HomeVideos

내 멘토는 피터틸 (1시간 30분 전체번역)

Now Playing

내 멘토는 피터틸 (1시간 30분 전체번역)

Transcript

3073 segments

0:00

I think there was an Elon metaphor he

0:01

told me once and this is like I I don't

0:03

remember when it was like maybe 2010

0:05

2011 and it was it was it was sort of

0:09

like uh there were there were these

0:10

three or four aerospace companies in the

0:14

US that were like really badly managed

0:16

and badly run and it was sort of they

0:19

were these very fat cows on this grassy

0:22

enclosed field and if you ever got in

0:24

there there was just going to be a blood

0:26

bath.

0:26

>> Oh yeah. [laughter]

0:27

So it's kind of

0:28

>> it's like he's like this this there's

0:31

this this carnivore and there's a big

0:32

fence and you got to climb this really

0:34

big fence. Eventually you got in there.

0:35

It's going to be it's but so so this is

0:37

a very optimistic thing though because

0:39

both Elon and and what we built together

0:42

uh with Palanteer were able to be great

0:44

enough and persistent enough to overcome

0:46

unnatural barriers and and do something

0:47

very optimistic and fix things. Is

0:49

should we also be trying to do that in

0:51

healthcare and education and

0:52

universities in some sense? Is it cuz it

0:54

was almost impossible but it turned out

0:55

after like years of work that we did it.

0:57

Shouldn't we also do that now in other

0:59

areas? It is it is worth trying in a lot

1:01

of areas. Look, I I always think it's

1:03

better to be optimistic than

1:05

pessimistic. It is um it's probably not

1:10

I the the place where I quibble with the

1:12

name of your show is that uh I I don't

1:14

think it's I don't think it's good to be

1:17

uh insanely optimistic or insanely

1:20

pessimistic. uh at the extreme optimism

1:22

and pessimism are the same thing because

1:25

if you are extremely pessimistic there's

1:28

nothing you can do. If you're extremely

1:31

optimistic there's nothing you need to

1:33

do. Um you just need to um you know the

1:35

singularity is near. It's the Curtzsw

1:37

Wild uh book title very optimistic title

1:40

and all you need to do is sit back and

1:43

eat some popcorn and watch the movie of

1:45

the future unfold. So both extreme

1:47

optimism and extreme pessimism. um

1:50

converge on laziness and on people not

1:53

doing anything. So uh so I so I I yeah

1:56

so I think moderate optimism is the

1:58

correct thing and then and then I I I

2:00

think you have to always pick pick your

2:02

battles. the the part that's um

2:07

the part that's

2:10

I think more difficult about education

2:12

and healthcare

2:14

um or historically was more difficult

2:17

was that these were these were areas

2:19

that lots of people were trying to

2:21

reform and there were lots of different

2:23

attempts people had uh if um whereas

2:27

let's say the in 2004 2003 2004 2005

2:31

when we started Palunteer

2:32

um the idea of trying to uh to do a

2:36

military related uh software startup.

2:40

There was nobody even wanted to give you

2:42

money as a venture capitalist. People

2:43

thought we were insane.

2:44

>> And so there was sort of an element

2:45

where

2:46

>> if you could get through you would have

2:48

no competition and then you know a

2:51

decade later we still had no

2:53

competition. There's probably a lot more

2:55

going on in the in the defense space now

2:58

than there was in the in the mid 2000s.

3:00

But and and again and again having zero

3:04

competition

3:05

it it's good if it works maybe it just

3:07

tells you you're insane. So even that's

3:09

an ambiguous thing

3:10

>> but then the um

3:13

>> the the place where I've had we probably

3:14

the two areas where I've had the you

3:17

know we we've tried to invest I at this

3:19

venture capital thing we started

3:20

founders fund 2005 in different ways

3:22

I've been been at this and I' done some

3:25

angel investing before that but I've

3:27

been been doing this for 20 plus years

3:30

and the area where we've probably had

3:34

the least success from a returnsbased is

3:38

in software. The two worst areas are um

3:42

are educational software related

3:44

companies and healthcare IT companies

3:48

and um and pro like education we we felt

3:52

was so hard we we didn't even uh we did

3:56

very little healthcare we tried a few

3:58

more things but um it's and and and the

4:02

the the rough version of what's gone

4:04

wrong in the health care sectors we've

4:06

done so little in education I feel I

4:07

can't even comment on it but but

4:08

healthcare What's what's roughly gone

4:10

wrong is um you have all sorts of things

4:13

that sound like a pretty good plan for

4:16

reforming the system and it's okay it's

4:18

this one angle on the pricing model you

4:20

need to change but then there are 10

4:23

different ways you can do it and it's

4:25

these complicated uh ways you have to

4:28

get the payments to shift from the

4:30

insurance companies or the government or

4:33

uh the hospitals which are very cluji if

4:35

you can just get consumers or doctors to

4:37

change that might be easier but the they

4:39

typically are not the people who pay.

4:41

And so there sort of are there's this,

4:43

you know, very cluji messed up system.

4:46

Lots of people try try to do it. My my

4:48

sociological model is always that I'm

4:51

I'm deeply skeptical of um of

4:54

investments where a lot of VCs uh would

4:59

go to some silly VC networking event and

5:02

uh want to brag about it. And I I I feel

5:05

there there's a lot of that with the

5:08

healthc care education startups. Um

5:12

>> there's like the green tech stuff too.

5:14

It's too popular.

5:15

>> Green tech stuff. Yeah. I mean at some

5:16

point it it loses money and then it's

5:18

not popular anymore. But but while it

5:20

lasts it's very popular because there's

5:22

there's no fast term accountability in

5:24

venture capital. You know sometimes you

5:26

have momentum and the valuations go up

5:28

and that that does give you a feel

5:30

whether things are working. But in

5:33

practice uh you of you know it's 5 10

5:36

years later you get liquidity on on

5:38

these businesses or even longer. it it

5:40

takes incredibly long and so people tend

5:42

to overindex [clears throat]

5:43

on these sociological factors and what's

5:46

what's popular and that's and so there

5:48

was there was a there's a way that uh

5:50

the healthc care education sectors even

5:53

though they're very broken um people

5:56

have these often too conventional ideas

5:58

on how to fix them and um and so yeah

6:02

it's I know you know another trend where

6:05

I' I've tried investing some over over

6:08

the last decade is that you know is is

6:10

are all these marijuana related

6:12

businesses and it's like yeah it's

6:13

obviously like a big trend. Um but it's

6:17

it's an obvious trend and then it turned

6:20

out to be extremely hard to figure out

6:22

what what to do with

6:23

>> Yeah. As I said some of the things

6:25

things that are too popular and too

6:27

obvious. It's just you have to have a

6:28

totally different take on them or it's

6:29

not going to work.

6:30

>> Uh yeah you like if you have an

6:33

unpopular take on it that that that

6:35

might work.

6:36

>> You were at Palanteer 04 to09. We were

6:38

also at PayPal and they they categorized

6:42

you as being part of the PayPal mafia.

6:44

Second, you know,

6:45

>> I I'm 15 years younger than Peter. I'm

6:46

12 years younger than Elon. I was a

6:48

junior kid there. I don't get any credit

6:49

for PayPal.

6:50

>> Well, you were still there. You were

6:51

still there learning. I was learning

6:52

from these guys.

6:52

>> And were you in meetings? Were you

6:54

watching them? Were you there? And what

6:55

was that like? What was Elon like in

6:56

meetings? What was Peter like in

6:57

meetings? What was that like? give us

6:59

some of the I mean I usually wasn't in

7:01

meetings with Elon be honest but I mean

7:03

these these guys were were just very

7:06

opinionated, very interested, very

7:08

ambitious, very fast. No tolerance for

7:10

things that are broken. You fix it right

7:11

away. Uh you know, you work through the

7:13

problems. You get you get everything

7:14

done today. You don't talk about what

7:16

you're going to do next week. You just

7:17

you just move. You know, people would be

7:18

there late at night fixing problems. You

7:20

know, passionate about their work.

7:22

>> What what was the demo of the age? What

7:25

what would you say the average is? would

7:26

you say 25 guys in their 20s mostly

7:29

>> in their 20s where you can as you can be

7:31

demanding of getting them to now at the

7:34

same time was Elon and Peter also there

7:36

driving working with them

7:37

>> yeah in different context right Peter's

7:39

more of the strategist the philosopher

7:41

the the the thinker he's not the one

7:43

he's not the technical guy himself um I

7:45

think I think Elon is more of the

7:47

operator he's more they're in the room I

7:48

mean I was I was in Paul Alto a couple

7:51

weeks ago right and I was meeting a

7:52

friend and Elon was in the back doing

7:53

engineering reviews at at xi just he's

7:56

just he just there working working he

7:57

he's kind of the more the workhorse just

7:59

pushed through solve the details of the

8:00

technical problems

8:01

>> was the level of intent like who was the

8:03

most intense guy at Apple the most

8:06

intense

8:08

>> I mean I think Elon's always been one of

8:10

the very most intense people I' I've

8:11

ever seen in terms of working but

8:13

there's other engineers who are just

8:14

there all the time pushing hard right I

8:16

think I think when when you're in

8:17

operation mode guys like Max Le and

8:19

others were just as far as I could tell

8:20

just always working some of the guys

8:22

that you you worked with that were maybe

8:24

junior guys like you that came out

8:26

became big stars as well.

8:27

>> Yeah. So, so I mean a lot of the guys

8:29

who were there, there were 16 different

8:31

companies that were started after PayPal

8:33

that quickly became billiond dollarar

8:34

companies, right? So, and a lot of these

8:35

guys were older than me, but it was the

8:37

guys that you know Chad and Steve who

8:38

built YouTube. It was Reed Hoffman who

8:40

built LinkedIn. It was obviously Elon

8:42

did Tesla and SpaceX. I mean, you know,

8:44

there's guys who built Iron Port.

8:46

There's just there's just so many things

8:47

that came out of there.

8:47

>> Got it. And so, from there, how did the

8:50

opportunity to be a co-founder of a

8:52

Palunteer come up? Well, I was working

8:54

with Peter at his hedge fund and uh I

8:56

was, you know, the hedge fund was a

8:58

little bit disorganized and I I started

9:00

bringing in my smartest friends to help

9:02

and there weren't really other managers

9:03

there. So, I'd helped start start

9:04

building things and a bunch of my

9:06

smartest friends I brought in one summer

9:07

to help us. They uh they weren't

9:09

interested in finance. They thought it

9:10

was boring. And so, and Peter and I had

9:12

been talking a lot about, you know, at

9:14

PayPal, we had to stop the Chinese and

9:16

Russian mafia from stealing all of our

9:17

money. And uh so, we got to know the

9:20

guys who were helping us arrest the bad

9:21

guys. There was Secret Service and the

9:22

FBI. And right after this happened, it

9:25

was 9/11. And so these guys were

9:26

spending billions of dollars on stuff

9:28

that we thought didn't make any sense.

9:29

And they were really, they kept asking

9:30

us for advice. They were confused about

9:31

how to do things. It's it was it was I

9:34

mean, President Bush created what was

9:36

called the Department of Homeland

9:37

Security at the time. I shouldn't be too

9:39

mean about it, but you know how it works

9:40

in government is that when you create a

9:42

new department, um, people can't really

9:44

fire people in government easily. So

9:45

sometimes they'll have a lot of people

9:46

they wanted to fire and then instead

9:48

they just push them into the new

9:49

department. So it was a bit of a mess.

9:50

photograph into new departments in your

9:52

fire.

9:53

>> So there's so therefore the whole

9:55

department didn't really know very well

9:56

what it was doing at first was my

9:57

impression and they were spending money

9:59

on just nonsense stuff and it became

10:01

really obvious to us that Silicon

10:02

Valley, Google, PayPal, all these things

10:04

that were going on out there were just

10:05

way ahead technically of where the

10:07

government was at that point and this

10:09

was a problem because because the

10:10

government was using all the government

10:11

was spending $ 38 billion a year

10:13

gathering data, looking at the data,

10:15

failing to stop the terrorists but also

10:17

abusing our civil liberties. So it was a

10:18

mess. So, we said, "You know what?

10:20

There's actually a really important

10:21

problem here to solve. A, I'd like to

10:23

stop the bad guys from attacking us

10:24

again and go get them instead. B, I

10:26

don't want everyone in the government

10:27

seeing all of my data without any

10:29

controls. That's crazy." And

10:31

>> so, that's when you said, "Why don't we

10:32

go do it?" You guys said, "Let's go do

10:33

it ourselves."

10:34

>> And I started having my friends I

10:35

brought that summer to help build the

10:36

prototype. And which which it sounds

10:38

even crazier than finance, but at least

10:39

it was interesting. They enjoyed help

10:41

helping me build it. And and my

10:42

roommate, who I'm actually seeing later,

10:44

he moved down here to Miami. He and I

10:46

controlled the team. and uh Stephan

10:47

Cullen, my co-founder, and we started

10:49

building it up

10:50

>> and and CIA was apparently one of the

10:52

first investors in in the company.

10:54

>> They eventually gave us money. I think

10:55

it was the second second or third round,

10:57

but it was it was only $2 million and we

10:59

actually bought him back. The reason we

11:00

needed them is that Peter basically no

11:03

one would give us money. First of all,

11:04

Alex Carpenter went all over Sanchell

11:06

Road like you guys have all this talent

11:08

because they can measure talent even

11:09

back then like why aren't you doing

11:10

social media? Why aren't you doing

11:11

something new and exciting? Working with

11:13

the government's crazy what's wrong with

11:14

you guys? It's not possible. So he

11:16

didn't think it was a good idea to

11:17

>> No, this is Alex and I talking to the

11:19

investors. So all the investors were

11:20

making they're telling us you guys are

11:22

crazy. No one does this. Why are you

11:24

doing this? It's not possible. It it

11:26

doesn't make any sense.

11:27

>> These are some big names. Credible.

11:28

>> These are these are the big names. These

11:30

are we got turned down by everyone.

11:31

Turned down by Excel, by Seoia. The guy

11:33

at Kleiner Perkins at the time and

11:35

they're a great firm. I admire them very

11:36

much. But the guy at the time who's no

11:37

longer there, he started laughing at us

11:39

on the phone because Alex didn't have a

11:40

technical degree. He's like, "You guys

11:41

don't even know what you're doing. And

11:42

he's a it's a doctorate, but it's not

11:44

even a relevant doctorate. Like like

11:45

he's laughing us out of the room,

11:47

basically.

11:47

>> And what what is Alex's mindset like

11:49

when he's walking out of the room?

11:50

What's he telling you?

11:51

>> Oh, we really

11:51

>> cuz Alex is a he's a very unique type of

11:54

guy himself.

11:54

>> We were we were not happy with these

11:56

people. Peter Peter Teal told me it was

11:58

probably really good for me cuz it gave

11:59

me like an even bigger chip on my

12:00

shoulder to make sure we succeeded after

12:02

being like mocked and turned down by

12:03

like 30 of these guys. And and Peter

12:05

couldn't just only fund it himself. You

12:07

need other people to fund these things.

