HomeVideos

The Path from The Soviet Union to Building Multi-Billion Dollar Companies — Max Levchin

Now Playing

The Path from The Soviet Union to Building Multi-Billion Dollar Companies — Max Levchin

Transcript

3093 segments

0:00

I was asking myself actually a question

0:01

like why is our relationship as good as

0:04

it is and it is really good and is there

0:06

anything to learn about co-ounding

0:09

companies like you're co-ounding a

0:10

family when you marry someone you get

0:12

together with someone I've had many

0:13

co-founders over the years and some were

0:14

amazing relationships and others didn't

0:16

work out and it's always tempting to

0:18

sort of blame yourself but you obviously

0:20

have to examine it through the lens of

0:21

like what fell apart as a relationship

0:23

between two co-founders and I realized

0:26

that the secret for my marriage

0:29

is I'm still trying to impress this

0:31

girl. Every day I roll out of bed

0:33

thinking like one like how did I get

0:35

this? This is definitely better than I

0:36

should have gotten. So, the attempt to

0:38

impress your mate on a daily basis is

0:42

the secret to the best marriage.

0:43

>> Max, it's nice to see you. And I wanted

0:47

to kick this off with a thank you,

0:50

believe it or not, from what you put

0:54

into my last book, Tribe of Mentors. And

0:57

it was in response, you had several

0:59

responses to this. But it was in

1:01

response to the metaphorical billboard

1:04

question. What would you put on a giant

1:06

billboard anywhere with anything on it,

1:08

what would it say and why? Any quotes

1:10

you think of often or live your life by?

1:12

And there was one that I had never come

1:14

across because I had never seen the

1:15

movie Ronin. And the line was whenever

1:18

there is any doubt, there is no doubt.

1:20

>> Whenever there is any doubt, there is no

1:22

doubt. And I have come back to this

1:25

after watching the movie so many times

1:27

it's hard for me to even give you an

1:30

accurate account or even estimate of

1:32

account. But perhaps we could give

1:34

people some context. Why was this one of

1:38

your answers to that question? It's a

1:40

great quote. I love that quote.

1:42

>> So good. The next line in the movie,

1:44

that's the first thing they teach you.

1:46

I'm only showing off because I love that

1:48

quote so much, too.

1:49

>> It's the first thing they teach you.

1:50

It's a great quote. Encapsulates so many

1:53

different flavors, right? It's like

1:55

don't doubt yourself. The real meaning

1:57

here is you know the answer. You may be

2:00

too scared, too embarrassed, too

2:05

unprepared to sort of embrace the fact

2:07

that you know the answer, but you know

2:08

the answer. So like let's not mess

2:10

around. You know the answer. So that's

2:12

layer one. Layer two is make a decision

2:15

already. Layer three is

2:18

even if it's unpleasant, you're going to

2:20

have to do it anyway. So don't delay.

2:22

There's just like so many flavors of the

2:24

same idea over and over again

2:25

encapsulated beautifully in just a few

2:27

words. I think it's a great succinct

2:30

line. The other thing I I'm a huge fan

2:33

of

2:35

studying leadership. Fan of leadership

2:37

sounded weird. So I love leadership

2:40

related lessons. And there are two

2:44

movies where leadership I think is or

2:45

I'm sure there are more than two but

2:47

like two of the movies that I fall back

2:49

on for leadership content is Ronin and

2:51

Seven Samurai. And Seven Samurai is hard

2:54

to quote because it's in Japanese and

2:55

you can either translation unless you're

2:57

native speaker which I'm not. But Ronin

2:59

is in English and so it's great to

3:00

quote. There's a lot of other great

3:02

lines but that's the best one. And one

3:04

of the layers of explanation that you

3:07

added to it, which I ended up applying,

3:10

it made it very concrete for me, which I

3:13

think you've already done, but it was

3:15

when you aren't sure, I'm quoting from

3:17

what you put in the book here, is when

3:21

you aren't sure about a key employee or

3:23

a co-founder, odds are exceedingly low.

3:25

Your mind will be changed for the

3:26

better. And the way I translated that

3:28

for myself and I kept thinking of this

3:30

and I modified it slightly for myself

3:32

like if there is any doubt there is no

3:35

doubt just to say that if you are

3:38

analytically trying to convince yourself

3:40

of something that doesn't feel right and

3:44

you have to cross-examine that quote

3:46

unquote intuition sometimes but

3:50

chances are you know the answer

3:52

particularly at least in my experience

3:54

with employees and hires Right.

3:57

[laughter]

3:58

>> It's always I should have done that a

4:00

lot sooner, at least in my very limited

4:03

experience. You have infinitely more

4:04

experience with that. So, thank you for

4:06

that. And on the Japanophile angle of

4:10

things, having lived there. I'm still

4:12

very close to my host family who I

4:14

stayed with when I was 15. I am no

4:16

longer 15, I think some people will

4:18

realize. But I was wondering, have you

4:20

ever seen the movie Paprika? This is an

4:23

old anime. It's from 2006, I believe.

4:28

>> I don't think I have.

4:29

>> If you haven't, it might be interesting.

4:32

It will be not too much of a spoiler

4:34

because there's so much around it, but

4:36

here's the basic premise. Psychiatrists

4:38

invent a device called the DC Mini that

4:40

allows therapists to enter and record

4:42

patients dreams. And then when the

4:45

prototypes are stolen by an unknown

4:47

dream terrorist, A, B, and C starts to

4:50

happen. What's crazy about this is that

4:52

there are actually researchers in Kyoto

4:54

in Japan who have been using fMRI and

4:58

waking dream subjects to try to use

5:01

machine learning to then correlate brain

5:03

activity to content. And they've made

5:07

some surprising

5:09

progress with being able to do that.

5:11

It's still very very low resolution, but

5:13

I don't know how far away we are from

5:15

being able to do something like is

5:17

depicted in Papriko, which is pretty

5:19

crazy. I used to be a a voracious

5:21

consumer of both manga and anime and

5:24

then it turns out that time is still a

5:28

zero sum game for most things and so

5:30

I've drifted off my weekly

5:33

supply of Japanese content production

5:36

but I will definitely make room for that

5:37

one. The thing with fMRI and

5:40

brain modeling,

5:43

>> at least last I looked and before I sort

5:45

of picked to do the thing that I'm doing

5:48

now with a firm, I actually looked

5:49

pretty seriously into I'm a huge sort of

5:52

a closet fan of brain computer interface

5:56

and all sort of surrounding ideas and

6:00

opportunities. So I've looked on and off

6:01

at fMRI and all the other sort of stuff

6:03

in a domain and it just seems that your

6:07

cranium is too dense. It's hard to get

6:10

signal other than like very superficial

6:13

you know connectivity from and you have

6:15

to like shave your head which you know

6:16

you've done but

6:17

>> yeah I faced that one.

6:19

>> I'm not yet prepared. Of course you have

6:20

to experiment on yourself otherwise

6:21

what's the point? But I'm actually

6:23

pretty excited about some of the newer

6:25

things happening in brain computer

6:27

interface specifically in things like

6:28

ultrasound. I think we're finally

6:30

looking at penetrating is a is a scary

6:32

term. I I don't mean that like

6:34

physically penetrating. But getting into

6:36

few inches into your brain versus just

6:39

scanning the current of the top of the

6:41

skull is is exciting.

6:43

>> Yeah, for sure. And looking at neurom

6:45

modulation things like focused

6:47

ultrasound could hit say the nucleus

6:50

encumbent for various substance abuse

6:53

issues or addiction issues and so on. So

6:55

it's pretty exciting. That's definitely

6:57

front and center of a lot of really cool

6:59

research and I feel like we're hanging

7:02

around in the same uh [laughter]

7:04

scientific also sounds wrong but

7:05

probably hanging out in the same

7:06

circles. [gasps]

7:08

>> Yeah. I want to actually stay on the

7:11

fiction thread for a minute because I

7:12

was actually hanging out with one of our

7:17

I suppose many overlapping connections,

7:20

Luke Nosk. And I asked him what I should

7:23

ask you since you guys go way way way

7:25

back. And here's the question. This is

7:28

from him. Well, first of all, I mean,

7:29

obviously he's a huge fan. Oh, man. So

7:31

many early memories. I miss working with

7:33

him. But here's the question. And

7:35

something I always wondered is that we

7:37

all had cryptonnomicon as required

7:39

reading. So I want you to explain that.

7:41

I love that book.

7:42

>> It's a great book.

7:43

>> Went to the financial crypto conference.

7:45

Max wanted to use cryptography to make a

7:47

new money system. How close were we to

7:50

really solving the same problem as the

7:52

Bitcoin algorithm or something along

7:54

those lines. So maybe you could just

7:56

provide some backdrop here for people

7:59

who have absolutely no context to put

8:02

this together. So that was one of the

8:04

weirder [sighs]

8:07

dual track like am I in a dream? Is this

8:09

a simulation

8:11

experiences of my life? So full context.

8:13

So Luke and I were classmates. I think

8:15

it was ahead of me by a year in school,

8:17

but we were at University of Illinois

8:18

together and worked on a couple of

8:20

failed startups together there. So we go

8:22

way back. We're really close friends.

8:24

And as

8:26

we all made our way to Silicon Valley, I

8:28

hung around campus because I actually

8:30

really wanted to do research in

8:31

cryptography, which back then crypto

8:33

meant cryptography, not cryptocurrency.

8:35

But I've always thought that there's

8:37

interesting opportunities in currency to

8:39

be invented powered by cryptography. So

8:42

I I was obsessed with things like, you

8:45

know, David Ch's work in anonymizing

8:47

signatures and digash was a thing back

8:50

in prehistoric times. in prehistoric

8:52

times like what year roughly?

8:54

>> I graduated UVI in 1997. Digicash, which

8:58

was the granddaddy of all

8:59

cryptocurrencies.

9:01

I hope I'm not offending anyone's prior

9:03

art, but I I believe it was probably the

9:05

very first attempt to make a real

9:07

digital currency using cryptography. I

9:09

believe it went bankrupt in the summer

9:11

of 98, the year I moved to Silicon

9:13

Valley. I actually went to the pouring

9:15

one out on the curb party held on kind

9:20

of somewhat less explicably on one of

9:22

the picnic grounds at Stanford

9:24

University. [laughter] I happened to be

9:25

there. There's actually a really good

9:27

lesson like sort of took that away with

9:29

me to PayPal where there was a bunch of

9:31

people some people you know today people

9:33

who are very active in crypto

9:34

cryptocurrencies now but also just like

9:36

interesting famous people famous today

9:38

they're all we're all very young back

9:39

then and the conversation was around

9:41

this idea like how the world is not

9:42

ready for this yet like the digash and

9:45

digital currency like people aren't

9:47

ready 98 you know this is like 10 years

9:50

before the bitcoin paper so I'm sure

9:52

they were right but the conversation was

9:54

really around how like it's obvious

9:56

obviously the right thing to do, the

9:57

right way to do it. People just don't

9:58

get it yet. And as a sort of naive,

10:01

freshly minted computer science

10:02

graduate, I was like, well, or the user

10:04

interface sucks. It's really hard to

10:06

have this very, very slow RSA

10:10

computation happening on your crappy

10:12

laptop if you're trying to pay for a cup

10:14

of coffee. Like, it's it's just too

10:15

slow. And people are like, "Ah, heathen,

10:18

you don't understand what are you

10:19

talking about." Like, no time too long

10:21

to wait for a digital signature.

10:23

Non-repudiation. I'm like, okay, great.

10:25

But seriously, you know, like I just

10:27

want a cup of coffee. So, it was kind of

10:29

an early moment like I feel like some of

10:31

these things are not like the others.

10:32

Anyway, I had literally within days of

10:35

that event, I met Peter Teal at

10:36

Stanford. He was giving a lecture. I was

10:38

hanging around Stanford campus. Luke was

10:40

actually the connective tissue. He knew

10:42

Peter and he went to school with me. And

10:43

so, he was kind of like, "Oh yeah, like

10:45

you guys should know each other. This

10:46

was great that you ran into one

10:47

another." So, the origins of PayPal

10:48

story were literally being made that

10:50

summer. And I was still like, sure, I'm

10:55

going to start a company in Silicon

10:56

Valley doing something. I don't know

10:57

what, but I was still going to like

10:59

scientific conferences. So, there was

11:00

this one conference called Financial

11:02

Cryptography. It was held in Anguila,

11:03

which is a kind of a small island off

11:06

the sort of a British West Indies. And

11:09

there's some all sorts of interesting

11:10

characters. It probably deserves its own

11:11

podcast one day. But it it's a

11:13

interesting, hilarious, bizarre story.

11:15

I'm 99% confident whoever Satoshi is was

11:18

probably going to the same conference at

11:19

the same time as we were. But Luke and I

11:21

were there and then another year Peter

11:22

and I went there and after PayPal was

11:24

already a thing I presented something

11:26

basically saying, "Hey, so we built this

11:28

thing. It has a great user interface and

11:30

I was there when DigiCash failed." And

11:32

like I think it failed because of user

11:33

interface and this was now a couple of

11:35

years into PayPal and we were definitely

11:36

doing well and Digicash was definitely

11:38

dead and people were in the audience

11:40

going like boo you totally don't get it.

11:42

They're like okay clearly maybe there's

11:44

a reason why we're we're succeeding.

11:46

Anyway, as we were doing all this stuff,

11:48

Neil Stevenson publishes the first 100

11:51

pages of Cryptonicon. And so as we're

11:53

coding and like I'm I'm literally coding

11:56

24/7. Steam comes out of my ears every

11:58

day cuz like I don't eat, I don't sleep.

12:00

Like the six of us are just in the

12:02

office together and three of us are

12:04

basically just typing code as quickly as

12:06

we can because we have this thing we

12:07

have to build.

12:08

And the only thing we take breaks for is

12:11

we're trying to read Cryptonomicon. And

12:14

like the more we read, the more like we

12:17

think he's writing about us. The first

12:20

few chapters of Cryptonomicon is like

12:22

this development of digital currency

12:23

using cryptography. And he's like

12:25

literally talking about I think he

12:27

actually references digicashe and he's

12:28

referencing like user interface sucks

12:30

and somebody needs to build something

12:31

more user friendly. So like am I in a

12:34

dream? Is this real? remember sitting

12:36

next to Russ, one of our earliest

12:38

engineers who I also went to college

12:39

with, al another Luke Nosk friend, and

12:41

I'm like, if we didn't know any better,

12:44

this is all just like a big brain and a

12:46

bad experiment. We just don't know whose

12:47

brain it is cuz it's being described to

12:49

us in a book that's being published

12:51

online. And then we ran out of pages and

12:54

like it wasn't published for a little

12:55

while longer. Like, how will we know

12:56

what to build next? We don't have the

12:57

book anymore. [laughter] That's like the

12:59

extended cryptonomican story.

13:01

>> Wow. And for folks who have never heard

13:04

of this book, it's an incredible read. I

13:06

mean, it's in retrospect, I suppose,

13:08

quaint in some ways, but so preient in

13:12

so many other ways. Just phenomenal. I

13:14

remember ripping through that book, and

13:16

like all Neil Stevenson books, it's like

13:18

5,000 pages long.

13:20

>> This one is only 4,000.

13:22

>> Yeah. [laughter]

13:23

>> It's a great book.

