Podchaser Logo
Home
Balaji Srinivasan on The Network State

Balaji Srinivasan on The Network State

Released Tuesday, 15th November 2022
Good episode? Give it some love!
Balaji Srinivasan on The Network State

Balaji Srinivasan on The Network State

Balaji Srinivasan on The Network State

Balaji Srinivasan on The Network State

Tuesday, 15th November 2022
Good episode? Give it some love!
Rate Episode

Episode Transcript

Transcripts are displayed as originally observed. Some content, including advertisements may have changed.

Use Ctrl + F to search

0:00

had somebody actually said, oh, I'm gonna

0:02

found a currency in two thousand seventeen, two

0:04

thousand eight. Almost every VC would

0:06

have looked to them like they had four eyes.

0:10

Today, I had the absolute pleasure of

0:12

talking to who I think is one of the

0:14

most fascinating minds on the planet right now.

0:16

That is Balaji Srinivasan. who

0:18

you might recognize as the former CTO of Coinbase,

0:21

former general partner at Andres and Horowitz.

0:23

He was also the Co Founder of Genetic Testing

0:25

Company Council and earned dot com, which

0:27

was acquired by Coinbase, an annual

0:29

investor in many successful companies like Deal,

0:31

Replicant, Super Human. And today,

0:33

we brought him in to discuss his best selling

0:35

book the network state. And

0:38

as you can imagine, when you talk to biology for

0:40

three hours, you can cover a lot of ground.

0:42

In this episode, we cover the difference between

0:44

nation and a state how constant become

0:46

variables, the cloud content, the newest

0:49

Leviathan, digital power, your identity

0:51

stack, calibrating risk, polycentric law,

0:53

cloud regulations, building fast with atoms,

0:56

founding versus inheriting, the powerful

0:58

versus the powerless, and just about

1:00

everything in between. So I hope you

1:02

enjoy this conversation as much as I did.

1:05

The

1:07

content here is for informational purposes

1:09

only should not be taken as legal business

1:11

tax or investment advice or be used to

1:13

evaluate any investment or security

1:15

and is not directed at any investors or

1:18

potential investors in any ASIC sixteen

1:20

z fund. For more details, BCA

1:22

sixteen z dot com slash disclosures.

1:35

Alright. We've got Balaji on the line. Balaji,

1:37

thanks so much for being here. Great

1:38

to be here.

1:39

I have to confess. I am a Balaji

1:42

fan. I've probably consumed an

1:44

unhealthy amount of biology content, so

1:46

very excited about this. Thank you for invading

1:49

my brain. I also read your

1:51

book recently, which is what we're gonna be talking about

1:53

today, this idea of the network state. And

1:55

I also have to say that similar to

1:58

what I learned about cryptocurrency. It kind of

1:59

forced me down this rabbit hole to

2:02

reconcile what happened around me in the

2:04

outside world, how monetary policy worked

2:06

as an example. The network state

2:08

kind of did the same thing, but with other concepts.

2:11

Right? Like nation states, what are these things?

2:13

So I'm excited to dive into that, but I wanna

2:15

introduce the audience to a simple

2:17

frame that I think was, again,

2:19

an aha moment for me, which is that

2:21

technology has been able to

2:23

reshape our lives at an individual level.

2:25

I think many people can understand that. They can

2:27

grasp that. Also, at the company

2:29

level, right, we've been able to start new companies

2:32

that are digital. It's also reshaped our

2:34

ability to participate in that ecosystem. And

2:36

then more recently, currencies as well.

2:38

Right? The network has changed that.

2:41

And your book discusses how technology

2:43

can actually potentially create new

2:45

states. very interesting concept.

2:47

And in order for us to dive into that,

2:49

why don't we start off with what

2:52

we have today, which are nation

2:54

states? So how would you define

2:56

a nation? And how would you define a state?

2:58

And how do those kind of interplay together?

3:00

Actually, I have this at the beginning of chapter

3:03

five in the book. And the

3:05

thing is we hear the term nation state

3:08

as a compound, you know, like a

3:10

phrase really. And we don't really

3:12

give too much thought as to what it means, but

3:15

it's useful to think about distinction between

3:17

those two words because they're actually very different. The nation

3:20

that comes from the same fruit as natality,

3:22

like the Latin like nocities, and

3:25

it's a common birth, common descent. So with

3:27

the Japanese nation, they have shared ancestry,

3:29

shared culture going back, hundreds,

3:31

you know, of years and shared

3:33

language, all all the type of stuff. And then

3:36

separate from that is the state, which is the administrative

3:38

unit that thin layer that sits

3:40

over the nation. Okay? And the

3:42

state could be of, you know, the Japanese

3:45

people themselves or could be, for example, nineteen

3:47

forty six after World War two. the

3:49

American government was basically the

3:51

state over the Japanese nation.

3:53

Right? And so once you think

3:55

of this distinction between the nation

3:57

and the state above it, the

3:59

cons of the nation state was every

4:02

ethnic group has the right

4:04

to have its own government. Now

4:07

the thing about this, of course, is

4:09

that's like an abstraction, but in the terms,

4:11

many of the ethnic groups on the planet

4:13

do not have their own government and

4:15

or they do not have their own territory, which is

4:18

another part of it. And so if you're talking about the

4:20

Catalonians or you know,

4:22

the the Kurds. These are groups

4:24

that have long histories

4:26

and they have, like, a legitimate nature

4:28

from the perspective, like, the seventeen hundreds,

4:30

eight centers person who think about, you know, the

4:32

nation state, they'd be considered a real

4:34

nation, but they do not have

4:36

territory. in the global game of musical

4:38

chairs, they didn't get a seat.

4:41

Okay? So, like, you have this map of

4:43

the world and everybody's kind of moving around

4:45

it, and they just lost out in the seat. They're

4:47

a stateless nation. Okay? And

4:49

so, like, the Kurds, for example, their ethnic

4:51

group overlaps or their historical lands overlap

4:54

like Turkey and other tenants' place Okay?

4:56

And the Californians, for example, they overlap

4:59

modern Spain. The reason for this is

5:01

that the maps that we have

5:03

are actually very abstract

5:05

constructs. You know, if you think about

5:06

a map of

5:08

Africa or actually even in the US, you have

5:10

these straight lines on a map.

5:12

whenever you see straight lines on a map, that's like

5:14

some surveyor said, okay. I'm gonna put this here

5:16

because of the latitude. That's some like abstract

5:19

political decision. In a sense, a map

5:21

is like a digital layer above the physical

5:23

world. You don't see the lines of the map in the

5:25

physical world. Right? Often,

5:27

those boundaries on a map do not

5:29

reflect the ancestral, you

5:31

know, long running boundaries of language

5:34

and culture and so on. There is however

5:36

a feedback loop where what happens

5:38

is those borders or those

5:40

boundaries that humans impose, then

5:42

in turn, change how

5:45

humans start living. And so

5:47

you will have something where there's now a sharp

5:49

transition from French to German

5:51

as you cross the Franco German border. Signs

5:53

change, languages change, that

5:55

human digital thing has caused

5:57

a digital crease in the physical

5:59

world, whereas before it might have been more continuous.