12:08

So when we got Inqel which is CIA's

12:10

venture capital arm at the time to give

12:12

us a little bit of money and Peter gave

12:13

us even more that was really critical.

12:15

>> And then how long later did you guys pay

12:16

the 2 million back?

12:18

>> I think we bought it for a much higher

12:19

amount. We had some option. I forget. It

12:21

must have been five or six years later.

12:22

>> Okay. So they still made some money on

12:24

it.

12:24

>> Oh they made plenty of money on it. They

12:26

and you know and and and more

12:27

importantly we saved the government

12:30

billions of dollars versus what they

12:31

were doing. I mean you literally had

12:32

things going on the DHS guys. cuz I

12:34

remember there was some Unicist thing

12:35

where it was like $3 billion to

12:37

integrate all the data and we came and

12:39

we showed them we could have done the

12:40

same thing for them out of the box in a

12:41

month like don't waste billions of

12:43

dollars. So it's it turns out competence

12:44

saves the government lots of money and

12:46

protects civil liberties. People don't

12:47

realize it's both of those things.

12:48

>> What what is the the the desire to

12:50

constantly use names that are from I

12:52

believe Lord of the Rings, right? What

12:54

where does that come from?

12:55

>> Peter gets credit for naming

12:56

>> Peter gets credit,

12:57

>> you know, but I'm a fan of it. I think

12:59

listen and and we wrote about this at

13:00

the time. So, at the time when we're

13:02

creating this, we said this is a

13:03

dangerous thing to create, you know, and

13:05

but we believe it's a it's we believe

13:07

it's a worthy thing to create and it

13:08

took it took a lot of coverage. Like,

13:10

you know, there's a lot of different

13:11

things we could have worked on. It could

13:12

have made a lot of money doing a lot of

13:13

other things, but it was really

13:15

important to both help. You know, we

13:18

helped eliminate probably about to 10 up

13:20

to 10,000 terrorists that might not got

13:21

eliminated. We worked with the we work

13:23

with all sorts of amazing groups in the

13:24

government alongside them with the

13:26

technology and with the ability and we

13:27

help protect civil liberties and make

13:29

sure the government watchers are being

13:30

watched. Now

13:33

of you know of course you create this

13:34

technology it's it's if it gets into the

13:36

wrong hands and they turn off the audit

13:37

trails who knows what what bad could be

13:38

done with it. So there's there's good

13:39

and bad here. Do you think that we could

13:41

bring maybe using AI some more advanced

13:43

manufacturing here that was in China? Is

13:44

that a real thing the next decade to do

13:46

a bunch more of it here thanks to our

13:48

more efficiencies and productivity? Uh

13:50

there's there's there's probably always

13:51

a decent amount of things like that. You

13:53

you have to um you you have to always be

13:57

extremely granular on how expensive the

13:59

robots are, how expensive the equipment

14:01

is because in some sense manufacturing

14:04

has been getting steadily more automated

14:07

for 250 years. And you know and and

14:10

you've had the machines that make the

14:11

machines, the machines that make the

14:13

machines are getting sort of smarter and

14:15

more complicated. Um and and there are

14:19

ways in which AI is a quantum leap.

14:22

There's a way in which it's just you

14:23

know a natural continuation of this you

14:25

know two century two century long long

14:28

process. Um I think um but yeah I think

14:32

I think there are some parts that can be

14:34

moved to the US um with AI maybe also if

14:38

you if you change some of the

14:39

environmental rules and some some of the

14:41

other anti-industrial policies we have

14:43

in the US. Um and then but then if if

14:46

parts of this are moved to um other

14:49

emerging market countries um you know I

14:52

I don't Vietnam is look it's a communist

14:56

country it's it's it's a it has bad

14:59

mercantalist policies but it's not

15:01

planning to take over the world and so

15:04

at the margins if we can move things

15:05

from China to Vietnam that's a big win.

15:08

>> I want to ask you about a theory I have

15:09

that's based on how you see things. So,

15:11

one of the things you've been talking a

15:12

lot about lately is is is is how our

15:15

culture I want to put words in your

15:17

mouth, but things are broken to such an

15:18

extent that like we're not advancing as

15:20

quickly as we were in the in the

15:21

industrial revolution, right?

15:22

Something's wrong with our culture.

15:23

Something's wrong with science. It's not

15:25

not advancing. Yeah. I would Yeah. The

15:27

there's it's generally been stagnant in

15:30

the world of atoms, not the world of

15:32

bets for something like 50 years. And um

15:34

and then yeah, why why it's happened is

15:36

overdetermined, but there are cultural

15:38

factors. Maybe we just ran out of ideas,

15:41

which is a very pessimistic version.

15:43

Maybe there's something wrong in our

15:44

culture. Maybe we're too riskaverse.

15:46

Maybe we're we're scared about how

15:49

apocalyptic some of the science and tech

15:51

are. Um, but for for a range of

15:54

different reasons, it's it's slowed down

15:55

since the 1970s. Mhm. And I I I tend to

15:58

think of things and I'm not sure if this

15:59

is how you see it, but I take I take

16:01

some of maybe your interest in Gerard

16:02

and mimemetic theory. And I say people

16:04

tend to all think of things the same way

16:06

even when it's incorrect because it's

16:07

natural to be mimed and to think of

16:09

things the wrong way together. That's

16:10

kind of why cultures move in the wrong

16:11

way. And so I I tend to see that there's

16:13

a lot of opportunities right now thanks

16:15

to the fact that you do have this kind

16:17

of broken cultural situation where

16:19

people are thinking of it wrong. So for

16:20

example, I'm particularly bullish. I

16:22

know it's random, but on Elon's boring

16:23

company I'm excited. I was working on

16:24

with some of the guys yesterday on

16:26

tunnels because I think it's just

16:27

obvious we could create trillions of

16:28

dollars of value by building tunnels and

16:30

getting rid of traffic around our cities

16:31

and and the reason that opportunity

16:32

exists is because of this thing that's

16:34

really broken because it's an obvious

16:35

thing but just it's so broken that no

16:36

one's doing it. Is there are there

16:38

things like that where there's

16:38

opportunities tied to the fact that

16:40

things are broken? Do do you see it that

16:41

way?

16:42

>> Well, those those anything that fits

16:44

that pattern is a great opportunity. So

16:46

if if it's um if you have some if if you

16:51

know it's it's always if it's broken not

16:54

because of the science or tech. So it

16:56

doesn't require a miraculous

16:59

you know you don't have to have a dozen

17:00

miracles in terms of tech breakthroughs.

17:03

you have to just do something in this,

17:07

you know, in this non-mafia controlled

17:10

drilling way or whatever where, you

17:12

know, I think so many of the um

17:13

tunneling is this a super corrupt um

17:17

contracting. It's mafia adjacent, you

17:20

know, sort of super corrupt labor union

17:22

adjacent. And so I don't know, it costs,

17:25

you know, what it cost to build a subway

17:27

in Manhattan. It's like a billion

17:28

dollars mile.

17:29

>> Billions. Yeah. 10 I think it's 10 times

17:32

as much as Paris or something like that

17:33

which we don't think exactly capitalist

17:36

efficiency

17:36

>> it's insane

17:37

>> so

17:38

>> and Elon's Elon's a 10th the cost of

17:40

Paris right so it's like a 100x multiple

17:42

>> crazy

17:42

>> but you know but then the the the place

17:45

where I I always think it's non-trivial

17:47

is uh is that there's there's some non-

17:51

tech reason we're not even at the Paris

17:52

level.

17:53

>> Yep. Yep. So, but if you could push

17:55

through, then you could disagree.

17:56

>> You could do it even you can do you

17:58

could do something big even if you could

17:59

get to the Paris level and uh and so you

18:02

have to think really hard about what

18:04

that what that political blocker is and

18:06

how to how to overcome that

18:07

>> and and so there's all these things that

18:09

are that are dysfunctional and broken

18:11

and not working and like and like if

18:13

things right now I think you're more

18:15

optimistic than you have been. said like

18:17

like should we be trying to use our

18:19

voice and fix things more? Should like I

18:21

know a lot of friends who just decided

18:22

to exit, right? And a lot of friends are

18:24

in Singapore and Switzerland, they gave

18:25

up. They think it's going to be too hard

18:26

to fix it. I feel like you I feel like

18:28

you're in between this where you you've

18:30

done a lot of things to fight for our

18:31

society and to fix a lot of things, but

18:33

then you're like skeptical. You should

18:34

be pushing too hard in a bunch of areas.

18:36

How do you think about that trade-off

18:37

with voice and exit on these things?

18:39

>> I think it's I think it's very important

18:41

to try. It's very important to pick It's

18:44

very important to pick the right battles

18:46

is what I is is what I would always what

18:48

I would always say. And so I don't know.

18:50

Yeah. The the the in between answer is

18:52

um you know I I don't think you can just

18:55

be I don't know shooting at bad guys

18:58

when they're two miles away and firing

19:00

your guns at random and it's it's yeah

19:03

you get credit for some kind of you know

19:04

Marxist labor theory of value. You're

19:06

trying hard but um we're not we're not

19:09

Marxists here and we're it's not

19:10

>> about trying. It's about actually it's

19:12

it's not we're not measuring uh input.

19:14

We're measuring you know where where you

19:16

can actually uh where you can actually

19:17

get output. I I um I I keep trying in a

19:22

lot of these areas. I you know we we've

19:23

we've we've made a number of investments

19:26

over the years in um in nuclear

19:29

technology. Um and it it always it

19:32

always when you look at the science

19:34

there there are all these much better

19:37

ways that in theory you could do it. You

19:39

can build small modular fision reactors.

19:42

Maybe the fusion technology is not that

19:44

far away. There are different verticals

19:46

within the industry where things are

19:49

extremely broken and in theory should

19:52

find a way to to um to build a

19:55

successful company that overcomes some

19:56

of these these problems. And then um

19:59

there's you know then there's a sobering

20:01

reality that uh you know the Nuclear

20:03

Regulatory Commission basically hasn't

20:05

approved a new reactor design in in 50

20:07

plus years. And so there's so so it's

20:11

always it's it's always these these

20:13

these these things that you're you're

20:15

trading off.

20:15

>> What do you learn from your time with

20:16

Peter? What are the things that have

20:17

stuck with you?

20:18

>> Oh gosh, so many things. He's always

20:20

he's always approaching the world from

20:22

some kind of like orthogonal perspective

20:24

and finding new ways to pick apart the

20:26

most important reasons for things. Every

20:28

time I see him, I learn something. Uh

20:29

you know, I I I wrote this piece online

20:31

a while ago, like my team about 15 years

20:33

ago. It's like main lessons from Peter

20:35

Teal. So I won't I won't repeat all of

20:37

them here but there were nine key

20:38

lessons. I think one of them was to

20:39

really value intelligence really highly.

20:41

I think that was absolutely key. And so

20:43

it just turns out the very brightest

20:45

people matter a lot. Um one of them was

20:47

you have to break down like their actual

20:49

reasons for things and their core

20:51

components. And usually the number one

20:52

reason should be like much bigger than

20:54

everything else. So if you tell me I

20:55

have four reasons for doing this

20:56

business thing, it means you haven't

20:58

really thought about it enough. There's

20:59

probably like one thing that's dominant.

21:01

>> Those were really big. One thing you

21:03

always talked about was that uh effort

21:05

on any project is convex. And what

21:07

convex means, it's a shape of a curve

21:09

where if you spend like 80% of your time

21:11

focused on something, that's maybe half

21:13

as good as spending 90% of your time

21:14

focused on something because those that

21:15

last bit of effort and like is one of

21:17

those things where like I think being 99

21:19

percentile is worth so much more than

21:21

being 90th percentile also because that

21:23

means you're number one and being number

21:24

one is worth a lot. So there's there's a

21:25

lot of things like that that you just

21:27

kind of gave me all these concepts that

21:28

we all kind of learn when we work with

21:29

them. Do you find it difficult to not

21:31

divide your attention in that way? You

21:33

>> It's very very hard and I think the most

21:37

important things I've accomplished has

21:38

been when I've been able to really focus

21:40

on something for a while and whether

21:42

that's Palunteer, whether that's Adapar,

21:44

whether that's spending months on a

21:45

thesis at HVC and a framework at HVC

21:47

that we're going to work on for our

21:48

investing. Yeah, it it is really

21:50

important to focus and and you know, you

21:52

get to a certain point where I have

21:54

obviously a lot of financial resources

21:55

now, a lot of influence in the world and

21:57

so I'm able to help others who are

21:58

focusing, but anything I invest in or do

22:00

it has to be someone really amazing is

22:03

making it their main thing and some

22:05

someone the CEO has to be all in and you

22:07

know for for the things I do,

22:09

>> right? So an advice for talent is to not

22:12

divide your focus at all.

22:14

>> You really just need to like be

22:15

courageous. I think I think the a lot of

22:17

people in our culture nowadays a lot of

22:19

them want to do incubators or they just

22:20

want to do a fund never having built

22:22

something or they just want to say I'm

22:23

going to help five different projects

22:25

and that that's actually kind of like a

22:27

form it's a it's a type of cowardice.

22:29

It's a type of saying I'm afraid I'm

22:32

afraid to go all in. I'm afraid to say

22:33

this is the best and I'm going to crush

22:35

it. And like 99.9% of the people who are

22:37

crushing it and who are changing the

22:38

world and who are really you know

22:40

building the future of our civilization

22:41

they're they're focusing on something.

22:44

How do you come to think about risk?

22:46

It's sort of u built into the

22:48

conversation around courage is fear and

22:50

uncertainty and risk and dealing with

22:52

risk and stuff like that. How do you

22:54

assess it?

22:55

>> You know, I think we're all really lucky

22:56

today versus the past. I think it is

22:58

true that the conditions under which all

23:00

of us evolved. If you go all in on

23:02

something and you fail, you might have

23:03

starved to death. You might have been

23:05

eaten by lions or or some kind of giant

23:07

old cave bear. You might have been

23:08

crushed by the local tribe. So, I think

23:10

we all evolved uh to have existential

23:13

risk and to be really afraid. And it's

23:15

not that it's great not to have money in

23:17

our society. I can't I can't speak to

23:18

that. Obviously, it's not not great. But

23:20

come on. Like, we're it's not it's not

23:22

like 3,000 years ago where you might

23:23

just get you might just die if you don't

23:25

succeed. So, I think I think there is

23:26

enough of a safety net. And listen, it's

23:28

easier for me to say that coming from a

23:29

middle class family when I grew up that

23:31

I knew my parents would be able to take

23:32

care of me if something didn't work out.

23:33

So, obviously, I had some privilege, but

23:35

I think a lot of people with that

23:36

privilege still aren't willing, you

23:37

know, to take the risk they should be.

23:39

What about obsessing over perfection?

23:41

Something else that I think Peter that's

23:43

very big on

23:43

>> 100% that really ties into the 99.9

23:46

percentile thing is like getting

23:47

something just to be the absolute best.