13:24

>> It's so good. It is so good. My college

13:27

life was shaped by Snow Crash with

13:29

probably the thing that really put him

13:30

on the map. And I read that there was

13:32

like a dog eared version of it hanging

13:34

around one of the offices that I sort of

13:36

hung out in college and I was like this

13:38

is the most amazing thing ever. Like it

13:40

this this this book sort of shaped my at

13:43

least software engineering life and then

13:45

cryogan was the next one. I was like all

13:48

I need to do is just read more Neil

13:49

Stevenson and he'll show me the way.

13:51

There were some really long books after

13:53

which sort of were quite challenging to

13:55

to get through. But then we got to red

13:58

which was another gift. So I'm reviewing

14:00

my preferences here.

14:02

>> He's like the digital Nostradamus in the

14:04

world of fiction.

14:05

>> Totally.

14:06

>> Snow Crash for people who also are not

14:09

familiar is where the term I believe

14:11

it's where the term metaverse was first

14:14

coined is in that book.

14:15

>> He coined Metverse.

14:16

>> Exactly.

14:17

>> And those Boston Dynamics, I think

14:19

that's the company with the

14:21

cute/terrifying possibly soon to be

14:23

weaponized robotic dogs. Go check out

14:26

Snow Crash. there in that book.

14:29

[laughter]

14:29

>> So, Thorrash basically presented an

14:31

alternative to Neurommancer.

14:33

Neurommancer was this dark dystopian.

14:35

Also, by the way, first book I read when

14:37

I came to the US.

14:38

>> Neurommancer.

14:39

>> Yeah. Yeah. So, two things. So, actually

14:41

this is extremely uh Are we in a

14:44

simulation? So, I come to the US July

14:46

16th, 1991,

14:49

6 weeks, I think before Soviet Union

14:50

fails. But I leave Soviet Union and I

14:52

come into this dreamland of America. And

14:56

then six weeks later, my passport is a

14:58

passport that doesn't have a country

14:59

anymore. And so I become a man with no

15:01

state and etc., etc. Makes for an

15:03

interesting travel existence. But I'm

15:05

like an ultra ultra nerd. And I was like

15:07

the king of the nerds, you know, in in

15:09

my prior life. And this is a summer, so

15:11

I don't know

15:13

how and where I'm going to find my

15:16

people. But like six weeks go by, so a

15:18

union fails, and I start a public high

15:20

school in Chicago. And I'm sort of like

15:22

casting around for my my people. And I

15:25

run into these guys who are still really

15:26

good friends of mine and they're like,

15:28

"Oh, you're one of us. We have a word

15:30

for this. You're a nerd. It's great. You

15:32

like science fiction? Great. You love

15:33

science fiction. Have you read

15:35

Neurommancer? Have you seen Akira?" And

15:37

they're like, "No. Hasn't made it to

15:38

Soviet Union yet." I was fluent in

15:40

English. This was not a difficult

15:41

conversation to have. But they're like,

15:43

"Okay, here's a VHS of Akira. Here's a

15:46

copy of Neurommancer. Go home. Don't

15:48

come back until you've been completely

15:50

filled with these things." And I was

15:51

like, okay, like one, these these are

15:54

going to be my friends for life, which

15:55

turned out to be true. It's been 30 odd

15:56

years, and we're still talking to each

15:58

other every day. And it's the beginnings

16:00

of a completely different science

16:02

fiction consumption experience. That

16:03

said, for the next few years, I was

16:05

like, Neurommancer and everything that

16:07

came after Neurommancer is the dark sort

16:10

of plug your brain in future. And Snow

16:13

Crash offers this hilarious, completely

16:16

different, consumerist,

16:19

satirical version of the plug your brain

16:22

in versus this dark corporations rule

16:25

the world ninjas in skin suits and kill

16:29

you while you're plugged into a deck

16:32

trying to hack your way into uh some

16:34

corporations. Anyway, you're like

16:36

walking me down the memory lane to some

16:37

fun recesses.

16:39

>> Well, there are a few pieces of

16:40

commentary on that. Number one, you got

16:42

to watch Paprika.

16:43

>> I'm on it.

16:43

>> It is very adorable in its attempt

16:47

manually to do what we could do with CGI

16:52

and so on now, but it is all the more

16:55

impressive for what it is able to

16:58

accomplish with handdrawn cells. I mean,

17:00

it's just bananas. So, it's worth

17:02

checking out. Neurommancer also, and I

17:05

believe, and I'm not alone in this, that

17:08

sci-fi is a great place to go looking

17:10

for the future, but it's also a great

17:12

place to go looking for philosophy. So,

17:15

I wanted to share a couple things from

17:16

William Gibson. And these are the kind

17:19

of quotes that will stick with you.

17:20

Like, when the past is always with you,

17:22

it may as well be the present. And if it

17:24

is the present, it will be the future as

17:26

well. Noodle on that for a minute.

17:28

Right? And I'm going to get this

17:29

paraphrase wrong, but I'm going to get

17:31

it, as they say in Silicon Valley, sort

17:33

of directionally correct, which is the

17:36

future is already here. It's just not

17:37

evenly distributed. Something along

17:39

those lines. In any case, I'm getting um

17:41

off track. I want to This is You

17:44

probably weren't anticipating, nor was

17:45

I, that this was going to be like a roll

17:48

call of fiction, but I want to come back

17:51

to something that you had in Tribe of

17:54

Mentors, and it's related to books. I

17:57

asked you

17:59

what book or books you've given the most

18:02

as a gift and why. And one that I've

18:06

still not read. I'm very embarrassed to

18:08

say this is the Master and Margarita.

18:11

>> You're missing out.

18:12

>> Yeah. So, this is now on my to-do list.

18:15

I'm going to let you fill in the blanks

18:16

here, but you you said I usually buy

18:19

M&M, which even though literally five

18:21

lines above it, it says the Master

18:23

Margarita. I was like M&M's in batches

18:25

of five or [laughter]

18:27

but you give it as gifts to new friends.

18:29

>> I do.

18:29

>> You always have copies on your desk at

18:31

work in case someone wants to borrow

18:33

one. Why? What makes this book so

18:36

special? Who's the author? What's the

18:37

background?

18:38

>> It's a great book. It is well translated

18:40

by a couple of different translator

18:42

teams. So, you can almost not go wrong

18:45

buying

18:46

one of the last three translations.

18:49

There's about six that I remember exist

18:52

in English anyway. I would not bother

18:54

with the first three.

18:56

>> I think one is Pevear. I'm not sure I'm

18:58

pronouncing

18:59

>> Pevear and Balonsky is the kind of

19:02

canonical like great translation into

19:04

English. Book is obviously in Russian

19:05

originally. There's now a new one, but

19:08

Peavar Balconsky is a fantastic very

19:11

very very good translation. Anyway, it

19:13

was written by Mikail Bulgakov, who is

19:18

arguably the greatest Russian language

19:22

writer of the 20th century in my biased

19:24

opinion. There there's a lot of great

19:26

Russian language writers, so it's a bit

19:27

of a fish in a barrel type situation,

19:28

like who's a great Russian author, and

19:30

there's some fantastic science fiction

19:31

Russian authors, incidentally. But he

19:35

wrote many books, and all of them are

19:37

really good. The premise for Astro

19:39

Margarita is so it's set in kind of an

19:42

early 1920s Moscow. Soviet

19:46

revolution happened. The Russian civil

19:48

war ended. The place is now socialist

19:51

and sort of slipping into the socialist

19:54

bureaucracy building mode. And yet it's

19:58

kind of, you know, it's a socialist

20:00

paradise in the making. And so the

20:02

devil, like the prince of darkness,

20:04

visits socialist Moscow in a summer of

20:06

1925. let's say with a codery of

20:11

characters from what you would expect

20:12

devil to hang out with and reres

20:14

absolute havoc on the socialist

20:16

paradise. So first of all it's a book

20:19

within a book which is like one of my

20:20

favorite sort of plot devices. The main

20:23

character the aonomous master is a

20:26

writer who wrote a book about the last

20:30

days of Christ. It's a book about

20:32

crucifixion and it's beautifully so it a

20:35

lot of it is in the book. it's

20:36

interspersed with the chapters of the

20:38

original and obviously Abulaka wrote

20:40

both but he is very much referencing

20:43

himself so he was in Moscow during the

20:45

revolution and and on and very quickly

20:48

ran a file of Stalin and

20:51

was also because he was such a prolific

20:53

and brilliant author was Stalin's

20:55

favorite author was persecuted

20:57

ultimately asked for permission to leave

21:00

the country because he couldn't be

21:01

himself he couldn't write Stalin

21:03

apparently permitted it but then he died

21:05

before this could So this it's a little

21:07

autobiographical story of the of the

21:09

writer himself. It's a story of devil

21:13

mocking making a mockery of socialist

21:15

paradise. It's a story of Christ and his

21:17

last days on earth told by an amazing

21:20

writer both the author of the book and

21:22

the book's author also known as the

21:24

master. It would take many hours and at

21:26

some point I'd be like you know well you

21:27

should go read the book cuz it's so

21:28

amazing. But it is one of the best

21:31

portrayals, certainly the earliest

21:33

brilliant portrayal of homosveticus.

21:36

What happens to the human mind when

21:40

exposed to rampant socialism? And it is

21:45

funny as hell because having grown up

21:47

there, spent 16 years there, the

21:49

insanity of living in a socialist

21:52

paradise and yet you don't have enough

21:55

to eat. It's tragic obviously, but it's

21:57

also very funny and the book does it

21:59

amazing justice and it's like super

22:01

duper supernatural sort of science

22:03

fictiony in what's possible. There's

22:05

there are flights, there are witches,

22:07

there is a devil Sabbath, there's all

22:10

kinds of crazy [ __ ] that happens there.

22:11

It's amazing.

22:12

>> Something for everyone.

22:13

>> It's an amazing book

22:14

>> and it's shorter than most Neil

22:16

Stevenson books is at least is my

22:18

impression.

22:19

>> It's like a quarter length or a fifth of

22:21

a a short Neil Stevenson product. So it

22:23

it's definitely very accessible.

22:26

It's also incidentally, not to

22:28

overshare, it is the reason I am married

22:31

to the woman I am.

22:32

>> Okay, I'm not going to let that pass.

22:34

Yeah, please expand.

22:36

>> We met completely randomly through like

22:38

a a sheer accident and both were

22:41

predisposed to be like, who is this

22:44

weird person? Why am I talking to

22:46

whoever this girl is, whoever this guy

22:47

is? And then I thought she was quoting

22:49

Master Margarita and I was like before I

22:51

like bow my way out of this

22:53

conversation. I'm going to find out.

22:55

It's my favorite book. It's always been

22:56

my favorite book. Like how can I, you

22:58

know, not so because I know the book as

23:00

well as I do because we're speaking in

23:01

Russian. I quoted the next line and

23:03

she's like, "Oh my god, Master

23:04

Margarita. That's my favorite book."

23:06

Like no, it's my favorite book. And it's

23:09

been 27

23:12

years and two children and and a very

23:14

happy life together.

23:15

>> Incredible. That's incredible. It's a

23:18

book that many things in my life are are

23:20

owed to. By the way, to give credit

23:22

where it's due, she's the one who said,

23:24

"Oh my god, that is the best line I've

23:27

ever heard in a movie referencing

23:29

whenever there's doubt, there's no

23:30

doubt." She So, she's actually She

23:32

deserves full credit. [laughter]

23:34

>> Full circle. Oh, boy. Yeah. The brain in

23:37

the vet question just keeps popping up.

23:39

So, let's actually talk about your wife

23:41

for a second and your relationship for a

23:43

second because I was texting with

23:45

another mutual connection, Sami Inken,

23:49

who is also a absolute monster endurance

23:52

fan and we were texting back and forth

23:55

about that. I was [clears throat]

23:57

chatting with him and I know you're an

23:59

investor in Verta Health, which people

24:01

should check out also. Let me just pull

24:03

it up here because he shot me over a

24:06

couple of different options. for

24:09

questions, right? I asked him the same

24:10

thing and

24:14

I'll give you a sampling and then I'll

24:16

indicate the one that most caught my

24:17

attention, right? So, I asked him I

24:19

asked him about what it's like to cycle

24:22

with you. And he said, "Well, if you're

24:23

going to be riding with him, just know

24:25

he pretty much goes all out always for

24:27

the first 90 minutes, no matter who he's

24:29

riding with, but enjoy." And I was like,

24:31

"I'm not going to be riding with him.

24:32

Don't worry." And then he had a number

24:34

of questions. grew up a super intense

24:35

engineer now a well-rounded CEO leader.

24:38

What did it take to go from one to the

24:39

other cycling? Why did he go like fully

24:42

allin a few years back? Which is

24:43

actually kind of an interesting question

24:44

cuz I know that's been part of your life

24:46

really really consistently on a near

24:48

daily basis for much longer than a few

24:50

years. But third, this is the one that

24:53

caught my attention. He's essentially

24:54

working with his wife on sci-fi, you

24:56

know, his family office VC. What's his

24:59

advice after years of experience

25:01

regarding working with a spouse? So

25:03

that's the one that I wanted to hit

25:05

next.

25:06

>> Totally. I mean, you got to marry right

25:08

otherwise it's hard to work. Uh, no.

25:10

Well, what are your thoughts on marrying

25:12

right? If you have thoughts on that. I

25:13

do actually. I've spent a lot of time

25:16

thinking I mean I spent a lot of time

25:17

thinking about all kinds of random stuff

25:19

and this is no exception. The line marry

25:21

up is a line people use all the time and

25:24

I think that's cute and I agree with it.

25:27

I just think it deserves unpacking. I

25:29

was asking myself actually a question

25:30

like why is our relationship as good as

25:33

it is and it it is really good and is

25:35

there anything to learn about co-ounding

25:38

companies like you're co-ounding a

25:39

family when you when you marry someone

25:41

you know you get together with someone

25:43

and are there parallels and I I was

25:45

contemplating all this stuff and I sort

25:47

of realized that I mean some I've had

25:49

many co-ounders over the years and some

25:50

were amazing relationships and others

25:52

didn't work out and it's always tempting

25:54

to sort of blame yourself but you

25:56

obviously have to examine it through the

25:57

lens of like what felt apart as a

25:59

relationship between two co-founders.

