HomeVideos

Qualcomm Jumps on $15 Billion Data Center Sales Projection | Bloomberg Businessweek

Now Playing

Qualcomm Jumps on $15 Billion Data Center Sales Projection | Bloomberg Businessweek

Transcript

1223 segments

0:00

[music]

0:02

Bloomberg Audio Studios, podcasts,

0:05

radio, news.

0:08

This is Bloomberg Business Week Daily,

0:11

reporting from the magazine that helps

0:13

global leaders stay ahead with insight

0:16

on the people, companies, and trends

0:18

shaping today's complex [music] economy.

0:20

Plus, global business, finance, and tech

0:23

news as it happens. The Bloomberg

0:25

Business Week Daily podcast [music] with

0:27

Carol Masser and Tim Stenbec on

0:30

Bloomberg radio.

0:32

>> This is a different conversation. It's a

0:34

different data center right now. This is

0:35

about AI and but it's a valid question,

0:38

right? So I will give you the same

0:40

answer that I gave when we entered the

0:43

automotive business when we entered the

0:44

industrial business and everything that

0:46

we have been doing to diversifi

0:48

diversify the company. Those are fastm

0:51

moving markets now changing with

0:53

technology. So where we really focus is

0:55

not what the data center is doing today

0:58

but what the data center is going to be

0:59

doing tomorrow. The data center is now

1:01

moving to a gentech. That's why you see

1:03

CPU companies on the rise. You see

1:06

demand on the rise.

1:07

>> Good stuff there. Cristiano Aman

1:09

Qualcomm CEO there with Roma Bastik and

1:12

of course the host of Bloomberg Tech Ed

1:14

Ledo who kindly is joining us right now

1:16

from our bureau in San Francisco. Do

1:18

want to point out Qualcomm shares are up

1:20

about 7%. They're off their highs of the

1:22

session. Uh, Micron, we're going to get

1:23

to that, too. It's up about 17 and a

1:25

half%. I think it's at an all-time high

1:26

as we speak. Whoa. Ed Lello, where do we

1:30

begin? Talk to us.

1:31

>> Yeah, it's quite a lot. [laughter]

1:32

>> What do you think? Let's talk about

1:34

Qualcomm. We just heard a snippet uh of

1:36

the CEO.

1:37

>> Let's do that for you.

1:38

>> Credit where Yeah, credit where credit's

1:39

due, right? You know, Qualcomm is the

1:41

world's biggest provider of smartphone

1:43

processors, a market that is not doing

1:45

great right now. And they have made this

1:48

entry into data centers. And what you

1:50

heard from Cristiano on there was the

1:52

answer to the question, didn't you

1:55

already give up on data centers? And

1:57

secondly, if you look at what they

1:59

announced, Wall Street's clearly

2:01

cheering what they've said about the

2:03

future financial opportunity, but

2:05

they're a little bit late to the game

2:07

because these CPUs come online in 2028.

2:10

Their accelerators, their custom um AI

2:13

chips for running inference workloads

2:15

aren't here yet. And the problem is

2:17

they're in this environment where Nvidia

2:19

absolutely dominates the market and

2:21

whatever's left AMD come comes up and

2:24

gets the rest. But both of those

2:26

companies have a new chip every single

2:27

year. Qualcomm's chips are two or three

2:29

years away. Um but you know he was

2:32

honest about that like the the market

2:34

opportunity is absolutely ginormous and

2:36

they are very confident they're going

2:38

after a very specific bit of it because

2:39

their chips are nothing to be sniffed

2:41

at. They're like highly performing um

2:44

very efficient chips on a dollar per

2:46

token basis or a dollar per kilowatt

2:47

basis.

2:48

>> Ed, you mentioned AMD, you mentioned

2:51

Nvidia. What about Intel, Broadcom, uh

2:54

Google, Amazon?

2:56

>> Can we can we throw Cerebras in there?

2:59

>> Yeah, you you absolutely can. I mean the

3:01

field is becoming more crowded in terms

3:03

of the number of players but based on

3:06

current data which you can say is just

3:08

simply sales or its deployments Nvidia

3:11

has a technical monopoly which at

3:13

Bloomberg we believe a technical

3:14

monopoly is more than 70% of a market.

3:17

Um so you know the opportunity is there

3:20

and clearly based on the share

3:21

performance and also like what they said

3:22

right you know like they're going to do

3:24

$5 billion of data center sales um next

3:27

year fiscal or calendar 27 and that will

3:30

go up to 15 by 2029

3:33

that shows progress for a company that

3:36

has historically not been in that

3:37

market. All right. So,

3:42

>> Micron,

3:43

>> yeah,

3:43

>> let's talk Micron. [laughter]

3:44

>> Yeah.

3:45

>> It was a thing that leading up to

3:47

yesterday's results after the closing

3:48

bell, every market guest that we had on

3:51

said, "This is going to be really

3:53

important." So, what was so important

3:54

about the results that has this stock I

3:56

think at an all-time high as we speak.

3:58

>> Yeah. It's not about beating the

4:00

expectations of the street. Micron reset

4:03

the expectations of the entire industry.

4:05

And by the way guys, I felt you both did

4:07

a really excellent job to try and get

4:09

the story from the earnings statement

4:11

because that's all you had to go on at

4:13

the time. It was that supply was tight

4:15

and it's going to be tight for a long

4:17

time and clearly higher prices is what

4:20

moved the needle for Micron, not

4:22

necessarily selling more units. And like

4:24

that's exactly how the call played out.

4:26

So the big wow number was that in the

4:28

fiscal fourth quarter of the current

4:29

period, Micron saw $51 billion of sales.

4:33

Um, and it highly relates to data center

4:35

and high bandwidth memory and the street

4:37

was seeing maybe 43. But the bigger

4:40

picture is like that's the best position

4:41

for Micron to be in where there's very

4:44

tight supply. They have pricing power

4:45

and everyone wants the thing you're

4:46

selling.

4:47

>> Compare and contrast. It's not good for

4:49

the companies who are in the market that

4:51

are trying to get hold of those chips.

4:53

>> Uh, I just want to know I'm leaving uh

4:55

731 cuz love the praise. I feel like

4:58

I've done it when Ed Lelo says we did a

5:00

great job on complicated company like

5:02

Micron. I'm done.

5:04

>> I thought we were going to talk about

5:05

soccer, but that's okay.

