HomeVideos

Ruth Porat - President of Alphabet | Podcast | In Good Company | Norges Bank Investment Management

Now Playing

Ruth Porat - President of Alphabet | Podcast | In Good Company | Norges Bank Investment Management

Transcript

1218 segments

0:00

[Music]

0:01

hi everybody I'm Nicola tangan the CEO

0:03

of the Norwegian SN wealth fund and

0:05

today I'm in really good company because

0:07

I'm here with Ruth porat the president

0:09

and chief investment officer of alphabet

0:12

which is the parent company of Google

0:13

now of course Google needs no

0:15

introduction but alphabet in addition

0:17

also makes self-driving cars AI chips it

0:20

owns YouTube and many more things and we

0:22

own 1.3% of the company totaling roughly

0:26

$30 billion so big welcome Ru it's great

0:29

to be with you thank you who is going to

0:31

win the AI race well we feel we're um in

0:34

a really strong position for a number of

0:38

reasons I think when you look at the

0:40

history of Google focused on

0:44

AI uh we started many many years ago and

0:48

at this point we have a very

0:49

differentiated approach which starts

0:52

with the extraordinary team we have led

0:54

by Demis aabas who obviously just won

0:56

the Nobel Prize that goes to the

0:59

strength of the mod models you look at

1:01

what we're doing on chips with our tpus

1:04

and then we're already really using it

1:07

across our various platforms so billions

1:10

of people are benefiting from AI so this

1:12

full stack approach we think is an

1:14

important element that being said what's

1:17

exciting to see is how much Innovation

1:20

there is broadly and so what we're

1:22

looking at is the opportunity I think

1:25

collectively globally to have an unlock

1:28

from the upside given the Innovation

1:31

that we're seeing not just at Google but

1:33

at other places so is it now a matter of

1:36

who's got the best people or the most

1:38

money or the best data or the best chips

1:40

what is it what is it about well the

1:42

reason I started with what we're really

1:44

proud of our full stack approach from

1:46

talent and models to the platforms and

1:49

chips is it really is taking this full

1:51

stack approach and I think very

1:54

importantly um this intense focus on

1:56

Innovation and continuing to push forth

1:59

you know one of the things that really

2:00

struck me when Demus was awarded the

2:03

Nobel Prize this last fall and he was

2:07

asked what was it that really started

2:10

him on this journey and he said what

2:12

motivates him is to take on the most

2:15

intractable what were thought to be the

2:17

most intractable problems facing

2:19

humanity and I think that ethos is

2:22

another really important element of how

2:25

one continues to drive forth and have

2:27

the biggest impact with AI do you think

2:30

generative AI can be monetized the same

2:32

way as search is I think generative AI

2:35

is continuing to enable us and others to

2:40

uh expand kind of the quality of what is

2:43

being delivered across a a various a

2:47

wide set of opportunities so just to

2:49

give you two examples you know I'm often

2:51

asked where do we where are we investing

2:54

ex you know extensively and probably not

2:58

surprisingly you know this well given

2:59

how close you are to all that we've been

3:01

doing but search remains so core to who

3:04

we are and we've evolved search

3:06

meaningfully over the years and we're

3:08

continuing to apply generative AI to

3:10

what is that experience that one has um

3:14

when you search and what we're finding

3:15

is that it opens the the types of

3:18

queries that are being explored it goes

3:22

for a deeper richer engagement so that's

3:24

one area the other area that's really

3:26

important is what we're seeing on the

3:28

Enterprise side and the abil to help

3:30

companies and the public sector

3:32

transform the businesses and their

3:35

approach whether it's engaging with

3:37

customers or on the efficiency side or

3:39

on risk analytics or with constituents

3:42

each of those provides opportunity for

3:44

monetization some of your competitors

3:46

say that you were a bit slow out of the

3:47

box when it cames to AI models but um

3:51

your latest Gemini is uh it's it's

3:54

phenomenal so so what happened here well

3:57

a search Google search for decades has

4:00

really stood for Extraordinary quality

4:03

it's what everybody around the globe

4:05

expects of us when you go to Google

4:07

quality answer uh very rapidly surfaced

4:11

for you in the most potent way and one

4:15

of the very important questions for us

4:17

slind talked about this in the early

4:20

days of generative AI internally we were

4:23

all talking about uh the risk of

4:26

hallucination and that term now is very

4:28

well known quite broadly and one of the

4:31

concerns is if you in the middle of the

4:33

night wake up your child is sick you

4:36

want to figure out how much TI andol to

4:38

give to a three-year-old there can be no

4:40

margin for error yeah that's what our

4:43

brand stands for and so it's very

4:45

important to us to make sure that as we

4:47

were evolving and applying generative AI

4:50

we did it in a way that was consistent

4:52

with the quality that's expected

4:55

appropriately from Google and I

4:57

appreciate your question because what

4:59

you've seen

5:00

is the really ongoing momentum in models

5:03

and introducing models more broadly

5:06

externally what we've done with for

5:08

example AI overviews where when you

5:10

search you'll get this kind of an AI

5:13

cockpit is the way I like to think about

5:14

it you're seeing more and things like a

5:18

model advancement that can be applied in

5:20

in other uh in other applications and so

5:23

we're excited about the momentum that

5:24

people have seen through 2024 and what

5:27

is ahead in 2025 talking a you told the

5:30

late you call the latest model for um

5:33

the model for the agentic area what what

5:35

does that mean or what what do you put

5:37

in that so that's one area that I I know

5:39

SAR and Demis and the team are really

5:41

excited about which is how can AI

5:44

actually be an agent working on our

5:46

behalf how can it help you do some of

5:49

the administrative tasks that are going

5:51

to make your life easier and actually

5:53

free you up to do something else book a

5:56

reservation for you um research

5:58

something for you how can it even be

6:00

applied for example in science to do

6:03

some of the basic inquiries so set it

6:05

out to do something to work on your

6:07

behalf so three three years from now

6:09

what does your day look like how is it

6:10

changing your life you know that's it's

6:13

a great question as it relates broadly

6:16

to AI I think one of the most um

6:20