26:01

And I realized that the secret for my

26:05

marriage is I'm still trying to impress

26:07

this girl. Every day I roll out of bed

26:09

thinking like one like how did I get

26:11

this? This is definitely better than I

26:13

should have gotten. So clearly

26:15

I got to work on this. Like I can't let

26:17

her find out that she married down. And

26:19

so the attempt to impress your mate on a

26:23

daily basis is the secret to the best

26:25

marriage. And I think that the good ones

26:27

are where both sides are kind of

26:28

secretly thinking, I definitely lucked

26:31

out. Like this is so much better than I

26:33

should have gotten. And so if the

26:34

feeling is mutual, you're going to go

26:35

very far together just constantly trying

26:37

to grow. And

26:39

you know, come up with new tricks, new

26:41

ideas, new know, you don't let yourself

26:43

be stale because you're watching the

26:46

person next to you grow and become more

26:49

well-rounded, more intelligent, more

26:50

successful, etc., etc. So that's sort of

26:52

the secret to our relationship and

26:54

everything I've accomplished in my life

26:57

since we've met has been modulated

27:02

by her. We keep on referring to this

27:05

her. So her name is Nelly. I suspect

27:07

many of our mutual connections refer to

27:09

her as the sort of actual reason behind

27:11

anything that I've done successfully in

27:13

my life. So and it is true. She's been

27:15

extraordinary at both calling my BS out

27:19

when when there is some and pushing me

27:23

at the you know things like whenever

27:25

there's doubt there's no doubt situation

27:27

saying like hey trust your gut make a

27:29

decision this isn't the right person

27:31

this isn't the right idea etc and so on

27:35

one fun fact after PayPal so we met in

27:38

the very early days of PayPal so that

27:39

same summer of 98 is when the seeds were

27:42

in the ground for PayPal early 99 when

27:44

we started My wife and I met in mid99,

27:47

so exactly a year after I moved to

27:48

Silicon Valley. And so she she witnessed

27:50

the entire PayPal saga from the front

27:52

row seat. After PayPal, I was like,

27:54

"Anything but financial services. I'm

27:55

going to go build some other stuff." But

27:56

like I I can't allow my sophomore act to

27:58

be in the shadow of my freshman act. And

28:01

she sort of said, you know, do whatever

28:02

you want. Like obviously you have a

28:03

right to to come up with anything, but

28:05

I'm just gonna tell you, man. That's

28:06

clearly what you're meant for. And so

28:08

like, you know, you go on. So, [snorts]

28:09

seven years later, after I've been

28:10

meandering through the desert, she's

28:11

like, you know, I think I said it

28:12

before, but I just want you to know,

28:14

like, you were really good. Like, that's

28:15

still true. Maybe you should go do what

28:17

you're good at. And [laughter]

28:20

a firm is very much a product of that

28:21

conversation. I was like, I think she's

28:23

right. I think I'm a onetrick pony. And

28:25

the trick is a good trick, though. So,

28:27

here it is. Anyway, [laughter] but

28:29

trying to impress your mate, trying to

28:31

impress your co-founder every day

28:33

because you can see the other person

28:35

progressing in front of you and get

28:37

better and smarter and more intimidating

28:40

by way of just being so much sharper, so

28:43

much more interesting than they were

28:44

when you met is a great motivator. It

28:46

makes you keeps you on your toes, makes

28:48

you work harder at improving yourself.

28:50

>> And on top of that, what have you

28:52

learned about

28:54

working with a spouse or collaborating

28:56

with a spouse? Right? Because sometimes

28:57

spouses can be often very well matched

29:02

and have a dynamic like the dynamic you

29:05

describe, but they have some

29:07

compartmentalization or separation,

29:09

right? And I've I've met so many friends

29:11

and even have some neighbors where

29:14

they've just quote unquote attempted to

29:17

retire, sort of like a failed

29:18

retirement. And then [laughter]

29:20

one neighbor was describing this to me.

29:22

His wife is also very active, has got

29:25

tons of big projects she's working on,

29:27

and he doesn't know what to do with

29:29

himself. So he'd like sit down at the

29:30

breakfast table while she's on her

29:32

laptop and he'd be like, "So, what are

29:34

you up to today?" And she was like took

29:36

off her glasses very slowly and she

29:37

said, "You know, I'm not used to

29:40

checking in with you every morning.

29:41

[laughter]

29:42

You need to get a job." So working

29:44

together is a very different thing when

29:47

you have that. And maybe I'm overstating

29:50

the extent to which you guys sort of

29:54

collaborate in that way, but any

29:56

thoughts or recommendations for people

29:58

who are entering that type of

30:02

relationship with the spouse?

30:03

>> It can definitely get hairy, for lack of

30:06

a better term, if you're not thoughtful

30:07

about it. So, you have to think through,

30:09

you have to talk through kind of what

30:11

makes sense, what doesn't.

30:13

We're well matched because I am quite

30:16

technical, have lots of interest in sort

30:19

of the technical depths of matters that

30:22

we work on together where she works on I

30:24

work on etc. And she has an extremely

30:27

strong finance background and

30:29

understands business models and kind of

30:30

the sort of philosopher of business

30:32

these days. And so our conversations are

30:35

very rarely about checking each other's

30:37

work, which is where I think you could

30:38

really end up with like why are you

30:40

nitpicking

30:41

>> checking my homework,

30:42

>> right? micromanaging.

30:43

>> Yeah. Don't micromanage what I'm good

30:44

at. And also, she's an incredible empath

30:47

and and I'm not like I frequently be

30:48

like, "So, I'm I'm dealing with this in

30:50

my day job. Why is this person unhappy?

30:53

Can you explain to me?" And and she's

30:54

incredibly good at that. And so, I think

30:56

because we are such kind of non-over

30:58

overlapping areas of strength,

31:00

>> we end up being very complimentary. If

31:02

anything, we tend to be like, "Wait a

31:04

second, we're on your stuff. Why are you

31:06

bouncing to mine? We need to finish that

31:08

thread." And both of us have pretty

31:09

active professional lives. And so we're

31:11

we are constantly discussing them, but

31:14

we don't, I think, ever run into or step

31:17

on each other's toes. She doesn't ask me

31:20

what's your token budget anymore than I

31:23

ask her, how did you make this hiring

31:25

decision? Like I I I know she's very

31:27

good at or how did you decide that this

31:28

was a bad investment opportunity?

31:30

>> I'm pretty sure she looked at it with

31:32

all the tools she's gotten, they're

31:34

better than mine. But you have to be

31:35

prepared to communicate and it does not

31:38

hurt to be direct with one another. And

31:41

so you have to be able to take on

31:42

directness without sort of like

31:46

hurt feelings and a great oneliner.

31:48

Another gift from my wife. She didn't

31:50

come up with it, but she read it

31:51

somewhere and she's like, "Here's a new

31:53

motto for our family unit. Don't go to

31:56

bed angry. Stay up and fight."

31:58

[laughter]

31:59

You guys put that into practice. Uh we

32:02

don't fight too often. Yeah,

32:04

>> but we definitely don't let

32:07

>> things fester

32:08

>> curfew prevent us from hashing out

32:10

whatever it is our differences are at

32:11

the moment. But yeah, we try very hard

32:13

not to let things gesterure and that

32:15

works really well. If you're going to

32:16

overlap

32:18

familiarly and professionally,

32:22

you better

32:25

deal with conflict or disagreement

32:28

quickly because you're acrewing areas of

32:31

negative overlap faster than most people

32:33

do. Like, I hate the way you chew, but I

32:36

love you all the other ways. Like, we

32:37

got to talk about that, but it's not an

32:38

important thing. It's like, I hate the

32:39

way you chew, and by the way, you've

32:41

belittled me by checking on my token

32:43

budget. M is like you're compounding

32:45

frustrations and marriages fail because

32:47

people just don't say out loud like

32:49

here's a thing that I'm not sure I'm

32:51

prepared to put up with.

32:52

>> Yeah. And so just got to stay up and

32:54

fight. Got to get it out of there. Stay

32:56

up and fight. [laughter]

32:58

That'll be the title of this podcast.

32:59

Max left. Stay up and fight. You

33:01

mentioned earlier time is a zero sum

33:03

game. I want to talk about

33:06

obsession and [sighs and gasps]

33:09

analysis and tracking for a second

33:12

because ultimately the question is

33:16

how do you choose what to track right

33:18

and this is pulling from a men's journal

33:19

piece that came out in 2018 so some of

33:21

it is dated but today if you can name it

33:24

Levchin has quantified it food he spent

33:25

weeks using his iPhone to photograph all

33:27

of his meals and snacks and later gauge

33:28

their nutritional values sidebar from

33:31

Tim I did that too I still can't believe

33:34

that say the Dexcom app doesn't give you

33:37

some estimate of caloric content or

33:39

macros. It's just like you ate chicken.

33:41

I'm like that's not helpful. Anyway,

33:42

coming back to sleep. As an overtaxed

33:45

entrepreneur, as the parent of two

33:46

children under the age of five, Levchin

33:48

determined his minimum threshold by

33:49

shaving 5 minutes off of each successive

33:51

night. Interpersonal relations and time

33:54

management. Lean says he once spent a

33:55

month grading every business and

33:57

personal meeting on a scale of 1 to 10

33:59

on several metrics like usefulness,

34:00

intellectual stimulation, and social

34:02

stimulation. It goes on and on. [snorts]

34:04

>> It sounds like a weirdo, but I like

34:07

that's all true, man.

34:10

[laughter]

34:10

>> So, [snorts] I've done lots of this,

34:12

too. I'm just curious how you have over

34:16

time determined what to track and maybe

34:18

what you're paying attention to outside

34:20

of a firm business metrics and things

34:24

internal to a firm and we'll get to a

34:25

firm, but I'm just wondering for

34:28

yourself outside of that company

34:31

professional context, how do you choose

34:33

what to track? I've definitely gotten

34:37

much better at deciding what matters,

34:41

what doesn't. So, when this article was

34:43

written, I'm sure no metric looked dumb.

34:46

No metric looked superfluous. I was

34:47

like, great. Counting the number of

34:51

times I trimmed my fingernails. H, what

34:53

was there signal in that?

34:56

As you age, you replace raw compute and

35:00

willingness to engage in raw compute

35:02

with pattern recognition and olude paths

35:04

that you don't need to travel down

35:06

because you know there's nothing there.

35:08

Like there's absolutely no value in

35:09

tracking your fingernail clippings or

35:11

rate they're off. So I've become more

35:12

focused on tracking sleep because

35:15

there's more and more science showing

35:16

that that's really important. Even that

35:18

I sort of used to obsess over a bunch of

35:20

metrics and I think at this point the

35:22

sort of reasonably cutting edge and you

35:24

know better than I do I think these days

35:27

resting heart rate and heart rate

35:29

variability are the two really good

35:31

anchors like if if you you know keep one

35:33

down keep the other one up you know that

35:35

predicts your recovery predicts your

35:36

intellectual capacity for the day really

35:38

well. So I track that pretty

35:39

obsessively. I try to stay in bed some

35:44

number right number of hours. I'm a lot

35:46

less obsessed with I don't need to sleep

35:48

more than X, so better be, you know,

35:49

bulleted out of bed by X plus one

35:51

second. I'm also not precious about, you

35:54

know what, I don't feel like I need any

35:55

more sleep. I'm just going to go and

35:56

tool around and do more stuff. So, I

35:58

I've relaxed a bunch on things that I

36:00

found to be less impactful to my own

36:02

life, but I have the benefit of spending

36:04

a bunch of years obsessing over

36:05

everything and deciding, well, it

36:07

doesn't seem to matter that much what I

36:08

eat, for example. So, for a long time I

36:10

was obsessed with the exact sort of a

36:12

caloric and macro targets and then at

36:16

some point I had to travel a lot more

36:18

than I used to. And obviously that goes

36:19

out the window because you can't bring

36:21

your favorite yogurt with you everywhere

36:22

you go. And so you end up like, well,

36:24

I'm just going to have to have some

36:25

fermented foods from the assortment

36:27

available in this geography. And it

36:28

turns out that basically your gut health

36:30

is like going to be okay as long as

36:31

you're having enough fermented foods

36:33

versus like the fermented food you've

36:34

selected for yourself. So you relax some

36:36

constraints and move on to spending less

36:38

time thinking about them. I track sleep.

36:41

I'm still an obsessive cyclist, so I

36:43

track every metric you can imagine. I

36:45

become more obsessive as a cyclist.

36:46

Actually bothers me a little bit, but

36:48

because I'm unwilling to believe that

36:50

I'm ever going to age properly. I keep

36:53

on looking for marginal gains by

36:56

obsessing over the exact crank length

37:00

adjustments I can put myself through

37:02

just to return some of you marginal

37:05

watts that I lose with age. I've gone

37:08

from believing that all you need is

37:09

cycling to widening my physical training

37:13

routine too. So it's maybe slightly more

37:14

brain allocated to like what's the right

37:17

day to lift heavy weights versus the

37:19

right day to not do that. I'm curious

37:21

when, and I've never asked you this

37:23

before, but when cycling entered the

37:26

picture, because in doing some research

37:29

for this conversation,

37:31

well, I can I can just pull this is

37:32

from, I believe, an alumni magazine.

37:35

Behind the curtain of Soviet Ukraine,

37:38

and then your full name, which I cannot

37:39

pronounce, is born to a family of

37:41

physicists. As a child, you struggle to

37:43

overcome life-threatening respiratory

37:45

diseases. Your parents are told you

37:47

won't live past childhood and with your

37:49

mother's urging a determined young Max

37:50

begins playing the clarinet to build

37:52

lung capacity.

37:53

>> True story.

37:55

>> So I mean that doesn't seem to have the

37:57

makings of a serious cyclist. When did

38:00

you embrace cycling and what does it

38:03

give you? Also there's an interesting

38:05

origin story that actually kind of

38:07

intertwines Ukraine and clarinet and

38:09

everything. So, I did have some fairly

38:12

gnarly respiratory stuff when I was a

38:14

kid. And I think my parents probably

38:17

were scared into kind of throwing

38:20

everything in the kitchen sink into it.

38:22

As a nice Jewish boy from Ukraine, I

38:24

probably should have played the violin,

38:25

but clarinet is a a reasonable close

38:26

second. They're like, "Oh, but lung

38:28

capacity, great, let's do that." So,

38:29

that that's how I got into clarinet.

38:30

There was like literally something that

38:32

will, you know, feed the stereotype, but

38:35

also help with the lung capacity

38:36

development. And so I think my V2 max at

38:38

the time must have been like in the

38:39

single digits or something cuz like I

38:40

remember just like trying to open up my

38:42

lungs and breathe in like it didn't

38:43

work. And then like everything else I do

38:47

like ah well I can just obsess in

38:50

measuring it. And V2 max is definitely

38:52

not a thing Soviet Ukraine really

38:54

measured but they did have like the

38:55

basic approximation like blow into a

38:57

tube and see how far you can push the

38:58

ball like all the the physics

39:00

equivalents of lung capacity

39:01

measurements. So, I was already looking

39:03

at quantify itself from a lung capacity

39:05

perspective as like a 7-year-old kid or

39:07

something. All of this while living in

39:09

an apartment complex that overlooked

39:12

Kiev's only outdoor velroone. And so my

39:16

platonic ideal of a manly man who was

39:20

like a sports phenom was these dudes who

39:23

would like get on the boards of a Kia

39:26

Veladrome and just put their watts down

39:29

and like go for an hour or two just

39:32

ripping through this thing. I was like

39:34

wow like one day when I'm not a scrawny

39:37

little kid with a clarinet I'm going to

39:38

be like that. So, I used to sneak into

39:40

the Velro when it was closed and try to

39:43

ride my bike there except the cranks

39:45

would touch the woodboards. Like, one of

39:46

the worst accidents of my childhood was

39:48

I had a giant wood splinter in my butt

39:51

>> because I crashed at the top of the drum

39:54

trying to do loops and slid down. Of

39:56

course, these weren't well finished

39:57

boards and it was outdoors, so like the

39:59

elements rained on it.

40:01

>> Probably not PG-13 after this, but

40:03

anyway. So, I didn't grow up with a ton

40:04

of sport in my life. Like soccer is

40:07

prevalent obviously in Eastern Europe,

40:09

but like cycling was right there on

40:11

display every day. Like if I wanted to

40:13

see someone who looked like just full of

40:16

life and vigor, I would see a cyclist.

40:19

And so I always rode a bike as a kid and

40:21

always rode a bike in college. And sort

40:24

of as I got into sort of busier and

40:27

busier where this notion of

40:31

how do you stay sane became an important

40:34

part of the consideration like if you're

40:36

going to work for 3 days straight

40:37

without sleep like you got to offset it

40:39

somehow.

40:40

>> You can't just like sleep it off. You

40:41

also need to do something to your body

40:42

versus your brain.