5:06

>> I'm good. I'm good. Love you. Love you

5:08

as always. And you also helping us

5:11

always make sense of all of these

5:12

reports. Um Ed Ledlo, thank you so much.

5:14

Be well. Ed Ledllo, he's the host of

5:16

Bloomberg Tech. Check it out on

5:17

Bloomberg TV. 11:00 a.m. to noon Monday

5:20

through Friday. [music]

5:20

>> Stay with us. More from Bloomberg

5:22

Business Week Daily coming up after

5:24

this.

5:28

You're listening to the Bloomberg

5:30

Business Week [music] daily podcast.

5:32

Catch us live weekday afternoons from

5:34

2:00 to 5:00 Eastern.

5:35

>> Listen on Apple CarPlay and Android Auto

5:37

with the Bloomberg Business App or watch

5:40

us live on YouTube. [music]

5:43

>> We're going to stay on technology and

5:44

how it continues to transform amid the

5:45

AI backdrop. And it takes us to someone

5:47

who's been innovating and disrupting as

5:49

well and remembers all too well the boom

5:51

and bust and changing landscape of the

5:53

dot era. He's best known as the founder

5:55

and CEO of Zingga, the social games

5:57

company that IPOed in 2011. But he's

5:59

built and invested in so much more. And

6:01

through every iteration of the internet,

6:03

the '90s boom and bust, the social media

6:05

boom of web 2.0, and now the AI era.

6:08

Carol, he's also a prolific investor.

6:10

>> He is indeed. He's invested, you'll know

6:12

these names folks, Napster, SpaceX. And

6:15

he had he held on to his shares, his

6:17

$38,000 seed investment in Facebook, now

6:19

Meta, for about half a percentage of the

6:21

company would be worth many, many

6:23

billions of dollars right now. Uh, smart

6:25

investment, I would say.

6:26

>> We're talking about Mark Pinkis. He's

6:28

got a new book out, and we know all this

6:30

thanks to the new book. It's called Life

6:31

at the Speed of Play. Launch Products

6:33

People Love. Mark Pinkis joins us here

6:35

in the Bloomberg Interactive Brokers

6:37

Studio. Welcome. Welcome. How are you?

6:38

Congrats on the book.

6:39

>> Thanks. Thanks. By the way, I really am

6:42

happy about the memory stocks.

6:43

>> Okay.

6:45

>> Well, why?

6:47

>> Well, we're diving right into the AI

6:49

trade. [laughter]

6:51

Let's not mess around. Uh I I'd say that

6:55

that it's it's a pretty simple trade at

6:58

this point. It's a belief and either you

7:01

believe that the AI infrastructure

7:04

investment is going to pay off and keep

7:07

playing out in which case all of these

7:09

companies are generationally

7:13

undervalued. It's a generational buying

7:15

opportunity if you're getting a a PEG

7:18

ratio of.3,

7:20

you know, or less.

7:21

>> Or you think that it's not going to play

7:24

out and then you should stay away from

7:25

all of it. So,

7:27

>> I'm a believer. So, it's it's funny

7:29

because we just spoke with with Ted

7:30

Oakley. [clears throat] He manages money

7:31

for Oxbow Advisors and and he sent us a

7:34

bunch of stocks and missing from there

7:35

were tech names and memory names. He

7:36

does own some tech, but he said

7:39

>> the memory got too expensive. And he

7:40

said, "I've been doing this for decade,

7:42

many decades. I remember the 1990s. We

7:44

sold Intel and I watched Intel stock go

7:47

up for the next 12 months before then it

7:49

went down." So, he understands these

7:51

cycles. Are we in one of those cyclical

7:53

moments right now?

7:55

We'll we'll know in the future, but

7:57

it'll be it's only cyclical if this if

8:03

these AI growth rates and numbers don't

8:06

play out. If they play out, I think he

8:09

would even agree that they're still

8:11

undervalued.

8:12

>> You know, Mark, one of the things that

8:13

we're thinking, we want to get into the

8:14

book and talk about the, you know, life

8:16

at the speed of play and what it all

8:18

means, but we are curious. You have been

8:20

in Silicon Valley for all of the

8:22

iterations of the No, no. Seasoned like

8:26

a great wine. Like that's how I think

8:27

it. Love talking with people who have

8:29

seen cycles, right? And can sometimes

8:32

figure out the silly from the stuff that

8:34

really matters. How do you like how do

8:36

we make sense? Like we see the money

8:38

going in.

8:39

>> Um we see the circular financing that

8:41

makes us a little uncomfortable. Um we

8:44

see the narrative around AI changing. I

8:47

get it. Disruption. This is what

8:48

happens. But help us understand like is

8:51

this a boom cycle with no bust or is

8:53

there going to be a a a breaking point

8:56

at some point or for only maybe for some

9:00

>> the if I had the perfect answer I I

9:02

[laughter]

9:02

>> you wouldn't be talking to us.

9:04

>> Yeah. Um I would have held my meta stop

9:07

too. [laughter] Um

9:08

>> you did okay. You did okay.

9:09

>> Yeah.

9:10

>> Um

9:12

>> like how do you think about some of

9:14

this? I I built a boring enterprise

9:17

software company in the middle of the

9:19

dot boom. You know, it was called

9:21

support.com, but it was actually an

9:23

early SAS

9:24

>> company. We went public on the last day

9:27

of the IPO window.

9:29

>> I'd say that then versus now.

9:32

>> We, my peers and I thought it didn't

9:35

make any sense during the dot bubble. We

9:39

we thought it was crazy what we saw

9:41

going on

9:42

>> around us. And now my peers, my smartest

9:45

friends think this makes a lot of sense.

9:48

I mean, it's it was dark fiber then and

9:50

now it's like hot GPUs. I mean, it's

9:53

actually being

9:54

>> Yeah.