6:02

Right? Before we have modern map making,

6:04

before the world was mapped, you sort of had a gradual

6:07

climb, a gradual shift. Right?

6:09

So once you think of the difference to the nation

6:11

in the state, that the nation is

6:13

group of people with common birth and state administrative

6:15

layer. And you think about how that relates

6:17

to maps and so on, you can start thinking,

6:19

oh, okay. The exact configuration that we

6:21

have today is not only not

6:23

how it's not always being, but not how it

6:25

always will be. Yeah.

6:26

I think that was another

6:29

aha moment for me because in

6:31

my lifetime, those lines have been

6:33

constant. For the most part, right, there's exceptions,

6:35

but you kind of accept that this is the way the

6:37

world works. Right? These are the countries And

6:39

I think another interesting concept

6:41

that you mentioned in your book is even the United

6:43

Nations is perhaps named incorrectly. Right?

6:45

It should really be like the the chosen states

6:47

or something along those lines. The

6:49

selected states. The selected states.

6:51

Yeah. There's there's a guy actually the

6:53

Kazak san head. He actually made this

6:55

remark. He's like, Look, if we

6:58

allowed every group in the world with

7:00

legitimate right to self determination to

7:02

get their own territory, there'd be chaos.

7:04

We'd go from hundred ninety three

7:07

countries to six hundred is what you

7:09

said. Right? I think it's probably way more

7:11

than six hundred. But that just

7:13

gives you a sense of, oh, okay.

7:15

I used to be against

7:17

word games for the sake of word games,

7:20

but then I got

7:22

more into it. And one way of thinking about that

7:24

is if you you know, dolly two? Yep.

7:27

Exactly. Yeah. Mhmm. So dolly

7:29

2II had this tweet, like, the age of the

7:31

phrase. Right? The short phrase,

7:33

like whether it's a tweet, you know, on social

7:35

media, or whether it's twelve or

7:37

thirteen or fourteen word phrase to

7:40

go and reset your cryptocurrency, your wallet,

7:42

you know, passphrase. Or if

7:44

it's a short phrase with, like, just

7:46

a single some of his character or

7:48

single word that's different that changes

7:51

what the computer generates. You

7:53

can actually see that changing one

7:55

word changes what is

7:57

in somebody's brain. And the concept

7:59

of brussel conjugation, eye sweat,

8:02

uberspire, but she glows. The

8:04

same behavior can be flipped

8:07

neutrally, negatively, positively

8:10

without lying, what you're just

8:12

doing is just changing the tone of it.

8:14

And so you can flip all of those variables

8:16

negative. If you wanna make somebody look bad or flip

8:18

all them positive, he is, you

8:20

know, in an uncontrollable range,

8:22

but she claps back. preciously.

8:25

You know? That kind of thing. Right? And you just

8:27

characterize the situation differently, you wrestle conjugate

8:29

it. Right? And so the point

8:31

being that that slight trace

8:33

word now with dolly two, we can see

8:36

a tiny change of folks literally a

8:38

different visual in the computer brain.

8:41

and it also evokes a different visual

8:43

probably in the human brain. So with

8:45

that, just to descend onto this

8:47

concept of the nation, the fact that people

8:49

have conflated the nation and the

8:51

state recently, like people just say a nation state.

8:53

They don't realize there's a difference in the

8:55

nation state. Many states

8:57

today are actually not nation states.

8:59

Japan is a economical example of a nation

9:01

state. Israel is others

9:03

like, let's say India,

9:05

that's like more of a civilization state. There's

9:07

lots of different ethnic groups. India is more like

9:09

Europe, you know. Like the South Indian and

9:11

North Indian are like historically

9:13

is different as like, you

9:15

know, a Spaniard or somebody from

9:17

Finland. Okay? India was not

9:19

a country until Recently, it was

9:21

a civilization. They did have things in common, but

9:23

a civilization say is different from a nation say.

9:25

That's one example. Or, you know,

9:28

something like you know, Singapore or

9:30

or the United States, which is

9:32

a multinational state where there's many

9:34

different ethnic groups. under one

9:37

administrative zone. It's not a nation state.

9:39

It is a multinational state. Now people

9:41

could argue that there was an

9:43

American nation, like, by

9:45

mid nineteen hundreds, there's a lot of

9:47

effort to kind of pack down

9:49

people. Like, and tell you Roosevelt talked about he wanted the

9:51

end of the hyphenated American, you know, he

9:53

didn't want people to be. I'm a Polish american,

9:55

a dis everybody's just an American. Right? So it's

9:57

an enormous process where the state

9:59

was

9:59

trying to pack

10:00

in and centralize and

10:03

remove those points of synced in

10:05

a survey and just think of themselves as American

10:07

only. And by mid century, you could

10:09

argue that was somewhat successful and

10:10

that was like peak centralization. And

10:14

why is that important? Well,

10:16

then people start not

10:18

thinking about the difference between the nation and the

10:20

state where all the state itself

10:22

has formed the nation. Do you see what I'm

10:24

saying? The state now says the

10:26

German state is Germany. The French state

10:28

is France. Right? The Italian state is

10:30

Italy. The American state is

10:32

America, and it waves a flag and so on.

10:34

And it's to the state's advantage

10:36

to identify itself with the nation. One

10:38

analogy I have in the book, which I think is a

10:40

funny way of re thinking it's

10:42

like labor and management. And in

10:44

a small company, they're actually literally the

10:46

same. You know, the founder is both the CEO

10:48

and the person who's going and taking out the

10:50

garbage or fixing the computers

10:53

or whatever. like, every is done by one person. Right? There's

10:55

no, you know, division of labor

10:57

or what have you. And then as the

10:59

company scales, then management and labor

11:01

arguably get more alienated

11:03

or whatever, and then people can say, oh,

11:05

management is designed with labor. And what's

11:07

interesting is, those people who will

11:09

accept that are often folks who will

11:11

contest the idea that the state and the

11:13

nation are disciplined, but it's actually a very useful

11:15

analogy to see that they can't be aligned,

11:17

but they can also be disciplined. Right? I'm

11:19

obviously not somebody who's anti management,

11:21

but I recognize that you want equity

11:23

structures and things like that to align folks.

11:25

Right? Once you kind of realize

11:27

that the nation in the state are different. You can

11:29

reexamine all these words. Example,

11:31

multinational. Right? So multinational

11:34

for multinational corporations is actually

11:36

a misnomer. it really should be called multidurisdictional

11:38

or multistate because it's

11:40

not as if Google has like a

11:42

direct ambassador to the Catalonians and

11:44

the BaaS and the Kurds and so on.

11:46

It's actually talking to the states that

11:49

have the administrative units. Right? So

11:51

it's multi jurisdictional, multi state.