40:44

>> I sort of naturally looked at cycling

40:45

and then back to Nelly. When I left

40:48

PayPal after we got acquired by eBay, I

40:52

definitely hit like a very low point

40:54

where there's nothing going on in my

40:55

life that was worth staying up all night

40:57

for. And she pointed out that I have

40:59

this amazing bike hanging on the wall

41:01

that you know had since I think before

41:03

we met. She like why don't you like get

41:04

back on the bike and you know see see if

41:07

you can find joy in that. And and she

41:08

knew my stories of rolling around the

41:10

Kia drum. And so I got on a bike and it

41:12

was one of these moments where I went

41:14

for a ride with a bunch of people and

41:15

they were all like very serious road

41:18

jewelry as the saying goes.

41:20

>> I've never heard that.

41:21

>> If you're buying more of a bike than you

41:22

deserve to ride like that's nice road

41:24

jewelry got there. I was on like a

41:26

beater bike relatively speaking. I think

41:28

I brought it from college actually and I

41:30

sort of dropped everybody on the first

41:32

hill and I was like oh yeah I know how

41:34

to ride a bike. I have big lungs. That

41:36

clarinet really paid off. And so that

41:38

was like I should really invest some

41:41

time into getting better at this thing.

41:43

And that that's how I got into. But I am

41:45

exactly what they call a hammerhead. So

41:48

you know it's a if I see a road I can't

41:50

not go fast or at least try to go fast.

41:52

>> What does it give you in terms of

41:54

dividends just on an ongoing basis?

41:57

>> A couple of different things. So it's

41:59

definitely it's great. I mean it's a low

42:01

impact sport. So you can ride a bike for

42:04

6 hours and no part of your body will be

42:06

like in absolute shreds the day after if

42:09

you at least somewhat know what you're

42:11

doing. So it's it's a great just if you

42:13

have the time and if you're interested

42:14

in it, it's not a self-destructive

42:16

sport, which some sports are more and

42:17

others are less. This is definitely one

42:19

of the less self-destructive on the

42:20

actual physiology. Part of why I like to

42:23

go hard on the bike is it actually

42:25

allows me at least to clear my head.

42:27

It's one of the few times when I'm not

42:29

constantly playing with some work

42:31

rellated concept or thinking through

42:33

some future ideas or you know whatever.

42:35

It's a moment where sort of everything

42:37

gets tuned out and it's just me and the

42:41

the pain in the legs and sort of you're

42:43

telling yourself go harder or manage

42:44

your energy but there's not enough time

42:46

or cycles left and like also what if we

42:49

went to this market now? So it's a

42:51

moment to go into like a pure moment of

42:53

flushing whatever's left in the head and

42:55

it does keep you healthy. It's a great

42:57

way to stay alive and be healthy. And

43:00

then more and more these days, it is a

43:04

lifestyle sport of CEOs and leaders and

43:07

people who sort of read the same kind of

43:09

books. And it's another form of a

43:11

community where if you don't have the

43:13

time to sort of join

43:16

clubs and you know, I'm I'm not a big

43:20

foodie. I'm not a big drinker. I don't

43:22

have a ton of time for these protracted

43:24

social events. I try not to go to too

43:26

many conferences and yet I'm human. I

43:28

enjoy human company and so riding with a

43:31

pack of people who are interested in

43:34

similar things is is is another form of

43:36

staying engaged.

43:38

>> So I imagine that's a being near people,

43:40

not a talking with people type of

43:42

situation. If you're going as hard as I

43:44

suspect you're going, right?

43:45

>> Part of the trick is if you can still

43:47

talk and they cannot.

43:49

>> Yeah. Then you're doing it right.

43:51

[laughter]

43:52

Cycling takes place over hours and so

43:55

you can spend some time talking. There's

43:57

also the

44:00

an important part of cycling culture is

44:02

the coffee shop ride where you sort of

44:04

you hammer to a coffee shop and then you

44:06

sort of

44:07

>> everybody falls in their chairs and gets

44:09

their pinkies out and sips espresso. And

44:11

[laughter] I love coffee almost as much

44:13

as I like bikes.

44:13

>> If we have time, we'll come back to

44:15

coffee because I did dig into that in

44:18

the research. two other quotes that you

44:20

gave for your potential billboard

44:22

answers. You said if it were in Marin

44:24

County and I'm going to mispronounce

44:25

this so you can correct me, but it's

44:27

either Jen's or Yen's Voit. I'm not

44:29

sure, but you can correct me in a

44:31

second. So, two quotes from him.

44:33

Legendary cyclist. Number one, when my

44:35

legs hurt, I say, "Shut up legs. Do what

44:37

I tell you to do." That's the first. The

44:38

second is, "If it hurts me, it must hurt

44:40

the other ones twice as much." That's a

44:43

great one. It's so applicable in so many

44:45

contexts. Am I saying that name

44:47

correctly? Probably not.

44:48

>> I think it's Yens.

44:50

>> Yens.

44:51

>> I think he's East German. Yens V.

44:54

>> V. Okay, got it.

44:55

>> One of my proudest moments in the bike

44:57

is I rode with Yens a few years ago now,

44:59

but he used to hang around Northern

45:01

California. He has a lot of fans

45:02

actually. He's transcended the very

45:04

specific sport of cycling into having a

45:06

fan base worldwide because he's such a

45:09

character. The shutup legs. I own

45:10

multiple pieces of clothing with shutup

45:13

legs that I did not personally produce.

45:16

He's famous for that line, but he's very

45:17

funny. He's now primarily a commentator

45:20

in bike races, but he is a great

45:24

embodiment of this culture of endurance.

45:28

You know, pain with a smile, if you

45:29

will. Pain with a smile. Before we move

45:33

on, cuz I do want to ask you a bunch

45:35

about a firm and broadly how things

45:39

you've done with a firm reflect

45:42

your thinking and ability to see things

45:45

that others don't see perhaps. But

45:47

before we get there, I wanted to come

45:50

back to Eminem, [laughter]

45:53

this book we were discussing earlier.

45:55

And for people who are like, "Oh yeah, I

45:57

meant to write that down." on the master

45:58

in Margarita. You wrote here, just to

46:01

reiterate some of what you already said,

46:03

it's a fairly short novel, remarkable in

46:05

its exceptional depth, exploring

46:07

everything from fundamentals of

46:08

Christian philosophy to the fantastical

46:10

and hilarious satire, soulcrupting 20th

46:13

century Soviet socialism. My god, that's

46:15

really well put. Yeah, I don't think I'm

46:17

alone in watching some of what's

46:19

happening in the US and elsewhere with

46:23

respect to

46:25

this embrace of socialism or things that

46:30

go by the name socialism perhaps you

46:32

could apply other labels and while you

46:36

know capitalism is not a panacea for all

46:38

things and there are plenty of warts and

46:39

risks and you know humans respond to

46:42

incentives

46:43

as someone who came here at 15 or 16. I

46:47

would just love to hear you speak for a

46:50

few minutes on what you're seeing and

46:53

what your thoughts are.

46:54

>> I'm pretty worried, not to sort of go

46:57

all dark immediately, but I think

47:01

people without the benefit of [snorts]

47:04

questionable benefit of growing up in

47:07

the collectivist paradise to quote the

47:10

current mayor of New York don't

47:12

understand just how corrupting it is.

47:14

The ideas of socialism are amazing.

47:17

This, you know, me for my fellow men and

47:20

share and share a like and do the right

47:23

thing because it's the right thing to

47:24

do, not because there's a financial

47:25

incentive to do it. All those things

47:27

sound amazing and so it's seductive. the

47:29

idea of a workers's paradise without the

47:34

greedy

47:36

lenders, capitalists, bankers, sort of

47:39

all all the sort of things we were fed

47:40

as children or I was fed as a child in

47:42

Soviet Union. At first, Blushers are

47:45

like, "Yeah, it makes sense. People who

47:47

work hard should have more. And people

47:49

who are

47:52

great even though they're not capable of

47:54

producing valuable things in society

47:56

because they're good people should have

47:58

everything that I have even though I am

48:00

working harder than they are. And those

48:02

things sound pretty good, but they

48:04

inevitably require a bunch of structural

48:07

change that just does not work thanks to

48:10

human nature and many obvious things

48:12

>> like it. For example,

48:15

one of the greatest kind of mental

48:18

images I have from growing up in Soviet

48:20

Union is if you went to a

48:22

governmentowned store, and every store

48:24

was of course owned by the government,

48:25

you would very quickly notice that the

48:27

people whose job was to sell you things

48:29

from behind the counter were always very

48:31

fat

48:32

>> while everyone you knew in your life was

48:35

always very skinny. And you sort of like

48:37

this doesn't make any sense. like how do

48:39

these people who happen to work in food

48:40

stores are always really well fed? It's

48:44

like, well, because they're stealing,

48:45

like they get access to the food. And it

48:48

expands that notion of this idea of

48:51

everybody just pulls all their work

48:53

product into one big pool and then

48:55

someone's in charge of fairly

48:56

distributing it to everyone who needs it

48:58

from each according to their abilities

49:00

to each according to their need. Is kind

49:02

of the socialist/communist

49:04

motto. And it works great except people

49:07

who are doing the actual redistribution

49:09

get to keep a lot for themselves. And

49:11

doesn't matter how

49:13

honest they begin that journey, by the

49:16

time they get to real power, they become

49:18

profoundly corrupt and steal and keep

49:21

and redistribute primarily to

49:22

themselves. And so that alone is just

49:26

like an indictment of socialism that you

49:28

cannot get around. It gets worse because

49:33

markets, free markets, capitalism, you

49:35

know, whatever term you want to use,

49:37

inherently forces competition. If you

49:41

believe you have a thing that you can

49:43

sell at a lower price than the other

49:44

guy, you get the market's attention. You

49:47

get the business that's available to you

49:48

because the thing is worth buying by

49:51

someone, you're going to work on

49:53

creating a margin for yourself by

49:54

lowering the cost of production, finding

49:56

efficiencies, finding some ways of

49:58

making the thing cheaper for you. and

49:59

we're cheaper for the buyer so you can

50:00

compete on price. So all those things

50:02

cannot do not exist in socialism because

50:04

it's all centrally planned. We're going

50:05

to make this many widgets and then we're

50:07

going to have someone redistribute them

50:09

to all the right people full stop. And

50:11

what happens is there's never the

50:13

pressure to improve and so you always

50:16

make stuff at the governmentmandated

50:18

cost and the governmentmandated

50:21

price is what's being used to sell it.

50:23

And so you have natural stagnation. You

50:25

have a system that rewards graft, gives

50:29

power to people who are most likely to

50:30

become

50:32

graphdriven

50:34

and prevents anyone else talented or

50:36

otherwise from trying to innovate and

50:38

improve the efficiency of the system

50:40

itself. And so it just does not work.

50:41

And the thing that worries me in the US

50:44

and and the world, this isn't unique to

50:47

to us here is the headlines that make

50:51

you feel like, "Wow, I would love to

50:52

live in the world where everyone is fed

50:55

and everyone gets more or less the

50:57

baseline of good living is very

50:59

compelling. Like it it is easy to agree

51:01

with. The recipe for that remains to be

51:04

the best recipe we know, the best recipe

51:06

we've discovered as humanity is

51:08

capitalism. a force for constant

51:11

creative destruction where you build the

51:13

next thing better than the other guy.

51:14

And yes, his business may be put out

51:16

because your business thrives, but the

51:19

beneficiary is the buyer or the the user

51:21

of the product, the buyer of the widget

51:24

because you're constantly working to

51:25

make the whole thing more efficient. If

51:27

you eliminate the ability to or you

51:29

eliminate the need to create

51:31

efficiencies, you just stagnate. And the

51:33

death of Soviet Union was surprised to

51:34

exactly zero people because we knew as

51:38

people who lived there more than most

51:41

that nothing ever changed for the

51:42

better. Like everything was the same

51:44

price, the government mandated price.

51:46

And we were still using phones that

51:47

looked like they were made in 1950s. And

51:50

when I moved to the US like, "Oh, you

51:52

have buttons on your phones? That's

51:53

amazing. Like we still do

51:55

>> like the the clickity rotary."

51:57

>> Yeah. Yeah. And so anyway, I can't say

51:59

enough how the seductive story of

52:01

socialism really appeals and I can

52:03

understand why. But [snorts] you know,

52:06

if I if I didn't have my day job, I'd

52:08

spend a lot of time screaming from every

52:09

street corner, don't don't fall for the

52:11

trap. It's a surefire way of getting to

52:15

no progress at all.

52:16

>> So a firm, you know, on some levels is a

52:20

I don't want to say response, but it's a

52:22

solution that replaces potentially

52:24

predatory practices, right? And we're

52:27

going to get to that, but I'm wondering

52:29

how you would explain the current

52:31

apparent wellspring of attraction to,

52:36

just to keep it simple, socialism in the

52:38

United States. Are there structural or

52:40

systemic problems that have contributed

52:43

to that? Is it mostly sort of demagogues

52:46

using whatever talking points they think

52:49

they can

52:51

leverage to attract votes, etc. What is

52:56

the cocktail? What are the ingredients

52:58

in the cocktail that are contributing to

53:00

this current phenomenon in your mind?

53:02

>> I'm confident I don't know every strand

53:06

of this particular disease. So, it's

53:08

it's a complicated issue and

53:10

>> I think

53:13

it is definitely the case that the

53:15

brochure looks great. The storytelling

53:17

of do you know someone who's poor? Do

53:20

you know someone who deserves better? Do

53:22

you know someone who's been laid off? It

53:24

is the case that capitalism can be

53:28

profoundly unfair at least to an

53:30

individual as as a system as a way of

53:33

improving the world. It works amazingly

53:35

well. We've not created anything better.

53:38

But you get thrown off the bus if you

53:42

build something that gets outmoded or

53:45

out competed by another participant in

53:48

the market. And it can hurt. And for the

53:51

entrepreneur, it looks like failure,

53:52

which we are wired to accept, but it

53:55

still hurts. For a worker, it looks like

53:57

being laid off, which could be

53:58

catastrophic for a family. For people

54:02

who sort of misinvested their money or

54:04

something went wrong, it doesn't taste

54:06

good. And so, the idea of social safety

54:09

net, the idea of welfare

54:12

is natural. You hate seeing your fellow

54:14

men and women starve or not do well. So

54:17

I understand the definition of the

54:19

problem and the notion of income

54:21

inequality and like all the things that

54:22

we see in Silicon Valley firsthand is

54:25

very real. It's not like how are you

54:27

complaining like there's a lot of people

54:28

who are not doing well and every new

54:31

disruptive change from you know the

54:34

industrial revolution to the AI

54:35

revolution there's always someone on the

54:37

receiving end of the efficiency because

54:39

their work is no longer required or

54:41

their products or their ideas are no

54:43

longer relevant. And so socialism is a

54:45

compelling story of how to protect those

54:47

people, how to prevent them from

54:49

suddenly being on the outskirts of

54:52

society or worse. And I think there's

54:55

definitely a lot to do to make sure we

54:59

don't leave our fellow humans behind.