important things for all of us is that

6:23

AI can really be an assist it can be

6:26

operating leverage for every one of us

6:28

you when I started my career on Wall

6:30

Street how was it as you were moving

6:32

from Bas you know big computers to

6:35

laptops to phone this becomes operating

6:38

leverage if appropriately applied and so

6:41

what one scientist once said to me that

6:43

they view this as augmented intelligence

6:45

not artificial intelligence and wish

6:48

that had been the word that was used

6:49

that's what it is for each one of us and

6:51

I think what's really important is to

6:54

understand that in fact it's not that we

6:56

will be replaced by AI in our roles but

6:59

we can be replaced by someone who's

7:01

using AI who's getting that operating

7:03

leverage if we're not and so it's so

7:05

important for people to just start

7:07

experimenting just start playing so that

7:09

you're on that Learning Journey now you

7:11

make your own AI chips the tpus why

7:15

could you explain just how they are

7:16

different from other chips and why why

7:18

you do this well we work very closely

7:20

with Nvidia they're a strong partner of

7:22

ours so we use both gpus and tpus and

7:25

many years ago the team started on this

7:27

journey of developing our our own chips

7:30

tpus to address some of the specific

7:33

requirements that we felt we needed and

7:35

what we found is that we continue to

7:37

have the type of performance that we're

7:39

looking for in particular in training um

7:42

and as we've Advanced the tpus what

7:45

we're seeing is we're continuing to

7:47

drive greater efficiency in all elements

7:50

of uh energy requirements with tpus so

7:52

they there's a cost benefit for us and

7:55

and impact benefit and so we're using

7:57

both in the fleet talking about various

7:59

versions of things so now we've had deep

8:01

seek um being launched which is uh an

8:05

open- Source model uh now just how do

8:08

you look at open versus closed

8:12

Source models now so open source has

8:15

been core to Google really for since

8:19

Inception if you think about something

8:21

like Android uh which is uh such an

8:24

important operating system globally it

8:26

reflects who we are when you think about

8:28

the trans former paper which has been so

8:32

critical for anybody who's thinking

8:34

about how do you build in this AI world

8:36

like what's the next step to the next

8:39

iteration so open source has been

8:41

important to us at the same time this is

8:44

a really powerful technology and so what

8:46

we're looking at is how does one

8:50

maximize upside but have the appropriate

8:52

guard rails and controls over certain

8:55

elements of it to ensure that you're

8:58

protecting how it's used used and where

8:59

it's used and getting that balance right

9:02

will continue to be very important for

9:04

all of us what are you personally most

9:05

excited about in terms of usage I mean

9:08

you you talked about um um medicine

9:12

Alpha fold and so on you you personally

9:15

what do you what do you feel more

9:16

strongly about oh I so when we look at

9:19

at AI I think there are four primary

9:23

areas that are really

9:25

exciting um first is the economic upside

9:28

you know it globally the estimates are

9:31

there can be 20 trillion added to GDP in

9:34

the next decade again if appropriately

9:36

executed if adopted across industry when

9:40

you think about the benefit to society

9:42

if in fact we have anything close to

9:44

that economic uplift I get excited about

9:47

it it's not to be taken for granted as I

9:49

said it requires what economists call

9:51

diffusion across industry adoption a

9:54

radical rethink of many elements of the

9:55

way the public sector and the private

9:57

sector Works excited about that very

9:59

excited about the breakthroughs in

10:00

science what Demis aavas and John jumper

10:04

awarded the Nobel this past fall you

10:06

know that was for something called Alpha

10:08

fold which you're familiar with which

10:09

has been described as the greatest

10:11

contribution to drug discovery that is

10:14

exciting then just even sitting here

10:16

today the Practical applications of AI

10:20

in health care in education are really

10:23

exciting um so I would put those at the

10:26

top of the list now talking about

10:28

something uh um slightly different you

10:31

uh before the holiday season sent shocks

10:33

waves through the world with your willow

10:35

Quantum chip so why why is that such a

10:38

leap forward so we've been working on

10:40

Quantum AI Quantum Computing for quite

10:42

some time uh well over I think a decade

10:45

at this point uh and what's really

10:48

exciting is the computational

10:50

capabilities with Quantum so the willow

10:53

chip is able to um to handle a

10:57

computation in less than 5 minutes

11:00

that previously on the best

11:01

supercomputers on the planet today would

11:04

take 10 septian years which even I had

11:06

to Google it's it's 24 zeros on the back

11:10

of it so what that means in terms of the

11:12

ability to see and analyze more whether

11:15

it's in biology or other areas is

11:18

exciting and profound and so we see this

11:20

as another path that we will continue to

11:23

execute against how do you view uh

11:27

Microsoft's version of it uh we're

11:29

really proud of our own we think that uh

11:33

that what you've seen time and time

11:34

again is the breakthroughs from that

11:37

team led by an extraordinary leader

11:39

Hartman Nevan um U published in all

11:42

sorts of different places and we just

11:44

continue to build success to success

11:46

with the willow chip being the most

11:47

recent when do you think this will be

11:49

commercial yeah that's a great question

11:51

not surprisingly I ask that question all

11:53

the time as well and have been for um

11:55

some time you know I think it's there's

11:57

still a number of years ahead it's

11:58

starting to to point to different

12:00

applications that we're excited about

12:01

but whether we're three years five years

12:04

you know hard to say uh but it's it's

12:06

getting closer since I started asking

12:08

that

12:10

question and when when you talk about

12:12

these kind of 24 zeros uh uh and so on

12:15

just what are the what are the

12:17

implications for what are the kind of

12:18

things we can do when we have that kind

12:20

of compute power well when you think

12:22

about it take something like the human

12:24

body and the complexity of the human

12:26

body or anything in nature that's you

12:28

know multi-dimensional um the ability to

12:31

actually crunch data more efficiently to

12:33

have better insights is one area that's

12:36

very exciting for for us to think about

12:39

and the possibilities that come from