11:53

The United Nations, as we just discussed, is really

11:55

best called the selected states

11:57

because it doesn't have all the

11:59

nations of the

11:59

world. Right? And in fact,

12:01

many nations missed out in the

12:04

game of musical chairs and they

12:06

don't have a seat at the accommodations. And

12:08

now you can actually realize, oh, there's

12:10

a whole queue

12:11

of people people,

12:13

really, I should say groups. Right?

12:15

With long histories and cultures that

12:18

are boxed out of the United

12:20

Nations. And, you know,

12:22

one that's very prominent is like

12:24

the, you know, mention the Kurds,

12:26

Californians, one that's like sort of

12:28

becoming a nation or like the Taiwanese.

12:30

where there wasn't like historically this

12:32

huge necessarily difference from mainland China or

12:34

at least there's some difference in a lot

12:36

of folks in Taiwan came from

12:38

Mainland China because, you know, the Kuomintang,

12:40

the the national sources of civil war, you know, a

12:42

bunch of them moved out. Now,

12:44

Taiwan is like this that that's a

12:46

very famous example something that's sort of on the

12:48

boundary between just a group of people

12:50

and like a full fledged country that's

12:52

recognized for everybody because China leads

12:54

on people to not recognize Taiwan.

12:57

Right? That's like a famous example, but there's lots

12:59

of other things that are in this sort of boundary zone in

13:01

different ways where some countries recognize

13:03

them but not all, you know. So

13:05

it's not part of the club fully, but

13:07

it's also not totally not part of the club. It's

13:09

like in this another zone. This is a

13:11

book called Invisible Countries on this. I'm

13:13

getting at is once you distinguish nation

13:15

the state, you reexamine words like multinational. You

13:18

reexamine words like the United

13:20

Nations. you reexam words like national

13:22

security or phrases like national security.

13:24

Really, that's like federal security.

13:26

Right? Nation is often used to mean

13:28

the whole thing but it's like

13:30

federal, like the overall government. Right?

13:33

And you realize you've actually there's a base that's

13:35

being stolen, you know. The nation and the

13:37

state are different. And once you

13:39

realize nations are different, you're like, oh,

13:41

well, America actually is

13:43

not a single nation at a minimum it's

13:45

by national.

13:45

as you said, the nation in the state are not always

13:47

aligned and there is this gray zone.

13:49

Many people accept that these lines

13:51

are very concrete they're very binary.

13:53

And I think many of the examples you've

13:55

given show that they're not necessarily. And

13:57

I think that relates directly

13:59

into this idea of a network

14:01

state because if you accept that the lines

14:04

are binary, they're set. You

14:06

can't change them. There's no point

14:08

of even going into this

14:10

concept of a network state. But as you

14:12

dive into these gray

14:14

zones, and actually just to bring up one quick example

14:16

from your book is even just naming countries

14:18

that many people didn't realize have

14:20

emerged in the last couple of decades. Right? You

14:22

think back to, like, the largest countries

14:24

like, America, China, etcetera.

14:26

They have much longer histories, but

14:28

there are new countries on the map that

14:30

have happened within our lifetime. So that

14:32

was another example

14:34

of gray becoming part

14:36

of the existing

14:37

infrastructure. This whole concept

14:39

of concepts becoming variables,

14:41

I think is a very important

14:44

characteristic of our age, fourteen years

14:46

ago, there wasn't, like, fiat

14:48

currency and cryptocurrency. Right?

14:50

There wasn't a distinct it was just, like, you

14:52

know, it's currency. If you had walked into a

14:55

VC's office in two

14:57

thousand eight, two thousand seven,

14:59

and you said, I'm gonna have found a

15:01

new currency. Okay? Well, actually Peter

15:04

Teal and Paypal, they kinda tried to do

15:06

that. Okay? but they didn't really say

15:08

that that was the thing. They said, you know, it's

15:10

like online payments and so on, they they did their founder

15:12

currency. But the

15:14

fundamental issue, you know, which

15:16

Toshi later put a finger on, but it wasn't

15:18

obvious was decentralization. And then how you actually

15:20

make it decentralized back end and the blockchain, all

15:22

the stuff we had somebody actually said, oh, I'm

15:24

gonna found a currency in two thousand

15:26

seven, two thousand eight. Almost every

15:28

VC would have looked to them like they had four

15:30

eyes. And they would have said, where

15:32

you're gonna go petition like the IMF

15:34

and the World Bank? Oh, it's like

15:36

and it's deflationary to

15:38

haven't you read Econ 101

15:41

Paul Krugman proved that

15:43

deflation is bad and

15:45

you can never make, you know, that they just

15:47

quote these like priests to you basically

15:49

to say that it's impossible. It should be a joke. try

15:51

and do some act and so those people were thought

15:53

of this, Joe. But Cytoshi

15:55

figured out a new way that

15:58

literally started on a message board.

16:00

you know, like, you posted this on

16:02

the, like, the Metz Dow, cryptography,

16:04

message board. And what

16:06

still she realized that I don't think most people

16:08

realize is even if

16:10

this is implicit,

16:12

you can think of the

16:14

internet as basically like giving rise to a

16:16

new continent. Okay? Imagine an Atlantis

16:18

that just arose

16:20

out of the middle of the ocean.

16:22

And people were just taking commuter flights

16:24

there back and forth each day.

16:26

Okay? to spend eight hours in

16:28

Atlantis and sixteen hours at home.

16:30

That's really what the Internet is. You know how

16:32

I can prove that? Well, we're in it right now.

16:34

Well, right. Exactly. Like, one way of thinking about it is

16:37

ask themselves what percentage of their

16:39

time? They spend their

16:41

waking hours. They spend looking at a

16:43

screen of some kind. Okay? Whether

16:45

it's laptop, mobile phone,

16:47

tablet, you know, they're a smartwatch, something

16:49

like that. Right? What percentage of that time is that

16:51

for you, sir? I, unfortunately, have

16:53

to say it's probably, like, fourteen

16:56

hours a day, but I'm I'm probably not Something

16:58

like that. I'd say probably the average person, though.

17:01

Right? It's it's a third of their day, maybe eight hours.

17:02

That's right. So what that means and

17:05

that's up from basically zero

17:07

in nineteen ninety one. Yes.

17:09

Right? So, you know, this this

17:12

Atlantis, this cloud continent.

17:14

Right? So just to extend the metaphor, we're

17:16

taking these commuter planes

17:19

to the cloud continent fourteen hours a day

17:21

and coming back. And we're only

17:24

spending two hours of our waking

17:26

lives, in your case, on the land and

17:28

fourteen hours in the cloud. Right? For other

17:30

people, it might only be a few, like, three or

17:32

four hours, but, like, that's

17:34

amazing. Millions of people have

17:37

migrated huge chunks of their

17:39

lives to this cloud continent. Okay?

17:41

When I say billions, I mean,

17:43

like, three something billion just on Facebook. Right?

17:45

And you add all the people with smartphones and so on.

17:47

So let's say it's on there to three four billion people

17:49

in the world, half the people

17:51

of have are now spending half

17:54

their lives in this cloud

17:56

continent, half their waking hours. Okay?