55:01

Certainly my love of capitalism and free

55:03

market does not obiate my humanity. I

55:07

care about people just because I am also

55:09

human and I see it as part of my job to

55:12

ask the question, can we do better? Can

55:14

we provide the social safety net for

55:17

everyone? We're creating more and more

55:18

efficiency. We're creating more and more

55:19

value. There have to be ways of helping

55:22

people who are thrown off the bus to

55:24

make sure they don't end up in the

55:26

gutter. And I happen to have some

55:28

solutions that I love and solutions that

55:30

I think are worthless. the idea of let's

55:33

concentrate it all in the hands of the

55:34

government, give a handful of people the

55:35

right to redistribute it all and we'll

55:38

do better doesn't work and I lived to

55:40

tell the tale. I do think that things

55:43

like philanthropy is profoundly

55:44

important and the more disruption we're

55:46

going through as a society the more

55:47

important philanthropy becomes. And I

55:49

think in that sense I'm not an

55:51

especially outwardly religious but I

55:53

think religion provides a really useful

55:57

set of frameworks around how to think

55:59

about it. You know, every organized

56:01

religion has a version of what it means

56:03

to be philanthropic. And the older I

56:04

get, the more I start respecting the

56:08

thought that's sort of been aggregated

56:09

over the thousands of years that most

56:10

religions existed. So I think that's

56:13

super interesting and important and

56:15

worth

56:17

participating in.

56:19

On the other side of it is I think

56:21

capitalism in particular has created

56:25

opportunities for products to develop

56:27

where the sort of optimizing up and to

56:31

the right of the most profit at lowest

56:35

cost ends up building products that are

56:37

actually devolved from not just their

56:40

economic platonic ideal but societal

56:42

ideal. And I don't think we have to

56:45

compromise by saying well you know what

56:47

I don't actually care that this is

56:49

harmful to someone because it's the most

56:51

profitable marketefficient thing to do.

56:53

And my theory for a long time has been

56:55

that you can build products certainly

56:56

financial services that are optimal not

57:00

just from the lens of most profitable

57:02

but also most society beneficial. And

57:05

that's not uniformly distributed. There

57:07

are plenty of products that look like

57:09

predatory lending and all the other

57:11

things or just sort of make money

57:13

because the underlying customer doesn't

57:15

understand what's going on, doesn't have

57:17

the capacity to deal with the complexity

57:20

of the math involved. A firm is the

57:21

answer to that question. In some ways,

57:23

what we are doing here is trying to say,

57:25

look, you can optimize a product beyond

57:29

just making the most money. In fact,

57:30

it's okay to make a little bit less

57:32

money if you are able to create

57:35

something that's society more

57:37

successful, more important because in

57:40

the long term, I think you will have

57:41

retention that consumers will not go to

57:44

a cheaper product or a product that is

57:47

less profitable for shareholders because

57:48

it's so societally important. And so I

57:51

am to at least some extent embracing the

57:53

ideas of pro-social product and

57:57

engineering. I just choose to do it

57:58

strictly through the lens of capitalism.

58:00

>> I'd like to go to early early a firm,

58:04

right? And this parallel to PayPal,

58:07

which looking back at say early

58:09

fundraising for PayPal, it wasn't

58:11

exactly like every door at Sand Hill

58:14

Road was opened with a red carpet,

58:16

right? I mean, it was incredibly

58:18

hardgoing. People were like, "Yeah, no

58:19

way. Possible. Not going to work." I

58:21

mean, I've had conversations with Reed

58:22

Hoffen about this just like, "Yeah, no."

58:24

Like, at every turn, no chance. And I

58:27

mean ultimately with Peter pulling from

58:29

the hedge fund and stuff I mean it's a

58:31

hilarious story and incredible a firm

58:35

you know early on similarly right it was

58:38

not obvious a lot of folks would be of

58:40

the opinion credit cards just fine late

58:43

fees too profitable

58:45

and what I'm wondering and this applies

58:48

elsewhere to your investing and

58:50

involvement with companies like Yelp and

58:52

so on.

58:54

What are you seeing? What gives you

58:57

confidence? And of course there's a bit

58:58

of a in the phrasing of the question a

59:01

survivorship bias. I mean but so putting

59:03

that aside for the moment what gave you

59:06

we could use affirm as the example but

59:08

like what gave you the confidence to

59:10

keep driving? What did you see that

59:12

other people didn't see or what were

59:14

assumptions you made or hypotheses

59:17

that you felt just were worth continuing

59:21

to hammer on and test because you had

59:23

some degree of confidence. And what does

59:25

a firm do? Maybe you should explain that

59:26

first.

59:27

>> Well, there's this acronym BNPL, buy

59:29

now, pay later. We kind of invented the

59:31

genre about 15 years ago. And the basic

59:34

idea is at the point of sale, most of us

59:37

pull out a debit card or a credit card.

59:40

And those who pull out a debit card

59:42

generally speaking, think themselves

59:43

financially responsible and think, you

59:45

know, if I can't afford it, I'm not

59:46

going to pay for it. So, like, if I

59:48

don't don't have enough in my bank

59:49

account, just not going to buy right

59:50

now. I'm going to save up or maybe buy

59:52

something cheaper. If you pull out a

59:54

credit card, the minority of us who are

59:58

for all intents and purposes are

59:59

independently wealthy just say, "I'm

60:00

going to put on my credit card because

60:01

it's easy, and then I'm going to pay it

60:02

off at the end of the month, and I'll

60:04

get some free float on that, but I'm not

60:05

going to pay interest or pay over time."

60:08

There's a huge percentage of Americans

60:09

that pull out their credit card and say,

60:10

"I'm going to add it to the giant pile

60:12

of revolving debt that I already have

60:14

and I'm going to pay interest on it and

60:15

I don't really know how to calculate

60:16

that and I have no idea when I'm going

60:18

to be out of debt, but I need that thing

60:19

and I'm not disciplined enough to pay

60:22

for it with cash or I just don't have

60:23

the cash and I need the thing." And

60:24

sometimes the thing is like, well, I

60:25

need a new bicycle which we can maybe

60:27

postpone and sometimes the thing is

60:28

like, I need a pram to put my baby in

60:30

and yeah, I'm just going to have to buy

60:32

it one way or the other. And so the idea

60:35

of a firm was and is and it's worked

60:37

amazingly 15 years in that what if there

60:40

was a third way where what if you could

60:42

have the same transparency and

60:45

responsibility and clarity of a

60:47

financial decision-m with a payover time

60:51

product that you do have with a debit

60:53

card. So, what if we added the idea of I

60:58

want to pay for this over time, but I

61:00

don't want it to revolve and I

61:01

definitely don't want to be confused as

61:03

to when I'll be out of debt. Essentially

61:05

turning every transaction into a simple

61:06

plan. Pay for my baby pram for over 6

61:10

months, let's say $100 at a time,

61:13

doesn't break the bank. Exactly 6 months

61:15

out, I'm done. That's it. Not revolving.

61:18

I'm not paying any more interest or or

61:20

anything like that. And as we sort of

61:22

described the idea sort of thought well

61:24

like if you're going to do it right how

61:26

about you build it so that there is no

61:27

revolving possible there's a fixed

61:30

schedule only that every transaction is

61:32

pre-ric if you're paying interest you

61:34

know exactly how many dollars of

61:35

interest you'll ever pay even if you

61:37

take longer that can't change that

61:38

shouldn't change and if you're late

61:41

there shouldn't be any late fees like

61:42

you should just be motivated to pay us

61:44

on time because if you take too long and

61:47

you're late all the time maybe we won't

61:49

lend you next time. So, it's sort of

61:51

almost like a too simple, too black and

61:53

white a way to lend money as an

61:55

alternative to credit cards, which is a

61:56

very long way of answering the question.

61:58

But, it is available now at threearters

62:00

of all e-commerce checkouts in the US

62:02

and Canada and rapidly expanding into

62:05

UK. And we've announced we're going to

62:07

go to a bunch of European countries and

62:08

there plenty of competitors to do the

62:10

same thing. But we were kind of the

62:11

first I think there were some obviously

62:13

logical and spiritual predecessors to

62:16

this idea but we sort of purified it

62:17

down to no fees, no revolving, simple

62:19

schedule, everything is pre-priced,

62:22

everything is super transparent and this

62:24

year we'll do almost $50 billion of

62:26

these transactions. So it's worked out.

62:29

>> Seems to be working.

62:30

>> It seems to be working. We publicly

62:32

trade it quite profitable. Have been

62:34

profitable for a bunch of time. the last

62:36

10 quarters we grew 30 plus percent

62:38

year-over-year or faster. So, it's like

62:41

it's both very big, growing really well,

62:44

still never charged a penny of late

62:46

fees, never charged a penny of revolving

62:48

interest. So, we're still true to the

62:50

idealistic nature. And so, as I was

62:52

describing the product, I talked to a

62:53

bunch of banking people who laughed at

62:55

me like, "Okay, dude, you don't get it.

62:57

You've obviously never lent money for a

62:58

living. Late fees is where the profit

63:00

is. Half the profit comes from late

63:01

fees. So, like, you just cut your

63:03

profitability opportunity by half. So,

63:05

putting that aside, if you don't

63:06

revolve, what are you going to do when

63:08

people take too long to pay you back?

63:10

You want them to take forever. You want

63:11

them to make minimum payments so the the

63:13

principle keeps being big and then

63:15

interest will compound into principle.

63:17

Be amazing. Like, yeah, all that sounds

63:19

crappy. Like, none of this sounds

63:20

transparent. None of this makes any

63:22

sense. And people in Silicon Valley who

63:23

generally speaking don't need these

63:25

services don't get just how significant

63:27

the cost is on your regular consumer

63:30

that doesn't have a seven-digit salary,

63:32

etc. And so as I was sort of trying to

63:35

raise money for this idea, I would talk

63:37

to people, you know, VCs and they're

63:39

like, I don't know why why can't they

63:40

use their platinum AMX? Like they don't

63:42

have a platinum. [laughter] Like, yeah,

63:44

like why why don't they get a platinum?

63:46

I'm exaggerating, but only slightly.

63:47

>> Right. Right. Right.

63:48

>> Anyways, the difficulty raising money

63:50

for this financial services company,

63:52

unlike PayPal, was I couldn't convince

63:55

anybody this was a a thing that normal

63:57

people would use.

63:58

>> How did you land on this, though? Like

64:00

how did you even get to the point where

64:02

you're like this is the deck I'm going

64:04

out with and this is what I'm pitching.

64:05

>> So the earliest version of this thing is

64:07

actually pretty funny. So

64:10

a bridged since it it's a colorful story

64:12

and actually Luke is involved like weird

64:14

Luke makes another appearance.

64:15

>> I like colorful stories. We're not in a

64:17

too much of a rush. Yeah. Take your

64:19

time.

64:20

>> Right after PayPal went public. Luke and

64:22

I I mean Luke is a wild and crazy guy.

64:24

You've met him. [laughter]

64:25

I think he talked me I'm going to go

64:27

with he talks me into. But I'm I'm

64:30

already dating Nelly and I'm trying to

64:31

impress Nelly. I'm still trying to

64:32

impress Nelly every day. But I decide

64:34

the way to impress Nelly is to get a

64:36

really cool convertible. And so Luke and

64:38

I like find this convertible. Mercedes

64:42

just launched their first hardtop

64:43

convertible. So we fly to LA with a view

64:46

to buy these two identical like we

64:49

debate. We can't have the same color. So

64:51

we need to like take we anyway.

64:52

[laughter]

64:53

So we get to LA. We go to the dealership

64:54

like, "We want to buy this cool hardtop

64:56

convertible from Mercedes, like a

64:58

mechanical folding hardtop." So, it's

65:00

like a thing that like transformer car

65:03

becomes a hardtop or a convertible if

65:05

you press a button. Super cool. So, we

65:06

got there and they're like, "Okay, Mr.

65:08

Nosk, here's your auto loan. Go. Here

65:11

are your keys. Mr. Leftchin, your credit

65:13

sucks, so we're not going to sell you a

65:15

car today." I'm like, "Wait a second. We

65:17

like literally just took PayPal public.

65:18

I'm okay. I've paid off my student loans

65:20

and everything." like, "Yeah, but your

65:22

credit score is really, really bad. So,

65:25

if you want to buy a car today, you're

65:26

going to have to wire the total amount

65:27

of money." Which is like, "My future

65:30

wife then girlfriend is like," oo, okay,

65:33

you didn't tell me that part of your

65:34

biography when [snorts] we just started

65:36

getting to know each other. What did you

65:37

do? Like, I started this company in

65:39

college with Luke Nos like who's in the

65:40

room. He knew and I didn't that if you

65:42

don't pay your credit card bills on

65:44

time, it'll come and, you know, show up

65:45

on your permanent record. So anyway, I

65:48

had to in fact wire money and like as

65:51

this dealership was wrapping up of the

65:52

day, they're like, "Oh yeah, we finally

65:53

got your wire. Okay, here are your

65:54

keys." So we drove from LA to San

65:58

Francisco on five, basically racing each

66:00

other at some ungodly speeds and got

66:03

pulled over at 2:00 in the morning and

66:05

the cops were very upset with us, but

66:06

actually let us go because they were

66:08

similarly fascinated by these cool new

66:09

cars. So, we didn't actually get into a

66:11

lot of trouble, but got into a minor

66:13

amount of trouble for going way too fast

66:14

at 2 o'clock in the morning. But the

66:17

story stuck with me because I mean, I

66:20

was fine financially, better than fine.

66:23

I was my mid20s and just, you know,

66:27

never needed to work a day in my life.

66:29

And yet, the car dealer who knew who I

66:31

was, he looked us up. He was like, "Oh,

66:32

yeah, I know you guys. You're PayPal

66:33

kids. Cool. Yeah, you can buy a car.

66:35

Sorry, your credit sucks."

66:37

And like, why does my credit report or

66:40

my credit score not reflect the fact

66:42

that I'm basically independently wealthy

66:43

now?

66:44

>> And that really stuck with me. And so I

66:46

had this a version of this conversation

66:49

including the And then Luke Nosk had a

66:52

sweater wrapped around his head because

66:54

it was really cold on five middle of

66:56

[laughter] the night and cops were

66:58

really worried about him anyway. And so

67:01

the story comes up or came up over and

67:03

over and over again. And I was like, I

67:05

should do something about this. Like

67:06

there's got to be a credit score to be

67:08

built around like came to the US as a

67:11

teenager with no record of any kind and

67:14

$600 for a family of five and now

67:17

independently wealthy Silicon Valley

67:19

entrepreneur. Like it was like a five or

67:21

sevenyear window between those two

67:23

events. So it's it's not weird, but it's

67:25

super weird. Seven years. Like hello, I

67:27

got a computer science degree. I was

67:29

imminently employable. I started a bunch

67:30

of companies. Not all of them failed.

67:32

and why it did not catch up to me at

67:34

all. In fact, it seems to have had no

67:36

bearing on my ability to borrow a fairly

67:38

modest amount of money relative to, you

67:40

know, everything else. And I that's been

67:43

like sloshing around my head for years.

67:45

And I finally sat down with another Luc

67:48

friend from U of I like what if we built

67:51

a better credit score and so like yeah,

67:53

it wouldn't be too hard. So like all of

67:55

us are CS majors. We all have some

67:56

number theory background, therefore

67:58

machine learning, therefore AI. And so

68:00

we built a score and it was like just

68:01

use some publicly available data a

68:03

little bit sort of some other secret

68:04

sauce. We built a score and like oh well

68:06

now we should make people lend money

68:08

using our score because it's cool. And

68:10

so I talked to some bankers which how I

68:12

began the sort of like oh you're doing

68:14

it wrong kid conversations and like yeah

68:16

no one's going to lend money using a

68:18

score that no one else is lending money

68:20

against. What if it's a bad score? What

68:21

if it doesn't work? And so if you tell

68:23

an entrepreneur over and over again this

68:25

thing will never work unless someone

68:27

does X. The natural response is, well, I

68:29

will do X and see if it works. And so

68:32

like, all right, I guess we're going to

68:33

lend money. And so like, I'm just going

68:35

to go and get into a lending business.