9:54

>> used. It was betting on a whole consumer

10:00

um that that didn't show up or or didn't

10:03

show up yet. And now the the the

10:06

investment the infrastructure is going

10:08

to enterprises who are bottom

10:11

lineoriented and you know then it was uh

10:15

eyeballs and now it's ARR you know it's

10:17

it's actual revenues. So it's definitely

10:20

not the same. I think what we do have is

10:24

extreme volatility and I think that the

10:26

volatility comes from where it's hard to

10:30

remember this level of growth market and

10:33

when there's this level of growth like

10:35

we've seen just this week and today when

10:38

there's a negative data point or even

10:41

absence of more positives it starts to

10:44

be run for the hills and doom and bubble

10:48

and then when we see like microns

10:50

numbers and guidance all of a sudden

10:53

everyone's bullish again and we're going

10:54

to keep seeing that it's my take

10:57

>> one thing that you write about in the

10:58

book repeatedly is that

11:00

>> you know and you get into your whole

11:01

background Wharton Harvard Business

11:03

School going and working for some

11:05

investors that are household names to

11:07

the Bloomberg audience I mean Steve

11:09

Ratner John Malone I mean these are

11:12

these are legends um

11:15

>> you are constantly saying you don't know

11:17

how to code you don't know how to write

11:18

code yet you built all these companies

11:20

What struck me about about reading that

11:22

was that that kind of doesn't matter now

11:25

in a way that it that mattered when you

11:27

were building all these companies that

11:28

you write about.

11:30

>> How has the idea of like OpenAI's codeex

11:33

or Claude Code from Anthropic? How does

11:36

that completely change the game moving

11:38

forward? Well, if you pull the camera

11:41

back and even think more broadly like

11:43

how has the game of startups just

11:47

changed and how is this another step

11:48

function changed in each chapter? The

11:52

amount of capital that you needed to get

11:54

going has gone down, you know, not up.

11:58

When I was starting my first company, I

12:01

had to recruit these government um

12:03

mainframe programmers, have them learn

12:06

C++ and HTML and and I I needed a fair

12:11

amount of money to convince them to to

12:13

quit their jobs. And and each of my

12:15

subsequent companies today, you know,

12:19

you don't need that, right? You don't

12:20

need as much capital. You don't need to

12:22

go and necessarily recruit a whole team

12:25

of the world's best engineers. It's

12:27

being democratized

12:29

>> today really. Um there's it's more

12:33

possible than ever for somebody with an

12:36

amazing idea who's willing to move on it

12:39

now to really get somewhere um in a far

12:42

more capital efficient way.

12:45

>> Yeah. I mean I feel like we also saw

12:46

that in the pandemic of people just

12:48

being able to start things um while they

12:50

were home. Um I am curious about um the

12:55

speed of play because that is on your

12:56

book. Talk to us about that and the

12:58

importance of it.

13:00

>> Sure. So there was a lot of debate with

13:04

uh my my co-author Carly, my amazing

13:07

editor Hollis on the title to the book.

13:09

It was originally called Proven Better

13:11

New, which we can get into, but that's

13:13

this that's kind of some of the core

13:15

value in the book. And I eventually

13:16

said, you know, it's just that's not the

13:19

whole gist of the book. It's life at the

13:22

speed of play. And that's because it's

13:24

both it's both the the place that we are

13:28

moving into in this AI era. It's it's

13:31

it's really um I like I start the book

13:35

by saying that that this is Elon is the

13:38

one person who's already been living his

13:40

life at the speed of play and it I

13:42

definitely think he's having more fun

13:43

than the rest of us. Okay.

13:44

>> He's also working harder.

13:45

>> I want to push back on this cuz Carol

13:46

and I talked about this a lot.

13:47

>> He doesn't always look like he's having

13:48

fun.

13:49

>> Yeah. He he really I mean I think he's

13:51

been really public about having this

13:52

tortured existence and how difficult it

13:54

is to be Elon Musk and the challenges

13:56

that he struggled with. Uh

13:58

>> personally, I think more so than

14:00

professionally. Do you actually think

14:01

he's having more fun than the rest of

14:03

us?

14:04

>> I do. He's He's got an amazing sense of

14:06

humor. He's Every time I see him, he's

14:11

joking, laughing. I I was with him um at

14:16

a friend's house one night and it wasn't

14:18

long after he had bought Twitter and it

14:21

was basically like an hour of the best

14:23

stand-up comedy I've ever seen. I mean,

14:25

he the way he talked about his

14:27

experience of coming in to Twitter um

14:30

and how insane the company was was just

14:35

was just really funny. So So my point of

14:37

view, I mean whether he's having more

14:40

fun, who knows? Um, I think that that

14:44

what I what we can see, you know, I said

14:46

maybe he's the one who solved the

14:47

simulation. What we can see is that he

14:51

can almost tweet something into

14:53

existence that he said there this

14:56

traffic in LA is terrible. There should

14:57

be a boring tunneling company and then a

15:00

few months later and a few billion

15:01

dollars raised. There was and my point

15:04

is to some degree we all are on the

15:07

brink of living a a a part of that. And

15:10

and the reason I call this life of the

15:12

speed of play is that to me what the

15:15

book is really about is a product

15:18

mindset and and I think that every so

15:22

many of us have an idea but we don't

15:24

know how to pursue it or we are pursuing

15:26

an idea but the odds of success are too

15:29

low. Too many too many founders are

15:31

failing for the wrong reasons. So a good

15:34

idea is a good idea, but that's not

15:36

enough necessarily to run with something

15:38

and build something that lasts for

15:40

longer or has some significant impact.

15:42

Correct.

15:43

>> Yeah. And and it might be that you have

15:45

a good idea, but it's behind some some

15:50

like obvious mistakes that you're making

15:52

that the more junior a product maker or

15:55

founder is the and the less experienced

15:57

they are, the more likely they are to do

16:00

too much new, to just reinvent

16:02

everything. Steve Jobs talked about

16:04

this. So the the point of the book is I

16:08

I like to think that this book is like a

16:10

cheat code that you can whatever it is

16:13

you're doing, you can change your odds

16:16

of success and getting to a hit. It's

16:18

the book I wish I had early in my

16:20

career.

16:21

>> So is this for founders? Like who do you

16:22

like think about? It sounds like it is

16:24

for people who have an idea or want to

16:26

run with something, right? This book is

16:28

every one of your listeners right now

16:30

probably has some instinct. They have

16:33

some sense. It's either a specific idea

16:36

or a sense that something could be

16:39

better. And the the book is the point of

16:42

the book is that they should what is

16:46

when you think what should they do with

16:49

that instinct? Very few of them will act

16:52

on it and and turn that into a product

16:54

or a company. And then their odds of

16:57

success will be so low because they're

16:59

going to take one shot on goal and it's

17:01

probably going to miss.