17:59

Up from nothing in nineteen

18:01

ninety something. when we think about that, that

18:03

is actually a different way

18:05

of visualizing the whole thing. And

18:07

you realize the incident is actually on

18:09

par with the discovery of the

18:11

Americas for the Europeans. Right? Yes.

18:13

Of course, there were people in the Americas

18:15

before the Europeans got there. I talked about this

18:17

in the book actually that like, if you go and look

18:19

at the Bantou expansion or

18:21

the mongols sweeping across the

18:23

world, there's essentially no

18:26

ethnic group that has ever had some location since

18:28

time in memorial. They just killed

18:30

the previous folks and kinda took over

18:32

their territory or whatever. Right? So

18:35

leaving that whole part of things aside

18:37

from the perspective of Europeans, like,

18:39

quote, the discovery of the new world was

18:41

this huge thing. similarly like the folks

18:43

who went over the baring street.

18:45

Their discovery of the Americas was this

18:47

huge thing. It was this new frontier. Right? Which is

18:49

obviously thousands of years earlier. this Internet

18:51

frontier where we've migrated

18:53

to will over time

18:55

give rise to new countries just like

18:57

the Americas did. Right? The

19:00

people came there and

19:02

they didn't think of themselves as

19:05

American or Brazilian or

19:08

Mexican or Canadian or something

19:10

like that. Nowadays, North and South America have

19:12

they're all, you know, slodged into the same

19:14

grid as like the old world. Right? But

19:16

initially, they taught themselves as English or French

19:18

or, you know, they were colonists, they were settlers.

19:21

Right? They didn't identify with

19:23

the new land as primary

19:25

and the old world is secondary.

19:27

Right? They didn't think of themselves as

19:29

a Polish american or

19:31

English american. Right? That's also

19:34

just English. And that's similar to folks who

19:36

spend all of this time, this cloud continent, but

19:38

have not made the flip. Right? You're

19:40

spending the majority of your time in the

19:43

cloud continent But you're not thinking

19:45

yourself as a cloud person -- Yes. --

19:47

first. Yes. Yeah. It's

19:47

the keyword. Yes. I think one is

19:49

the keyword.

19:50

Right? Interesting thing

19:51

for you to share would be

19:53

I think many people, like the individual you

19:55

might have just described,

19:58

they, in reality, spend

19:59

a lot of

19:59

time in the cloud, but they don't actually

20:02

see the cloud or the

20:04

network as a true challenger

20:06

to something like the state or prior

20:08

to that religion. and you

20:10

talk about this as this idea of a new Leviathan. And

20:13

I think what would actually help people

20:15

wrap their head around this is is what

20:17

are some examples of where the network

20:20

did challenge the state where there was almost

20:22

like a head to head or

20:25

strong sway that the network

20:27

had in our wider world. So

20:29

one thing I've heard you talk about is Wall Street

20:31

bets, which people might think is kind of a

20:33

silly example, but there are larger ones. Right?

20:35

So so what are some of those examples?

20:37

So consternible

20:38

leviathan generalizes.

20:40

And, you know, this is the

20:42

least common word in Silicon Valley is

20:44

God. Okay? Or in technology is

20:46

God. Right? So basically, I'm talk I'm gonna talk about God for a second. So

20:48

the concept of Leviathan generalizes God.

20:51

And what it basically means is in this

20:53

context, it's like Hobbs of Leviathan and I'm just

20:56

kind of you know, taking that continent and taking a little bit

20:58

further. And the leviathan

21:00

is that all powerful

21:02

force that stands

21:04

above all men and

21:07

makes antisocial people behave in

21:09

pro social ways. Okay? So

21:12

I have introduced three Leviathan's

21:14

God's State and Network. And it was actually

21:16

gratifying I have this passage in there and one

21:18

of the footnotes by a guy named Jacob

21:21

Burkhardt from Force and Freedom. Did you see that

21:23

passage? I'm not sure. It's like footnote

21:25

n or whatever of it. There's a lot of footnotes. A lot

21:27

of links. but it it's great. Why don't you why

21:29

don't you share what that footnote is? Yeah.

21:31

So here is here is this pretty cool

21:33

footnote. I think it's cool. At least it's kind of thing I think

21:35

it's So this is from a book that was

21:37

written almost two hundred years ago,

21:39

okay, Force and Freedom. And

21:42

I I found this after I'd written much

21:44

of the book or whatever just like like, you

21:46

know, he's flipping through old stuff and I was like, and he identifies

21:48

essentially three forces in the world. He calls

21:50

him state, church,

21:52

And

21:53

then the third force, he calls

21:56

culture. Right? And this is way before the

21:58

Internet or anything like that,

22:00

obviously. And by the state and church are kind of

22:02

you can kinda guess what those are. Right?

22:04

Culture, he defines as all the peer

22:06

to peer interactions between people.

22:08

he didn't say quite peer peer, he says, I was gonna give

22:10

the exact quote. Our theme is the

22:13

state religion and culture in their

22:15

mutual bearings. We are fully aware

22:17

of arbitrariness of this division into three

22:19

powers. The division hour is a mere device to

22:21

enable us to cover the ground. Indeed, any

22:23

historical subject must proceed this way.

22:26

Right? The three powers are supremely heterogeneous to each other

22:28

and cannot be coordinated. And even if we were

22:30

to coordinate the two constant state and religion

22:32

culture would be still something essentially different.

22:36

The same religion, the expressions of political and

22:38

metaphysical need may claim authority at least over

22:40

their particular peoples and indeed over the

22:42

world. For our special purpose however,

22:44

culture which means material and spiritual need in

22:46

neuroscience is the sum of all that has spontaneously

22:49

arisen for the advancement of material life as

22:51

an expression of spiritual and moral life all social intercourse

22:53

technologies, arts, literature, and sciences is a

22:55

realm of variable, free, not

22:57

necessarily universal of all that cannot, late

22:59

claim, to compulsive authority.

23:02

Wow. Okay. That's

23:04

like two hundred years ago. And what he's talking

23:06

about in modern language would be

23:08

culture is the network of

23:10

peer to peer volitional interactions

23:13

between people as opposed to

23:15

the top down and positions of

23:17

the church and the state respect And this

23:19

relates to constant leviathans. So the three leviathans I describe as

23:22

god's data network. Right? And

23:24

how do you think about this? So in the eighteen hundreds,

23:26

what's the most powerful force in the

23:28

world? God. Why? You do something wrong? You

23:30

steal? God will punish you. Right? That's why

23:32

you didn't steal. Okay? By the nineteen

23:34

hundreds, enough people didn't believe in God.

23:37

you had nature writing about how God was dead and,

23:39

you know, essentially the the basis for a

23:41

lot of civilization. You had this you

23:43

went from this decentralized law enforcer

23:46

who had hit you with lightning bolts. If you did something bad, people

23:48

actually believe in that as like the super cop.