23:49

I remember working with him. I was 21

23:51

years old and there was some speech that

23:52

was going to go on in New York the next

23:54

day and we were like basically pulling

23:55

an allnighter with a few of the guys in

23:57

the office to be ready to like have this

23:59

thing was it was about inflation versus

24:00

deflation and the risk for both of

24:02

those. And it was just it was just a

24:03

natural thing to do to just try to make

24:05

it like absolutely perfect before we're

24:06

going to go present to the investors.

24:07

And it was really funny. I actually was

24:09

with Ken Howry who's an ambassador now

24:11

to Denmark. He was ambassador of Sweden

24:12

last time. He's a good friend. He's

24:13

who's our age and very successful guy in

24:15

the background. We were just like going

24:16

back and forth with him and a few others

24:18

just like working hard and then jumping

24:19

on the plane and sleeping on the plane

24:20

on the way over and it's just like

24:22

everything has to be as good as possible

24:23

and pushes you push as hard as possible

24:25

which is if something was wrong like

24:27

this totally unacceptable.

24:29

>> How do you avoid that from holding you

24:30

back? Uh because perfectionism can be

24:32

procrastination sort of masquerading as

24:34

quality control like you know the

24:37

classic

24:38

west coast move fast break things

24:41

mentality.

24:43

Is there a tension between these?

24:44

>> 100%. And I think if you have really

24:46

tight fast deadlines, it's probably

24:47

good. So it had to be as perfect as

24:48

possible given that it was coming due in

24:50

the next day, but we we weren't going to

24:52

be able to work on it for 5 weeks,

24:53

right? So I think I think there is

24:55

something about about really making

24:56

things as strong as possible, but but

24:58

sprinting and having really tight

24:59

deadlines and getting it done right

25:00

away. I think I think if you use

25:01

perfection as procrastination, then it

25:03

becomes a problem.

25:05

>> Yeah. Well, I'm I'm still interested in

25:07

this sense of not getting distracted and

25:10

trying to keep the main thing the main

25:11

thing, especially if your main thing

25:12

becomes a varied thing, right? Like

25:15

built into a lot of people's lives,

25:16

especially as they end up getting to the

25:18

kind of place that they want to, is

25:20

well, you don't have to do things you

25:22

don't want to do that much anymore. No

25:23

one tells you what to do. So you end up

25:26

in a world where you think, well, I get

25:28

to choose. But with that comes a lot of

25:30

responsibility cuz I have to choose now.

25:32

As opposed to before where I just sat on

25:34

the set of train tracks, it's God, do I

25:35

want to go left, do I want to go right?

25:36

The same for yourself. Do I want to

25:37

invest in or should I sit down and spend

25:39

6 months working on a thesis? Um

25:42

what about the skill set of learning to

25:44

sort of let go of what was there of how

25:48

you operated previously? The sort of

25:50

courage to to do something uh new even

25:53

as you've got something that's given you

25:55

success in the past.

25:56

>> No, it's that's totally right. You do

25:57

have to constantly keep adjusting for

25:59

what makes sense today. And it's it's

26:00

interesting. There's different versions

26:01

of this. One version is as you're

26:03

successful something you would have been

26:04

really excited about before, you have to

26:07

be like, I don't have time for that now.

26:08

And because all of a sudden you could

26:09

just have things you were really excited

26:10

about before 10 times a day. And I do

26:12

fall into this myself sometimes because

26:13

there's lots of really exciting things

26:14

to do and you have to so it's like

26:16

really hard to say no enough when you go

26:17

through periods where you don't say no

26:18

enough.

26:19

>> You might be feel like you're getting

26:20

stuff done but not actually getting

26:22

things done. And that's it's really

26:23

tough. And

26:24

>> but you know what you said what you said

26:25

earlier about like things you don't want

26:27

to do. To me that's like one of the most

26:29

important things that we focus on is

26:32

what do you like to do? And as you're

26:34

successful, you should probably mostly

26:35

only do things you like to do because

26:37

when what would like to do means to me

26:39

anyway is that it's like stimulating

26:40

your entire brain. Right? So if you look

26:42

at like a grandmaster chess player and

26:44

the very best chess players, when you

26:45

map out their brains when they're

26:46

playing, there's always emotions that

26:48

are turned on. There's always full parts

26:49

of their brains that are turned on and

26:50

it lets them be a lot better what

26:52

they're doing. And I think this is true

26:53

in anything we do. If you really love

26:55

it, then your whole mind is engaged and

26:57

and you're just able to bring like this

26:59

this power to bear on things that if

27:01

it's something you don't really like,

27:02

you're probably never going to have just

27:04

that top top ability there. You know,

27:06

>> it's almost matching up with what you

27:07

said about being sort of the 99th

27:09

percentile within an industry

27:11

accumulating the 99% of your brain power

27:14

onto this.

27:15

>> You have to you have to be obsessed and

27:16

love something. And like this is not to

27:18

say that like everyone should only do

27:20

things they love to be successful cuz

27:21

you got to do all the grunt work too.

27:23

But then once you have a certain level

27:25

of where you are, you should structure

27:27

your company and structure your life

27:28

where you do the parts that you love and

27:30

you're good at and other people could do

27:31

the parts, you know, that you're not as

27:32

good at that you don't love. Joe Hudson,

27:35

uh, who has just become the head of

27:39

human performance at OpenAI. Uh, he's

27:42

like kind of an underground hero coach

27:44

type person. I'm aware that coach has

27:46

got a lot of icky uh, associations with

27:48

it, but this guy's legit.

27:50

Really, really great. He says,

27:51

"Enjoyment is efficiency." And that's

27:54

kind of I think referencing what you're

27:57

talking about here, which is if you

27:58

absolutely love something, it takes

28:00

fewer inputs to get more outputs.

28:02

>> 100%. You can get into the flow. You

28:04

could just be great if you love it. And

28:05

so that's that's how you should

28:06

structure your life as much as possible

28:08

is what are the things that you love

28:10

that you're good at that are worth

28:11

doing? And that's and then do more of

28:12

those and do less of the things you

28:13

don't like, but have someone else do

28:15

them if they're necessary.

28:16

>> How do you avoid cynicism? It's a very

28:19

easy trap in the modern world.

28:21

>> This is uh this is the debate I was

28:22

having with Peter Teal earlier about

28:24

stuff. He tells me I'm too naively

28:25

optimistic and is like you want to be

28:27

kind of optimistic in general, but you

28:29

don't want to be you don't want to be

28:30

like overly so. And you know, in

28:32

general, I think it's easier to be

28:34

pessimistic and and cynical. I think

28:36

it's like an easier thing to be. I think

28:37

I think it's like you can just always

28:38

say why things won't work. And it

28:40

actually takes it's a little bit of a

28:42

challenge to say, "Okay, this is really

28:43

broken. The system is really broken.

28:45

Other people haven't been able to do it.

28:47

How are we going to make it work despite

28:48

that? It's it's kind of like a it's kind

28:50

of like being like the hero, warrior,

28:52

champion to say even though even though

28:54

this is a mess, what are we gonna do to

28:56

make it work? And and and that's it's

28:57

it's a it's a leadership quality that I

28:59

think if you bias towards that it can be

29:01

figured out often I've just found often

29:03

times things can be.

29:04

>> Have you ever read uh Endurance by

29:06

Alfred Lansing? It's about Sir Ernest

29:08

Shackleton's crossing of the Antarctic.

29:10

Very cool. So it's the best I think the

29:12

best retelling of that. And um it's

29:15

really interesting cuz all of the guys

29:16

had their their own individual journals

29:18

or diaries that they were writing in.

29:20

And what you hear from everybody else

29:24

except for Shackleton is what

29:26

Shackleton's saying. But what you read

29:28

in Shackleton's diary is what Shackleton

29:30

was thinking. And it's this really

29:32

interesting dichotomy between what he

29:36

says and how he needs to show up as a

29:38

leader.

29:38

>> Yeah.

29:39

>> And what he's thinking privately. And

29:41

it's almost like a Bruce Wayne Batman

29:44

type split personality.

29:46

>> What was he thinking privately?

29:47

>> He is just swimming in self-doubt and

29:50

uncertainty and fear. He has no idea if

29:52

it's going to work. He he doesn't even

29:54

know if he if this is the right. But he

29:56

goes out there and he needs to say to

29:57

the guys, "This is exactly the way that

29:59

we're going to go and we know that this

30:00

is going to work and we're such and

30:02

such." And um it it was the first time

30:05

that I'd ever really thought because

30:06

obviously the

30:07

>> the consequences are so dire. But it

30:10

really made me think about huh there are

30:12

prices that leaders pay that nobody else

30:16

pays and that you can't share the burden

30:18

of and everybody everybody has main

30:21

character energy in their own life.

30:23

>> Right? Everybody is the lead star.

30:25

They're the front man woman of their own

30:27

existence. Um, and I think a lot of the

30:30

time we want to port that across onto

30:31

the teams that we work in, the

30:32

organizations that we're a part of. You

30:34

go, okay, there's going to be some

30:36

prices that you're going to have to pay

30:37

for that.

30:37

>> As as a as a leader, you have to suffer

30:39

things that no one else suffers, and you

30:41

have to deal with the things no one else

30:42

deals with. And it's actually really

30:44

interesting cuz I I I invest in a lot of

30:46

great leaders now and try to help them

30:48

and try to mentor them. And it's very

30:49

funny cuz you end up sometimes having to

30:51

be their therapist a little bit because

30:52

there's no one else they can talk to the

30:54

company, what's going on. You know, we

30:56

we didn't really grow up with therapy in

30:58

my house. It's not something I do at

30:59

all, but I think I imagine it's

31:00

something similar when people are

31:01

dealing with struggling through

31:02

something really hard like this. One

31:04

thing I I kind of wonder is you've been

31:06

around a bunch of special people, Peter

31:07

Teal, Elon Musk, all this stuff.

31:09

>> Is it clear in early like in the early

31:12

days? Like, do you remember the moment

31:13

where you're like, "Oh, this person's

31:14

kind of different. They are that five

31:15

standard deviations." Yeah.

31:17

>> And intelligence or conviction or

31:18

whatever.

31:19

>> One of the I mean, we all try to like

31:20

set strategies for ourselves and goals

31:22

and whatnot. Like I remember in college

31:24

I thought, "Wow, this is a really good

31:25

opportunity at Stanford to reach out to

31:27

a bunch of successful people and kind of

31:29

learn from them, see who I see who I

31:31

admire, who I want to be like." And you

31:33

know, I met some of the legends of

31:35

global macro finance who I'm still in

31:36

touch with. I met a bunch of other big

31:38

business people and run Fortune 500

31:39

companies. And Peter to me had by far

31:41

the most interesting intellect. He was

31:42

he was clearly just one of the smartest

31:44

people I'd ever met. And he was

31:45

interested in some of the same things I

31:46

was interested in. And he taken some of

31:48

the logic farther than I ever had. So So

31:50

I always learned the most from

31:51

conversations with him. So that's that

31:52

that to me was one of the reasons he was

31:54

most attractive to try to work with and

31:55

try to build with.

31:56

>> And so before you start uh multiple

31:58

other companies, you do Clarium, which

32:00

is his hedge fund, I guess is the way

32:02

>> global macro fund. Yeah. It's it's like

32:04

the most fun part of finance. There's

32:05

always there's lots of parts of finance.

32:06

Finance is like all sorts of areas.

32:08

Global macro is looking from like the

32:10

highest possible perspective at

32:11

everything going on in finance. It's

32:13

like the the bond yields around the

32:15

world, the flows of money around the

32:16

world, how equities are being valued.

32:18

you basically it's like just from very

32:20

top down like what is happening in the

32:22

world of finance and how do you map it

32:23

out how you and you look at

32:24

relationships and you understand you

32:26

know of course like you know the

32:27

Australian currency is going to be

32:28

really correlated with the price of

32:30

metals and there's different ways in

32:31

which they can get way off and you can

32:32

create a regression trading system if

32:34

they're way too far off where you can

32:35

make money based on the reversion but

32:37

then you say oh normally I would trade

32:38

this but actually this is happening cuz

32:39

because China just like you know invaded

32:41

this thing over here so actually it's

32:43

not a good time to do it and it's just

32:44

it's this really really fun map of the

32:46

world and we we're really interested in

32:48

longdated oil prices, what was happening

32:50

with oil. We're very interested in how

32:52

uh Fanny May and Freddy Mack, which are

32:54

the two big giant housing groups, they'd

32:55

gotten to be much much bigger in the US

32:57

in the early 2000s. And and they they're

33:00

giant mortgage back security portfolios

33:01

that are worth trillions of dollars. And

33:03

they had to hedge these portfolios

33:04

basically and I won't go into all the

33:06

details. It's a little bit esoteric, but

33:08

based on how they were hedging their

33:09

portfolios was completely changing the

33:11

global fixed income market, completely

33:12

changing the pricing of bonds. We found

33:13

some really interesting ways to do

33:15

things there. I my f my my favorite

33:17

story though I'll give you is just to

33:18

give so so we're working at the hedge

33:20

fund market's open we're in San

33:21

Francisco we're in 555 California right

33:23

it's that big building downtown like

33:26

after after 911 I think Peter had a

33:27

parachute in his room I always told him

33:28

have parachutes for us too but anyway

33:30

[laughter]

33:30

there's one parachute okay that was just

33:32

a joke but it's probably anyway but no

33:35

so it's cool office on the 42nd floor

33:37

and and and the market's open at 6:30 so

33:39

the traders are supposed to be in at

33:40

6:30 except the first Friday of every

33:42

month you had to come in at 5:30 because

33:43

the non-farm payrolls come out still now

33:46

the first Friday of every month at 8:30

33:48

a.m. on the east coast. And it turns out

33:50

this is a very important number because

33:52

it shows you what's going on in job

33:53

growth in the whole economy. Uh you know

33:55

based on the Bureau of Labor Statistics

33:56

numbers and and what was really cool is

33:58

Kevin Harrington was our head of

33:59

research. He's like genius crazy physics

34:01

biology interesting guy who's who's

34:03

ended up being on the national security

34:04

council later like super interesting

34:05

smart guy. And he like had mapped out

34:07

this thing I'd helped him with where we

34:09

realized that they were adjusting the

34:12

numbers wrong uh based on seasonal

34:13

adjustments. instance, we were able to

34:15

predict whether the number would hit or

34:16

miss. And this thing would move bond

34:18

markets like crazy, like like hundreds

34:20

of billions of dollars and move around

34:21

based on this number. And we figured out

34:22

there systematically mis misdoing it.

34:24

This is totally legal. You're allowed to

34:25

do this in finance. So we figured, okay,

34:26

this is a mistake. 75% of the time we

34:29

can make money trading this. And then we

34:31

can adjust it, make it even better based

34:32

on where the market was position. So the

34:34

market's going to be surprised and we

34:35

can kind of tell there's a good chance

34:36

it might be. But take a big position,

34:37

right,

34:38

>> like the night before or whatever and

34:39

then like make a ton of money that

34:40

morning when it worked. And it was at

34:41

4:30 in the morning about 75% chance.