12:40

that and other things where you are

12:42

where you are strong is a self uh

12:44

self-driving taxi company now I'm not

12:47

sure what the latest number is in terms

12:48

of rides per week where where are you

12:50

now it's uh 200,000 rides paid rides so

12:54

what will the city look like in 5 years

12:56

time or 10 years time when do you think

12:58

it would be properly rolled out well I

12:59

think if I just step back and talk a bit

13:01

about wayo because we are very excited

13:03

about it you know we started on that

13:04

Journey more than a decade ago as well

13:07

and the original thesis is that more

13:09

than a million people die on the road

13:11

every year in accidents and if AI can

13:14

help improve the safety of driving

13:17

because our wayo self- driver the AI

13:21

does not get tired it does not get

13:23

distracted it you know it stays focused

13:25

on the road you've got camera sensors

13:28

everywhere we can improve safety we can

13:30

help save lives and that was a really

13:32

exciting motivator for the team we've

13:35

been rolling it out it's been

13:37

extraordinary to see the takeup it's now

13:39

one of the top attractions in San

13:41

Francisco if anyone's out this way but

13:43

we're also in LA in Phoenix and Austin

13:45

and expanding and we're going to

13:47

continue to expand because you see both

13:49

the um the reaction to it when people

13:52

get in some people are anxious about it

13:54

at first and then within literally under

13:56

a minute they see they they just go

13:58

right into to whatever it is they wanted

14:00

to be doing there's a safety element

14:01

around it uh so we think it'll continue

14:03

to be rolled out uh we're doing a pilot

14:06

in Japan right now and there's an

14:09

opportunity we think to help save lives

14:11

and are excited about doing that so 10

14:13

years from now who is going to own a car

14:15

you know I think it's too early to call

14:17

um I you know a couple of reasons one

14:21

many people in particular in the states

14:23

other developed markets own more than

14:24

one car they're good reasons to own a

14:26

car uh and so we'll see human human

14:30

nature and human behavior changes over

14:31

time the other is that weo is primarily

14:34

now focused on what are these Robo taxis

14:37

as you described it but it's it's

14:40

reasonable to assume that this

14:41

capability this extraordinary technology

14:43

can also be an assist with um with cars

14:46

that are owned and so there are a lot of

14:47

execution paths which enable you me

14:50

others to decide how they actually want

14:52

to get around now you're also the chief

14:54

investment officer and you are uh in a

14:57

very fortunate position in that you are

14:59

sitting on roughly hundred billion

15:01

dollars and you have all these

15:02

interesting areas just how do you

15:04

allocate Capital between them so Capital

15:06

allocation like I think one of the core

15:09

elements in capital allocation is that

15:12

you need to ensure that you're

15:14

continuing to invest aggressively for

15:16

the long run um if you don't invest for

15:18

the long run you're sewing the seeds of

15:20

your own destruction and that's a lesson

15:23

I've seen throughout my career over and

15:24

over and so very important it's been

15:26

core to the ethos obviously of this

15:28

company since Inception the early days

15:31

there was a a a kind of a mantra which

15:33

is 70 in your core 20% adjacent 10%

15:37

moonshots and that's evolved over time

15:39

but this core sense of you've got to

15:42

continue to invest aggressively for the

15:44

long run remains core to who we are and

15:47

we're at such an exciting time in

15:49

history given the opportunity with AI

15:52

you know both on the consumer side on

15:53

the Enterprise side so we're certainly

15:55

making Investments there I end up

15:57

spending a lot of my time globally

15:59

because what is really key is every head

16:02

of state is saying the same type of

16:04

thing which is I want to be a part of

16:06

this digital transformation it is key

16:09

for all the reasons that we've already

16:10

talked about and so what we spend time

16:13

looking at is as we're continuing to

16:15

invest in our technical infrastructure

16:18

globally in other words data centers

16:20

subc cables how do we link up the world

16:23

the the the opportunity is how do we

16:25

engage more deeply how what work are we

16:28

doing on their behalf to help accelerate

16:30

their digital transformation how do we

16:32

engage on the public C side and so

16:34

there's a really global view to this now

16:36

you you have this moonshot Factory in

16:39

house what are the type of things you're

16:40

working on there so the moonshot factory

16:42

as it was lovingly named years ago um uh

16:46

is the core one of the core areas within

16:50

other bets it's called X and out of X

16:53

came for example whmo it was incubated

16:56

in X and then got um moved out to become

16:59

an independent uh company they've also

17:02

worked on for example Wing or incubated

17:04

Wing which is our drone business excited

17:07

about Wing they're doing work as an

17:09

example with Walmart and we're seeing uh

17:12

really exciting results there and see

17:15

the upside they're also incub they've

17:17

been incubating and you'll hear more

17:18

about a company called intrinsic which

17:21

is a robotics um operating system and so

17:24

they have a number of different things

17:25

that they've been working on I think one

17:27

of the very important ele elements there

17:30

is that when um they are approaching any

17:33

incubation their mindset is obviously

17:36

you can't incubate everything and have

17:38

it work well and so kill things fast in

17:41

order to move into the areas that are

17:43

the most promising that sounds really

17:45

fun I think it was Charlie Monger who

17:47

said that Google appeared like a very

17:48

rich

17:49

kindergarten well it's we've got

17:52

everything from kindergarten all the way

17:53

through to postgraduate

17:56

robotics now you mentioned um uh sub

17:59

cables the fact that you also do subsy

18:00

cables and so on does it make you more

18:02

resilient you think as a tech company I

18:04

think there are a lot of elements that

18:06

build resilience and one of the are one

18:09

of the important areas is continuing to

18:11

invest in subc cables but it's also the

18:15

what we call technical infrastructure

18:17

more broadly so it's data center

18:20

resilience data center redundancy so

18:22

that you can actually be positioned to

18:24

serve customers when and where needed

18:27

and in the event of something that

18:28

happens and things unfortunately always

18:30