23:50

Right? People didn't believe in that anymore since

23:52

said, you had the rise to a greater

23:54

extent of the state filling that

23:56

void. Right? The uniformed police

23:58

force is the poison blue in the

24:00

extremist. It's the Soviet Union of a

24:02

totally godless state and so on. And in fact, that was huge

24:04

collision in the twentieth century between the Soviet

24:06

Union, which is a pure state, and America,

24:08

which is a god state combination

24:10

like marine corps of mid century would

24:12

say, you know, for god and country. So in the nineteen hundreds,

24:14

why don't you steal because the state would punish you?

24:16

Right? Even if you didn't believe in god,

24:18

the Boysen Blue would get you. That's why there's so

24:20

many police procedurals on TV. Right?

24:23

Because the state is portrayed

24:25

as omnipotent domestically. all of the things, even like

24:27

a small little boy, you know,

24:30

who knows what a cop and a robber is.

24:32

Right? They can see the uniforms and

24:34

so and so forth. They are

24:36

also, you know, taught,

24:38

oh, the US military abroad

24:40

can go and invade any country, blow anything

24:42

up. It's big, bad, all powerful, blah,

24:45

blah. Right? And until I think about twenty nineteen,

24:47

those kinds of things, while Ricky,

24:49

they like, people kind of believe in them.

24:51

And now what we're getting into is

24:53

the third Leviathan, which is a network. And

24:55

what that is is that's a cryptocurrency network,

24:58

that's a social network. And now the

25:00

network, you have a third thing, you you don't

25:02

seal because network won't let you. either you'll get canceled

25:04

by the social network or

25:06

you will not be able to take it

25:08

because the encryption prevents you.

25:11

And it is a third way of thinking about it. God would

25:13

fight you eighteen hundreds. State will punish

25:15

you nineteen hundreds. Never won't let you

25:17

two thousands. Okay? These are three

25:20

different theories of a prime mover. What

25:22

is the most powerful force in the world? Is

25:24

it God? Is it the US military?

25:26

Or is it encryption? All kinds of

25:28

political arguments. moral arguments,

25:30

cultural arguments, social arguments,

25:32

a lot of them reduced down

25:35

to who is

25:36

my Leviathan? Like, what

25:38

is what is the final thing that I'm

25:41

invoking that says

25:43

basically, like, My dad can be of your

25:45

dad. Like, my god is stronger than your

25:47

god. What do I think of as that, you know,

25:49

thing? And of course, there could be conflicts

25:51

within people of God. Right? And there can be that have

25:53

mentioned between people of God and people of the states.

25:55

Christians and the state like the Soviet

25:57

Union persecuted people of God.

26:00

there can be conflict between people, the state, and people of the network, like

26:02

the antitrust cases against

26:04

tech companies or the

26:07

state department versus tornado cache and and so on and so forth.

26:09

Right? Once you kinda see this, it's actually like

26:11

a vocabulary for parsing the world.

26:14

And just to generalize this this is a part that's not in v one

26:16

of the book that's coming in v two.

26:19

So with, you know, God,

26:20

we're familiar

26:21

with the concept of Atheus

26:23

monothesis, polytheus. Right? Okay.

26:26

So Atheus doesn't believe in God. There's

26:28

also agnostic and so on which is for our

26:30

purposes. Right? So Atheus

26:32

monotheus smelthias smelthias. A monotheus doesn't believe

26:35

in God. A

26:38

monotheus like a Hindu believes in

26:40

many gods. zero, one,

26:42

one and we can now

26:43

generalize this to the a

26:46

status, the monostatist,

26:48

and the polystatist. And

26:50

let's say a coin ish, monoclonist,

26:53

polycoinist, or you might say a

26:55

numerist like numerist is like a sodium

26:57

coins. Right? So let's say A status. A

26:59

status is an anarchist. Okay? They don't

27:01

believe in the state at all. So it could be a

27:03

crypto anarchist. It could be in an archicomist. They

27:05

just don't believe in the state of all. Okay.

27:07

Zero states. monostatists. That's

27:09

somebody who believes that their empire should run the

27:11

world. Right? Like a

27:13

national greatness, NeoCon, or like

27:15

somebody who thinks China should dominate the whole

27:17

world or whatever. Right? Or back in the

27:19

day, like, someone the Roman Empire, the Soviet

27:21

right, Soviet imperialist. That's like a monocetist.

27:24

Our state should run everything. And

27:26

then the poly sadist is

27:28

like somebody who's into competitive

27:30

government, digital nomads. Okay?

27:32

Switching between countries, passports. Okay?

27:35

Now, you can also apply the same framework to

27:37

coins to the network. A

27:39

coin is the no coiner. They don't believe in

27:41

coins at all. They hate coins. Right?

27:43

you know, web three critics blah

27:45

blah. Okay. Monoclonus is

27:48

like the bitcoin maximalist or

27:50

a maximalist of anyone they think their

27:52

digital currency or their network

27:55

is just the number one, everybody. Right? The

27:57

Polycoinist, I mean, there's

27:59

actually this

27:59

PolyChain. Right? There's multi coin.

28:02

Those

28:02

are literally funds that were

28:05

set up on the premise that multiple

28:07

coins will exist, which was controversial

28:09

at the time that they set it up. Right? Like

28:11

in, you know, I think PolyChain was set up in

28:13

twenty fifteen ish. They're about multi client on the

28:15

exact time frame, but around that time. I'm just naming those

28:17

as two funds which are poly chain

28:19

and multi coin. So the poly coin is,

28:22

okay? Or the polypneumist is another term.

28:24

Numus is like numerism is the

28:26

study of coins like pneumatic. Right? They believe multiple

28:28

coins. Now here's what's interesting. Once you

28:30

have this cleavage of the world, right,

28:32

you can realize so that's like

28:34

three times three times three, many

28:37

ideologies can be further slotted

28:39

into like combinations of these.

28:41

For example, you have somebody who's

28:43

the atheist, monostatist,

28:45

a coin is. Right? That

28:47

is somebody who is like your

28:50

secular East Coast establishment

28:52

person who doesn't believe in God, believes

28:54

in the US government and especially the regulatory

28:56

state and hits coins. You know, crypto

28:58

energy is bitcoin, Maximus. they're an

29:00

atheist often. Sometimes they're a monothesis

29:02

or a polytheist. Right? Let's say, often an

29:04

atheist, they are an

29:06

a status. they don't believe the state should

29:08

exist, but they're a mononomist.

29:10

okay Okay?

29:11

So that's a real conflict

29:13

that's gonna happen between the

29:16

atheist, monostatist,

29:18

a minutes newist, and the

29:20

atheist, a status, mononemus.