68:37

At which point I was like, oh, I got to

68:40

understand how this stuff works. I've

68:41

never actually looked at lending very

68:43

carefully. And that began the journey of

68:44

like, okay, so how do credit cards

68:46

really work? How did I manage to mess up

68:47

my credit at the ripe age of 18 by

68:50

missing three payments? How did that

68:52

happen? And so

68:55

a year later, I was like, "Oh my god,

68:57

this whole thing is so broken and no one

68:58

even knows it." Revolving on a $1,000

69:04

draw and finding yourself with a $3,000

69:08

debt a couple years later and you can't

69:10

explain it or getting a 0% loan but not

69:13

realizing that if you're a dollar short

69:15

or a minute late, it will compound

69:18

retroactively from the time it was

69:20

written to you. So there all these

69:21

things that the industry built over the

69:23

years that are just profoundly anti-c

69:25

customer. It's one of the only

69:27

industries in the world actually like a

69:28

great way of thinking about it. I think

69:30

I heard someone else explain this to me

69:31

this way, but it's the industry where

69:34

you and your service provider face each

69:37

other and you're like, "Hey, I want to

69:39

borrow some money." Great. I'm going to

69:41

bet on you failing. I'm going to

69:44

[snorts] give you a loan, but I want to

69:46

believe that you're you're not going to

69:47

fail completely, but you're going to be

69:49

really late. Ideally, you take forever

69:51

to pay the loan back because I make more

69:53

money that way. Optimally, we sneak in a

69:56

few things you don't notice. And it's

69:58

all in a fine print, but I hope you

69:59

don't read it. It is a very, very

70:01

strange vestigial

70:03

like capitalism supposed to get to

70:05

efficiency. Like, this got us into a bad

70:07

part of the decision tree. Like, what if

70:08

I took a bunch of steps back into the

70:11

branching point where you could have

70:13

done the more consumer-friendly thing

70:15

and just did that over and over and over

70:16

again? What product would I end up with?

70:17

Like, oh, ended up with a firm. And

70:19

that's how we got here.

70:20

>> And what was your confidence in

70:25

market adoption? Right. So going out,

70:27

you're pitching a firm for fundraising.

70:31

And I mean, there's so many questions I

70:33

could ask, right? You're at at a point

70:35

where you could probably self-fund for a

70:38

while, but what were some of your

70:40

assumptions, correct or otherwise, that

70:43

underpinned turning it into a business?

70:46

So you have identified perverse

70:48

incentives and predatory fine print and

70:52

so on. So there's a problem, right? But

70:55

then there are the bankers who are

70:56

saying, "Hey, look, this will never

70:58

work." But you're like, "Well, if if I

71:00

have to do X and get into the lending

71:02

business, I'll get into the lending

71:03

business." So maybe that answers the

71:04

fundraising question actually. But how

71:07

did you have confidence that there was a

71:10

there from a business perspective?

71:12

>> I would pitch this idea in its various

71:14

forms. So, first of all, I did have to

71:16

fund for a while on my own, which I

71:18

didn't mind doing because I was getting

71:20

completely obsessed with the idea. So,

71:21

you know, as such things go, that was a

71:23

small price to pay, relatively speaking.

71:25

But I would keep on telling the story to

71:27

somebody who was like a CEO or former

71:29

CEO of a bank or somebody who managed a

71:30

credit card portfolio, and they'd be

71:31

like, "Okay, you're doing it wrong. All

71:33

the money's in the late fees, the

71:34

money's in delinquencies, just eh, come

71:36

on." I had two ideas that both turn out

71:40

to be amazingly right. By the way, every

71:42

business there's like a moment of luck

71:43

where people like, "Yeah, I don't really

71:44

want to admit to it." A firm had two.

71:47

And there were ideas that I would I'll

71:50

claim some ownership over, but I had no

71:52

idea if I was right. And the first one,

71:55

I read this study that said that

71:58

millennials hate banks. And like I

72:01

didn't really explain why. It was just

72:02

like millennials hate banks. Like they

72:03

hate a lot of things. Like millennials

72:04

and every generation has a list of

72:06

things they wants to complain about.

72:07

Like ah these complaining millennials.

72:09

I'm not a millennial. I'm a relatively,

72:11

I guess, young Gen Xer, but millennials

72:13

were definitely like, ah, they're such

72:15

whiny people. But they were willing to

72:17

participate in some study where they

72:18

ranked all the things they hate. And the

72:20

number one thing they hated, like 70% of

72:21

millennials, according to that study,

72:23

hated banks. And so, every time I would

72:25

ask a banker, why can't this exist? Why

72:27

can't there be a lending product that

72:28

doesn't take advantage of people's

72:30

unwillingness to read fine print and do

72:32

exponential math, they're like, because

72:34

banking is like the stickiest thing in

72:36

the world. You bank where you live. And

72:37

yeah, you you kind of like your bank.

72:39

It's where your parents banked. Like you

72:41

go to Marble Hall and the the vault and

72:43

you you like it. It's it's likable. I'm

72:45

reading a study here that says that they

72:47

hate you. Like all the people that are

72:49

going to be shopping and buying stuff 10

72:51

years from now, 5 years from now, they

72:52

hate you. 78% of them think banks are

72:54

terrible. They should go kill

72:55

themselves. And none [snorts] of these

72:56

people are like, "Yeah, I don't know.

72:57

Like, I don't believe that study." Like,

72:58

okay. I mean, they surveyed like 10,000

73:00

people. That's a pretty stat

73:03

study. So if I'm right and I do a thing

73:06

that's even slightly better than the

73:08

existing thing, I have a readym made

73:10

audience of people that are going to be

73:11

like cool, I hate the other one anyway.

73:13

And at some point like I need an

73:15

explanation beyond just they hate it.

73:17

And so what I realized and I think this

73:19

is the part that was lucky cuz I didn't

73:21

know any better in the beginning, but

73:22

I've since sort of confirmed this to be

73:24

true.

73:25

>> Millennials were early teens during the

73:28

great financial crisis. M.

73:30

>> And so if you are getting booted out of

73:32

your house in '08 and you're like an

73:34

impressionable youth and you're asking

73:36

your parents like, "What's going on? Why

73:37

do we have to live in a motel now?"

73:39

Parents are like, "Because the bank they

73:41

just took the money, they took my house

73:44

away." I mean, the embarrassment and the

73:46

horror of telling your kids, "We're

73:48

going to move to a smaller house or to a

73:49

to a rental or to like we're going to

73:51

live on the streets, god forbid." Like

73:52

that's a horrible impression to have in

73:54

your most impressionable years. And so I

73:57

think there's a lot of people who are

73:58

like actually quite genuine in the I am

74:00

willing to try anything but the thing

74:02

that my parents got the the sharpen the

74:03

stick for. And so that was kind of luck

74:06

number one. And I I really believed that

74:08

from the very beginning and it turned

74:09

out to be true. But like if we build

74:11

something that's good. Normally you have

74:12

to market it. You have to convince

74:13

people to trust you and all. Like I

74:15

think they don't need any of that. They

74:17

just need the belief that there's a

74:19

better thing out there that exists and

74:20

they'll give it a try because they are

74:22

predisposed to not liking the other

74:23

part. And that turned out to be

74:24

completely true. The other part, so if

74:27

you are unwilling to profit from all the

74:29

sort of wacky externalities of financial

74:32

services, you better be good at

74:33

underwriting. So if you're lending money

74:35

to someone, you are paving over

74:38

underwriting mistakes or lack of

74:40

underwriting or poor credit score.

74:42

>> Can you define underwriting here?

74:43

>> When somebody shows up and says, "I'd

74:46

like to borrow $100." And you want to

74:47

respond to them in real time. You have

74:49

some giant machine learning model or a

74:51

small machine learning model. Decide yes

74:53

or no. And what that practically means

74:55

is it has to take whatever available

74:57

data that person is willing to share be

74:58

it public or private data run it through

75:01

some form of classifier or these days

75:04

all kinds of interesting architectural

75:05

sort of ideas exist in underwriting and

75:07

say yes or no but yes or no is actually

75:09

not enough. What do you really want is a

75:10

price of risk. So it's some for

75:13

simplicity expected value or expected

75:15

loss. If I give you money today what are

75:19

the odds? what's the expected value of

75:20

you bringing it back to me with the

75:22

appropriate amount of interest if that's

75:23

what the price of the transaction is and

75:24

so underwriting discipline of doing that

75:27

at scale in our case in completely in

75:30

real time

75:30

>> right

75:31

>> that's the score we built the

75:32

underwriting score the the credit score

75:34

was that

75:35

>> and the theory that I had was credit

75:38

scoring is hard and you would have to

75:40

have the absolute best people working on

75:42

it and yet most of the time when you

75:44

encounter people working on underwriting

75:45

or credit scoring they're not as

75:47

impressive as people I'd met in my

75:49

wondering Silicon Valley up until that

75:50

point and like why is that? It's an

75:52

interesting problem. It's quite

75:53

mathematical. There's some very smart

75:55

mathematically inclined people I know

75:57

>> and the best I could get to was it's

76:00

embarrassing to talk about that you're

76:02

working on loan and credit underwriting

76:04

stuff at cocktail parties. Like when

76:06

somebody asked you what do you do?

76:07

You're a math genius. You went to school

76:09

for applied math or computer science.

76:11

What do you do? One answer is well I

76:13

went to Wall Street and I build

76:15

real-time trading models. Like that's

76:16

kind of cool. Yeah, maybe it's like not

76:18

society important, but you're doing

76:20

something interesting and there's all

76:21

the other answers. I work for the NSA

76:23

and I break codes. Like that's cool.

76:24

That's what I wanted to do when I was in

76:25

college. If you're like, I make lots of

76:28

loans and I make sure that they're

76:30

underwritten well. And by the way, I

76:32

make most of my profits by charging

76:34

people late fees and sneaking in nasty

76:36

terms. So, like there's [snorts] got to

76:38

be a latent pocket of talent of people

76:40

who would absolutely work on

76:41

underwriting because it's a really hard

76:42

and really interesting problem, but they

76:45

won't join the industry until someone

76:46

shows up and says, "I'm going to strip

76:48

it of all the gunk. I'm going to make it

76:50

completely transparent. I'm going to be

76:51

super proconsumer. I'm going to take a

76:53

lot of pride in the brand that we have,"

76:56

which stands for transparency and

76:57

honesty and all the good things you can

76:58

do if you would just take a broom to the

77:00

whole thing. And so as I sort of

77:02

formulated the product, like I'm going

77:04

to get my unfair share of really

77:06

brilliant mathematicians because they're

77:08

not going to go to Wall Street because

77:09

it's kind of a, you know, soul hollowing

77:13

thing to do to squeeze pennies out of

77:14

the market. They'd rather come to work

77:16

for me and build underwriting models

77:18

with me trying to help people in normal

77:21

America borrow money and not feel

77:22

screwed or not get screwed. And they're

77:24

totally exactly right. Like we have

77:25

people who've been here for 10, 11, 12

77:27

years doing that job who are like still

77:30

I'm so proud of what I do. I'm a

77:32

mathematician and I'm putting my big

77:35

brain to work on making honest financial

77:37

products. So those two things were like

77:40

I mean they're they're good ideas but

77:41

they're also a lot of luck. Those

77:43

quarteron quarter growth rates that you

77:45

were mentioning are kind of nuts. I

77:48

mean, seems really remarkable. And I'm

77:51

wondering looking through your scrying

77:54

ball into the future, what do you think

77:57

e-commerce looks like? This could be as

78:00

it relates to a firm. It could be more

78:01

broadly speaking in 2 to 3 years. Who

78:05

knows, right? Let's just pick that as an

78:07

arbitrary timeline. And you know, does a

78:10

gentic commerce if that's actually going

78:13

to be something that takes hold quickly

78:15

if you think really affect a firm or do

78:17

people state preferences in advance? So,

78:19

not really. It's just maybe the way that

78:22

people purchase things change. But

78:24

honestly, for you structurally, things

78:25

don't really change. I mean, how do you

78:27

think about the future of e-commerce?

78:29

>> First of all, the reason the growth

78:30

rates are compounding at a almost $50

78:33

billion mark, 30 plus% year-over-year,

78:36

quarter after quarter after quarter is

78:38

pretty awesome. like it is a

78:39

staggeringly good growth rate. It is

78:41

made better

78:43

by way of noticing that our credit

78:45

related losses are super consistent.

78:48

Like it's easy to grow. I mean the the

78:50

top line, you know, I've lent more money

78:52

this quarter than I did last quarter.

78:53

Great. [laughter] But will it all come

78:54

back is a really important question. And

78:57

the way you measure the success of these

78:58

underwriting models is you ask the

79:00

question, okay, so you made a bunch of

79:01

loans last quarter, last year. Average

79:04

halfife of our loan is five months. So 5

79:06

months back, you get a pretty good

79:08

picture of like how how good are you

79:10

relative to today's macroeconomic

79:11

reality. And if you look at our loss

79:13

numbers or delinquency numbers, whatever

79:14

metric you want to choose, they're

79:15

really really consistent. So we are in

79:18

fact very good at underwriting.

79:20

And yet even at 50ish billion, we are

79:24

about a footnote. So the overall credit

79:26

card debt in the US last I looked was

79:28

like 1.3 trillion. So, we're like

79:31

scratching at the total size of just the

79:35

credit card debt and like not all credit

79:37

card transactions become debt and

79:39

there's plenty of debit card

79:40

transactions and so the overall size of

79:42

commerce is enormous relative to where

79:45

we are and so our growth is not a

79:48

surprise in a sense that what may have

79:50

been a surprise at some point was people

79:52

actually like this product. There are

79:54

reasons to believe or reasons to say

79:56

that this product is harder to use. So,

79:59

credit cards and debit cards for that

80:01

matter is the single best financial

80:03

interface ever created. Like, if you

80:05

think of the level of complexity that

80:07

takes place when you take out your card

80:08

and tap a checkout counter and just walk

80:10

off with your cup of coffee, there's a

80:12

lot of stuff that happens underneath.

80:14

Like, there's an acquiring bank, there's

80:15

an issuing bank, they're talking to

80:17

through a network, there's all kinds of

80:19

complexity just at a technical level,

80:20

and then there's credit and blah blah.

80:22

There's like layers and layers and

80:23

layers of many decades worth of

80:25

innovation that make an incredible

80:28

single transaction happen in like a tap

80:31

and and off you go. And so when you

80:34

introduce buy now pay later, which is

80:36

this idea every transaction is separate,

80:38

you're conscious of every transaction.

80:39

You understand how long it's going to

80:41

take to pay this transaction off. You're

80:43

actually creating friction. You're

80:45

giving people more steps. So, one

80:48

counter-argument to why this thing makes

80:49

any sense at all is like why would you

80:51

want to trade the beauty of tap and go

80:54

with this like open your app and do some

80:56

stuff and the answer is simple. The

80:58

industry devolved and you don't really

81:00

trust what's going to happen to you

81:02

after you tap. So, you hesitate and

81:03

you're like h am I going to really be

81:05

able to afford it or is my card going to

81:06

work? Am I going to get declined or am I

81:08

going to go into debt that I probably

81:10

shouldn't go into? A firm offers the

81:12

antidote of you go into the app, you

81:13

look at your purchasing power, you know

81:15

exactly what we are comfortable lending

81:16

to you, you get explicit approval, next

81:19

transaction is guaranteed to work, and

81:21

by way of saying we're not going to

81:22

charge you late fees or change any of

81:24

the pricing. We're actually telling you,

81:26

you should feel very good about this

81:27

transaction. We are going to make less

81:28

money if you're a minute late. We'll

81:30

make no money if you don't pay us. And

81:31

we feel good about lending you this

81:33

money.