34:44

And so if you do it enough, you're going

34:45

to do it well over time. And it was the

34:46

most fun thing cuz we take these like

34:48

giant positions on this thing. And you'd

34:50

be there with like Peter and Kevin.

34:52

>> This just like the highest high stakes

34:53

game of poker.

34:54

>> It's the most fun I think of poker. And

34:56

you're like literally personally have

34:57

like a few million dollars writing on it

34:59

at a time as early 20s. These guys, you

35:00

know, Peter probably had hundred million

35:01

dollars riding on it every time. And and

35:03

and you're like high-fiving if it goes

35:05

right or you're depressed if it goes

35:06

wrong. It was the most it was it's the

35:08

most fun heist example.

35:09

>> Is he sweating or what's his uh

35:10

everyone's like everyone's like

35:12

[laughter] what's going to happen? It's

35:13

like so exciting and then you're like I

35:15

remember so many like celebrations. We'd

35:17

all like go to like our favorite

35:18

breakfast place afterwards if it worked

35:19

out. It was [laughter] it was pretty

35:20

cool.

35:21

>> Yeah. What you do if it didn't work out?

35:23

>> Yeah. Sometimes we go there anyway. But

35:25

but uh for some reason in my mind, this

35:27

is maybe who I am. I remember all these

35:28

times it worked out. I barely remember

35:29

the time it didn't. I know it didn't

35:30

sometimes but like maybe this is like a

35:32

serial entrepreneur sort of thing where

35:34

you just have the good memories, you

35:35

know?

35:35

>> Right. I was like, but I remember it was

35:36

it was it was a fun time.

35:37

>> That's a entrepreneur survival uh

35:39

tactic.

35:41

>> All the traumas too because if you know

35:43

if you remember how hard something is,

35:45

you're very less much less likely to

35:46

have started again or do another

35:48

>> way. I would have built all these

35:48

companies if I had to force myself to

35:50

like think about that too much. You

35:51

know,

35:52

>> you the way you just described like the

35:53

global macro fund, you're pretty young

35:55

when you're doing this and that's not

35:57

what you were doing just before. So,

35:59

it's not like you had tons of

36:00

experience.

36:00

>> My my little brother got me reading like

36:02

a bunch of economics and finance stuff

36:04

when I was young. So I kind of had a

36:05

weird both computer science and finance

36:07

background. So and I had opinions about

36:08

it. So I think it's a weird those are

36:10

like my two things and so so you know

36:13

so I knew that world pretty well and

36:15

yeah you learn a lot being in it. You

36:16

just do. You just have to be in it to do

36:17

it.

36:18

>> But it's I mean credit to them for

36:19

betting on you to be able to to do that

36:20

versus you know they could have hired

36:21

somebody who's been doing that for 20

36:23

years. And

36:24

>> this is we talk about lessons from Peter

36:26

Teal. I think one of the most important

36:28

ones I've seen from him that I've

36:29

definitely taken to heart in what I've

36:30

done is I'd much rather hire hire I'd

36:33

much rather hire for talent and ambition

36:36

and hard work then hire for expertise.

36:38

So almost and there's exceptions, right?

36:40

So if if you have a company that's

36:42

already really working and already

36:43

really scaling and there's like a

36:45

machine that has to be run, you want to

36:47

hire the guy who knows how to run the

36:48

machine. That's fine. That's like that's

36:49

like that's the operator who knows what

36:51

they're doing already. But when you

36:52

before you've actually built the machine

36:54

or when you're doing something with a

36:56

group and like kind of high finance or

36:57

whatever, you want to hire for raw IQ

37:00

and hard work and just iterating and

37:01

figuring it out and that's that's a

37:02

different skill set than the people who

37:04

run the machine.

37:04

>> What's your way of filtering for that

37:06

IQ? What you know how do you interview

37:08

or what are the signals you look for?

37:10

>> This is like super hard. I think you

37:12

start with

37:12

>> throw hard math problems at people. What

37:14

are you doing?

37:14

>> We used to do that at Clarium. We had

37:15

these really hard problems. I remember

37:17

uh my friend Sujio ran Dropbox one of

37:19

his partners there his new his new

37:21

ventures channel Lee was like the

37:22

highest scorer on on these tests I gave

37:24

this smart smart guy I uh yeah we had to

37:26

do really hard tests but you know once I

37:29

already had around me a bunch of people

37:30

who were really top talent it's like

37:33

they knew people like oh who is the

37:34

smartest guy in your physics PhD at MIT

37:36

who is you know things like that so it's

37:38

it's like you kind of you kind of get a

37:39

network that's already really solid and

37:40

then you go out from there

37:42

>> and so let's go more on that so uh two

37:44

things one Ben told me something that

37:46

like there's some number I don't know

37:48

the exact number 20 30 people who have

37:50

worked at your companies that have then

37:52

gone on to start like unicorn companies

37:54

dude what is what is that number roughly

37:56

>> oh gosh you know I don't actually keep

37:57

track of the official number but there's

37:58

over 20 people who've worked directly

38:00

for me who've started huge companies and

38:01

it's something I'm really proud of is

38:03

that we've had a lot of super talented

38:04

people and you know in some ways it's

38:06

like I like to like back my mind like oh

38:08

yeah I'm a great guy take credit for

38:09

shaping them but in some ways it's like

38:11

they were already really amazing and and

38:13

you know and and so I was like maybe I'm

38:15

just good at convincing them have worked

38:16

with me ahead of time, you know. So,

38:17

it's like there's a

38:18

>> Well, there's a nature nurture thing,

38:19

right? So,

38:19

>> yeah, there's and there's some of both.

38:20

I I do like to think that we've given

38:21

them useful frameworks just like I

38:22

learned from Peter. They might have

38:24

learned some of these things from me,

38:25

too. But, but, uh, I mean, one of the

38:27

one, you know, not all of them I've even

38:28

backed. Like, some of them I back, some

38:29

of them went off and then did it

38:30

themselves and Right., you know, which,

38:32

you know, usually they let me back them,

38:33

though. [laughter]

38:35

>> Yeah, that's the worst when you don't

38:36

you didn't get the call.

38:37

>> Actually, I'm still proud. I was still I

38:38

was still proud of like I'll tell you

38:39

the guy who came right out of school to

38:41

help run adapar with me and and and I

38:44

was like he became president of the

38:45

company super talented guy and uh I did

38:48

not know what I was doing uh even though

38:50

I just already done Palunteer there's

38:51

always mistakes I was making cuz there's

38:53

like things I didn't realize I had

38:54

advantages of Palunteer and you know I

38:55

had a part multillion dollar company now

38:56

it's great but there's things that took

38:57

longer there's mistakes but there's also

38:59

obviously good things we did cuz the

39:00

company worked out really well and I

39:02

think he had he had like went off on his

39:04

own and he had like uh you know his own

39:06

tough time or whatever for a little

39:07

while afterwards and then he ended up

39:08

partnering with someone and he started a

39:10

company that became worth $3 billion in

39:11

like four years and he sold it.

39:12

>> Wow.

39:13

>> Which I'm like, "Oh, wow. That's pretty

39:14

cool." So, but I was like and

39:16

you know and I and I he didn't have they

39:17

didn't bring anyone else into that

39:18

company, but I was still really proud of

39:20

the fact that I'd worked with this guy

39:21

and that he'd gone on and had like crazy

39:23

success. I mean, that's a cool feeling,

39:24

you know?

39:24

>> But for the record, it's always I'm more

39:26

proud if I got to invest, too, right?

39:28

>> Come on. Let's take care of your

39:29

buddies. Exactly. Most of the like 95%

39:32

of them it's like people will bring me

39:34

in and I'll still like help and we'll

39:35

coach and work with them together. So

39:36

let's do uh big hit number two. Adapar.

39:40

Um you explain like

39:42

>> explain in layman's terms what even is

39:44

adapar and then I want I have some

39:45

questions for you but let's first do the

39:46

the explain like I'm 5 years old. What

39:48

is adapar?

39:49

>> So adapar is uh is the leading wealth

39:51

management technology company in the US.

39:53

And so there's people who work at what

39:55

are called raas registered investment

39:56

advisors. There's tens of thousands of

39:58

these. And they will help you or your

40:01

parents or whoever kind of manage their

40:03

wealth. They'll give you reports.

40:04

They'll do your trusts in estate.

40:05

They'll just do everything around kind

40:07

of the wealth planning. Man, it was

40:08

really freaking hard. We built all this

40:10

stuff and we like hooked up to Schwab

40:11

and Fidel because those are like two of

40:12

the biggest everyone's heard of those,

40:14

right? The biggest custodians and we had

40:16

all this wrong data in the reports like

40:17

what are we doing wrong? And it turned

40:19

out the data coming from Schwab and

40:20

Fidelli was often times wrong like what

40:22

is wrong with this space? It's like the

40:23

whole thing's a mess. So it was it was a

40:25

hard company to build. It took about

40:26

three or four years before we really got

40:28

to the point where where it was useful

40:29

to people. And

40:30

>> so you didn't give up though during that

40:31

time. Uh,

40:32

>> oh no, that's like the last like you

40:34

have to have a personality where you're

40:35

not going to give up. I mean, I'll give

40:36

up if it turns out someone else has

40:38

already solved it perfectly. Like if a

40:39

already existed, it would have been a

40:40

stupid thing to do then and I learned

40:41

that then then maybe you give up and

40:42

pivot. But the fact that it's just a

40:44

hard problem, you're not going to give

40:45

up. You're just going to like double

40:46

down and like fix it, you know?

40:47

>> Interesting. So you're you're uh you you

40:49

don't give up if it's too hard, too.

40:51

>> But like that's better. That means once

40:52

I do solve it, it's going to have this

40:54

giant moat. Like Adabar has the biggest

40:55

freaking moat. We were talking earlier

40:56

about guys like Rocket Internet will

40:58

like go around copying things and annoy

40:59

people. Like good luck. If you want to

41:01

try to copy that, you got to take a

41:02

billion dollars in 10 years, you know,

41:03

like no chance. No chance.

41:05

>> You talk about hard work. I when I was

41:07

doing research for this, I I called

41:08

somebody who works on your team and I

41:10

said, "Tell me about it. What's he like?

41:11

What's it like to work with him?" One of

41:12

the things I said was like, "He's always

41:14

on. I've never seen him just go like go

41:17

dark, go off, or just like turn his

41:19

brain off or be like, I just want to

41:21

chill on the beach and just drink my

41:22

ties and not do anything." Is that

41:24

>> Well, this is like accurate for you.

41:25

>> This is like my game, though. I mean, I

41:26

think you have to love what you're

41:27

doing, right? So, like you shouldn't

41:29

play games. you don't you don't like you

41:30

know if you're not playing you're not

41:32

enjoying the game you got to switch it

41:33

up.

41:33

>> Yeah.

41:33

>> So this is this this is this is this is

41:34

what I'm doing this because I really

41:36

enjoy it. It's really fun. I'm really

41:37

good. I think one of the you know

41:38

Aristotle would talk about how like one

41:41

of the highest pleasures is like is like

41:43

using your mind in its highest capacity.

41:45

Right. So it's like you explain like you

41:46

actually enjoy chess more than checkers

41:48

if you're good at both because chess is

41:50

has more levels to it. It's more

41:51

complicated game and so you you want

41:52

your mind to be engaged in interesting

41:54

complex useful things that are getting

41:56

things done. It's it's like a positive

41:57

part. I think it's how humanity works,

41:59

>> right? How do you still um recharge or

42:02

just get clarity of thought? Cuz one of

42:04

the challenges is if you're always on,

42:07

uh you might not have that chance to

42:08

step back and reassess.

42:10

>> You do need open time for creativity. I

42:11

I've noticed when I overchedule myself,

42:13

I'm not nearly as creative then. And I

42:15

think the cold plunge helped, right? I

42:16

don't know if that clear mind.

42:17

>> That was a quick you were in there for

42:19

60 seconds 51 second reset we had.

42:21

>> I think I stayed in for the full

42:22

>> Yeah, you you you actually did two

42:23

plunges. That was pretty good.

42:24

[laughter]

42:25

>> What is a typical schedule like? Uh

42:27

what's a day in the life? Cuz you got

42:28

your fund, you got companies, you got

42:30

all kinds of stuff. You got your

42:31

podcast.

42:31

>> Yeah. My number one my number one is AVC

42:33

and AVC has a build program which is my

42:35

favorite part of it. A third of our

42:36

money now goes towards building new

42:37

companies. So my favorite thing is like

42:39

the strategy for recruiting talent and

42:41

and then strategy sessions with with you

42:42

know to figure out what we're going to

42:43

build.

42:44

>> Right. And so when you you said strategy

42:46

for talent, what was the other one?

42:47

Strategy for building the company. Is

42:48

that

42:49

>> Yeah. You're basically strategy when

42:50

you're the way I look at the way I look

42:52

at opportunities for new businesses is

42:53

you're looking for conceptual gaps in

42:55

the world. you're saying like

42:57

>> here's here's where this industry is now

42:59

like here's how this thing is working

43:00

here's how this process is done in the

43:01

economy and here's where it could be if

43:04

it was using like the right new

43:06

technology or the right new incentives

43:07

or this or this right new effect of

43:08

cultures or what whatever it is that

43:10

like could be better about it and the

43:12

question is not just it's actually very

43:14

easy to find conceptual gaps I'll put

43:16

forward if you spend some time with like

43:17

a few smart friends and you like study

43:18

industry and talk to people you're like

43:19

you usually will find oh wow this thing

43:21

could be done this way and so the the

43:23

harder question is not just what it

43:25

could be done, but like what's the path

43:27

to get there? What could we put in place

43:28

that would allow us to get it there? How

43:30

do we build this thing? How much is it

43:31

going to cost? How do we how do we make

43:33

the people want to? So, there's a lot of

43:34

things, for example, in like healthcare

43:35

and hospitals that are done completely

43:37

wrong, but it makes the hospitals more

43:40

money. So, so, so, so it's like, so even

43:42

though it's like clearly not the right

43:42

way of doing things, they're not going

43:44

to just fix it. You have to find some

43:45

way to get from here to there with the

43:47

actors being willing to do it, right?

43:48

>> So, let's do an example. So, um, give me

43:51

an example. either maybe something that

43:52

you're that's top of mind because you're

43:54

thinking about it right now, maybe it's

43:55

not all figured out or a recent one.