29:23

Right? that's like that's a collision because they

29:25

they have different gods. Right? That's gonna be,

29:27

like, the US government and the

29:29

US dollar versus Bitcoin is as

29:31

big a clash as, like, you

29:33

know, I don't know, Christian Muslim was during the crusades. I

29:35

mean, that's a that's a clash of, you know,

29:37

or like the Soviet University US. Those

29:39

are two fundamentally different class of

29:42

ideologies. I

29:42

think too many people that

29:45

would seem a little outlandish, you

29:47

know, being the devil's advocate, what

29:49

comes to mind is these are all

29:51

collective terms or

29:53

structures or things that we all believe in

29:55

or don't believe in. Right? Some level of

29:57

belief depending on where you fit in

29:59

that matrix. And for many

30:01

people, let's say if we use God as Bill

30:03

Bison, the repercussions

30:05

as you're saying of, like, not behaving in certain

30:07

ways. For them, the people who believe in that,

30:09

they're like, I'm to hell. That's like a a

30:11

lot of force or at least imagined force.

30:13

The same thing is true with the state. Yes. Right?

30:15

So if if I do something wrong, I'm going

30:18

to jail. a lot of force What's interesting about the

30:20

new Leviathan that you're describing

30:22

as the network, there's

30:24

force, but it's it's not

30:26

approach the same way if if that makes sense. Like,

30:28

it's it's actually

30:29

-- Yes. -- a lack of

30:30

force in some ways. It's it's cryptography and

30:32

it's using the network to

30:34

enforce certain things, but

30:37

without the same force that you imagine

30:39

through god or the

30:41

state. And I think that's maybe

30:43

why it's hard for some people to imagine this being a

30:45

Leviathan because a lot of people view the

30:47

network as what what is just a bunch of

30:49

semiconductors and servers and

30:51

and bytes running back and forth. But I think that's why

30:53

it's important to think through, like or to to

30:55

give these examples of where the network has

30:58

actually been incredibly strong.

31:01

relative to other lebizans?

31:03

Well, so here's

31:04

the thing. It's basically one of the things

31:06

I touched on there and I talk talk about more

31:08

in the book is fusions of

31:10

all kinds. Right? So you introduce

31:12

these pure forms of like god's seed network, and

31:14

you can have fusion. So I mentioned

31:16

like god plus state is the

31:18

American state. Right? And there's different

31:21

versions of network plus state. One

31:23

is a particular version of a book,

31:25

the eponymous network

31:27

state, which is a fusion of network

31:29

and state that you can think of as

31:31

the network. Is the state

31:34

like it's it's the government?

31:36

Right? to network is the nation that underpins the

31:38

state because you have a social network online that

31:40

is actually giving legitimacy as opposed

31:42

to a physical physically based

31:45

nation where everybody lived together It's a

31:47

digitally based nation where it thinks the same,

31:49

and they're aligned that way rather than the same

31:51

language. They have the same culture, but there's culture

31:53

is online. Different ways of

31:55

parsing But to your point, in terms of

31:57

force, well, one fusion

31:59

of a kind of network state

32:01

is sort of a China's bill.

32:03

Right? And actually with the

32:05

US Cellular's building. And so you connect

32:08

the network to drones. Right? You

32:10

connect the network to robots. for this physical

32:12

actuators, and that's, like, one way of thinking

32:14

about it. So now that absolutely is force

32:16

in the physical world. And in fact, that's already being

32:18

deployed and and so on. Right? Another

32:20

way of thinking about it is, which is already there. So

32:23

this thread that I just pasted into chat. So

32:25

this is a while back, but basically,

32:27

I was thinking about digital power and how

32:29

to articulate it. Digital power is not

32:32

really soft power. Okay? If you're a

32:34

deep platform and seeing all your money

32:36

frozen, that's more than just

32:38

influence. But it's also not what we

32:40

traditionally think of as hard power because it's

32:42

invisible. It's intangible. You can use it

32:44

on a hundred million people. There's

32:46

no optics. There's no, like, fireworks,

32:48

something blows up, there's no nuclear explosion.

32:50

Like, the human brain is not

32:52

trained to react to the use of digital power.

32:54

It's not dramatic. In that way, it's like a

32:56

bit slipping on the server. Right? So I was

32:58

trying to think about how to classify it, and then I

33:00

eventually was like, oh, you know, here's a four

33:02

part classification. Analog

33:05

soft power. analog hard power,

33:07

digital soft power, digital hard power. Right?

33:09

You could also say physical soft power, physical

33:11

hard power, digital soft power, and

33:13

what is that? Analog soft power, that's culture

33:15

and influence. Okay? Analog heart

33:17

power, that's bombs and bullets. This is the

33:19

classic soft heart power. Digital

33:22

digital soft power is ranking

33:24

and recommendation. Digital hard

33:27

power is de platforming, freezing, and seizing.

33:29

And so the key differentiation is Soft

33:31

power is probabilistic and hard power is

33:33

deterministic. Soft power you're

33:35

persuading, hard power you're

33:37

compelling. Okay? So now you start to

33:39

actually see, okay, that's another

33:41

way. Besides the obvious thing of drones and

33:43

so on or, you know, that's, like, compelling in a

33:45

very gun gunpoint that you kind

33:47

of weigh. you can compel in a

33:49

different way with the network where

33:51

you lock people out. Either

33:53

of their accounts online or

33:55

they can't access the building. Their keycard is disabled. Right?

33:58

Their funds are frozen. Their

34:00

account has permissions

34:02

reduced. Their suspended, etcetera.

34:04

that's actually a very significant punishment. In

34:06

China, for example, like people, their COVID codes

34:08

are used, the red and green zones are

34:10

often marked as like COVID red,

34:14

and then they can't travel. Their WeChat doesn't work.

34:16

And, yeah, in theory, like, it doesn't actually

34:18

affect their physical body, like, they're not

34:22

hurt. Okay? But they're basically unpersoned. They are

34:24

disconnected from huge chunks of the Chinese network. They have

34:26

to rely on a friend to do

34:28

things and so on. They are just

34:30

much less

34:32

like, independent that they were. Right? It's similar, though it's

34:34

not quite there to, like, exile. Like,

34:36

when you're exiled from the community and, like, grease sort

34:38

of. Right? So it's, like, digital exile.

34:41

think digital power should not be

34:44

underestimated. Even if it's like flips on

34:46

bits on servers, it's

34:48

growing in power. Drones are the

34:50

most having example, but locking you out of an account or digital soft power

34:52

like downranking you. These are big things

34:54

that can totally crash somebody's, you

34:56

know, company can make

34:58

or break you socially, etcetera,

35:00

etcetera. Yeah. And I think it's

35:01

it's directly correlated to our

35:04

dependency on

35:06

the network. So as more people go online, as more people spend more time

35:08

online, as we talked about before, if people are

35:10

spending eight hours of their day online, that's a

35:12

third of their

35:14

living hours.

35:16

half of their awake hours in which that

35:18

is important to them to some degree. And as

35:20

it gets more integrated with work, let's

35:22

say, that's that is a meaningful

35:24

is a meaningful sense of power

35:26

to your point. One interesting aspect to consider as well

35:29

is that with all of these forms of

35:31

powers, there's tears. Right? So,

35:33

like, you could think of the most extreme

35:35

tier of hard analog power

35:37

or force is like the

35:39

dust penalty. Right? but you can actually dissuade someone

35:41

from doing something by saying, you know,

35:44

applying a night in jail or maybe it's a month in

35:46

jail. Right?