81:33

what we sell or what we offer to the

81:36

consumer and of this is certainty and

81:38

sense of control by way of creating some

81:40

incremental friction. Fast forward a

81:42

couple years, I don't have enough of a

81:44

clear crystal ball to know exactly what

81:45

agent commerce looks like, but I know a

81:47

lot of it is happening already and I

81:48

think a lot more will happen. The

81:50

discretization of transactions and this

81:52

incremental friction will be reduced.

81:55

>> You don't actually have to do all the

81:57

manual labor we are putting you through

81:59

right now because your agents will do it

82:02

for you. And so today we're offering a

82:04

thing in exchange for some friction. In

82:07

tomorrow's world, we're offering the

82:09

thing that obviously works. $50 billion

82:11

can't be wrong. But the friction will go

82:13

away or largely go away, which I think

82:16

is just going to accelerate this whole

82:17

approach to financial transactions. And

82:19

the way people think about money will

82:21

change towards I know exactly what the

82:24

total financial state I'm in. I have

82:26

agents that are looking out for me. They

82:29

won't get me into debt that I shouldn't

82:30

be in. They will get me out of debt the

82:33

second I should be out. I will have a

82:35

PhD in consumer finance embedded in my

82:38

phone looking out for every penny I got.

82:40

There will be no slop. And by God, no

82:42

one's going to fool me again with fine

82:44

print driven business model. So like the

82:46

world in which you have AI looking out

82:48

for all of your financial concerns is a

82:50

beautiful world because we are already

82:52

there by way of not having any

82:54

dependency on. You're too dumb to know

82:55

what's happening to you. So just pay up.

82:57

>> That's how a lot of the industry works

82:59

today. And so I'm very excited about the

83:00

world where you don't have to change

83:02

your business model because suddenly no

83:04

one's getting fooled because that's, you

83:05

know, was never a part of our business

83:06

model. [clears throat]

83:08

>> So I'm going to ask you, it's might be

83:12

naive, probably annoying, but it's going

83:14

to be a how long until X happens type of

83:17

question. So I apologize in advance for

83:18

that, but sure. I look at say China and

83:21

WeChat, right? It's like the interface

83:23

to everything in terms of purchasing,

83:25

you name it. It's incredible what

83:28

they've been able to do and there are a

83:29

lot of reasons for that. And then I look

83:31

at say chat GPT I look at claude and as

83:36

you mentioned with that tap to pay like

83:38

a lot has to happen under the surface

83:40

right so when people say oh well chat

83:43

GPD can just put ads into the LLM

83:48

responses I'm like you may be

83:50

underestimating the relationship

83:52

building that Google has done over

83:54

decades and decades and decades. There's

83:55

a lot that goes into it, but I am

83:57

curious to know how far away you think

84:01

the reality is wherein someone can pull

84:04

up a claude or chat GPT and ask

84:08

questions and make purchases directly

84:10

from a single interface. I don't know if

84:13

that's the form it'll take. I suspect at

84:15

least there'll be attempts made to

84:17

create something like that and then

84:19

everything will happen in the background

84:21

including presumably some type of a

84:24

firmlike option right if you select to

84:26

do that. How far off do you think that

84:28

is? It's very close. I think it's very

84:31

dangerous to consider commerce as this

84:34

uniform

84:36

fabric that just kind of want a thing

84:39

buy a thing. One of the dangers, by the

84:41

way, of building on mass market products

84:43

in Silicon Valley, if you're not

84:45

inventing the future, in which case

84:47

you're green field, you know, blank

84:49

sheet of paper, if you're trying to

84:50

improve something that exists and has

84:52

existed for a while,

84:54

>> it's essential you travel to where

84:56

normal people live and see what they do

84:58

because we are not normal in Silicon

85:00

Valley in a global Silicon Valley. This

85:01

is not limited to San Francisco and the

85:03

Bay Area.

85:05

We think you just want a thing and buy a

85:08

thing. And if that thing is $10, great.

85:10

And if it's $10,000, well, you know, we

85:12

get paid a lot. So, okay, just buy the

85:15

$10,000 thing. And you care about speed

85:17

and efficiency and less distraction. And

85:19

normal people actually hang back and

85:21

like, whoa, it's $10,000. First of all,

85:23

that's an incredible amount of money.

85:25

And by the way, for someone who is not

85:28

within these hot, you know, beds of

85:31

growth and opportunity, maybe $1,000 is

85:34

an incredible amount of money. Maybe

85:36

$500 is an incredible amount of money.

85:37

Depends on where you are. And the sort

85:39

of, ooh, I got to think about this. I

85:41

want to make sure I'm not exposing

85:43

myself to a bad financial decision. I

85:45

also want to make sure I'm getting a

85:46

good deal. How am I going to pay for

85:47

this? How long is it going to be on my

85:50

personal balance sheet before I'm done

85:52

paying for it? Like those are all

85:53

questions people are very conscious of.

85:54

And so for a lot of people in every part

85:58

of the US, Agent Commerce is here when

86:00

you tell Instacart bring me a sandwich

86:02

or bring bring back your groceries and

86:04

you know Door Dash bring me a sandwich

86:06

or vice versa.

86:08

>> And

86:09

that's like the threshold of is this

86:12

enough money for me to sort of hang back

86:13

and ask these questions? For most people

86:15

on who it is too much, they're probably

86:17

not using these services just yet. For

86:20

whom it's sort of like yeah, I want a

86:21

sandwich. I want it now. and like I'm

86:22

going to outsource the rest of this to

86:24

an agent be it a human or a robot don't

86:26

care like you have my credit card number

86:27

bring me a sandwich and I think that's

86:30

here today and we will see more of that

86:33

become a thing and many things like

86:35

real-time delivery or near realtime

86:37

delivery if you go to other markets you

86:39

have I think in India somebody launched

86:41

a 6 minutes or less delivery service

86:44

which boggles my mind but their storage

86:47

of mass market goods to buy has got to

86:50

be just like an exercise extraordinary

86:52

logistical prediction. Anyway, so I

86:55

think that stuff is here. It's coming

86:57

and more will happen.

86:58

>> The I want to know what I'm getting

87:00

into. I want to have agency in selecting

87:03

a thing. My taste matters to me is a

87:07

thing that probably will always involve

87:09

humans. I need a pair of pants is not

87:12

actually a thing 95%. If you're a

87:14

Silicon Valley engineer and you're too

87:15

busy coding and you just, you know, you

87:17

have a big hole on your butt, you got to

87:18

need some pants. Agent, bring me a pair

87:20

of pants. Sure, for most people like I

87:22

want them to make sure they fit and I

87:24

want them to be my favorite color and I

87:26

also want it to match the rest of my

87:27

ensemble hair and and so on and so and

87:29

it goes up from there like I need a

87:31

bicycle but I have brand preferences and

87:33

my components are important to me etc

87:35

etc and so AI will not replace the need

87:38

for decision making and thoughtfulness

87:41

in consumption. It will obiate some

87:44

pieces like who has the best price on

87:47

the bicycle I want with the component

87:49

tree that I prefer is a thing that you

87:51

will gladly outsource like you need to

87:53

make sure it's real. You need to make

87:54

sure that the data you you know LMS are

87:56

fetching for you is current. But that's

87:58

what we're working on today. But at some

88:00

point very soon you'll say, "Hey Chadbot

88:03

of the moment or of my browser, of my

88:05

desktop, I would like to buy a beautiful

88:08

looking Italian make made bike with

88:12

Shimano components. I want a good deal.

88:15

I'm probably going to pay for it over

88:16

time because it's expensive and I

88:17

definitely want to pay no late fees and

88:19

ideally I don't want to pay any

88:20

interest.

88:21

>> I'm bringing my business to someone who

88:22

should be so lucky because multiple

88:24

people will sell me a beautiful

88:25

Italianmade bike." And go do some

88:27

comparison shopping for me. bring me

88:29

some images of beautiful bikes and I

88:31

want to lust after all of them and then

88:33

pick the one that gets my business.

88:35

>> That's a great task for AI.

88:37

>> The building blocks for that will

88:39

include something like a firm where the

88:41

AI will say, "All right, so all these

88:42

people offer some ways of paying. Some

88:44

of them offer a firm which has no late

88:46

fees and by the way has negotiated a

88:48

special deal with a manufacturer or the

88:50

seller where they will pay your interest

88:52

for you. So they're actually covering

88:53

the time value of money. So the plan is

88:56

interest free which a lot of our

88:57

transactions are interest free." exactly

88:58

this way. They're funded by the retailer

89:00

or or the manufacturer. And we're

89:03

grasping at that future right now. We're

89:04

certainly building a lot of the pieces.

89:06

So, you know, I I tend to be slightly

89:08

ahead of schedule as far as the future I

89:11

want to live in. I'm trying to pull it

89:12

in here, but this is quarters, not not

89:16

years.

89:17

>> Incredible. Man,

89:19

>> for the foreseeable future, there's so

89:22

much to build. The thing that irks me

89:23

the most in today's media is the SAS

89:26

apocalypse, whatever. But the job

89:28

apocalypse people are proclaiming is so

89:30

goofy. There's so much opportunity to

89:32

build so many exciting things for

89:34

everyone, not just Silicon Valley

89:35

startup. Like everyone from the oldest

89:36

companies to the youngest ones are like,

89:38

"Oh my god, there all these pieces of

89:40

software that I had to buy from someone

89:41

who did a bad job or that I could never

89:44

buy anywhere and I wouldn't know how to

89:45

start with." Like, well, now you have

89:47

all these amazing tools that just birthe

89:50

it for you straight from your head.

89:51

Yeah. Speaking of like things you can

89:54

put on your brain, very soon we're going

89:56

to have ultrasound

89:58

mind reading that spits code on the

90:01

other side of it. That's what I'm

90:02

waiting for.

90:03

>> Seems a lot closer than I would have

90:06

predicted if you asked me just a few

90:08

years ago. It's wild.

90:10

>> The world changed a bunch of times over

90:11

the last 12 months, but I think like the

90:13

Cloud Code moment last December was like

90:15

a big like whoa, it really is here. The

90:18

ability to produce something from a

90:20

glimmer in your mind to a thing that

90:22

actually works reasonably well and has

90:25

only improved since was like a big

90:27

breaking point. The one before was

90:28

obviously CH GPT and and so on.

90:30

>> A couple of rapid fire questions for

90:33

you.

90:34

>> Yeah, coffee. We're going to short

90:36

change this one. I apologize. Maybe we

90:38

do another podcast solely dedicated to

90:40

coffee. But for people who want to

90:41

improve their coffee experience,

90:45

cheap option, intermediate option,

90:48

Bugatti option. Any thoughts on

90:51

improving coffee experience?

90:53

>> I'm not paid by any of these brands, so

90:55

I just I just want to make it very

90:57

clear. I I have some very strong

90:58

opinions. You strong opinions. Come on,

91:00

Max.

91:01

>> Yeah, [laughter] exactly.

91:02

>> All right. So, first of all, this is

91:05

primarily relevant to espresso. So, I

91:07

love all coffee in all forms.

91:10

I happen to prefer espresso as a

91:12

beverage of choice, but I don't judge.

91:14

You can have your lattes, your

91:15

americanos, you know, whatever. But from

91:17

the point of view of espresso and

91:18

espresso based drinks, the single most

91:20

important thing is the grinder. And so,

91:22

you can get a fantastic grinder like a

91:25

niche or these $600, $700 grinders that

91:29

will elevate your game to an incredible

91:32

level. So, if you have a grinder that

91:34

you didn't pay a couple hundred for, got

91:36

to go do that. If you want to get to

91:38

like, you know, a demonstrabably better

91:40

grinder than some of the sort of the the

91:43

$600 range, go get something that looks

91:46

like a Kaya Orbit. That's like a $1,600

91:49

to $2,000 grinder. It's a great grinder.

91:52

You can swap out burrs. So, switching

91:54

out burrs is really important. But even

91:56

if you're never going to replace your

91:57

burrs, you still want something that's a

91:58

a fall through grinder. And a good one

92:01

will cost you on the order of $2,000

92:03

plus or minus. And then if you're like,

92:05

you know what, no price too large. I'm

92:06

going to get me the absolute fantastic

92:09

home grinder. I mean, obviously there's

92:10

like industrial strength stuff that's

92:11

that's even out out there. There's this

92:14

thing called Weber Workshops that

92:15

produces unbelievably expensive but

92:17

unbelievably good stuff. So if you want

92:19

a Weber, God bless. That's a great

92:21

product.

92:21

>> Same company that makes the grills. No,

92:23

different. Got to be [laughter] totally

92:26

different. I had this exact question

92:27

yesterday. Anyway, so better grinder

92:30

goes a very long way. After that, go

92:34

invest in skills. So, there's an

92:36

incredible number of online resources

92:39

that'll teach you how to do a great job.

92:41

I'm but a student of those people. There

92:44

are many, many, many great videos you

92:46

can watch on YouTube that'll teach you

92:47

how to make a great cup of espresso.

92:49

Before you buy anything else, just go

92:50

watch all that stuff because it'll teach

92:52

you a lot of things you don't know

92:52

about. After that, you're probably going

92:55

to want to buy a good espresso machine.

92:57

The great news is that there are many

92:59

many very good espresso machines.

93:03

I'm don't want to begin religious wars

93:05

here, so I'll state my preferences, but

93:07

these are preferences. I prefer 58 mm

93:10

porter filter diameter.

93:14

Many people now swear by 54, 53 mm. You

93:17

can go up or down, but I stick to the

93:20

classics. I think the finest, most

93:23

reliable home machine is made by

93:25

Lamarzoko. that is definitely in the

93:27

multi,000 range, but that's what I have

93:29

on my counter and it's never failed me

93:31

and it's beautiful and both inside and

93:33

outside. There's lots of cool like if

93:35

you're a super nerd and you really want

93:36

to nerd out on the metrics, decent

93:38

espresso is a fantastic product, really

93:40

fun, but it's like an Android tablet

93:42

that happens to be attached to an

93:43

espresso machine. Basically, Lamarzoko

93:45

is old school. That that is the logo

93:46

you'll find in every self-respecting

93:48

coffee shop. Like the long red typically

93:51

spelled out Lamarzoko, that's a

93:52

granddaddy of them all. It's amazing.

93:54

and so on. And there are many, many

93:55

other brands that will sell you a good

93:57

espresso machine. If you're trying to go

93:58

down market a little bit, but like

94:00

bulletproof will will make a good cup

94:02

every time. Brevel,

94:03

>> people tend to sort of hate on it a

94:05

little bit because it's so consumer, but

94:06

it makes great coffee. Brevel's

94:08

surprisingly good.

94:09

>> It's very, very good. I mostly just

94:11

wanted you to showcase my obsession.

94:14

>> Yeah, your obsession. So, I'm going to

94:17

force your hand though on one here.

94:19

We're going to take a a side quest away

94:21

from pure espresso territory. ChemX,

94:24

French press, or Aerrow Press? If you

94:26

had to do coffee with one of those

94:28

three,

94:28

>> ChemX. ChemX. Why ChemX? If you're going

94:31

to go for a light roast, each type of

94:33

roast, each type of bean speaks to a

94:35

method. So, this is like another heresy,

94:37

by the way, in the modern day. I like

94:40

medium dark roast for my espresso. I

94:42

love the honeyed viscous,

94:45

you know, feels like it's too thick. You

94:47

can almost chew it. I love that in

94:49

espresso. Like that's a great cup of

94:50

espresso.