17:03

>> Well, I I like how you write about that

17:04

in the book because you use the idea of

17:06

Uber as an example in the book. You you

17:08

had the idea for Uber in 2002.

17:11

SMS dispatching to a cab. Uh you're not

17:14

Travis Kalanick. Uh you did not start

17:16

Uber.

17:17

>> No.

17:17

>> So you you you make this argument that I

17:19

think a lot of people everybody has

17:20

these ideas.

17:21

>> I had that idea. Just because you had

17:23

that [laughter] idea doesn't mean you

17:24

actually create the company. Why did Why

17:26

was Travis able to do it, but Mark

17:28

Pinkis, who was in Silicon Valley in

17:29

2002, wasn't able to do it.

17:32

>> I can't tell you why Travis was, but I

17:35

can tell you why I wasn't. I didn't.

17:38

First of all, I looked at the world as

17:40

it existed, not as it could exist the

17:43

way that Travis did. And and by the way,

17:45

in 2002, there wasn't a mobile, you

17:47

know, smartphone. Then, and I thought

17:49

about it in these conventional ways. Oh,

17:52

I'm going to deliver your order to the

17:55

taxi broker who will call a cab. I

17:58

didn't ever think Travis's idea of the

18:00

gig economy. I'm going to let anybody

18:02

become a driver and have a driver within

18:05

a minute of you.

18:06

>> That was that was brilliant. But but I

18:10

had an instinct an instinct that we

18:13

should be able to

18:15

>> order a phone through our order a taxi

18:18

through our phone. And you know the the

18:21

the importance the point I'm trying to

18:24

get people to focus on in the book is

18:26

that if you assume your instincts are

18:28

right you know 95% of the time and your

18:31

ideas right at best 25% of the time what

18:34

would you do with that information you

18:36

know it's like a time machine.

18:38

>> You write about these instinct veins

18:40

deep sources of insight about human

18:42

needs and behavior that can spawn

18:44

multiple product ideas even whole new

18:47

industries. Is that what AI is right

18:49

now? Like what we're doing? Is that what

18:51

that is? Or is that

18:52

>> AI is beyond an instinct vein? I mean a

18:54

AI is is a fundamental shift in in

18:59

computing. I mean the way that the

19:01

internet was and so so I wouldn't

19:04

>> it's not apples to apples. Yeah. Yeah.

19:06

No, it's interesting. Um you know, one

19:08

of the things that I find also is some

19:12

of the things you talk about like

19:14

leadership. You talk about

19:16

micromanagement is beautiful. And I

19:17

think about how many times when you

19:19

think about leaders that it's like don't

19:23

micromanage people.

19:24

>> Yeah. I was told that so many times,

19:26

[laughter]

19:26

>> right? No, think about it. Like you

19:28

bring in, you know, you know, experts or

19:31

consultants and they're like don't

19:32

micromanage your people. Why is it so

19:34

important?

19:35

>> Why is it so important that you do do

19:37

that you do micromanage? Yeah. Because

19:40

the the point I'm trying to make here is

19:43

that at the end of the day, what matters

19:46

most is your customer experience.

19:48

>> Yeah.

19:49

>> Not how you delivered it. And so the

19:52

point to me is deliver the best possible

19:56

customer experience any way that you get

19:59

there. And if it's through

20:03

micromanagement, if if you can guarantee

20:05

the quality, if you can guarantee the

20:07

delivery because you micromanaged,

20:09

>> then by all means do that. I'm like, I

20:12

don't care that McDonald's served 15

20:14

million burgers today. That doesn't make

20:16

mine any better. I if I want the Colonel

20:18

cooking mine, you know,

20:20

>> yeah,

20:20

>> individually.

20:21

>> Well, that that's a major theme in the

20:23

book is is sort of throwing out this

20:24

idea of the minimum minimal viable

20:26

product. Yeah. I'm not going to repeat

20:28

what the chapter's called because no,

20:30

>> we're FCC regulated here, but um

20:32

>> we're family.

20:33

>> Well, that chapter you can say was an

20:34

MVP trap. So, [laughter] other chapters,

20:37

no.

20:37

>> Okay, cool. Maybe. Yeah, great. There it

20:39

is. Um why we've been sold this idea

20:43

though of iterating and Silicon Valley

20:45

sort of throws something, sees, you

20:46

know, see if what sticks and iterates on

20:48

that over and over again. Why wasn't

20:50

that ever right for you? Well, the the

20:54

original concept that Eric Reese had of

20:57

minimum viable product and moving fast

20:59

and being in the market is is a great

21:02

concept. It's just that we now we don't

21:06

the the point I make is that we no

21:08

longer have time to go wait the the

21:13

learning is too slow. If we build a

21:15

minimum viable product, there's hope in

21:18

the word viable that this might be your

21:20

launch product and then you're going to

21:21

invest more in that product than you

21:23

should. And I like to say just build it

21:25

wrong before you build it right. Just

21:28

build to learn. be whatever gets you

21:32

signal from your customer the fastest is

21:36

is and and in the age of AI we can

21:40

>> prototype something or test something so

21:42

much faster but it's dangerous what I'm

21:46

seeing with AI is less that people are

21:49

using AI to test and learn faster but

21:52

more build faster and so if I can build

21:55

something now in 3 months instead of

21:56

three years that's so alluring ing that

22:00

I might go do that but I don't have

22:03

three months to to learn I'm wrong. Does

22:05

that make sense? So I don't have that

22:07

viable word is is tricky.

22:09

>> Well I think you you had also shared in

22:11

the book uh the example of Twitch and

22:14

building you know the founder of Twitch

22:16

their team was was changing their

22:17

product every two or three days at that

22:19

point and

22:19

>> yeah they were twitchy

22:20

>> and that was that ended up being a good

22:22

thing for them.

22:23

>> Yeah

22:23

>> they got immediate feedback.

22:25

>> Yeah. Well, they I I don't even know if

22:27

it was getting any customer feedback.

22:29

The feedback was just from themselves.

22:30

It was I don't like this product idea

22:33

anymore. Let's switch. And so that's

22:37

also important. And part of I guess part

22:41

of the the what makes it so hard to be a

22:44

founder, to be a CEO, is that we're

22:46

supposed to express confidence when we

22:49

don't personally feel confidence, right?