35:48

So there's tiers to it. And I think there's going to be as well power.

35:50

Right? Where it's not just the

35:52

extreme version at all times, but

35:54

also it can be where

35:57

you lose all access to your digital life, and that's that's quite significant. And the

35:59

thing is your

35:59

digital life just becomes your life, you know,

36:02

in a sense of it's your login to

36:04

all kinds of things. Right?

36:05

Yeah. Why don't we return to this idea that, you know,

36:07

it sounds like this digital power

36:10

concept may not resonate with

36:12

everyone, but wanna give a couple of these examples because they really resonated with

36:14

me that you've shared in in your book and

36:16

otherwise. So, I mean, I mentioned Wall Street Pets. It's

36:18

maybe a

36:20

silly example. Another example of a network

36:22

utilizing some degree of force

36:24

is Amazon H22 That's an example

36:26

that you gave where actually like

36:28

some concept

36:29

Yeah. Persuasion, though, that's the sport. Exactly.

36:32

Yes. So that that so, you know, there's

36:34

convincing and there's compelling. Yeah. I mean, something else

36:36

that you've

36:36

said is your immigration policy is your firewall.

36:39

you've also shown how networks can be

36:42

utilized in unison with

36:44

humans, of course. So another example that you've

36:46

given is the power

36:48

of twelve Instagram engineers

36:50

to beat twelve thousand from Kodak.

36:52

Right? So

36:52

Kodak. Yeah. Not not an example. that's

36:54

original to me, but yes, it's an important example. Right. And so think it

36:56

it's it's important to recognize how

36:58

the network does change

37:02

the game. it changes people's ability to participate within

37:04

these systems, to challenge the state, to

37:06

challenge previous leviathans. And

37:08

I think with that

37:10

understanding, let's return back to the idea

37:12

of the network state and

37:14

actually building these

37:16

new lines that many people assume are

37:19

very much set. Yeah.

37:20

So so I actually I start off

37:22

the book with this because people

37:25

skim nowadays, and so never

37:27

state in one informal sentence is a

37:29

highly aligned online community with a capacity

37:31

for collective action

37:34

that crowd from surgery around the world and eventually gain symptomatic recognition

37:37

from preexisting states. That's like an

37:39

informal sentence. We just kind of

37:42

describe that. highly aligned online community. Okay? So it's not

37:44

just, you know, a Game of

37:46

Thrones Facebook group. Okay? Such as

37:48

people who

37:50

are there pop popcorn. They are highly aligned. They all think of themselves

37:52

literally part of the same

37:54

community. Okay? And they've

37:56

got leadership,

37:58

and they've got an org chart of some kind and they've got

38:00

probably membership dues and a cryptocurrency

38:02

and, like, this is at

38:04

the top of their identity stack.

38:06

Okay? You know, what you do is not what you are. You may not think

38:09

of yourself as a left hander, even if you are left handed. It used

38:11

to be I'm Jim. I'm thirty two.

38:13

I like the Steelers. live

38:17

in Philly, whatever, something like that. Those are less and less common

38:19

as bios. Right? Instead, what people do

38:21

is they put pound

38:24

x you know, like pound BTC or

38:26

pound this movement, pound that movement in their

38:28

vials, they'll basically put their tribal

38:32

flags there. both their attack and their defense. Like, you you kind of

38:34

instantly know where somebody's coming from when you know

38:36

their tribe. Right?

38:38

There's various shibblets,

38:40

various words and things people put in

38:42

there that identify what gang they're

38:44

informally a part of, what social they are self

38:46

identifying with and that probably

38:48

also accepts them to some extent. They are

38:49

also ready to sort of defend that

38:52

identity. They're putting it out there on their sleeve

38:54

and often the attacks

38:56

and somebody We'll attack them. Lal, another Bitcoin

38:58

nutter or something like that. You know, the people

39:00

will say that. Right? Or,

39:02

oh, you know, how

39:04

is mess you know, some people call it Ethereum math. Right? And and, of

39:06

course, there's a zillion other things like this. So it's

39:08

a, you know, then the next

39:10

step is it

39:12

has a capacity for collective action. Right? So it's aligned and it has common

39:14

beliefs. Capacity for collective action, that's part of a

39:16

segment with respect to the org chart and the

39:19

leadership. So again, this is an extremely

39:22

selective filter. The vast majority

39:24

of online communities do not

39:26

have a capacity for collective

39:28

action. Why? because

39:30

I can you know, if you look at

39:32

a a Twitter following. Right? Let's

39:34

say somebody has a million followers on

39:38

Twitter. Okay? how many likes does this does a typical tweet

39:40

get? Actually, a pretty

39:42

good thing would be if you have a million

39:44

followers to have a thousand or

39:46

two thousand That's actually pretty good. All in the order of point one percent to point

39:48

two percent are actually engaging with it. That is not

39:50

what I'm calling a capacity for collect fraction. That

39:52

is a capacity for

39:54

popcorn action. Okay?

39:55

Yeah. It's a cool

39:56

three d boot like this. Right? Okay. A

39:58

capacity of collective action would

39:59

mean that when you put something out

40:02

to a thousand people, you get a

40:04

thousand likes. you don't

40:05

get fun. Right? You put

40:07

something out to a million

40:09

people. You get a million people

40:11

hitting the button, basically. that's

40:13

what a capacity for collective action is.

40:15

It's something where the group moves

40:17

as one. If you go and

40:19

look at the web two Internet.

40:22

Okay? You look at Hacker News, and you

40:24

look at Reddit, and you look at Twitter, there's an

40:26

aspect of it. You can't unseat once you

40:28

see it, which is

40:30

it's entropy. It's just thirty random links. Every time you

40:32

refresh the page, thirty random

40:34

links. Okay? And what that

40:36

means is, intellectually,

40:38

you're like, oh, hey, you

40:40

know, a shiny object. Oh, there's another red Dead

40:42

direction. No. That's cool. Right?

40:44

And so what happens if

40:46

you randomly

40:48

move, like, a meter in this direction and randomly in this direction and then

40:50

randomly in this direction. Right? Do you

40:52

make any progress?

40:54

Probably not. Right? That's

40:56

actually what's like the so called spherical random

40:58

walk. You you just kind of just

41:00

drift away from the origin or what

41:02

have you. And By

41:04

contrast, if you have a focused direction and

41:06

you're like, I'm learning

41:08

AI and I'm learning this today and

41:10

that tomorrow and that and that that then you're

41:12

moving in direction. Now there's ways of

41:14

combining these by the way. I'm not saying

41:16

serendipity is always bad. The

41:18

flagella of E. coli famously that have

41:20

the the tumble mode where it, like,

41:22

randomly seeks it out, and then it

41:24

finds the nearest food supply and then it

41:26

runs along that thing.