43:57

>> Yeah. So, a gap and a path that you kind

43:59

of figured out

43:59

>> 100%. So, uh it turns out that there's

44:03

tens of thousands of loading yards

44:04

around the country in logistics. These

44:05

are places where trucks come, they get

44:07

in line, they drop things off, they move

44:08

things around, they they you pick things

44:10

up, and it's it's a pretty interesting

44:12

area. And like the biggest warehouse

44:13

company in the country is Prologice. And

44:15

it turns out the founders investor with

44:17

me is based in based in, you know,

44:19

nearby in California. were back there

44:20

and then it turns out there's like a

44:22

handful of the biggest carrier companies

44:24

like Ryder you've probably heard of and

44:25

others and we've actually sold a company

44:26

to Ryder so we're close to those guys

44:27

too and so so because we knew a bunch of

44:29

these guys we started analyzing what can

44:31

we do with AI to make your yards better

44:33

and what we figured out is that if you

44:36

put a just a couple cameras in first we

44:37

thought we need six but just a couple

44:38

cameras you can kind of watch and map

44:40

out where everything is map out how

44:42

efficient the gate is map out the

44:43

processes create accountability and

44:44

incentives for people to go faster uh

44:46

create ways of pinging the carriers if

44:48

it's backed up to like to use other

44:50

loading yards nearby. Just like basic

44:52

little ways on the margins of like

44:53

making these things work better and and

44:54

and timing them and creating because you

44:56

want a little bit of competition. People

44:57

always respond well, you know, if you do

44:58

it properly. And so we've mapped this

45:00

out and we ended up founding a company

45:02

with a really strong AI team. Uh and and

45:04

we ended up kind of co-founding it with

45:06

a this like, you know, the warehouse

45:08

companies, the carriers and others based

45:10

on their needs. And it's really cool

45:11

because the way we built it with them is

45:13

we have like a few hundred million

45:14

revenue already built in that we can

45:16

scale up into their into their yards

45:17

they already own. And then we can have

45:18

something all the other yards want to

45:19

use as well.

45:20

>> Let's do more details. So you you're

45:22

already kind of in adjacent to that

45:24

world because you've done stuff, but how

45:25

does the brainstorm work inside your

45:27

office or where like where when does

45:29

somebody ask the question, hey, could

45:30

you use AI for these loading yards? Like

45:33

how how do like take me all the way to

45:34

that? Like where does a question like

45:35

that even come from? Where does the

45:36

nugget of insight come from?

45:38

>> So my partner Jake's in charge of

45:39

logistics and supply chain. He's the

45:40

he's the kind of biggest driver here and

45:42

he's, you know, done some strategy

45:44

sessions with, you know, our our friends

45:46

running these different companies we're

45:47

close to. get together. I I have a

45:48

vineyard in in Napa Valley and we like

45:50

bring every year like the 100 top

45:51

logistics CEOs and you know drink and

45:54

strategize and hang out you know it's

45:55

like a good way

45:56

>> and you kind of are asking them

45:57

questions to try to find their

45:59

>> what are your problems what works you

46:00

spending money on what are you seeing

46:02

and we'll bring like new entrepreneurs

46:03

and we're backing and like get their

46:05

take on things and we'll bring what are

46:07

called e entrepreneurs and residents you

46:08

know who will talk with them and iterate

46:10

on the idea so we kind of knew loading

46:12

yards is a big thing let's go talk and

46:13

iterate on ideas what we could do to

46:14

apply new things to it and uh you You

46:17

know, in this case, there's a really

46:18

strong AI team that happens to be in the

46:19

UK that we'd met through other friends

46:21

that like had worked on some really cool

46:22

visual problems we thought was very very

46:24

good for this and ended up recruiting

46:26

them as co-founders, which which worked

46:27

out really well. So, we're constantly on

46:28

the lookout for like separately like

46:30

napkin math. You're like, "Okay, so if

46:32

we can improve the loading yards by

46:33

this, that's the size of the prize and

46:35

then that kind of clears a hurdle for

46:37

you."

46:37

>> Yeah, this this is this is and you know,

46:38

the other thing here which I I don't

46:40

love but is definitely helpful is

46:41

California had some new regulatory rules

46:43

about things you had to actually map out

46:44

as well. So they were going to have to

46:46

do something anyway, but why not create

46:47

more value based on then but then like

46:49

be the thing they put in when they're

46:51

going to have to do it anyway in the

46:52

next year. They have the rules coming

46:53

up. There's like, you know, logistics

46:54

has had a few of these things. There had

46:55

some rule on telematics where you had to

46:57

like I think for purposes of not making

46:58

people drive too long, you had to like

47:00

record things differently, which was

47:01

kind of an insertion point for a couple

47:02

TMAX companies we held back as well,

47:04

which is the thing tracks the trucks and

47:05

stuff. So there's things like that where

47:06

the regulatory thing was helpful, too.

47:08

>> And then yeah, you just you sit down,

47:10

you map it out, you know, you've known

47:11

the people for a long I mean, the the

47:13

one of the unfair secrets of business is

47:15

the more success you have, the more it's

47:17

like a platform for a bigger thing you

47:18

can do next time. And that's just the

47:19

way it is. You just got to keep in the

47:21

game cuz like it's like the I think when

47:23

I was really young, I used to think, oh,

47:24

I can start a business and make a lot of

47:25

money and then go to the beach and then

47:26

that would be like fun. And that's like

47:28

that's not nearly as fun as the fact

47:29

that once you've already started and

47:30

successful once, now you get to do like

47:32

bigger things more easily, right?

47:33

>> So, it's like it's like and that's why

47:34

like I think when you're young, partner

47:36

with someone else who already has that

47:37

unfair advantage, help them use their

47:38

unfair advantage, build something with

47:40

them. Now you have all these extra

47:42

unfair advantages. You could do it

47:43

again. So this is this is like there's

47:44

like this naive thing where like you're

47:45

like this lone business founder. I think

47:47

you can do that but it's still hard,

47:49

>> right? Why play the game on hard mode?

47:50

>> Yeah. Why play game on hard mode? Why

47:51

not why not find there's so many places

47:53

in the world where there's unfair

47:54

advantages? And an unfair advantage what

47:56

I mean by that is like every time you

47:57

have a success, you now have the respect

47:59

and attention and understanding of all

48:02

these people who will then help you do

48:03

the next thing, right? And and once you

48:04

prove you're competent once then you

48:06

know there's you know I have a bunch

48:07

more things I could be doing now if I

48:08

had more competent like young people

48:09

around me who want to be entrepreneurs

48:11

and partner with me. We have a bunch of

48:12

them but we always need more cuz there's

48:13

the thing we lack. We don't lack ideas.

48:15

We don't like money. We lack really

48:16

talented hardworking people with

48:18

persistence you know. So at the time I

48:19

was we working I was working for you

48:21

working on the hedge fund you were

48:22

invested in Facebook you're doing other

48:24

investing and we decided to really go

48:26

all in and spend a lot of time on

48:27

Palunteer. Why why was that the right

48:29

choice then it's tech but go more

48:30

specifically like what what what was

48:32

what was your high level around that at

48:34

that time? How were you thinking about

48:35

it? What what gap were we going after?

48:37

It was very broadly uh government and

48:40

then it was within that it was defense

48:43

uh more narrowly it was sort of the the

48:46

um intelligence community that sort of

48:47

adj you know defense adjacent uh I

48:53

you know it was it was a couple of years

48:54

after 9/11 we're sort of starting it in

48:56

2003 2004

48:59

and uh there were all these ways that uh

49:02

I felt that we didn't you know that we

49:06

that we had these very low tech and very

49:09

privacy violating solutions for the

49:12

terrorism problem. You know, it was it

49:14

was annoying that you had to take off

49:15

your shoes at every airport, which was

49:17

like just not security, but it was this

49:20

incredibly intrusive security theater.

49:23

And uh and then yeah, the the question

49:25

was could you do something where you'd

49:27

have, you know, way more security with

49:29

way fewer um you know, technolog is

49:32

doing more with less. And so the the

49:34

terrorism context was could we have more

49:36

security against terrorists with less

49:38

intrusive um civil liberties violations?

49:42

And then um and then and then there was

49:44

broadly this question, you know, was was

49:47

there some way to to reset all these

49:49

things in the government? And uh it was

49:51

a black box and you know we did not know

49:54

whether places like the CIA and NSA and

49:58

places like that had you know some super

50:00

duper computer working in the background

50:02

that was catching all the terrorists. Um

50:04

we sort of suspected they didn't um and

50:07

um and then it took it took a while to

50:09

figure that out but uh I I think our

50:11

suspicions were probably correct. Yeah,

50:12

we saw them doing a lot of things that

50:14

made us suspect that they actually

50:15

gotten way behind Silicon Valley and

50:17

there were I mean there were details

50:18

that you know it was hard to say you

50:20

know with our I remember our first trip

50:21

to Langley in uh in 2000 2005 we got

50:25

inel to invest and you and I were there

50:26

and I think it took us about a day or

50:28

two to process but you know at the end

50:30

of the trip it was like do you want to

50:31

come do you want to buy some stuff from

50:32

the CIA gift gift shop and it was like

50:34

you can get a windbreaker and a CIA cap

50:37

>> coins they were kind of cool

50:38

>> I think they were half of them were made

50:39

in China which was you know But but um

50:43

and then of course there was probably

50:44

some you know there's some steelman

50:46

version where the the it was just a

50:47

decoy and there was really a lot of

50:48

stuff going on in the background. But uh

50:50

>> do you remember the doors all had these

50:52

like eye detectors on each of the door

50:53

that you're supposed to scan your eye to

50:55

get in but none of them worked. So there

50:56

so wherever you go around there would be

50:57

like the eye detector but they were

50:58

broken. they turn them all off which I

51:00

thought was kind of like a kind of a

51:01

typical

51:02

>> I don't know there you know there yeah

51:05

there all these all all all these play

51:07

and so in that context you you you would

51:10

think okay there's there's some very big

51:12

problem that can be can be solved it you

51:15

know in some ways it turned out to be

51:17

more difficult than uh than one would

51:19

have thought uh you know there obviously

51:22

are all these pro all these sectors

51:23

where you have that are very very broken

51:26

where um you know education healthare,

51:29

all all these sort of deeply broken

51:31

sectors where the solution um you know

51:36

there's some obvious solutions and the

51:37

fact they have not been implemented

51:39

tells you there's some big barriers to

51:40

implementation.

51:41

>> They're set up not to allow you to break

51:42

them. This is kind of what I think the

51:43

palanteer version is you know there was

51:45

uh there was this uh forget what the

51:46

acronym stands for. was this a software

51:48

program. This was once we moved from

51:50

intelligence community into the broader

51:52

military called DSyncs which was just

51:53

sort of this um in theory was a system

51:56

we were supposed to figure out ways to

51:58

integrate the data across all these uh

52:00

units in the field and headquarters and

52:03

you know integrate what was going on

52:04

which is incredibly important thing to

52:07

do in practice. Uh it was a sort of a

52:09

cluji consultant

52:12

billion-dollar boondoggle that didn't

52:14

work and uh and we naturally thought we

52:17

would replace it in in 6 months and it

52:19

took it took something like 10 years. We

52:22

finally replaced it in in 2019 and you

52:25

know we had to sue the army which and

52:27

you know we had all our all our

52:28

consultants quit and said you know um

52:31

this is just the most ridiculous thing.

52:32

you'll never get a you can't sue your

52:34

customer and um you definitely can't sue

52:36

the army and uh not only will you not

52:38

get them as a customer, you will not get

52:40

anybody else in DC. You'll just be black

52:42

ballalled by everybody. And then but

52:44

then you know um and it was under we we

52:47

it was in 2016 that we sued them. It was

52:49

under a 1994 law where um you know the

52:53

military was supposed to have an open

52:54

procurement process and then when you uh

52:58

and then what you can do when you sue

52:59

somebody is you you get discovery and

53:01

you discover and you sort of discover

53:03

what's going on in the sausage making

53:05

factory that no one's investigated for

53:06

22 years and there's a lot of stuff and

53:08

there were sort of you know there there

53:10

were internal reports that Palanteer

53:11

worked better which didn't matter

53:13

because obviously nobody can assess the

53:14

software but then more scandalously

53:17

there were um you know [clears throat]

53:18

there were um orders to suppress the

53:20

reports. There were emails that people

53:22

had to hide the findings and there was a

53:25

cover up and this this did did not sound

53:27

like the open procurement process and uh

53:29

and then you know and you you won at the

53:32

court level then you won at the district

53:35

at the DC circuit level and uh you know

53:38

you were forced to they were forced to

53:39

reopen the process and then you know

53:41

after after a decade we we sort of got

53:43

the contract. So yeah, so there was

53:47

something that was doable, but it was it

53:48

was not as straightforward as it as it

53:50

sounded.

53:51

>> It feels like there's these and SpaceX

53:52

had to sue as well, of course. It feels

53:53

like there's these parts of our society

53:54

that are set up to just maintain their

53:57

broken nature, and you have to just

53:58

punch really, really, really hard

54:00

through them. And it's very rare people

54:01

do that, you know.

54:02

>> Yeah, SpaceX was the first successful

54:03

template to do it.

54:05

>> Uh and

54:08

and there's, you know, there's obviously

54:09

something like this. you know, the

54:12

university system seems as broken or or

54:16

more than the military procurement

54:17

process. And then especially with the,

54:19

you know, crediting agencies and all

54:21

these things where where they where they

54:23

just uh are designed to to prevent

54:25

newcomers and the it's it's like if you

54:27

actually had a newcomer, they would just

54:30

eat the field.

54:31

>> They try to crush you. They try to crush

54:33

you. It took us thousands of pages of

54:34

regulation. There's this kind of like

54:36

trope that you've had throughout your

54:37

life or like thoughtful intentionality

54:39

around going towards talent dense

54:41

clusters. This is probably a function of

54:43

why you were rejected PayPal but still

54:45

were like no I have to work here. Like

54:46

it doesn't

54:47

>> This is becoming more obvious to

54:48

everyone now, right? It wasn't as

54:50

obvious 25 years ago.

54:51

>> Um but but the question being why was it

54:53

to you and why was it so clear that you

54:54

have to work at PayPal?

54:56

>> To me, you want to be around the most

54:57

interesting people that I could learn

54:58

something from.

54:59

>> I like had some other internship

55:01

freshman year and I just learned

55:03

basically nothing. If anything, I was

55:04

teaching the people their stuff, which

55:05

is terrible cuz I was 18.

55:07

>> I I don't want to attack them. It was

55:09

some like silly startup. This is like

55:10

this is the year 2000. So, it was like I

55:12

don't want them to It's like they'll be

55:13

the one mentioning them online. They'll

55:14

some they probably I haven't seen them

55:16

in 20 years. They'll probably be really

55:17

sad. I'm like, "Yeah, those dumb people.

55:19

>> Terrible." [laughter]

55:19

>> I mean, they're very nice people. I just

55:21

wasn't learning anything.

55:22

>> I've always been interested in ideas and

55:25

improvement. And like I was at Stamford

55:27

taking really advanced things to push

55:28

myself. It made sense like follow the

55:30

most interesting, most dynamic people.

55:32

like where are the people I admire the

55:33

most going and go try to work with them

55:36

and and and learn and build together,

55:37

you know?

55:38

>> What were most of your friends doing at

55:39

that time?