22:51

And so, how do you come in on Monday and

22:54

what if you just what if you learned

22:57

something in the past week that just

22:58

said, "This product isn't quite right or

23:01

it's finally part built and you're like,

23:03

I'm just not that into it."

23:05

>> Right?

23:05

>> Do you go to your team on Monday and

23:07

say, "Guys, I know I I got you to work

23:10

nights and weekends for this, but I

23:12

don't think this is right anymore." Or

23:14

do you say, "Well,

23:16

I don't want to demotivate my team and

23:18

so I'm just going to continue on this

23:20

path for this whole product cycle or I

23:23

can't I'm afraid to tell my investors

23:24

who backed me that I was wrong." Right?

23:27

And so the question is, are you more

23:29

committed to the intellectual honesty or

23:32

to harmony?

23:33

>> I would say intellectual honesty. I

23:35

mean,

23:35

>> most people would not most people would

23:37

not act on it.

23:37

>> I know. I know. But it's like I don't

23:39

know. At some point you need to to be

23:40

doing that. Hey, we're talking with Mark

23:42

Pinkis, founder of Zingga. He's got a

23:43

new book out. It is entitled titled Life

23:46

at the Speed of Play. Launch Products

23:48

that people um love. Are there products

23:51

I mean the product cycle is it getting

23:53

shorter or or longer especially when it

23:55

comes to technology because I feel like

23:57

there are things that people are so into

23:59

and then they move on to the next thing

24:00

and there's so much out there. How

24:02

>> from a consumer standpoint? Well, I I

24:05

say in the book that I think there's a

24:07

metric that I don't know anyone but me

24:10

who focuses on it and my former teams.

24:13

Day 365 retention. So, if a 100 people

24:16

were using your product today, you know,

24:18

or a year ago today, how many would

24:20

still be using it today? And and it's

24:24

such a hard thing to build against

24:26

because we don't have a year to wait,

24:28

>> right? But you never would have made

24:30

your [snorts] that that maybe works from

24:32

a product perspective of building a

24:33

company but from a venture capital

24:34

perspective that doesn't work. You write

24:36

about what attracted you to Mark

24:37

Zuckerberg when he was a teenager still

24:40

was that they were able to sign up you

24:42

know schools 20% on the first day the

24:45

next 80% the next week like that

24:48

happened instantly.

24:49

>> Yes. didn't know but but it turns out

24:53

that this is um necessary but but not

24:57

necessarily complete and sufficient is

24:59

that that if you have 60% engagement 60%

25:04

of your users show up every day which

25:07

has been true to this day with Facebook

25:10

>> the likelihood that you have high day

25:13

365 retention is is highly correlated it

25:17

may not be the case um but it's highly

25:20

highly correlated.

25:21

>> Did you knew know the minute you like

25:23

spoke with him that this was just

25:25

something remarkable meaning Facebook?

25:27

>> Yes. Yes. And and by the way so would

25:29

you both of you? So people who point to

25:34

the fact that they invested you know

25:36

early on in in Facebook or these

25:38

companies as a sign that they're a great

25:40

investor. It's it does not necessarily

25:42

mean that they are a great

25:45

you know that their judgment is so great

25:47

because we all would have said yes their

25:50

access is very impressive you know

25:52

>> that that they had and you didn't have

25:54

so um I but but but here's what's so

25:58

painful if you think about it that story

26:00

is less like kudos to Mark that he

26:02

invested and it's more like like Mark

26:06

how is it like think about this how is

26:08

it that in 2004 before when I met him um

26:12

I was doing tribe I was doing one of the

26:14

first social networks right before I

26:16

started before he did

26:18

>> how did I manage to fail it's an act of

26:23

willpower that I failed it wasn't just

26:25

Facebook there was eight or nine social

26:28

networks there was BBO tagged MySpace

26:32

>> Fster which I invested in they all

26:35

worked I had to I had to pick one idea

26:38

that didn't work and and stoically,

26:41

heroically stick with it no matter what.

26:43

I don't care.

26:45

>> We are with Mark Pinkis. He's the

26:47

founder of Zingga. He's the author of

26:49

the new book. It's out this week. Life

26:51

at the speed of play. Launch products

26:53

people love. I want to pick up with a

26:54

headline that we just heard from Amy

26:57

just now. It's coming from the New York

26:58

Times about how Open AI is leaning

27:01

toward waiting until next year for its

27:04

IPO. Rob Copeland and Mike Isaac writing

27:06

this over at the New York Times saying

27:07

that they're holding off on their

27:09

initial public offering until next year.

27:10

Three people involved in the company's

27:12

deliberations said uh it punctuates an

27:14

uncertain future for fast rising AI

27:17

giants. That's again coming from the New

27:18

York Times. Mark, you're an investor in

27:21

in open AI. You understand also what

27:22

it's like to take a company public. You

27:24

did that with several companies and

27:25

different periods of your professional

27:27

life.

27:27

>> That means everything sometimes, right?

27:29

>> Sure. Yes. Like the last day of the IPO

27:32

window. Yeah. [laughter]

27:34

um just your your thoughts on on Open AI

27:36

and its path to becoming a public

27:38

company.

27:38

>> I I

27:40

I'd say

27:42

I I the one hand I'm I'm not sure how

27:46

much the timing matters other than if it

27:50

changes their access to capital. So I

27:53

don't think they can afford to get you

27:55

know significantly behind in the you

27:58

know capital and buildout race but it's

28:02

not clear that you know you have to

28:05

necessarily go public these days if you

28:07

look at the the sizes of investments. So

28:10

I I think OpenAI is just an amazing kind

28:15

of generational

28:16

>> company and I think people who count

28:18

them out uh and say this is all

28:20

anthropic uh are shortsighted and I

28:24

think that open AI has to catch up on

28:28

the coding side. I think we'll see that

28:32

happen with codeex and they have an they

28:35

have a clear advantage on the consumer

28:36

side which I think is currently being

28:39

under uh weighted. I think people

28:42

wrongly think this is all just about

28:44

coding and enterprise

28:46

>> and I just think that is where the

28:49

action is right now. That's where the

28:51

most revenue growth is right now. But

28:54

there's no question to me that consumer

28:57

the consumer side of AI will be just as

29:01

big if not bigger. So, so I know CO2 put

29:04

out a report calling a $6 trillion

29:05

market and I thought it was interesting

29:07

that they they capped consumer at 500

29:10

billion consumer AI out of 6 trillion.