seem to happen around the world you've

18:32

got the resilience needed to continue to

18:34

operate at a high level it's our cyber

18:35

security defenses one of the things I've

18:38

learned in my career is is much easier

18:40

to prevent than to fix a problem and so

18:42

building in this strength upfront is key

18:45

and it's technical infrastructure I can

18:47

I'm happy to talk more about cables

18:48

because really proud of the network we

18:50

have globally but it's the other

18:52

elements like a zero trust approach on

18:54

cyber security so that you're fortifying

18:57

defenses in a world that is continuing

18:59

to become ever more challenging and I

19:01

think that's one of the key additional

19:03

applications of AI that's important and

19:05

how do you look at these additional

19:07

challenges I mean we are seeing attacks

19:09

on subsy cables we are seeing of course

19:11

accelerating cyber attacks how do you

19:14

how do you assess the state of the world

19:17

uh it's better to remain paranoid which

19:19

we do and build in resilience as you

19:21

said and so when we're looking at our

19:24

subc cable network we uh do build in

19:27

resilience to have alternate approaches

19:30

we're building globally we've been

19:32

building now in the southern hemisphere

19:34

as well um when we build cables as an

19:37

example one of the things that has

19:39

multiple benefits is we will build what

19:42

I will call a trunk from say Africa to

19:45

Australia and then we will have it in

19:49

chunks with nodes along the way that

19:51

builds in a resilience for the cable

19:53

itself but it also provides something

19:55

else that we're really proud of it

19:57

provides the ability off of a node to

19:59

light up island nations along the way

20:02

oftentimes working with governments that

20:05

say we want to help light up a nation

20:07

which maybe Google otherwise wouldn't so

20:10

as an example one of the ones I'm really

20:12

proud of is when we lit up Fiji uh and I

20:15

was talking about this at the Asia

20:16

Pacific economic um conference and we

20:19

had the prime minister of Fiji happen to

20:20

be in the room I said we lit up Fiji and

20:23

his comment was it was a gift to

20:25

everyone in Fiji because what that does

20:27

is link them to all the opportunities

20:29

that come otherwise from being having

20:32

access and what's really important to

20:34

remember is a third of the globe still

20:36

is not online still is not connected so

20:39

this gives us that added ability to

20:42

provide something that as he described

20:44

is an economic opportunity for Fiji for

20:47

generations to come with all um these

20:49

tech companies chasing the opportunities

20:51

here do you think tech companies are now

20:54

accepting lower returns for their

20:55

Investments well I think that at this

20:57

point what we're all looking at is this

21:02

theoretical economic upside which

21:05

assuming it

21:06

materializes creates the opportunity for

21:10

attractive returns and I think there's a

21:13

self-calibrating element to the pace of

21:15

investment that comes from that thesis

21:18

now when I said you know 20 trillion

21:20

theoretical economic upside globally

21:22

it's four trillion in the us alone what

21:26

is it require to get that you know it's

21:28

not applying using chat Bots it's

21:30

actually radically rethinking every

21:32

element of your business it's how do you

21:34

interact with with customers to drive

21:37

more Revenue it's what are operating

21:39

efficiencies in the business it's what

21:40

are risk analytics thinking front to

21:43

back on your operating processes because

21:45

this gives you an opportunity to

21:47

approach them differently that's what

21:49

leads to that the economic upside which

21:52

then drives the returns and so I think

21:54

that uh it's we've each got to make sure

21:57

that we're we're calibrating as we go

22:00

regarding the upside and then the other

22:02

part to your question about returns that

22:04

we're very focused on and others are as

22:07

well is how do you increase the

22:09

efficiency of that denominator all the

22:11

capex that it's being invested and so

22:14

we're approaching that in a number of

22:15

ways Sundar talked about it the

22:18

meaningful improvements in model

22:20

efficiency the meaningful improvements

22:22

you've already asked about in tpus our

22:24

Trillium chip our most recent chip is

22:28

about 67% more efficient than the prior

22:31

one so we're continuing to drive

22:33

efficiency in the cap capex utilization

22:36

while also trying to help unlock um

22:39

really that monetization upside now

22:41

related to this is is the corpor culture

22:43

and how you structure Innovation is

22:45

there a particular way that you

22:46

structure Innovation at Google it's been

22:49

such a core part of who we are I think

22:52

the driver of everything that we do and

22:54

there are elements of it that are about

22:56

how we bring in bring in Talent and and

22:59

apply them to explore uh different

23:01

Avenues we have an extraordinary

23:03

research team as an example everything

23:05

that's being done in Google deepmind the

23:07

demises leading and then on top of that

23:10

we have efforts I think more to your

23:12

question for example in the early days

23:14

of Google there was something called

23:15

Google Labs small Scrappy teams and told

23:19

go find something and Sundar

23:20

reconstituted that a couple of years ago

23:23

and you've seen already a number of

23:25

really exciting things come out of a

23:27

small Scrappy f focused empowered team

23:29

so for example hopefully you and others

23:32

listening have used notebook LM it's one

23:35

of my favorite it's the opportunity to

23:37

take your own your content speeches

23:39

you're interested in anything you're

23:40

interested in ingested into notebook LM

23:43

it can give you some of the information

23:45

that you need but you can also listen to

23:47

it as an AI podcast that can make things

23:50

sort of more accessible maybe within an

23:52

Enterprise or I've talked to some people

23:53

in the public sector and their wow

23:56

moment was wow you mean I can take all

23:58

these things that we publish that people

24:00

probably don't read and make it easier

24:02

for them to ingest the information the

24:03

answer is yes it was the time Innovation

24:06

product innovation of the year last year

24:09

and that's an example of something that

24:10

came out of of lab so we're continuing

24:13

to add- on in a lot of different ways to

24:15

make sure we're inspiring people to

24:17

dream big and think big but if I walk

24:20

around at Google and let's say now you

24:22

took down all the Google signs I mean

24:24