55:40

>> Teaching computer science at Stanford

55:41

and going way ahead in math and CS and

55:44

getting their PhDs in MIT and and

55:47

Berkeley and Stanford and Caltech or

55:49

whatever. You know, I had some friends

55:51

who were more into history, but it was

55:52

very few. Most most of the ones I was

55:54

closer to were kind of going deep on the

55:56

on the tech side. It's funny looking

55:57

back, but it wasn't that hard getting

55:59

all of our smartest friends to come work

56:00

with us. I think because we just had a

56:01

really interesting place and it was a

56:03

natural place once we had like some of

56:04

our smartest friends. Everyone just it

56:06

was just more fun to be there and be

56:07

somewhere else. And

56:08

>> someone was telling me the other day

56:09

they want to hire people who don't know

56:10

each other and I just for me I totally

56:11

disagree. I think it's like really good

56:12

if they like like each other and like

56:15

like admire each other and want and want

56:17

to like work hard together and then and

56:18

be at the office late together. I think

56:19

it's a positive thing.

56:20

>> Yeah, probably.

56:21

>> Is that what drew you towards PayPal?

56:24

Really? Just like close most your

56:25

friends were there or was

56:26

>> No, it was people I admired who were

56:28

there. That was like there everyone

56:29

there was older than me. Mhm.

56:30

>> So, when people talk about the PayPal

56:31

mafia, it's like, you know, Peter's 15

56:33

years older than me. Uh Elon's 12 years

56:35

older than me. They're both friends, but

56:37

they're both people who I I get no

56:39

credit for any of that stuff. I was just

56:40

like a kid who was learning there

56:42

basically. And and and and I did get to

56:43

know some more people there. And I guess

56:44

there were people like Bob Mcgru who

56:46

were working there who I didn't know

56:47

from my fraternity and from Stanford

56:49

review and stuff. I ended up recruiting

56:50

him to paler later as a key person. He

56:51

ended up building OpenAI after that

56:53

after after a decade. So, there were

56:54

some people there who I guess who I

56:55

knew, but it was mostly like getting to

56:57

know them there. But but kind of having

56:58

watched from the computer science

57:00

department, people like had like kind of

57:01

heard about it and like knew through

57:03

friends who were going there was was

57:04

more of the impetus cuz I was pretty

57:05

young at the time.

57:06

>> Yeah. Why do you think teal took this

57:08

like you'd eventually let you join

57:09

Clarium and be a very meaningful

57:11

partner?

57:11

>> He gave he gave a lot of young people

57:13

>> I think he gave a lot of young people a

57:14

chance who are smart. Uh I think he

57:16

likes people who are smart and who argue

57:18

with him and who bring him fresh ideas

57:19

and who uh and who are intellectually

57:21

intellectually dynamic. for to really

57:23

get along with him. I think you have to

57:25

have like an interest in economics and

57:27

history and philosophy. Like I said, I

57:29

think having that and then being able to

57:30

build stuff obviously was able to be

57:32

useful. My kind of self-startedness was

57:34

like off the charts. And so when I was

57:36

there, I like bring smart friends to

57:37

help us with things and push things

57:38

ahead and and be like, "We should be

57:40

doing this, you're doing that." And so I

57:41

ended up there wasn't really another

57:42

manager there. So I ended up kind of

57:44

running a lot of the place uh just just

57:46

by being confident, which was very

57:47

productive, but also I think there's

57:49

some people who were like 10 or 15 years

57:50

older who were pretty resentful at the

57:51

time. But looking back, I can see why

57:53

it'd be obnoxious. But but it but but it

57:55

just it's like not like he put me in

57:56

charge. I just started doing stuff and

57:58

creating things.

57:58

>> What was Elon like back then?

58:00

>> So I only knew Elon a little bit that

58:02

far back because he had just been kind

58:03

of like moved out of the company when I

58:05

was going in. So I only actually a

58:06

little bit. He's very smart, very

58:08

intense, very hardworking guy. But I

58:10

I've really gotten to know him better in

58:11

the last decade to be honest.

58:13

>> Makes sense. And sort of like around

58:14

this time you're you're working at

58:15

Clarium and and you had just graduated.

58:18

What were you doing there? I know this

58:19

is like a common story you've told where

58:21

you like there's some inefficiency I

58:22

forget what it wasn't had and made large

58:23

amount of money. Um [laughter]

58:26

>> there's all sorts of these things. I

58:27

mean it was kind of fun cuz it was

58:29

Peter's family office. There's all sorts

58:30

of companies at the time. He was

58:31

investing

58:32

>> in Facebook. He's investing in like I

58:33

remember Zoom with Kevin Harts and I was

58:35

helping out there. It's actually very

58:37

funny. The first time Kevin tried to

58:38

hire me away it was like Peter like then

58:40

took like like value me more which was

58:42

like very funny. There's always there's

58:44

always like indicators. It's very funny.

58:45

I think everyone has to do these things

58:47

and and yeah, there was like this macro

58:49

trading fund I would do I did a bunch of

58:50

research projects

58:52

>> on like all sorts of like commodities

58:54

and currencies and like fixed income

58:57

around the world and you know I think

58:59

the smartest guy there was Kevin

59:01

Harrington who I think he's still part

59:02

of the National Security Council

59:03

actually because Peter hooked him in

59:04

there as a genius kind of physics PhD

59:07

who studed physics and biology and other

59:09

things and uh and he would come up with

59:11

these like crazy conspiracy theories

59:13

about how the world worked and like more

59:15

than half of it would like clearly be

59:17

wrong. Although maybe there's like

59:18

something there stories.

59:20

>> Oh gosh, I'm going to get him. He's

59:21

going to be pissed at me cuz he's still

59:22

like into power. I got to be careful cuz

59:24

he's still in power. But it'd be like

59:25

it'd be like connecting all this stuff

59:27

to other things happening to the people

59:29

running a part of a world, but every

59:31

once in a while there'd be like these

59:32

genius patterns you'd find. Uh uh one of

59:34

them it was like really clear it was

59:38

really clear that like the hedging that

59:40

Fanny May and Freddy Mack were doing

59:42

these are called GSC's G plus enterprise

59:44

is called convexity hedging but

59:45

basically the hedging they had to do was

59:46

like completely destroying the global

59:47

fixed income markets and we can we can

59:49

walk if you want to walk through stuff

59:50

so they made it minute it's like okay so

59:51

so basically you have like the interest

59:54

rate is like 5 point something in late

59:57

2002 and you start and it starts to go

60:00

down because the economy starts looking

60:01

like it's doing worse still cuz they're

60:03

still in kind of throws after the

60:04

bubble, right? It's a mess and and then

60:06

and then Beni Bi at the times is uh his

60:09

Greenspan of course was cutting rates

60:11

and and so because he cuts rates because

60:13

it goes down more more people refinance

60:15

and as more people refinance so so

60:17

here's what you have. You have the the

60:20

GS are like these big leverage entities.

60:21

They have tons of cash. They have a

60:22

certain amount of equity and so they

60:24

have what are called assets and

60:25

liabilities. Their assets are mortgage

60:28

back securities. They're mortgage back

60:30

securities. their liabilities or bonds

60:31

they issue and a lot of the money from

60:32

the bonds comes from the world and so

60:34

they have to match the interest the

60:36

matched interest rate is in the duration

60:38

as much as possible these assets and

60:40

liabilities so duration is how long a

60:41

bond last and so basically if people

60:42

start to refinance your assets start to

60:45

have lower duration and in order to like

60:47

have these things match so you're not

60:49

exposed to interest rate changes you

60:50

have to go basically like buy tenure

60:53

notes

60:53

>> and you have to it pulls the duration up

60:55

>> and so these things were so big that in

60:58

order to in order to like rebalance

60:59

their duration They go when the interest

61:01

rate start going down where people start

61:02

refinancing you start they start going

61:04

buy a tenure notes now when you buy a

61:05

10ear note

61:06

>> you actually make interest rates go

61:07

lower

61:07

>> right bond price go up lower so so and

61:10

so they so go so more people refinance

61:12

said to buy more tenure notes to

61:13

refinance and it was literally like a

61:15

it's like a it's like a feedback loop in

61:16

the economy and so interest rate would

61:18

go all was just going steadily down all

61:20

the way from like 5% to like 3.1 in June

61:21

of 2003 and then finally when like

61:24

everything that had refied could refi

61:26

and and the really not everything refied

61:28

but basically the duration got out of

61:29

like one year and it's like about as

61:31

much as you can go. It's like extra

61:32

extra force. So there's like natural

61:33

force pushing back and then at the same

61:35

time there's all this money coming out

61:37

when you refy you spend more. So all of

61:39

a sudden there's like tons of more

61:40

expenditures happening in the economy.

61:41

Everything looks like it's doing better.

61:43

Inflation's going up, growth going up.

61:44

Fed's not going to Fed's going to raise

61:46

rates. So suddenly they're going to

61:47

start raising rates. And so all of a

61:48

sudden rates go up

61:50

>> and then like and then like they have to

61:52

sell 10ear notes. And so it went from 5

61:54

point something to 3.1 and then

61:57

shot up and all the money hit the

61:58

economy as rates were shooting up. So it

62:00

was like 7.3% GDP growth on the on the

62:02

on the consumption expenditure side. And

62:04

so so anyways, this is like a crazy bond

62:06

trade because you can like basically

62:07

watch it. You can be long and then you

62:08

could like as it starts to get extreme,

62:10

you start to short into it short short

62:11

and then you make a ton of money when it

62:12

comes back. And this is this is an

62:14

unwind of the forest fire effect. And

62:16

this happened like three times over four

62:17

years.

62:18

>> So we made like a ton of money trading

62:20

bonds. And it was really cool cuz it

62:21

turns out like global bonds are all

62:23

really correlated for various reasons.

62:25

So JGBs the Japanese government bonds

62:27

>> for the first time got down to like 40

62:29

to 50 basis points right around that

62:30

time like super super cheap and we're

62:32

able to like massively short. So I

62:33

remember having these like like we

62:34

looked all around the world and like

62:36

pricing like 6 months options on the

62:38

JGBs was like out of the money was like

62:40

so cheap. So we just made like a so I

62:42

made my account made like infinite but

62:44

time was infinite money for me just like

62:46

betting these things could shoot up a

62:47

lot more than people realize which they

62:48

did. So, it's a very it's like a very

62:50

interesting game and it's and

62:52

ultimately, you know, I've built a lot

62:53

of companies. I enjoy things that scale

62:55

over time, but it is like just

62:56

intellectually fascinating and I do

62:57

think it is good for the world to like

62:59

price global markets like more

63:01

accurately and have large financial

63:03

firms like balancing these things out so

63:04

they don't move as dramatically because

63:06

it actually helps make plans for finance

63:08

and stuff. So, actually finance is a

63:09

good thing for the economy, but it's so

63:10

abstract that it's like not really

63:12

nurturing for your soul. You don't feel

63:14

like you're building something. It is

63:15

fun, but you don't. So for me it doesn't

63:16

it's like not as satisfying as creating

63:18

something although I do think it should

63:19

exist.

63:20

>> How in the world were you actually good

63:22

at this right? You were a computer

63:24

programmer you studied economy.

63:25

>> Well I think I think because I was I

63:26

think to do the macro stuff you have to

63:28

be interested in finance interest

63:30

economics. You have to have read

63:32

thousands of articles. You have to have

63:33

like

63:34

>> just been into it and had opinions about

63:36

it. Like when I when I went to PayPal

63:38

actually as a as a kid, my first time

63:40

there and I was working there the first

63:41

summer, uh my friend Ralph Hoe uh was

63:44

working for Roloff in the Treasury

63:45

Department and they and he taught me how

63:47

to use Bloomberg which is like the like

63:49

the finance terminal because like he was

63:50

like rebound. He was like doing repo

63:52

trading on like I don't even know if

63:53

this is appropriate at the time. I love

63:54

Ralph but it was like PayPal is a

63:56

private company and they were like

63:57

trading bonds to like make extra money

63:59

like safely on on people's deposits

64:01

which is great. It was a very safe

64:02

version of trading bonds but still it

64:03

was like pretty fun. is I just learned

64:04

how it all worked, you know, and and

64:06

it's it's like a giant game that all

64:07

fits together and it's one of those

64:08

things if you don't understand finance,

64:10

>> you probably don't understand actually

64:12

how like a lot of things [clears throat]

64:13

work in the world. So, I think it's one

64:14

of those things that people should take

64:15

the time to learn. I don't know if if

64:17

you're interested in business, how the

64:18

world works, like if you want to run the

64:19

world, like the fact that there's people

64:20

who are politicians who don't even have

64:22

any idea how a bank works, which by the

64:24

way is most of them, it's kind of scary

64:25

to me cuz these people are regulating

64:26

them. They're making the rules. So, so

64:28

anyway, I think I think if if you're

64:29

interested as I am in how the world

64:30

works and where the world's going,

64:32

>> I think it's a natural thing to want to

64:33

study.

64:34

>> There's something about the cities that

64:36

I think is very hard to reform. Uh, and

64:39

um, and uh, you know, I'm I'm in I'm in

64:44

Los Angeles. We just had these

64:46

catastrophic fires. And what's

64:49

what's what's I think just depressing

64:51

about it is that, you know, it's there

64:54

are half a dozen reasons we can give for

64:55

the fires. It can be, you know, it can

64:57

be the California Coastal Commission

64:59

that didn't allow people to build modern

65:01

houses that were fireproof. It can be um

65:04

it can be the fire department that

65:06

didn't have, you know, where there was

65:08

no water in the reservoir. Uh it can be

65:10

the um you know, the three Kristen women

65:14

running the fire department who maybe

65:15

will be now replaced by three Karen

65:17

women or something like this. [laughter]

65:18

Or it can be the uh homeless people with

65:21

blowtorrches running around or you know

65:24

and or it can be the the mayor who is

65:26

you know out to lunch in Africa or

65:28

whatever. And

65:31

um and and somehow the pessimistic thing

65:34

is that uh I I maybe the mayor gets

65:38

turned to scapegoat and gets you know

65:41

doesn't get reelected or something like

65:43

this. But uh but that that's about the

65:46

that's about the limit. And then this

65:47

just looks like a a structurally hard

65:50

hard to reform thing.

65:51

>> It's very hard to fix. But couldn't

65:53

>> then this is where you know this is

65:54

where this is I where I would be deeply

65:58

skeptical even of of a place like

66:00

Austin, Texas. You know it's it is I

66:02

mean it's a government town. It's a bad

66:05

university college town. It is a good

66:08

university.

66:09

>> Well, University of Texas, not not

66:11

University of Austin.

66:12

>> We're doing well. But even UT has some

66:13

great students, too. It's not too bad.

66:15

you have the whole the whole the whole

66:17

structure.

66:17

>> The professors are going to be more

66:18

progressive. They're going to break

66:19

they're going to break the politics the

66:20

economic structure.