29:13

That doesn't make sense to you. Two

29:14

trillion just for the engineering the

29:16

coding side. No, I I that doesn't make

29:19

any sense to me. That's that's not the

29:21

way we've seen the internet play out.

29:22

You seem to be someone who can think

29:24

super big about things. Like how should

29:27

we be thinking about AI and how it

29:30

changes our word world as consumers?

29:32

Like how is it going to at home, at

29:34

work, at play? Like how is it going to

29:36

change our our world? Or is it just

29:38

another amped up way of communicating

29:42

online?

29:43

>> I don't know. What is it? How do you

29:44

think about it?

29:44

>> It's not a way to communicate online yet

29:46

because it's still oddly a singleplayer

29:49

experience, right? We're not we're not

29:51

>> in the AI, we're not in the GPTs and the

29:54

chats

29:55

>> together. But but if you break your

29:57

question down, I think first if you

29:58

think, you know, how will it impact uh

30:01

our jobs and the job market, I think

30:04

we're very very quickly uh transitioning

30:07

from a knowledge worker economy to

30:11

something else we don't have a name for,

30:13

but it's going to be, you know, a prompt

30:15

worker. It's going to be about

30:18

>> being generative and we're going to be

30:22

uh whatever our job is we're it's going

30:24

to be very quickly I mean it's probably

30:27

changed a lot for you guys and right the

30:29

amount that we have to rely on the AI

30:33

but the amount of leverage we get and

30:35

the amount of that it's it's moving from

30:39

a value on knowledge to a value on

30:43

questions and curiosity You also have to

30:46

have knowledge to ask a smart question

30:47

for sure,

30:48

>> right? Like, and we're learning too, and

30:50

we talk about this a lot, that um I

30:52

forget who the one of the interviews we

30:54

recently had that like a really good

30:55

question with AI is going to be several

30:58

paragraphs long, right? Like if you

31:01

really want to get a a smart useful

31:03

information,

31:03

>> you guys both impressively read my book

31:06

and you have great questions. And I

31:08

think if you'd relied on the AI, it you

31:10

know, and I'll say even writing the

31:12

book, I had this lovehate relationship

31:14

with AI that if at first it's magical

31:16

and you're like, "Oh my god, it can take

31:18

this long talk I just gave and condense

31:21

it or can or editing is so painful, word

31:24

smithing, and then you start reading it

31:26

and it's getting it starts to get

31:27

homogenized and it's like it starts to

31:30

feel soulless and it's like where do my

31:32

voice go in this?" And I even tried a

31:33

style guide and I'm like, "I want you to

31:35

sound like my voice." And eventually I

31:36

was like, I don't want you to change my

31:37

words unless you absolutely have to. And

31:40

then eventually I said, you know what,

31:41

it's really helpful for some things.

31:43

It's part of a process to take a long

31:45

talk track and condense it to make it

31:47

more organized. But I started to really

31:50

kind of like a vinyl record, appreciate

31:52

the imperfections of how I talk and say,

31:57

you know what, that's that's my voice

31:58

and that's how you know it is me,

32:01

>> right? You know, a major theme in the

32:03

book is is making decisions that your

32:05

future self will respect. You have this

32:06

this framework, the book of life. Um,

32:08

which is, I think, a really important

32:09

way to set up the book. I I want to talk

32:12

about that just in the last couple

32:13

minutes we have in the context of you

32:14

being so public in 2024, coming out and

32:17

saying after so many years and so many

32:20

millions of dollars donating to

32:21

Democrats, I am now coming out in

32:24

support of President Trump. It was

32:26

surprised, I think, a lot of people to

32:28

see you do that. How did your your book

32:30

of life framework and your idea of like

32:32

looking back at that decision inform

32:34

that decision at the time?

32:36

>> It's it's such a good question. You

32:38

know, it's it's definitely not something

32:39

I signed up for in my book of life.

32:43

However,

32:45

it's it is important to me. intellectual

32:48

honesty is important to me and being

32:52

willing to take an unpopular stand and

32:56

and I feel like if if I can't do that,

33:00

if I'm so scared because of group think

33:03

and the consequences of being uh taking

33:07

a very unpopular position in my, you

33:10

know, a lot of my communities,

33:12

>> then who can and and it was actually my

33:15

daughter Georgia. I I I didn't decide

33:17

until the Sunday before the election

33:19

that I was going to vote for Trump. I

33:22

decided I was definitely not going to

33:24

vote for Kamla and I had lost faith in

33:26

the Democrats um and the mainstream

33:30

media establishment

33:32

um that that I stopped trusting in that

33:36

whole process and and I but I was also

33:39

being very transparent and public on

33:41

Twitter. I I started I wrote a post in

33:43

the free press saying not that I support

33:46

Trump but that that Biden at that time

33:49

in late July of 24 felt even riskier

33:53

than Trump was right. And that alone

33:55

started a storm if I can say that

33:58

word.

33:58

>> It's out there.

33:59

>> Okay. So [laughter] it's too late to

34:00

take it back. But but Georgia came to me

34:03

on a Sunday before the election. She

34:05

said, "Dad,

34:07

you know you're going to vote for Trump

34:08

at this point." And I said, "Yeah, I

34:10

think you're probably right." and she

34:11

said, "Then you have to tweet that

34:13

because you've been so open and

34:15

transparent." And I said, "Yeah, you're

34:17

right." And so then I I wrote a whole

34:18

post. Um, it was, I think, the most

34:21

viral post I've ever put up because I

34:23

think it was it was a oddly a touchstone

34:26

for a lot of people because I was kind

34:28

of like this

34:30

>> de big Democratic donor and breaking

34:33

ranks, you know, was and I'm kind of

34:37

like, it's funny to call it, but I'm

34:39

kind of like part of the like rank and

34:41

file Silicon Valley founders. Um, and so

34:45

it was, I think, a little um, scary to

34:49

some of the establishment to see this

34:51

crumbling.

34:53

>> Um, and and I put the post out and I put

34:56

my reasons out and it was on the front

34:58

page of the New York Post like the next

35:01

day. I didn't think it would be news.

