inflation and AI, that a

25:22

lot of leaders that I'm talking

25:24

to are sort of, they're not

25:26

saying that they're resetting or they're

25:28

like, let's not be too aggressive,

25:30

but they're sort of reflecting a little bit

25:32

right now. And I don't

25:34

know, I wonder sometimes whether that's

25:36

complacency or that's fatigue or

25:38

that's wisdom. Well,

25:41

frequently with these choices, Bob, as you

25:43

know, it's a combination of all of

25:45

them and it depends on the different

25:47

leaders, right? To essentially go

25:49

to what people should be, as far

25:52

as the resetting is, should be saying, we

25:54

are at another wave of technological

25:56

transformation. We've had a bunch. internet,

26:00

we've had the iPhone, we've had

26:02

mobile computing, we've had cloud. Now

26:05

AI is going to take all of those

26:07

to another level. The reason why it's so

26:10

big is it's a compounder of all of

26:12

them together, and that's coming

26:15

in a small n number

26:17

of years. So if you're

26:19

not plotting part of

26:22

the technology strategy of your

26:24

business to be doing

26:26

this, I don't mean IT strategy, you know,

26:29

it's the technology underlying how your whole business

26:31

operates, what your supply chain is, how your

26:33

employees work together, how you sell

26:36

and market, how you build and constitute

26:38

your product, how you do innovations, how

26:40

you do financial analysis, all

26:42

of that is in evolution from

26:44

essentially the steam engine of the mind in

26:47

AI. And so the advice that I

26:49

give everyone, they say, well what should I do now? It's

26:51

like, well it's all in very dynamic flux, you can't just

26:54

say it's only X right now, and that's all

26:56

you need to do. So what you need

26:58

to do is start experimenting with it. So

27:00

the one thing about the reflective thing where

27:02

it's lazy is, no, no, you should

27:04

be diving into the experimentation, not necessarily committing

27:08

fully to a particular path right now because you don't really

27:10

know, but if you're not

27:12

experimenting with some vigor, you're

27:14

likely to be making a pretty

27:16

dangerous mistake. And if

27:19

you're sort of waiting for things to become quote

27:21

clear, which is what a lot of leaders I

27:23

think want to do, they're actually

27:25

losing ground because they're not getting comfortable

27:28

with how this new technology can change

27:30

the way they operate. Exactly.

27:32

You should not be waiting. You

27:35

should be experimenting, possibly

27:37

experimenting vigorously. Well

27:39

Reed, thanks for doing this. Always good to chat with you.

27:42

Always great to talk with you Bob. I look forward to the next.

27:48

What I came away with most from

27:50

this discussion with Reed is the potential

27:52

trap of waiting for the right moment

27:54

to engage with a new opportunity or

27:57

a new challenge. If we're

27:59

waiting for clarity... to emerge in

28:01

today's fast-changing world, then we're waiting

28:03

too long. We have to jump

28:05

in and experiment in the face

28:08

of ambiguity. I'm Bob Safian. Thanks

28:10

for listening. And

28:28

now, a final word from our brand

28:30

partner, Capital One Business. I

28:34

have never done anything close

28:36

to brand or marketing ever in my

28:38

life. And this

28:42

entire team, which has been around for years,

28:45

will they accept me as their leader? We're

28:47

back once more with Aparna Saran of Capital

28:50

One Business. That was the question

28:52

on her mind as she pivoted into marketing. But

28:54

thinking like an outsider led to a

28:56

series of big picture questions. But

28:59

my team was very gracious and patient

29:01

with me. And they went

29:03

out of their way to take the time

29:05

to help me. And I think that was

29:07

really critical. And before you know

29:09

it, what I had was like an initial

29:11

draft of what has become

29:14

now the vision for our next three

29:16

years of transformation journey in marketing.

29:18

You don't need to be a beginner to have

29:20

a beginner's mindset. You just need to be

29:22

able to step back, look at your business

29:24

with fresh eyes, and be open to new

29:26

solutions. And you'll find, as Aparna did, that

29:28

your team will follow your lead. I

29:31

love my current team. Everyone brings their unique

29:33

strengths and they are way smarter than I

29:35

am. And it just gives me

29:38

a new type of energy. Capital

29:40

One Business is proud to support

29:42

entrepreneurs and leaders working to scale

29:45

their impact from Fortune 500s to

29:47

first-time business owners. For more resources

29:49

to help drive your business forward,

29:52

visit capitalone.com/Business Hub. That's

29:54

capitalone.com/Business Hub. Wait

30:00

What Original. Our executive producer

30:02

is Eve Trowe. The production

30:04

team includes Chris Gautier, Adam

30:06

Skus, Alex Morris, Tucker Vodersky,

30:09

Masha Makutinina, and Brandon Klein.

30:12

Audio editing for this episode by Stephen

30:14

Wells. Mixing and mastering

30:16

by Aaron Vastinelli. Original

30:18

music by Eduardo Rivera and Ryan Holiday.

30:21

Our head of podcasts is Leetal

30:23

Malad. Visit mastersofscale.com to find the

30:26

transcript for this episode and to

30:28

subscribe to our email newsletter. Audio

30:33

editing by Chad

30:46

GPT. I produce a podcast called

30:48

Masters of Scale and

30:50

I'm looking for a way to get listeners

30:53

to leave us a five-star

30:55

review on Apple Podcasts.

30:57

Do you have any suggestions? Absolutely.

31:01

One effective way is to engage with

31:03

your listeners directly during your episodes. Remind

31:07

them how much a five-star review can support

31:09

your show. What about to

31:11

try to convince listeners to share one

31:13

of their favorite episodes with somebody in

31:16

their life? Do you have any suggestions

31:18

for that? Definitely.

31:21

People love to share content that has

31:23

made an impact on them. So remind

31:26

your listeners how sharing an episode can

31:28

start meaningful conversations with friends or help

31:30

others discover insights they've benefited from. Would

31:33

it be okay if I use maybe our

31:35

conversation as a segment on the show? That

31:38

sounds like a creative approach. Using

31:41

a meta-conversation about engaging your

31:43

audience as a segment could

31:45

be quite engaging. This

31:48

is executive producer Chris McLeod and

31:50

I urge you to take Chad GPT's advice

31:53

and leave us a

31:55

review on Apple Podcasts and share an episode with

31:57

a friend. It really, really helps.