you got a Google sign behind you right

24:25

you take away all the G's all the Google

24:27

signs just how would I how would I know

24:30

that I was in Google how do you do you

24:32

relate to each other in a different way

24:33

do you talk to each other in a different

24:35

way do you behave in a different way

24:36

compared to the other tech companies I

24:38

have so many different ways to answer

24:39

that when I when I first got here I saw

24:42

how different it was in the way we

24:43

worked and granted I was coming from a

24:45

financial institution not another tech

24:47

company but it's literally embedded in

24:50

the technology that we use so it just

24:52

starts right from square one with Google

24:54

Docs collaborative docs every every way

24:56

we're constantly interacting and it just

24:59

adds a velocity to the work that you do

25:01

U right across the street from me here

25:03

we have a wonderful building that we we

25:05

put up

25:06

recently uh that that is where our our

25:09

AI team gdm Google be mind team is

25:12

operating you just see people constantly

25:14

coming together in ways that are exactly

25:17

what the founders talked about at

25:18

Inception it's about serendipity and the

25:20

joy that comes from that but then it is

25:23

in the mindset that goes through our

25:25

people Ops and our workplace Services

25:27

teams to have s dipity in a lot of

25:29

places so that people do congregate and

25:32

exchange ideas but I think the core

25:34

point is it's the ethos from the people

25:36

come in there's a high bar that's set we

25:39

want to make the maximum difference to

25:40

humanity that's what we talk about a lot

25:43

it's you know I I'll give you one more

25:45

Demus story um when he was embarking on

25:50

Alpha fold and the concept of predicting

25:54

the protein structure for every protein

25:56

known to humanity previously it would

25:59

take a year or two just to do one he

26:01

wanted to do all 200 million it was one

26:03

of the grand challenges that had been

26:05

out there for a long time and some

26:07

scientists said how could this be

26:09

possible and his answer was why not and

26:13

that ethos that why not is a large part

26:16

of how I answer your question and you

26:20

hear it in meetings it's like why not

26:23

take it on so that to me is is the

26:26

explainer so um tell me about a time

26:29

recently when you learned something

26:31

about a project and you just thought wow

26:33

this is just like way cool or does it

26:35

happen all the time it does frankly

26:38

that's why I'm like where do I go you

26:39

know it's um you know when I saw

26:42

notebook LM I'm like wow this is really

26:45

amazing though the ability and starting

26:47

to think about what are all the

26:48

applications I remember calling the guy

26:51

who runs Google Labs is extraordinary

26:53

and I said if I'm if I'm running a

26:55

government can I ingest report from

26:58

across agencies and see where their

27:00

inconsistencies he's like why not yes

27:03

I'm going back to the why not Point uh

27:05

so that's really exciting uh you know a

27:08

little further back this was years ago

27:10

actually but we now have 20 billion

27:12

searches a month using your camera with

27:15

Google Lens if I want your shirt you

27:18

know I can take a photo of it and then I

27:21

can find where to buy it and priced and

27:23

everything else or I can I do with other

27:26

things like art who's that artist I

27:28

still get an ohow moment from that I

27:31

ride Whos all the time in San Francisco

27:34

and you know I always tell people when

27:36

I'm in the car watch the left turn

27:37

because the left turn that's like that's

27:39

really that's pretty cool and gives you

27:41

a sense of what the engineers have

27:43

accomplished so I think the awe of what

27:46

the team continues to do is there

27:50

constantly and then and then there are

27:52

the human moments like I am completely

27:56

in awe of what we're able to do with

28:00

Healthcare and you because you had you

28:02

had breast cancer right exactly had

28:05

breast cancer twice actually and

28:10

Google the amazing Engineers identified

28:14

um the opportunity to diagnose early

28:17

stage metastatic breast

28:20

cancer and what's extraordinary is in

28:24

the testing of it they found that

28:28

relative to the 80,000 sample set they

28:32

found 20% more cases more incidences of

28:36

cancer and no false positives and as we

28:39

all know the difference between survival

28:42

or not or a really difficult course of

28:45

treatment at stage four versus two is

28:49

really meaningful and so the ability I

28:52

still have a wow moment with that that

28:55

we with AI and with breakthroughs like

28:58

that can give people the opportunity for

29:01

the early diagnosis that's needed and

29:04

what's really important in discussing

29:06

this with my oncologist he said what it

29:09

does is it enables any doctor anywhere

29:12

across the us around the globe to be

29:14

operating at the highest level because

29:17

they have this this assist this

29:19

augmented intelligence to me that

29:22

becomes an extraordinary wow moment when

29:25

you think about what we can do and we're

29:27

doing it not just in breast cancer

29:29

there's early diagnosis in lung cancer

29:32

in something called diabetic retinopathy

29:34

which is Blindness from diabetes and in

29:37

that instance early detection leads to

29:41

early intervention that's manageable

29:43

around the globe many places you know

29:46

early detection isn't the Panacea it's

29:48

not an answer it's an assist but then

29:50

you need the rest of the treatment as

29:51

well so there are a lot of wow moments

29:55

that come um and I think it goes back to

29:57

the importance of each of us asking how

29:59

can we apply it well thanks for sharing

30:01

that now how has the culture de evolved

30:03

during the during your time at Google

30:07

you know when I got here I think the one

30:09

of the first questions I asked because I

30:11

I've always believed that culture is

30:13

more important than rules regulations

30:15

coming from a regulated Financial

30:16

Services environment it's such an

30:18

important explainer and so I that's

30:20

where I spent some of my early time

30:22

asking people so how do you define the

30:23

culture and I do think it comes back to

30:26

inquisitive people who are who believe

30:30

that with technology we can have a

30:32

positive impact on Humanity it's not a

30:34

Panacea but we're technology optimists

30:36

we always we often use that phrase and

30:40

you know it goes back to deus's comment

30:42

I want to take on the most intractable

30:44

problems in society that's very much the

30:47

ethos of the founders of what sundar's

30:49