94:52

For non-espresso drinks, I actually love

94:56

thin, high clarity, low body, just give

94:59

me the pure essence of not diluted, but

95:02

the bean in water

95:04

>> and chemics are probably the best one of

95:06

those. It's really the most

95:07

controllable.

95:07

>> Mhm.

95:08

>> Beyond that, French press is great.

95:09

That's a like a campfire type setup. And

95:12

[clears throat]

95:12

>> air press, if you're going to go there,

95:14

just get an espresso machine.

95:18

ChemX, for those people wondering, it

95:20

will will cost less than $600. And I

95:23

will say also that you can get, correct

95:27

me if I'm wrong, you're going to know

95:28

this better than I do, but you can get

95:30

they're not going to be anywhere near as

95:32

sophisticated as the devices you

95:33

describe, but if you are using, for

95:36

instance, a blade

95:40

grinder, right? something that is not a

95:42

burr grinder and you get a handheld burr

95:45

grinder, which you can probably buy for

95:47

$100 or less, the difference between

95:49

those two will still be noticeable. What

95:52

would you think?

95:52

>> You're like in the territory where I I

95:55

probably like, you know what, no coffee

95:57

today. [laughter]

95:59

I'm kidding. I've been known to exhaust

96:02

the local supply of pregrown pods in

96:06

hotel rooms. So

96:07

>> there's coffee and all the beauty of it

96:09

and then there's caffeine and you know I

96:11

need both

96:11

>> different things.

96:12

>> They're not the same.

96:13

>> The spice grinders as they're known, you

96:15

can definitely make a cup of coffee with

96:17

that too. [clears throat]

96:17

>> Mhm.

96:18

>> But that said, by the way, if you must

96:20

have a thing that is not an electric

96:22

grinder, get a manual grinder. Those are

96:25

very cheap.

96:27

>> I think

96:28

>> Lido, if I were correctly, is a brand

96:29

that'll sell you like a very very

96:31

high-end, but like still an order of

96:33

magnitude cheaper than any of the stuff

96:34

that I threw out. I mean, it's a little

96:36

bit of a, you know, good workout, but

96:38

you'll make beautiful coffee with that

96:40

because there's a really nice

96:42

correlation to the lower the RPM, the

96:45

less kind of you're damaging, if you

96:48

will, the bean as you're grinding it.

96:49

So, you can actually get some amazing

96:51

taste out of manually ground beans. It

96:54

does take like 10 minutes per cup,

96:55

though, so you got to you got to be

96:56

ready.

96:57

>> This is what I'm talking about. This is

96:58

exactly what I'm talking about. So, I

97:00

wasn't totally misspeaking. Yeah, the

97:01

Lido OG manual coffee grinder. Yeah,

97:04

>> this one's 273. There are some other

97:06

options that I found at slightly lower

97:08

prices which I thought did a nice job.

97:09

It is a workout. It is a workout.

97:11

>> Yeah, switch arms every once in a while.

97:13

>> But hey, if you're going to go into the

97:14

ritual of it, I know people who have

97:16

traveled with

97:18

>> something akin to that. An arrow press

97:21

and some other stuff.

97:22

>> Air press great campfire attraction.

97:25

>> Yeah, for sure.

97:25

>> All right. I'm gonna destroy any shred

97:27

of respect that you have for me with

97:28

respect to coffee by also saying if

97:30

people want something simple like

97:32

commenter actually I really enjoy some

97:34

of their stuff. Oh the face if you guys

97:36

aren't watching video the the [laughter]

97:37

response that I just got by the way. So

97:41

I grew up drinking not just instant

97:44

coffee which was hugely popular in

97:46

Eastern Europe. I occasionally was

97:49

exposed to chory coffee which I don't

97:52

know if you ever tried that.

97:53

>> Popular in New Orleans. They've added a

97:54

lot of sugar though

97:55

>> except there's no coffee in it. It's

97:57

just sugary root.

97:58

>> Oh god. So it's just like placebo.

98:00

>> Yeah. I mean it's like kind of sort of

98:02

tastes a little bit like coffee maybe

98:04

like pour enough hot water into it. If

98:05

it's hot enough you're like ah it

98:08

doesn't smell right but whatever. In

98:10

Soviet Union was one marketed under

98:13

coffee light.

98:14

>> [laughter]

98:14

>> the coffee

98:16

beverage

98:18

stock. Something meant to evoke some

98:22

plant component that both

98:26

coffee bush and chory

98:29

>> would share. So like, well, you know,

98:30

they both have it, so it's good

98:32

[laughter] enough. It was disgusting,

98:33

but it'll wake you up.

98:34

>> Yeah. Yeah. So there is that.

98:36

>> There's that.

98:37

>> Yeah. Chory is pretty interesting from a

98:38

fiber perspective, but I'll leave that

98:39

for another conversation. Max, last

98:41

question. This might be a terrible

98:43

question to end on, but I am curious for

98:45

someone who is technical

98:48

or just a kind of individual contributor

98:50

in the sense that you were very

98:52

technical and then grew into the CEO

98:54

role and running a public company. Let's

98:56

put that aside because that's that's a

98:58

whole kettle of fish by itself. But any

99:02

books or resources that you would

99:04

recommend to people who are hoping to

99:07

become a CEO for the first time

99:09

slashfounder?

99:10

>> Yes.

99:11

So, here's a couple. There's not a

99:13

panacea, but

99:16

I will preface this by saying I have

99:19

extremely low

99:22

degree, by which I mean like zero to

99:24

negative, of respect for business books.

99:27

Vast majority of them are far too long,

99:29

present company excluded. [laughter]

99:32

business books that are verbatim

99:33

anecdotes of people who've done it are

99:35

actually high value because I think it's

99:37

very hard to distill what is inherently

99:40

a collection of unique experiences. So

99:41

like you can edit it down, but you can't

99:43

make it

99:44

>> generalizable. Yeah. To generalize it,

99:46

you're robbing the reader of the true

99:48

history of what really happened. So, I'm

99:49

a huge fan actually of what you do in

99:51

your books because there's always so

99:53

much color

99:55

directly from the source versus I think

99:58

this person meant X when they talk to me

99:59

and I'm not going to print anything the

100:01

same the thing they said. Anyway, that

100:02

said, I'm a big fan of business books

100:04

that have been distilled to business

100:05

essays because then I feel like the

100:06

author did the work of like I'm going to

100:08

generalize. I'm going to really

100:09

generalize it.

100:10

>> So, there is this book called Seven

100:12

Powers by Hamilton Hemler. M

100:14

>> if you haven't read it, it's a really

100:16

worthwhile distillation of what it takes

100:19

to build a competitively

100:23

lasting business. It talks about

100:27

why network businesses are longer living

100:31

than non-etwork businesses and what

100:34

brand actually means. Like people, oh

100:35

you know, I'm going to build a brand.

100:36

Like why why do you care about having a

100:38

brand? And so it's a great book.

100:40

distills a ton of these kind of things

100:42

that you kind of think you know, but

100:44

like they need names and terminologies

100:46

and idiosyncratic in the sense that like

100:48

I didn't think of the word power. I

100:50

didn't think of the term power until I

100:51

read his book and I I found it to be a

100:53

great kind of a distillation. It's very

100:56

short. It's slightly shorter than Master

100:58

of Margarita. So anyway, that's a good

100:59

book. Another good book that is off the

101:03

beaten path, but a great one called A

101:06

Failure of Nerve. It's a book about

101:09

leadership and it really postulates this

101:11

concept of differentiated leader. So as

101:13

a CEO, as a founder, a lot of times you

101:15

find yourself at odds with your own team

101:18

where you're making an unpopular

101:19

decision. You're firing a beloved

101:21

employee. The sort of whenever there's

101:22

doubt, there's no doubt. Then you go do

101:24

it and you're like, "Holy crap, what

101:25

have you done?" This is a So you find

101:27

yourself in these moments where you have

101:29

to persevere before and then after a

101:32

decision that you made and many, many

101:34

flavors of it. It's a good book that

101:36

kind of teaches you how to think about

101:38

it, how to tolerate the stress that

101:40

comes with it, how to sort of put up

101:41

with the pressure you're going to get

101:43

and not lose your humanity and like not

101:45

become a tyrant, but also not be someone

101:48

who is easily bowled over into like,

101:50

okay, okay, I'll reverse my decision

101:51

because then you you lose the confidence

101:53

of the of the people you're supposed to

101:54

lead. So, that's a great book. I am sure

101:57

it's in your list and everyone's list,

101:59

but if you haven't read it, if you're

102:01

trying to start a business, you should

102:02

read Influence by Chelini because that

102:04

is probably the most important social

102:06

science book published in the last 50

102:08

years. So good. It's a great book. I'm

102:11

shocked that I had never heard of Seven

102:13

Powers and I just looked it up. The

102:15

forwards written by Reed Hastings of

102:16

Netflix. I have never even heard of it.

102:18

That's shocking.

102:19

>> There you go.

102:20

>> I'm excited. Yeah. If there's one thing

102:21

you find in this podcast, any

102:23

biographies that have, they don't have

102:25

to be business related, but biographies

102:28

that have informed

102:30

>> how you approach leadership or running

102:32

companies.

102:33

>> I love biographies,

102:34

>> could be indirect, too. It could be a

102:36

biography and how it informed your view

102:39

on persistence, right, through stories

102:41

or anything.

102:42

>> I think Cherno's biography, so Cherno

102:45

wrote a bunch of these like tombs

102:46

speaking of enormous books. Yeah.

102:48

There's everything from the JP Morgan,

102:51

>> Hamilton,

102:52

>> yeah,

102:53

>> Washington, a life, Grant, Mark Twain.

102:56

>> Mark Twain is a great one.

102:57

>> Titan is a huge one.

102:58

>> Titan is probably the one that's closest

103:00

to like business advice. It's a bit of a

103:02

like how to be ruthless basically, but

103:04

that's besides the point.

103:05

>> Yeah, it's the life of John D.

103:07

Rockefeller. The cover photo looks just

103:11

like the vampire and no feratu. It's

103:13

terrifying. But yes, [laughter]

103:15

>> exactly.

103:15

>> Kind of appropo.

103:17

>> Those are great books. There's a great

103:18

book called A Mind at Play, which is

103:21

Claude Shannon's biography.

103:22

>> By the way, when I was preparing for our

103:24

conversation, not nearly as well as you

103:26

are, I was listening to your

103:27

conversation with Ed Thorp.

103:29

>> Yeah.

103:30

>> Who I knew a little bit about, but I was

103:31

like, "Oh my god, that guy is flipping

103:33

amazing." Amazing. Like, I want to be

103:35

like him when I grew up. He's now one a

103:38

of the list of people I'd like to meet

103:40

at some point in my life. I've never

103:42

encountered him. So incredible. He just

103:44

seems like an unbelievable guy. But so

103:46

he was friends with Claude Shannon, let

103:48

alone makes him unbelievably

103:49

interesting. So Claude Shannon is kind

103:51

of like a platonic ideal of what I

103:53

thought I was going to be before I left

103:54

academia cuz like this guy is just like

103:56

brilliant in every way. Like everything

103:58

he touched became this flower of

103:59

intellectual brilliance.

104:01

>> American polymath and mathematician.

104:03

Yeah.

104:04

>> Why does Shannon stick out to you so

104:07

much?

104:07

>> Fun fact, you know Claude Code named

104:09

after Shannon.

104:10

>> I did not know that. That's incredible.

104:13

Well, actually, I didn't like learn this

104:14

from entropic people, but I would be

104:16

shocked if that's not the answer. Claude

104:18

is not a common name. And Shannon is

104:20

obviously the inventor, the postulator

104:22

of the information theory, right? Father

104:24

of information theory.

104:25

>> He is

104:28

a great example of someone who worked on

104:30

some very serious stuff and was always

104:35

playful. He's like Richard Feman of

104:36

Computer Science. So, Fineman was

104:38

famously brilliant. And there's a great

104:40

quote from Murray Yelman. Somebody asked

104:42

him like what's your algorithm for

104:43

solving really hard problems like

104:45

encounter really hard problem step two

104:47

give it to Richard Fineman step three

104:49

receive solution [laughter] Shannon was

104:51

kind of that in computer science proto

104:53

computer science obviously he was

104:55

building this in the 40s '50s60s and he

104:57

managed to remain fun he managed to have

105:00

fun like he was just this unbelievable

105:02

fountain of fun ideas and he was like a

105:04

hardware tinkerer and made card counting

105:06

devices and all kinds of hilarious stuff

105:08

>> makes sense that he would know Ed Thorp

105:10

right who beat the roulette table

105:11

>> of Exactly. And so that's a good book

105:14

about him. These days it's so hard to

105:16

read books because there's so much

105:18

content online in podcast and blog form.

105:21

>> So much content online feels ephemeral.

105:23

So you're like, I must read this now cuz

105:25

it might scroll off and I'll never see

105:27

it again. A book is in my hand. Like I

105:29

have books in my backpack that I travel

105:30

with that I've been meaning to read for

105:32

half a year now and I still haven't

105:33

opened them. Just like carry the weight.

105:36

I have a like a 500page graphic novel

105:39

that I'm going to be lugging to New York

105:41

City because still the digital

105:43

experience just doesn't do it. But you

105:45

know the [snorts] the burden I bear

105:46

>> I keep on looking for a great digital

105:48

graphic novel consumption experience but

105:52

it does not come close. [laughter]

105:54

>> It's not there. We'll be able to replay

105:56

our dreams before we get that. Uh Max,

105:58

[clears throat]

105:59

so much fun to hang. Thanks for taking

106:02

the time. And where would you like to

106:04

point people? We've got livechin.com. We

106:06

have sci-fi.vc. We have various social.

106:09

>> Affirm.

106:10

>> Affirm. Of course. Affirm.com. There's a

106:12

lot we could point people to. Anything

106:14

else you'd like to add? Whether it's

106:16

pointing people somewhere or closing

106:18

comments, anything at all?

106:20

>> No, my my work is out there. I try to

106:22

build in public. If you're interested in

106:25

a firm, obviously affirm.com is where

106:27

you find that. I tend to go deep, so my

106:29

project list is short at any given

106:31

moment, but leftchen.net, at le either

106:34

one is the list of those things.

106:36

>> Amazing. Max, thank you again for taking

106:39

the time and everybody listening. We

106:40

will link to all of the things we

106:42

mentioned, books and resources and

106:44

people and everything else. Affirm of

106:46

course and on and on and on in the show

106:49

notes as per usual at tim.blog/mpodcast.

106:53

Just search Levchin. I can promise you

106:55

he will be the only one. I might be the

106:56

only Max in fact on the podcast so far.

106:59

We'll find out. I might eat my words.

107:01

And until next time, of course, thank

107:03

you for tuning in. Be just a bit kinder

107:05

than is necessary to others, but also to

107:09

yourself.

107:10

And that's all for now. Thanks again,

107:12

Max.

107:13

>> Great to see you. Safe travels.

Interactive Summary

The video features a wide-ranging conversation with Max Levchin, covering his views on marriage, leadership, fiction, technology, and his entrepreneurial journey. Levchin shares insights into the secret of a successful marriage, his obsession with data and tracking, his passion for cycling, and the founding philosophy behind Affirm, a financial technology company that aims to provide consumer-friendly alternatives to traditional lending.

Suggested questions

4 ready-made prompts