66:22

>> But there is an answer here. There is an

66:23

answer here. Right. So what's

66:24

fascinating is in the Texas

66:25

constitution, cities only exist at the

66:27

behest of the state. And so if you

66:29

actually got the city to be a little

66:30

more if it became a little more broken

66:32

and you get the legislators and the

66:33

governor step up, you can create a

66:34

capital district and you actually run it

66:36

like in a competent way. So I I think

66:37

the very first governor to actually fix

66:40

a blue city will be like a hero and then

66:41

you'll do it in a bunch of other red

66:42

states. I think it's possible, right?

66:46

>> Um, am I am I just totally naive here?

66:49

>> What? You've had 20 years of Republican

66:51

governors. Why have they not pushed for

66:53

this entire

66:53

>> You had a bunch of Bush Republicans, as

66:54

you would call them, who and I and I I

66:56

like Greg Abbott. I think he's a good

66:57

guy, but there's just not like the

66:59

energy to like go in and be like, we are

67:00

just going to this and fix it and

67:02

it's unacceptable and we're getting rid

67:03

of this nonsense. There has not been

67:04

that energy yet in the party. So like if

67:06

I, for example, were to take over in 10

67:08

years and do that, I would just fix the

67:09

city. I say, "This is unacceptable. It's

67:11

broken. Here's what we're going to do

67:12

instead." like the fact there's more

67:13

people dying for not having enough

67:14

police, the fact it's just insane. We

67:15

just got to fix it. You know, that's

67:17

what I would do.

67:19

>> Okay. Anyway, this is this is where I

67:20

would push back where, you know, just

67:23

walking around Austin the last few days,

67:24

it seems like there are a lot of

67:26

homeless people,

67:27

>> right? Right now, the progressives are

67:28

in charge. Yeah, that's that's in the

67:29

city. You're right. You're right. His

67:30

problem.

67:31

>> And then and then somehow it doesn't

67:34

bother, you know, it doesn't bother the

67:36

Republican governor because, you know,

67:37

Republicans well don't have to live in

67:38

Austin. They live elsewhere. And maybe

67:40

>> I think it bothers him and they've

67:41

talked about making a capital district.

67:42

I think there's I think they need to

67:43

they need Yeah, there's a lot they're

67:44

trying to do well right now.

67:45

>> It's not it's not a top 10 issue.

67:47

>> Yeah, it's not yet and I think it should

67:48

be. So just going back to the question

67:50

then like so do you I think so we talk

67:52

about

67:53

>> so so yes I there's there's some there's

67:55

some point where it's you know it's it's

67:57

the right battle to to pick but

68:00

>> you can't you can't do all of them

68:01

obviously

68:01

>> you can't I don't know I know this is

68:03

probably not the best analogy but it's

68:05

if if you go surfing you know it's yeah

68:08

you you want to paddle just in front of

68:10

a wave you have to you have but the wave

68:12

has to be there

68:14

>> and if you're if you're paddling like

68:15

crazy on a still day where there are no

68:17

waves it doesn't work. Yeah. So, we live

68:19

in these complex systems and like a lot

68:21

of people they they basically are maybe

68:23

too pessimistic or too optimistic like

68:24

you said. They just think system's going

68:25

to happen no matter what. I guess for

68:27

the great man theory of history, you say

68:28

it's not just that a great man come and

68:30

does something. You have to understand

68:31

the system. You have to understand

68:32

what's possible. But you do believe

68:33

people can then make a big difference if

68:34

you if you understand it well enough.

68:36

>> Yeah. You have to there's there's

68:38

probably some sense of timing. There's

68:40

there's there's some point where where

68:43

these things where these things where

68:44

these things can break through. Um,

68:48

you know,

68:50

um, yeah, if if Austin can be fixed,

68:53

you'll you'll be the one to fix it. Uh,

68:56

if it if it can't be fixed, I'll be the

68:58

one to have told you told you so. And

69:00

maybe that's a lesser achievement.

69:02

>> We will we will we will see what we

69:03

could do. There's

69:04

>> like the optimistic thing is I think

69:06

there are some other things you could do

69:07

that might be bigger than fixing Austin.

69:09

>> I am working on many big things.

69:11

>> So if you if if you get too distracted

69:12

by Austin, this this could be a very

69:14

very costly mistake. Although although

69:16

studying

69:17

>> you're writing you're selling yourself

69:18

short you're just thinking fixing Austin

69:21

is is this really big thing you can do

69:23

[laughter]

69:23

and for an average for an average person

69:26

this would be a big thing but you know

69:28

so I think studying SF and studying

69:30

Austin led me to understand the NGO

69:32

thing which I then was very helpful with

69:34

with with DC and we actually are working

69:35

with people running a lot of agencies in

69:37

DC to go after the NOS's and based on

69:38

things we've learned the last seven

69:40

eight years. So I think I think

69:41

sometimes studying the problem around

69:42

you kind of you follow the trail and you

69:43

do find useful things to fight for. Tell

69:45

me a little bit about your relationship

69:46

with uh Peter who's also kind of a

69:48

luminary in Silicon Valley and how he

69:51

was kind of a role model and and some of

69:52

the the lessons that you learned from

69:54

him.

69:55

>> Sure. Peter had a family office that I

69:57

went to work for cuz he just sold PayPal

69:59

and he was starting a lot of companies

70:01

and he had just started the hedge fund

70:02

ended up being very involved in the

70:03

hedge fund and he was also the first

70:05

investor in Facebook at the time and

70:07

many other companies. So obviously it

70:09

was a really good place to learn. I I

70:11

mean Peter's one of the Peter's one of

70:12

the highest IQ people who just thinks

70:14

very very differently. He's always sort

70:16

of an outsider in my mind where he's

70:17

just coming at things from very

70:19

different angles and he he really always

70:21

focuses on narrowing things down to like

70:22

what's the factor that actually matters

70:24

here and is very good at being very

70:26

disciplined about that. Someone says

70:27

there's three reasons for something.

70:28

It's like what's the actual reason

70:29

because it's very rare in life that the

70:31

three are approximately the same.

70:32

There's usually one of them that matters

70:33

and he he valued talent very highly. He

70:35

just get things done right away. He

70:37

wouldn't he wouldn't put it off. Um,

70:39

there's there's a ton of lessons. I

70:40

actually wrote I actually wrote

70:42

something on the 10 lessons I learned

70:43

from Peter that I sent out a decade ago.

70:45

But he's he's he's just he's just a

70:46

really really smart guy and obviously

70:48

huge huge advantage to get to learn from

70:49

him for five or six years.

70:51

>> Yeah, absolutely. Did you ever share

70:52

that that that document in terms of the

70:54

10 10 lessons?

70:55

>> Yeah, it's out it's out there probably

70:57

on Kora and on my site. So, feel free to

71:00

look it up.

71:00

>> Cool. Yeah, we'll share that in the the

71:02

podcast notes afterwards. Um,

71:04

>> uh, amazing. Um and then tell us a

71:07

little bit about how you decided to

71:09

found um Palunteer. Like where did the

71:12

idea for that company, you know, come

71:14

from? What what kind of problems did it

71:15

solve? had been talking about it for a

71:17

while uh just in general because it was

71:20

it was kind of an obvious thing where

71:21

where PayPal had gotten really good at

71:24

this technology and the Secret Service

71:25

and FBI guys we work with had no idea

71:28

what to do with the technology and fraud

71:30

and and arresting people, you know, cuz

71:32

Palanteers, sorry, PayPal's big

71:33

challenge was a Chinese and Russian

71:35

mafia were stealing money from it and uh

71:38

and a lot of its competitors went

71:40

bankrupt. If you try to ask people for

71:41

too much information before you can send

71:42

money, no one wants to use your service.

71:44

you don't ask for enough information,

71:45

the mafia could buy stolen credit cards

71:47

and run things through the service,

71:48

right, and steal from you. So, you

71:50

basically ended up having to build AI to

71:52

detect the bad guys, but the AI would

71:54

the bad guys would always figure out

71:55

what the AI is detecting and keep

71:57

changing their behavior. And the AI

71:59

wasn't good enough to stop them because

72:01

they keep figuring out new new

72:02

solutions. And so, we ended up doing a

72:04

combination of AI plus like people using

72:06

investigative tools and the people

72:08

looking through. actually made a lot of

72:09

the customer service people in uh in

72:11

Nebraska for PayPal into investigators

72:14

and they built tools for the

72:15

investigators that combination of tools

72:17

with people and the people intelligence

72:19

plus the AI and iterating to improve the

72:21

AI iterating to improve the tools that

72:23

cut down from 90%. So it was clear that

72:25

model was very applicable to

72:27

investigations on large amounts of data.

72:30

And my roommate and I uh were talking

72:32

about this with Peter. I think he talked

72:33

about with Max. Max is from Russia. You

72:36

know the co co-founder. Max didn't want

72:37

anything to do with it. You know, you

72:38

only want to work with government and

72:39

spies and intelligence committee. That's

72:41

our people disappear, you know, back in

72:42

Russia. So probably certain wisdom to be

72:44

careful about these things. But in

72:46

America, you know, 911 happened and we

72:48

talked to these guys and we were giving

72:49

all this advice and they we saw how

72:51

their groups were spending literally

72:52

billions of dollars on really backwards

72:54

technology. The solutions really didn't

72:56

make any sense. It's really frustrating.

72:57

And we said, "Wow, we could actually

72:59

create something." And I I brought a

73:01

bunch of my friends to work at the hedge

73:03

fund in a summer of 2004. And we've been

73:05

talking about this and kind of sketching

73:06

things out for about six, seven months.

73:08

These guys were all coming from their

73:10

PhD programs in MIT and Caltech and etc.

73:13

and my roommate from Stanford who's a

73:15

senior because I he's one year behind

73:16

me. we were kind of sketching it all up

73:18

together and uh he kind of took the lead

73:21

along with me and and like building out

73:22

what a product would look like and these

73:25

PhD guys were really helpful and a

73:26

couple of them joined the company like a

73:28

couple years later but they thought we

73:29

were kind of crazy and went back to

73:30

school cuz they're like who are these

73:31

guys trying to be spies I think you know

73:33

and a couple of these guys are my

73:34

friends they know me since I was a kid

73:36

>> and so you know it's kind of hard to

73:37

picture your six-year-old friend

73:39

actually doing things that matter for

73:40

the CIA and the FBI and the NSA they

73:42

thought it was kind of silly and so

73:45

you know Stephan and I pushed it forward

73:47

already got up here to give us some

73:47

money. Peter introduced us to a guy who

73:49

had been at PayPal named Nathan uh

73:52

Gettings and he liked to stay in the

73:53

background but he was really key as well

73:55

cuz he had a lot of experience building

73:56

these teams and we started started

73:58

building it and you know actually you

73:59

know several months in Peter Peter tried

74:01

to hire a couple people from the DoD to

74:03

run it and they just didn't understand

74:04

the tech culture we wanted and we

74:05

thought they were pretty terrible uh as

74:07

leaders for our tech company. So, I

74:09

convinced Alex Karp, who was Peter's

74:10

friend, who was helping us in other

74:11

things, uh, to try to like pretend to be

74:13

CEO for a couple years so that we'd stop

74:15

having to have CEOs put in. And Alex was

74:17

so good at it that as he got up to speed

74:19

after a year or two, he became the real

74:20

CEO. And he's he's still running it 18

74:22

years later.

74:23

>> Amazing. Yeah. Um, it's always great to

74:26

hear sort of the the Genesis story. So,

74:28

I'm I'm sure you could have like

74:29

continued on in sort of the hedge fund

74:31

world, had a nice financially lucrative

74:33

career. What made you decide to take the

74:35

risk of, you know, going into

74:37

entrepreneurship, starting a company,

74:38

especially in a space that was, you

74:39

know, uh, at least back then, you know,

74:42

not regarded the same way as it is now.

74:44

>> Yeah, it just seemed like a really

74:46

important problem that needed to be

74:47

solved and it was a really fun

74:49

interesting problem to work on. So, it

74:51

was just one of those things. there's a

74:52

giant gap in the world and you need to

74:53

solve it and so there's of course you go

74:54

work on it and you know I ended up

74:57

obviously starting a few other companies

74:58

before becoming an investor and in each

75:00

case there was just like something that

75:01

clearly needed to be done clearly

75:03

valuable to do and uh and you know I I I

75:08

guess the hedge fun world is interesting

75:09

I I I think I did have the top trading

75:10

results for five or six years and I

75:12

really I got to know a bunch of the top

75:13

hedge fund guys and I really enjoyed the

75:15

markets I really enjoy the qualitative

75:17

and quantitative you know trying to

75:19

figure out what's going on there and but

75:20

it's hedge fun also a little bit

75:22

stressful. You have to you have to kill

75:23

what you eat every year, right? So So

75:24

every year it's starting from scratch

75:26

again. There's no sense of building over

75:27

time. There's no exponential growth in

75:29

the same way. I guess you'd have the

75:30

exponential growth of the money you're

75:31

managing. But it's not it's not I just

75:33

didn't get the sense of creating value

75:34

with it the same way I thought you could

75:36

create value by solving problems in the

75:37

world with companies. And uh and I guess

75:40

the other thing I'd say is the markets

75:41

around 2009 had gotten to be very

75:44

complicated and there was a lot more

75:45

driven by what the governments were

75:47

doing. So macro is a lot more driven by

75:50

if you're because we're doing macro a

75:51

lot more by what China might do, what

75:52

the central bank might do in America.

75:54

And it was getting like really just a

75:57

very different sort of game to trade it

75:59

and and I realized the macro bet that I

76:01

wanted to make for the next couple

76:03

decades was actually there's these giant

76:05

gaps and things to be built in techn in

76:06

the technology world and that was just a

76:08

much better use of my time. So I started

76:10

out a par started investing ended up

76:11

starting a fund in a couple years and

76:13

obviously it's been a great 10 years to

76:15

to to manage billions of dollars in the

76:17

venture world. So so I think that was

76:18

the right macro bet I was lucky to kind

76:21

of pick the highest return area to go

76:22

after. So I think that was a correct

76:23

macro thing to do you know a decade ago.

76:25

>> Yeah. Well I think it was very precient

76:27

that you were kind of even thinking

76:29

about you know 10 years out especially

76:31

you know someone in their early 20s. Uh

76:32

so obviously it was a great bet but I

76:34

but I love the lesson there about just

76:36

um working on what you thought was most

76:38

important in terms of solving problems

76:39

in the world and we definitely need more

76:41

entrepreneurs who are uh you know not

76:43

just like hacking an interesting um

76:46

opportunity in the market but but

76:47

solving real problems that affect the

76:49

world and I I think uh ABC does a great

76:51

job incubating investing which is why

76:53

you know I decided to work with you

76:55

guys. Um, so to tell me a little bit

76:57

about kind of especially the early years

76:58

of Palunteer, um, I'd love to hear, you

77:00

know, some of the lessons in terms of

77:02

scaling, uh, you know, an organization

77:05

like that and and I know especially that

77:07

that from what I've heard that part of

77:10

your secret to success was was around

77:12

talent. So, so tell me a little bit

77:14

about that.