been pushing and it's what I've seen

30:52

throughout and I think it's what is

30:54

inspiring um for a lot of what we've

30:57

done and I've seen

30:58

the application of it so when I got here

31:01

as a former Banker I was very excited to

31:03

look under the hood and see so what

31:05

really are the numbers the Mantra as you

31:07

may recall back in 2015 when I arrived

31:10

was the desktop is dead can Google

31:13

actually make the transition to mobile

31:15

and who knows about YouTube and it was

31:17

sort of we've been sort of sliding for a

31:19

bit and obviously history has

31:23

shown the application of Solutions on

31:27

mobile

31:28

has been it's been a pretty wonderful

31:31

decade and it's that intense focus on

31:34

Innovation and providing better ways for

31:37

advertisers for people around the globe

31:39

to use mobile get what they want access

31:42

how and where they it's this Innovation

31:45

cycle that we've continued to see and so

31:48

I got here 2015 the market cap was 400

31:50

billion it's now north of two trillion

31:53

and I think to that questioning back

31:55

then um and yet the whole series of

31:58

innovation that we've seen since then

32:00

and that is the evolution and it's been

32:02

this recurring let's keep pushing it and

32:04

there's another really important element

32:06

and I think it's two sides of the same

32:08

coin which is there's an incredible

32:10

humility here a sense every year that

32:13

this is the last great year and that

32:15

started the moment I got here and I

32:17

remember as I was listening to the

32:19

stories like well they're all these

32:20

headwinds they're always headwinds and

32:22

so that humility I think inspires people

32:25

to push higher on for themselves and

32:28

their team and try and figure out what

32:29

else can we do um and that's been an

32:32

ethos I think search for Innovation

32:34

coupled with humility how do you install

32:36

humility in an organization it starts

32:39

from the tone from the top I mean if you

32:41

look at Sundar as a leader he brings

32:44

this um integrity and Ethos about him

32:47

and that's the

32:48

expectation uh I think for the entire

32:51

team we are privileged to be at this

32:54

moment in history with the opportunity

32:57

set that we've talked about and we

32:59

better approach it

33:01

responsibly and boldly to have the right

33:05

to continue to execute in the way that

33:07

we are I think it's it's got to come

33:10

tone from the top do you think it makes

33:12

a difference that the founders are still

33:13

around well I think it's been you know

33:16

Sergey in particular um at the couple

33:19

years ago as the team was pushing forth

33:22

on on Gemini uh was very engaged and

33:26

back in

33:28

uh sitting with them and he made a

33:32

comment that this is the most exciting

33:35

time in computer science that we had

33:37

only just scratched the surface

33:40

previously and when you think Google's

33:43

done pretty well like when he first said

33:45

that to me I got chills because I think

33:47

about how much has been accomplished by

33:49

Google but that level of excitement and

33:52

the magic of what's to come you know has

33:54

to be infectious um for every one and so

33:59

it's been yeah it's been a joy I just

34:01

saw him um uh in an interview saying

34:04

that uh 60 hour work during the week is

34:07

perfect that's where you have your

34:09

optimal production and Innovation and

34:11

creativity so he's yes he's certainly

34:15

motivated and uh as I said he's vieing

34:17

this is the most exciting time in

34:19

computer science I think that when you

34:21

look at uh the that building over there

34:24

that I described with uh the Google

34:27

deepmind team team and the effort

34:28

everyone's focused in it people are

34:30

working because they're getting a real

34:31

charge out of what they're doing and you

34:33

don't need to tell you or me how many

34:36

hours to work you know there's a joy

34:38

that comes from it you know at at Google

34:40

we're still um of the view that the way

34:42

we get the best outcome is we want

34:45

people in three days or more a week and

34:49

teams can figure out what works best for

34:51

them to get maximum kind of Maximum

34:54

productivity progress who do you hire

34:56

really smart people

34:58

and uh it's a it remains a a rigorous

35:02

process the number of applicants remains

35:05

you know

35:06

daunting um I think that depending on

35:09

the area you know there's there's a mix

35:11

of um I you know what I would say what I

35:14

would look for is a mix of those people

35:17

who um come with what I will call

35:20

pattern recognition the experience that

35:22

gives you a sense that of predicting

35:23

what's going to come from here or very

35:26

specific domain expertise but then

35:28

evidence uh high performance and

35:31

creativity pushing themselves achieving

35:35

achieving remarkable things along the

35:36

way talking of which uh Ruth you've uh

35:39

achieved remarkable things uh and you

35:42

worked in Morgan Stanley before you

35:44

before you joined Google uh and you were

35:46

there during the financial crisis and uh

35:49

uh did a did an incredible job what did

35:50

you what did you learn from the

35:52

financial crisis because you were really

35:54

involved at the very top yeah I learned

35:56

a I learned a lot at the the time I was

35:58

running the financial institutions Group

36:00

which meant responsible for banks

36:02

insurance companies Asset Management I

36:04

was on the investment banking side and

36:06

one of the most privileged times in my

36:08

career is when secretary Hank pulson

36:10

Secretary of the Treasury gon said he

36:12

wanted a team to basically be Sanda down

36:16

um at US Treasury and initially we went

36:19

I went down in in July of' 08 to focus

36:23

on the housing crisis Fanny May Freddy

36:25

Mack and to try and understand

36:27

understand and diagnose what could be a

36:29

trigger for what we would call a run on

36:32

the bank run on on agencies so we went

36:35

through Fanny May Freddy Mack with him

36:37

and then I led the group that went um

36:39

through the AIG crisis as well and then

36:42

onward and there were a lot of really

36:45

important lessons that came out of it

36:46

and when I got to Google I was asked

36:48

about them which struck me as a bit odd

36:50

because Google had only seen sunny days

36:52

and I think really importantly the

36:53

lessons are good for good times and bad

36:56

and for said it is easier to to fortify

37:01

oneself ahead of an issue to prevent

37:03

than to deal with it in the moment the

37:05

first most the most important lesson

37:07

from the crisis is to identify your

37:10

greatest source of vulnerability ahead