35:03

Um, and you know, and it it really

35:07

really uh was much louder um, than I

35:11

anticipated. Um, and but I was like, you

35:15

know what? If if I'm gonna do it, I'm

35:16

I'm I'm not gonna It's silly that we

35:19

should have to hide our political views

35:22

because we're tagged with an identity.

35:24

And and I'll just say this.

35:26

>> I'm not on any team. And I said that

35:29

throughout. I said, I'm not on team

35:30

Democrats. I'm not on team MAGA. It

35:33

sounds cheesy. I'm team America. I'm

35:35

team my family community. And in this

35:38

environment where we just saw three

35:40

pretty extreme left Democratic

35:43

socialists win in New York, I don't know

35:46

that any of us can really say we're

35:48

identified with one party because the

35:50

who the party is is really shifting. I'm

35:53

a Chicago liberal. That's I've always

35:56

been that. I'm, you know, socially

35:58

liberal. I want people to have their own

36:01

rights and freedoms. And I'm fiscally

36:03

and economically

36:05

conservative. I want to see a

36:06

responsible government. I think I'm

36:08

stating some obvious things here that

36:11

80% of people 100 I don't know 90% of

36:14

people agree with. And so for that to

36:17

then define me as being right, you know,

36:19

far right and to hear journalists say I

36:22

need to get a I couldn't find a far

36:24

right founder in San Francisco. Can I

36:26

interview you? And I'm like, you want

36:27

me? I'm far right. I'm like I I'm still

36:30

here. Nothing's changed. So

36:31

>> it's a weird Yeah.

36:32

>> Yeah. It's so it's and and the

36:34

partnering with my future self is I I

36:37

even though it was painful and there was

36:40

some dislocations and I did um lose uh a

36:44

couple of dear friendships over it. Um I

36:49

I am happy that Mark 2024 No, I'm happy

36:53

Mark 2024 uh took a stand.

36:58

>> God, I feel like that's a perfect place

36:59

to to wrap. We don't really want to

37:01

wrap. We hope you will come back.

37:03

>> This is really fun and I was not

37:06

expecting you guys to have read my whole

37:08

book and have all of these uh really

37:11

insightful questions. So I I hope you

37:14

guys uh follow the the you know the

37:19

format my book. I hope that you take an

37:21

idea, prosecute it and as a side hustle

37:24

you build a huge business and then this

37:26

becomes your hobby.

37:28

>> I love it. I love it. Um, you give us a

37:30

lot to think about and I I have to tell

37:32

you we still have a bunch of questions

37:33

so you like I mean it. Please come back.

37:35

We would love it. Um, Mark Pinkis,

37:38

founder of Zinga, of course, founder of

37:39

several companies. Um, but his new book

37:41

is Life at the Speed of Play. Launch

37:43

Products People love. Um, there's a lot

37:45

in here and a lot of great stories.

37:47

>> Stay with us. More from Bloomberg

37:48

Business Week Daily coming up after

37:50

this. [music]

37:55

>> You're listening to the Bloomberg

37:56

Business [music] Week Daily podcast.

37:58

Catch us live weekday afternoons from

38:00

2:00 to 5:00 Eastern.

38:01

>> Listen on Apple CarPlay and Android Auto

38:04

with the Bloomberg Business [music] App

38:05

or watch us live on YouTube.

38:09

>> I'm looking at shares of Micron. They're

38:10

off their best levels but still up 15

38:12

1.5%. Remember this was yesterday. We

38:14

were just talking about this is the one

38:15

to watch. Cameron [music] Christ said

38:17

after uh Nvidia and Alphabet, this is

38:20

the most important publicly traded

38:21

company out there.

38:22

>> Record high intraday. I mean Yeah.

38:24

>> Yeah. So this year

38:26

>> Mhm. Micron after today's rally up 325%.

38:30

>> Just a little bit of a rally there.

38:32

>> Qualcomm shares jumping too after the

38:33

chipmaker forecast annual sales of more

38:35

than $15 billion from AI components and

38:37

data centers by fiscal 2029. We got to

38:40

bring back Mandep Singh. He's global

38:42

head of technology research for our

38:43

Bloomberg intelligence team. He joins us

38:45

from New York. I asked you the question

38:47

yesterday Mandeep. Does does Micron make

38:48

or break a cycle? You said no. It's it's

38:52

not big enough to do that. But I think

38:54

it's fair to say after those results, it

38:55

it gives us a little idea of where we

38:58

are in the cycle. Is that fair?

39:01

>> It is. [clears throat] And look, uh, the

39:03

picks and shovels continue to show very

39:06

robust growth. And in the case of

39:08

Micron, uh, they now have, you know, 100

39:12

billion worth of contracts with 16 of

39:16

their customers. These are take or pay

39:18

contracts. And uh it gives them a lot of

39:21

revenue visibility along with the

39:23

margins. I mean 85% gross margins, 81%

39:27

operating margins. The kind of leverage

39:30

these companies are showing is just

39:32

phenomenal. So these kind of quarters

39:35

are hard to come by, but when they do,

39:37

you just got to applaud, you know, uh

39:39

what these companies have delivered.

39:41

>> Well, based on what they reported, how

39:43

many more quarters can they have like

39:45

this? I mean, what's the what's the uh

39:48

runway that you see for companies like

39:50

Micron?

39:51

>> It feels like it's hard to top this one

39:54

in terms of, you know, the magnitude of

39:56

the beat. So, I would say in terms of

40:00

upward revisions, uh we probably won't

40:03

get these kind of beats. But look that's

40:05

where when a company is going growing

40:08

triple digits

40:10

it's hard to imagine a scenario where

40:12

there will be a bust and suddenly uh you

40:15

know they will have negative growth or

40:18

anything like that which has happened

40:20

with micron. So uh you know but the kind

40:23

of demand drivers we see and I know the

40:26

comparison with Micron uh memory being a

40:29

commodity right now it definitely

40:32

doesn't feel anywhere close to a

40:34

commodity in terms of what they are

40:36

selling and the pricing power they're

40:38

showing. So until we get a big supply

40:41

expansion I just don't see how they will

40:44

you know suddenly uh go bust and uh you

40:47

know stop growing uh in in a few

40:49

quarters. Uh, are you surprised by the

40:51

the stock reaction? I'm not I'm you

40:52

know, we're not going to ask you to give

40:54

a rating on the stock or a price target.

40:56

Bloomberg Intelligence doesn't do that,

40:57

but I'm I'm just curious about your view

40:59

on how the stock has reacted.