77:14

>> Yeah, you know, I mean, it was it's a

77:15

little talent's always a little bit

77:16

chicken and egg. We started off with a

77:19

lot of our friends from these PhD

77:20

programs who and from the world where

77:23

we'd won the national math and chess

77:24

championships. You start off with a

77:25

bunch of your bright friends and once

77:27

you have enough of the best people

77:28

there, you know, and and they're solving

77:30

really hard interesting problems.

77:31

They're working on really cool

77:32

interesting things, you know, obviously

77:34

you want to bring by people in the

77:36

industry to give credibility, to give

77:38

them feedback to to keep them excited

77:40

about what they're working on. But h

77:41

having a culture with the very best

77:43

people, they are people kind of want to

77:44

join. So if you can start off, it's

77:45

really your first four, five, 10 hires

77:47

are really important. The other thing we

77:49

did is spent a lot of time trying to

77:51

recruit friends who wanted to do their

77:52

own companies and convince them to join

77:53

us instead. And so if you if you can

77:55

get, you know, handful of people who

77:57

each want to be doing their own

77:58

companies, instead they're all working

77:59

together. And of course to do that, you

78:01

have to be generous on equity. You have

78:02

to be generous on how you respect them

78:04

and their role and the fact that what

78:06

they're doing is important. Um but but

78:08

getting getting a lot of people all

78:10

together, each could have done their own

78:11

company to work together was really

78:13

really key. We tried to learn from

78:14

technology culture at Google and at

78:16

PayPal. Back then, one of the things

78:18

that maybe is more obvious today, but

78:20

wasn't as obvious, is in most of the

78:22

world, you had business people ordering

78:23

engineers around. It was really

78:25

important at Palanteer, the engineers

78:26

who were the best were in charge of the

78:28

business people, not the other way

78:29

around. And which means you had to get

78:30

really really good engineers who who are

78:32

mult multi-talented for the for the

78:34

early leaders because they had to be

78:35

able to actually make, you know, you

78:37

basically want a lot of companies, it's

78:39

really good if the engineers can even

78:40

stand up to the business people and say

78:42

no. But the very best companies, you

78:43

have people who are great engineers who

78:45

are, you know, who work closely mapping

78:47

out the plans and and even even more in

78:49

charge in some ways. And so that that

78:50

was really key for for for a deep tech.

78:52

I don't think you want that at every

78:53

company, but for a deep tech company,

78:54

it's solving the hardest problems like

78:56

that what's possible from the

78:57

engineering side determines a lot of the

78:59

business. Um,

79:01

you know, and there's, you know, there

79:03

are a lot of things like I there's one

79:05

mistake I would have made is there was a

79:07

couple really compelling looking

79:08

partnerships early on and Peter Teal,

79:11

uh, one of the one of the really

79:13

valuable things he did early on is teach

79:14

us not to do these big partnerships

79:16

early on. It's very hard for startups to

79:18

to have big complex partnerships. They

79:19

often times break them and it's very

79:21

it's very hard, especially if there's

79:22

two things that are both sort of

79:23

startups. The partnerships very rarely

79:25

work out. Um, so he kind of kept us from

79:27

making some mistakes there, which I

79:28

thought was pretty, you know, obviously

79:30

it's pretty amazing. Um,

79:34

you know, I think a big a big thing we

79:35

did early on was tours of duty where

79:37

we'd actually bring the engineers on,

79:39

you know, and try to echo the culture of

79:40

the places we were working with and get

79:42

get them to know the problems. Another

79:44

thing I thought was really good is the

79:46

culture of the engineering side. We

79:47

considered engineers artists and

79:49

obviously you had a hardworking team but

79:51

you're a lot more tolerant of how you

79:52

interact with artists because you can't

79:54

you know for example you can't just like

79:55

tell an artist to do something between

79:57

certain times they might not be feeling

79:58

inspiration whereas the operational team

80:02

they would we call the office team they

80:04

would set up the office and set up the

80:05

desks and set up the conferences and any

80:08

any any other kind of operations we were

80:10

doing it all like that team was run like

80:11

a military it was like it was like

80:13

people who report to other people and

80:15

there's like you know perfect precision

80:17

and there's no tolerance at all for not

80:18

being on time and being late. It's not

80:20

it's not not that you're yelling at

80:21

people, but it's almost like that you

80:22

like the drill sergeant type thing, you

80:23

know, and so and that's a very drill

80:25

sergeant culture is a very different

80:26

culture. And I think being intentional

80:28

about those different cultures and and

80:30

and what you know and really telling the

80:32

engineers see, wow, there's this really

80:33

intense operational culture that really

80:35

respects me and the engineers are first

80:37

class citizens and they're doing all

80:38

these things for me. they're making sure

80:39

I don't ever waste time going to get my

80:41

cars

80:43

uh wheel fixed and oil change because

80:44

it's more important for me to be here

80:46

solving this hard problem and and it it

80:48

created a certain respect for the

80:50

culture, respect for the institution and

80:52

uh it was very untuitive to engineers at

80:54

first but but it be you know because a

80:55

lot of them are very uncomfortable

80:56

having people help them with things but

80:57

I think that culture really contributed

80:59

really strongly to to something very

81:00

special early on there.

81:02

>> Yeah, absolutely. And and I love this

81:03

analogy of sort of artists versus

81:05

soldiers and that you can have both

81:07

types at at a a company perhaps in

81:10

different divisions and almost have

81:12

different uh subcultures within a a

81:14

company that it doesn't need to be

81:15

absolutely uniform across which makes

81:17

sense especially at a place that that

81:19

has you know very different types of

81:20

people

81:21

>> um and and needs to scale. So uh yeah

81:25

amazing amazing sort of lessons there.

81:27

Um and then tell me a little bit about

81:29

the kind of the journey, you know, as

81:31

kind of Palunteer grew kind of what what

81:33

was your role at the company. Um and

81:36

>> yeah, so so I I was I was kind of like

81:38

the minister without portfolio in the

81:40

sense that I just jump in and build

81:42

things where was needed. So originally I

81:44

was helping run product, help come up

81:45

with a bunch of the early the early

81:47

product stuff uh and you know helped

81:49

recruit a lot of the first 50 to 100

81:51

engineers and others um including

81:54

including most of the early leaders

81:55

there. uh helped you know build the

81:58

business side initially, helped work and

82:00

get some of the initial contracts,

82:01

helped get some of the initial investors

82:02

other than Peter on board. So really

82:04

just kind of running everything in the

82:06

background and you know for the first I

82:08

think 150 people I just I issued all the

82:10

all the all all the offers and all the

82:12

options paperwork and help figure out

82:14

who we're giving what to. um you know

82:18

kind of offside reported them to me the

82:20

the sales you know for a sales I did for

82:22

a little bit but then very quick very

82:23

quickly we found people who were more

82:24

confident than me in each of these areas

82:25

and replace myself which is always the

82:27

always the right right way to handle

82:28

these things

82:29

>> and uh yeah when I did that I actually

82:31

started a separate division the finance

82:33

division about three years in and I was

82:35

still pretty involved in certain

82:36

strategy in the government side and

82:38

certain other business things there but

82:39

I was also really involved building out

82:41

the finance side as the first commercial

82:42

side and figuring out the figuring out

82:44

the strategy there as well. Yeah.

82:46

Amazing. So, you know, it's it's

82:47

interesting, especially someone who's

82:48

um, you know, majors in computer

82:50

science, can be very specialized. You

82:51

know, it sounds like you played a very

82:52

generalist role in the early days of the

82:54

company. Um,

82:55

>> yeah. No, I wouldn't I wouldn't label me

82:57

as a computer scientist as like my main

82:59

thing. I'm very good at computer

83:01

science,

83:02

>> but I was I I think I think I think I

83:04

think the humanities are probably my

83:05

strongest area, which is why I just was

83:07

so intolerant of of the humanities of

83:09

all these colleges that they talked

83:10

about because of how corrupt they become

83:12

and how there there really aren't too

83:14

many people with extremely high IQs

83:15

these days compared to 50 years ago in

83:18

those humanities areas. So, there's not

83:19

as fun to study for really bright people

83:21

is my experience.

83:22

>> Yeah. And I, you know, I think what

83:24

what's been shown over time is, you

83:25

know, you really became a company

83:26

builder, um, in terms of your your

83:28

specialty. Um, so so as kind of

83:31

Palunteer, you know, scaled, um, tell me

83:33

a little bit about how you got into the

83:35

venture capital/investing side of things

83:37

with formation 8, which the and which is

83:40

now.

83:41

>> Yeah. So, so we when we left Palunteer,

83:42

I started out of par. I was CEO of that

83:44

for a few years scaling that up wealth

83:46

manage technology platform and I was

83:48

also just you know had a small angel

83:51

investment fund and and I had a lot of

83:54

people I' I'd kind of brought a pounder

83:56

or who had helped run things there and I

83:58

was kind of guess a mentor to a handful

83:59

of these people or or else just a friend

84:01

for some of them and when they were

84:02

building something new or thinking about

84:04

building something new they' come see me

84:05

and we bounce ideas off you know on the

84:06

whiteboard and we'd sketch it up and uh

84:08

that's involved really early on and

84:10

people founding 10 or 15 other companies

84:12

you know over the over kind of like a

84:13

four or five year period and I was in

84:16

investing in a bunch of them. I was

84:17

advising some of them like I tell them

84:19

who to take their series A or B from. We

84:20

work together on it and I help them like

84:22

here's how you come across for

84:23

fundraising. So I was basically like one

84:25

of my mentors told me you're kind of Joe

84:26

already running a fund. You just don't

84:28

have very much money at all and you

84:29

don't have really anyone any not enough

84:31

people helping you. Had a couple friends

84:32

advising me. So decided to kind of you

84:35

know replace myself at Adapar and uh you

84:39

know I had an opportunity my friend from

84:41

Stanford his family in Korea built LG

84:43

and his dad gave us $50 million anchor

84:46

and I partner with him and one other guy

84:47

and started formation 8 and so I think

84:50

because that came up it just it just it

84:52

just made sense. I was doing it anyway

84:53

might as well be running a fund and once

84:55

I was running the fund of course I'm

84:56

very competitive so I just started

84:57

spending a ton of time you know really

84:59

going deep and learning and figuring out

85:01

what to do. that was different and we

85:03

started you know so we started having

85:04

like 150 events per year. We started

85:07

writing thesis on all sorts of things

85:09

and started just getting really really

85:10

intense on like what makes the very best

85:11

venture fund and and built up my own

85:13

team there and uh the team I built ended

85:16

up doing 70% of the deals the first fund

85:18

the next fund there's a little bit of

85:19

politics and we realized you know what

85:20

let's split up and so my team split off

85:22

after 82 to ABC1 uh just where we kind

85:25

of that way we can have 100% of our own

85:26

deals we're doing because for

85:27

information we only did about a third of

85:28

the deals and uh and I'm still friends

85:30

with the other guys but it made sense

85:32

for me to have my own my own firm cuz

85:33

that's such a different way of doing

85:34

things and uh we're on we're on the

85:37

third fund now, the fourth fund in the

85:39

coming months. So, it'll be my sixth

85:41

core fund we're putting together in the

85:42

next year.

85:43

>> Yeah, it's been an amazing journey. I

85:44

remember actually one time when I went

85:45

on a run with you, I was telling you

85:46

that uh you probably have the most

85:48

interesting team in my opinion in

85:50

venture capital in that it's probably

85:51

the most people that I think have worked

85:53

together in some capacity or or

85:54

connected in some way beforehand. Um and

85:57

so yeah, tell me a little bit about you

85:59

you mentioned kind of alluded to that

86:00

you you had a different way of doing

86:02

things uh about how to how to kind of

86:04

you know uh your philosophy about

86:06

forming ABC.

86:08

>> I think [clears throat and sighs]

86:10

the way these funds should work is you

86:11

have people with complimentary skills

86:13

that are different skills for a lot of

86:14

it. So rather than have six partners

86:16

that are each their own partner and each

86:18

kind of doing their own deals and they

86:19

get together and chat about things and

86:20

help each other, I like having we have

86:23

we have nine partners now, but but most

86:25

of them have very different skills than

86:27

each other. Some of them might be the

86:28

best in the world at growth hacking and

86:29

marketing. Some of them might be the

86:30

best in the world at knowing what's

86:31

going on in the Renaissance and bio.

86:33

another like you know was the lead

86:35

engineer uh who built a lot of the core

86:37

key engineering teams at LinkedIn and

86:38

Yahoo and other companies early and is a

86:40

great engineering mentor and they have

86:43

very different types of skills around me

86:45

Alex Moore who he was our first employee

86:48

at Palanteer actually ran the ops team I

86:49

talked about at first then he was COO of

86:51

two different companies and uh he's done

86:53

a lot of really good angel investing and

86:55

he's kind of goes in and and handles and

86:57

fixes situations if necessary and just

86:59

so just just a really different set of

87:00

people that all component each other

87:01

well you want you want I think you want

87:02

teams to rely on each other and they

87:04

need to rely on each other because it

87:05

makes them stronger and tighter. What

87:07

that also means is there's no fighting

87:08

about this is my deal, this is my deal.

87:10

Like the whole firm works on things and

87:11

one person might source it, but another

87:13

person might be the better person to

87:14

actually help and actually you know

87:16

actually spend more time on it or

87:18

whatever it is. And uh you know there's

87:21

45 people in the firm as a whole. Most

87:23

of them are builders. So we along with

87:25

doing series A and B and C, we've been

87:27

launching a lot of companies lately as

87:28

well. And that's that that actually

87:29

makes us a lot better investors in my

87:31

mind. Although it does keep us busy.

87:32

>> Yeah, absolutely. I think that was one

87:33

of the key differentiators that you know

87:35

when I was talking to to various firms

87:37

uh is is yeah you're one of the few that

87:38

actually incubates and continues to be

87:41

heavily uh involved with building uh cuz

87:43

I do think it keeps your your skill set

87:45

sharper versus you know something that

87:46

you might have done a decade or two ago

87:48

and you know the the game the rules of

87:49

the game have changed uh for sure. So

87:52

it's very very cool to see you know the

87:54

the unique kind of and contrarian

87:56

philosophy uh you know instantiated in

87:59

in in 8 now.

88:01

>> Yeah. Well, it's I mean and and not all

88:03

adventure can be of course totally

88:05

contrarian. I think you have to have

88:06

strong opinions about where are the gaps

88:07

in the world and you have to go after

88:09

those gaps and you have to fix them. And

88:11

there are some gaps we've identified

88:12

that others have identified too and we

88:14

have to invest in the best teams and

88:15

then multiple of us will make money on

88:17

it. But if you do if you do identify a

88:18

gap people are not working on that's

88:20

even better of

Interactive Summary

The video features a conversation with an early Palantir employee and venture capitalist, who discusses lessons learned from mentors like Peter Thiel and Elon Musk, the importance of focusing on difficult problems, the challenges and successes of his career in finance and startups, and his philosophy on hiring talent and building companies. The discussion covers themes of moderate optimism, the necessity of picking the right battles, and the value of surrounding oneself with high-intellect teams.

Suggested questions

4 ready-made prompts