37:12

of time and protect against it so for

37:15

financial institutions that would be

37:18

liquidity without liquidity nothing you

37:21

couldn't operate and in that moment in

37:24

that September October 2008 moment you

37:26

couldn't procure liquidity durable

37:28

liquidity if you tried and we did and

37:32

you couldn't um but six months prior you

37:34

could have and I think a really

37:36

important lesson for everyone is do that

37:39

question what is your greatest source of

37:40

vulnerability you can protect against it

37:42

early on but not in the moment the

37:44

second really came out of AIG as we now

37:47

all know the crisis in AIG really

37:49

started because of the derivative sub in

37:52

the UK not the op not the insurance

37:54

operation and the problem was there

37:57

wasn't visibility about the risk that

37:59

was being taken on by the derivative sub

38:02

and so the metaphor that I've used since

38:05

then is you would not drive a car with

38:07

mud on the windshield you cannot run a

38:09

business or a country with mud on the

38:11

windshield use data and analytics to

38:14

clear away that mud and then you can

38:16

actually go faster because you're taking

38:18

calculated risk you're taking swor risk

38:21

I think the other really important

38:23

lesson uh was that there are no good

38:26

choices in a crisis and so go for the

38:30

least worst and just keep moving because

38:33

standing still can actually just amplify

38:35

magnify and you're not going to end up

38:38

with a good solution in any event by

38:40

definition you're in a crisis say

38:43

actually the last one is make sure you

38:44

have a team with horizontal Vision

38:46

because you got to connect the dots

38:47

across a lot of different issues it's a

38:49

lot of wisdom here now how do you

38:51

protect the Google against future

38:52

vulnerability so I asked myself that I

38:55

then said okay I need to apply those

38:56

same rules here I think the greatest

38:59

source of

39:00

vulnerability for us and frankly really

39:04

across Industries is around Innovation

39:08

and long-term

39:09

investing and notably when you come to

39:13

campus um you will see that Larry and

39:15

Sergey put a dinosaur on our campus that

39:18

it's name Stan to remind us every day

39:22

that if we don't innovate we too can

39:24

become dinosaurs and I think it's really

39:27

important to

39:28

maintain um efforts like what sunar did

39:32

with Labs Google Labs other things that

39:33

are continual Catalyst to continue to

39:36

really SP spur small teams to think big

39:39

and and also from a capital allocation

39:42

perspective to make sure you're

39:43

investing for long-term long-term growth

39:47

talking a bit about the personal growth

39:49

you worked with what is called the

39:51

trillion dollar coach Bill

39:53

Campbell who has uh well worked with a

39:56

lot of in successful people but how did

39:59

you how did you end up working with him

40:00

and just what did you learn from him so

40:03

when I was at Morgan

40:05

Stanley I've always had this view in

40:08

life that I should ask what's my highest

40:11

and best use and keep learning and I

40:13

loved being CFO with James Gorman he was

40:16

an absolutely extraordinary CEO I joined

40:19

him as his CFO the day he began as C we

40:22

we we've had him on the podcast and I

40:24

fully agree extraordinary person EX so I

40:28

started his his CFO January 1st 2010 the

40:30

day he started as CEO but after about

40:33

five years I thought you know I feel

40:35

like I'm plateauing here and I'm

40:36

wondering what my next chapter is and as

40:39

you said Bill Campbell is um one of the

40:43

most extraordinary people that anyone

40:45

could meet trillion dollar coach because

40:47

he had coached Steve Jobs and Larry and

40:50

Sergey um and I sat down with him and

40:53

said I don't know what my next chapter

40:55

should be and I'd love to get your

40:56

thinking about it I was out at a

40:58

Stanford board meeting where I'd served

41:00

for years and so got together at his

41:01

house and we spent a couple hours

41:04

together and he started by saying so the

41:06

one thing you know is you won't leave

41:08

Morgan Stanley as CFO to be CFO anywhere

41:11

else and I'm like that's the only thing

41:12

I know but what's the next chapter and

41:14

at the end of the two hours he said I

41:16

have the perfect role for you CFO of

41:19

Google we both sort of laughed because I

41:21

had been so adamant the one thing I

41:22

wouldn't do was be CFO again I'm like if

41:24

it's Google of course I had run Tech

41:26

banking at Morgan Stanley involved with

41:28

the Google IPO loved Google for years

41:32

and um so that's how I ended up here and

41:35

then he was always an adviser who was

41:39

here on

41:40

campus and just the wise Voice

41:44

who um was blunt and clear and when

41:47

something didn't make sense you know he

41:48

was famous for saying kind of throwing

41:51

the flag on the field um and just truly

41:55

brought out the best in everyone made

41:57

told each of us make sure you always

41:59

have that human connection and then go

42:00

into the business item he tragically

42:03

passed away with cancer shortly after I

42:06

arrived uh but it was a he is a gift

42:09

that is a gift that keeps giving do you

42:12

Mentor people yes what's the key to good

42:15

mentorship well I think you know when

42:18

I when you asked that question I

42:20

immediately thought of one of the most

42:22

important discussions I had with someone

42:25

who ended up becoming with a mentor

42:28

sponsor uh a key person in my career and

42:32

it was back in

42:34

1996 and I was asked to uh lead

42:39

technology Equity Capital markets the

42:41

part within a bank that's between

42:44

Banking and sales and trading that

42:46

really launches IPOs and this was right

42:48

at the beginning of the whole internet

42:50

run of IPOs there were not very many

42:53

women on the trading floor and the guy

42:55

who ran institutional equities called me

42:58

into his office and he said I think

43:02

you're going to soar but if you stumble

43:05

I'm here I will back stop you I am your

43:08

senior air cover and what really struck

43:11

me is that we all need senior air cover

43:14

and we all need to be senior air cover

43:16

for someone and what he was saying is I

43:19

know you're going to run far and you're

43:21

going to run hard and but I'm here if

43:24

there's ever an issue if you need me for

43:26

something something and so for me on

43:30

people who work hard for me and I see

43:32

this is a star who needs to be sort of

43:35

unleash to run to to soar um I try and

43:39

give them the advice that I think has

43:41

served me best throughout my career and