41:01

>> No, I think these are strong

41:03

fundamentals and uh the stock should be

41:06

up on a print like this. So, if

41:09

anything, you know, they added more

41:11

visibility. They said they will report

41:13

RPO going forward, remaining performance

41:16

obligations and and that backlog number

41:19

should comfort a lot of the investors

41:21

who may think oh suddenly things may

41:24

change a couple of quarters from now and

41:26

we know the stock will be forwardlooking

41:28

but in this case uh you know it sounds

41:31

like uh the uh supply constraints will

41:34

exist through 2028 and so if a company

41:37

is printing you know earnings like this

41:40

then the stock deserves to be traded at

41:43

a higher multiple and uh I I think

41:45

that's what you're seeing in the

41:46

reaction today. Mandep you know speaking

41:48

of the semiconductor space the company

41:50

that we have talked about it feels like

41:52

non-stop for the last two two and a half

41:54

years is Nvidia and this one we have

41:57

seen it come down a lot um since about

42:01

mid let me just look at my numbers here

42:02

midMay it's down almost 20% a little bit

42:05

more than 17.5% here um I understand

42:08

we're talking a lot about memory but

42:10

what's going on with Nvidia

42:12

>> I mean look at you know Nvidia's supply

42:15

chain the fact that memory prices are

42:17

going up

42:18

>> that at some point has to reflect uh in

42:22

Nvidia's margins and they have to also

42:25

pay up uh you know Micron and Skhinix

42:28

and Samsung for higher memory costs even

42:30

if they are the largest player and they

42:33

have signed these long-term agreements

42:35

with Micron but I I feel like that's

42:38

where everyone is anticipating you know

42:40

the margins can't go any higher because

42:42

of all the inflation we are seeing in

42:45

the component pricing

42:46

and they are seeing more competition.

42:49

Open AAI just yesterday announced a new

42:51

chip with Broadcom. Granted, it will

42:54

take 9 months to tape out. But if you

42:57

start looking, you know, three four

42:58

quarters out, then you start to feel

43:01

like Nvidia will probably have more

43:03

competition and that's where I think

43:06

it's getting reflected in the valuation

43:08

and the stock reaction.

43:09

>> Well, speaking of stock reaction, Apple

43:11

shares fell as much as 6.6%

43:14

in today's session. this after the

43:16

company raised the prices of Macs,

43:17

iPads, home devices, and the Vision Pro

43:20

to offset cost hikes caused by a

43:21

shortage of the memory chips and storage

43:24

that we're talking about with regard to

43:26

Micron and others. Why is the stock

43:28

reacting this way? Does this does this

43:30

show or why are investors reacting this

43:32

way? Does it show that there's not

43:33

confidence that Apple has the pricing

43:35

power and there could be some demand

43:36

destruction?

43:38

Yes, I I think that demand destruction

43:40

aspect is definitely obvious here

43:43

because uh you can't expect to sell the

43:45

same units if your product is you know

43:48

20% more uh expensive uh now and so even

43:52

though Apple has the pricing power and

43:55

you know a lot of their customers will

43:58

uh be willing to pay that price but the

44:00

unit growth will have an impact and uh

44:02

look for some of the other hardware OEMs

44:05

they may not even have the option to

44:07

raise prices because the demand

44:09

destruction could be severe. I don't

44:11

think that's the case with Apple because

44:13

of their customer base. But this memory

44:16

pricing will trickle down to uh uh you

44:19

know the margins of a lot of hardware

44:22

OEMs and and that's where uh you know

44:25

there is no free lunch in terms of

44:27

paying up for memory and uh you know uh

44:29

the margins being stable for everyone

44:32

else out there and I I think you will

44:34

start to see the margin impact uh kind

44:36

of flow through for a lot of the

44:38

companies that don't have the pricing

44:39

power

44:40

>> mandep these higher prices do they

44:42

continue for a lot longer you know we

44:44

talked with you about this all the time,

44:46

the imbalances between supply and demand

44:48

right now and the semis always being a

44:51

little bit cautious about, you know,

44:52

further building out because they've

44:54

seen the booms and busts. Should we

44:55

assume that these higher prices are

44:57

going to persist longer um rather than a

45:00

shorter time period?

45:02

Well, the risk here that I see with that

45:05

uh you know continuous pricing story is

45:08

uh if let's say there is no big use case

45:11

after coding agent. Right now everyone

45:13

is bowled up about coding agent and how

45:16

large addressable market there is for

45:18

coding agents. What if there is no new

45:20

uh use case beyond coding agents? That's

45:23

when let's say a Meta or a Microsoft

45:25

pulls back on their capex then suddenly

45:29

I think that enthusiasm is is going to

45:32

settle down and and probably people

45:33

won't be as excited about their capex

45:36

because right now a lot of the

45:37

hyperscalers are eating the memory

45:39

prices in their capex

45:41

>> and I think that is the risk to that

45:43

story that if one of them pulls back

45:46

because they're not seeing the gains you

45:48

will start to see some impact on memory.

45:51

All you need to know folks right there

45:53

Mandep Singh [laughter] thank you so

45:55

much. Uh that of course is the global

45:57

head of uh technology research for our

45:59

Bloomberg intelligence team. He is Mand

46:01

Singh joining us right here in New York.

46:02

>> This is the Bloomberg Business Week

46:04

daily podcast available on Apple,

46:07

Spotify, and anywhere else you get your

46:09

podcasts. Listen live weekday afternoons

46:12

from 2 to 5:00 p.m. Eastern on

46:14

Bloomberg.com, the iHeart Radio app,

46:17

TuneIn, and the Bloomberg Business app.

46:20

You can also watch [music] us live every

46:22

weekday on YouTube and always on the

46:24

Bloomberg terminal. [music]

Interactive Summary

This episode of Bloomberg Business Week Daily covers the rapidly evolving AI landscape, featuring insights from Qualcomm CEO Cristiano Amon on data center strategies, expert analysis from Ed Ludlow on industry performance, and an interview with entrepreneur Mark Pincus regarding his new book 'Life at the Speed of Play.' The discussion explores the AI investment cycle, the impact of memory chip shortages on tech companies like Apple and Micron, and the broader implications for the future of enterprise and consumer AI.

Suggested questions

4 ready-made prompts