43:43

it's about continuing to learn and have

43:45

that senior air cover and really apply

43:48

data and Analysis to everything you do

43:51

and I just keep coming back to those

43:53

rules in particular I think it's so

43:55

important we each have SE air cover and

43:57

our senior air cover did you have a need

43:59

air cover you know there was I was I've

44:02

reflected on did I actually run into

44:06

issues there were small ones there

44:08

wasn't a crisis moment where I needed to

44:10

go to him but the insurance policy of

44:13

knowing he was there and the message

44:15

that he said I know you'll soar but I'm

44:18

here to back stop you in case there's an

44:20

issue is actually a real Catalyst it's

44:23

empowering you mentioned um the need and

44:26

wish to continue to learn what are you

44:29

what are you most exciting about

44:31

learning now well there's there's a

44:34

there's so much when I talk about AI in

44:38

this moment and the conversations I'm

44:40

having globally about the economic

44:42

unlock the real question is what does it

44:44

mean to radically rethink your business

44:47

or radically rethink the way you run the

44:50

public sector you know I've had quite a

44:53

number at this point of public sector

44:55

leaders Finance minist who said I'm not

44:57

going to get more budget how do I get

44:59

more out of my budget and really trying

45:01

to get drill into what can we do that

45:04

makes a difference how can we be

45:07

specific so for example one of my

45:09

favorite in the state of Minnesota they

45:11

wanted to do a better job delivering

45:13

services to their constituents and they

45:16

came to us and said um we can you help

45:19

us in four of our critical we have four

45:22

critical languages in Minnesota we said

45:24

of course Google translate 250 languages

45:26

no problem there we're going to help you

45:27

rethink how you engage within six months

45:30

they came back and they said actually we

45:31

need another 26 languages now that

45:34

Minesota experience for me can be

45:36

repeated over and over around the globe

45:38

like how do you interact better with

45:41

customers constituents but there's more

45:44

to it about efficiency unlock so I I'm

45:46

just a I think we're at this very early

45:49

stage of something that's huge and

45:51

possible and excited to make sure that I

45:54

understand how does it actually work and

45:55

what are the implications how do you

45:58

relax

46:00

um I love what I do so that's source of

46:03

relaxation and then beyond that you know

46:05

there's the biggest Joy is family and

46:08

kids U taking hikes with my kids I've

46:10

got a book club with my kids comparing

46:12

notes on books had dinner with one of my

46:15

kids last night it's um everything that

46:18

we can do together travel explore new

46:21

places um and I work out last question

46:26

so we both went to Wharton and if you

46:30

were giving the commencement

46:32

speech which I'm actually doing in May

46:35

uh what would you be telling The

46:37

Graduate students you know I actually

46:38

did do that a number of years ago and

46:42

what would you tell them now I mean have

46:44

you changed your mind probably not

46:46

because I think the core principles for

46:47

me never stop learning my father was a

46:52

holocaust Refugee he had no high school

46:54

or college education he ended up up

46:57

enlisting in the British Army and he

46:59

fought in the two battles of aliman and

47:01

he taught himself engineering and

47:03

physics because he knew or he assumed

47:07

that if he survived he wanted to get to

47:10

a place that was safe and he needed a

47:12

skill that people would value and he

47:15

thought engineering and physics would be

47:17

that skill and as a child he always told

47:20

me that his fellow soldiers would tease

47:23

him and say you're going to be dead

47:24

before you can ever use this and he

47:26

would say I'd rather die an educated man

47:29

and then the lesson for me as a child my

47:32

siblings was education as a passport for

47:35

life and I firmly believe that and I

47:38

think that one of the most important

47:39

things is never stop learning so that's

47:42

why I said when I found myself plateau

47:45

in my career I would go to somebody I

47:47

respected and say what is my highest and

47:49

best use and I was open to change and

47:52

continuing to grow I think that's one

47:54

really important one the other is Anchor

47:56

everything everything in data at Wharton

47:58

you clearly get great analytical skills

48:00

a lot of other places as well you can

48:03

argue with me in my approach to

48:05

something but you can't argue with data

48:07

and I always say don't give me flat data

48:09

put it into a sensitivity analysis so

48:11

that we can debate your assumptions

48:13

about the state of the world growth

48:15

rates how much do I need to invest let

48:17

me engage you with the data and that

48:20

should be the basis for the argument and

48:23

then the other thing I believe I told

48:25

them and I believe it is today as I did

48:27

then is Embrace Life as it comes don't

48:30

defer to a later time something which

48:32

you can do now because as I saw with

48:36

cancer you don't know if you're gonna

48:38

have that later time unfortunately that

48:40

was 20 years ago and I'm fine but in the

48:42

moment I didn't know but I was grateful

48:44

my bucket list wasn't very long I had

48:47

done what I wanted been married to the

48:48

same guy for many many decades have

48:50

three amazing kids I Professional Care I

48:53

wanted don't put it off because life

48:55

doesn't actually stick to your predes

48:59

predefined schedule well Ru that's uh uh

49:03

really motivational and beautiful place

49:05

to to end this podcast big thanks for

49:08

being with us and thanks for sharing

49:10

sharing all your thoughts and

49:11

backgrounds thank you so much you it is

49:13

always wonderful being with you thank

49:14

you for having me thank you

Interactive Summary

In this discussion, Ruth Porat, President and Chief Investment Officer of Alphabet, explores Google's strategic approach to the AI landscape. She emphasizes a 'full-stack' methodology, integrating top-tier talent, advanced models, custom TPU chips, and a commitment to solving intractable human problems. Porat shares insights on Google's cautious yet innovative stance, the evolution of its search and enterprise AI capabilities, the strategic role of Waymo, and her philosophy on capital allocation and long-term investment. Additionally, she reflects on her experiences during the 2008 financial crisis, the importance of data-driven decision-making, and her personal journey, offering advice on mentorship and continuous learning.

Suggested questions

4 ready-made prompts