41:28

It's like tumble and

41:30

run. Okay? to combine the random search with a

41:32

directed search, and you get a good combination.

41:34

Okay? Similar in somebody who likes to

41:36

cast grade, the center, or something

41:38

like that. And so

41:42

the point being though that if

41:44

you're just doing lots of entropic stuff, you're

41:46

not making progress on

41:48

an axis. And if you have lots of people who are

41:50

just online to hit random

41:52

buttons to doomscroll to

41:54

whatever to just click this and

41:56

that and popcorn. There's

41:58

no compulsion. They're that's

42:00

why they're wasting time. Right? They're

42:02

turning an hour of just looking

42:04

at the screen into like a few clicks. It's

42:06

a very low efficiency conversion of their intellectual energy

42:09

into output. The capacity of collection changes

42:10

that. So now you have this

42:14

mid organ. Okay. What's the third part of the definition? I never say it's a highly aligned, online

42:16

community with a capacity to collect of action that

42:18

crowd funds territory around the world.

42:20

Once you've

42:22

gotten people, a thousand people in a Discord or something like that. It might

42:24

turn out to be, by the way, another kind of app

42:26

because, for example, Discord

42:30

doesn't, like, enforced tasking.

42:32

Once you've done that, once you've shown a

42:34

capacity for digital collective action, once you can get

42:36

a thousand people to like something, which is hard,

42:38

by the way, very hard to do that. Okay?

42:40

and you might need new kinds of apps. Once you can do

42:43

that, well, you can get to a very high bar

42:45

of collective action, which is not just getting a

42:47

thousand people to like something, but

42:50

thousand people to buy something. And in fact, not just buy any old

42:52

thing, but to crowds and territory and move in together.

42:54

I'm gonna pause you there because Well, obviously, I wanna

42:56

get to the rest of the

42:57

definition, but but that is

43:00

something that, you know, is quite the leap.

43:02

Of course, yes. I agree with you getting a

43:04

thousand people to do anything even if it's

43:06

liking a tweet. That that is

43:08

hard, but get me from

43:10

there to crowd funding

43:13

land. Sure. Totally.

43:14

So I had this this

43:16

article calledsoftware's Jiragnes in the world, you know, almost ten years

43:18

ago. And I had a little table in

43:20

there, which I will see if I

43:23

can find it. classification of cloud formations taking

43:26

physical shape. Right? It's a very

43:28

picturesque term. A cloud formation, I

43:30

define it as a group of people

43:32

who meets on the

43:34

Internet. Okay? And taking

43:36

physical shape means they

43:38

materialize out of the cloud and

43:40

actually all come and aggregate

43:42

in person. And even in twenty

43:44

thirteen, almost ten years ago, I was

43:46

tracking these cloud formations, taking

43:48

physical shape, and here's a table. Right? There's

43:50

scale and duration. Right? Scale is the

43:52

number of people and duration. Okay? So

43:54

the simplest is two people, scale is two. Duration

43:56

is one day. So coffee

43:58

with a LinkedIn contact. Okay?

44:01

You someone LinkedIn, get off with them, but that was I don't

44:04

know if anybody's using LinkedIn today. No

44:06

offense to LinkedIn people listening to this. Alright?

44:08

But that was happening back in the day.

44:10

Alright? So

44:11

though Alright. Coffee

44:11

and LinkedIn content. Then two people for a

44:14

month is you meet a remote

44:16

engineer, you interview them online, okay, and you

44:18

bring them in for

44:20

an on-site. Okay? Come in for a month. Okay? Two people for a

44:22

year, that's like match dot com,

44:24

even in nineteen ninety I mean, obviously, now there's

44:26

Tinder or

44:28

whatever. But point is, like, mass dot com is, like, the beginning of that, let's say, nineteen

44:30

ninety five. Two people for ten years,

44:32

that's, like, eharmony dot com, like, oh,

44:34

they come together. You know,

44:36

it's, like, very high intention form a, you know, match from one of the bottom of

44:38

something. Right? Okay. Ten

44:40

people for a day. That's like hacked on. Obviously,

44:42

we've been going on for a

44:44

while. Right? ten people for

44:46

a month. That's like a

44:48

data science type program

44:50

or one of these sort of immersive courses where

44:52

people come and meet up. They take a course for a month

44:54

that comes ten people for a year, those are

44:56

like packer houses, which have started in the late

44:58

two thousand, you're probably aware of some people

45:00

who have kinda moved in

45:02

and just get a big house. And the

45:04

the downside is communal living, but

45:06

the upside is communal living. So the

45:08

downside is there's other

45:10

people around. soon as total privacy or whatever. But the upside

45:12

is you get a much nicer house than

45:14

you would otherwise, like, economies of

45:16

scale get really good quickly

45:18

with house there's

45:19

also the digital nomad version of that too.

45:21

Right? Where you don't actually have communal

45:23

living, but kind of as you talk about

45:25

cloud first land last, the digital nomad

45:27

hubs like Chengoo or Lisbon or Shanghai are equivalent

45:29

of this. Right? Like, I would put them probably on

45:31

the, like, ten thousand

45:34

people for couple

45:36

years.

45:36

Well, it's funny because people flip

45:38

through them, but they don't

45:40

stay it's interesting. It's one of those things where

45:43

it's like, maybe the ten thousand people in the cloud,

45:45

there's like a hundred people who are in that

45:47

location at any one given time or something like

45:49

that. Right? That community is kind

45:51

of there and then it's gone. It's like a cloud

45:53

like drifting through an area versus them

45:55

all actually being pressed. Right. Well, I think

45:57

so that's an interesting perspective.

46:00

I I

46:00

would say that's true for some of them. I would also say that some

46:02

of them are actually quite stable. And

46:04

there are many nomads

46:05

that spend, you know, many years in in

46:07

a given place,

46:10

if not, their intention is for But they're but they're but

46:11

ten thousand digital nomads

46:13

in one location while it's good

46:15

and it's on the kind

46:17

of track towards this is not what I would call a

46:19

cloud formation unless all of

46:22

those ten thousand people are basically friends

46:24

with each other. I see. Okay. And this is

46:26

actually this is an important concept, but one

46:28

of the things I developed and this will be more more in

46:30

the v two. I've got a lot of graphics and stuff.

46:32

Quantitative definitions of

46:34

terms from that we've been using sort

46:36

of verbally for a long time. Right? So for example, what is a

46:39

nation? There's like,

46:41

like, a dozen definitions that I've gotten in the book or something on

46:43

that order. You know, somebody's like, it's got a

46:46

common language. Oh, no. It's it's a the

46:48

folks who lived in a territory or no. It's a common

46:50

answer street. or you have,

46:52

like, for non definition, which I really like. It's,

46:54

like, a group of people who

46:56

have done great things and wish to do great

46:58

things together. My definition is it's a

47:00

densely connected sub graph in a social

47:02

network. What exactly does that mean? A

47:04

graph in in the math sense, it's like

47:06

nodes and

47:08

edges. Right? Density connected means all

47:10

nodes are connected to all

47:12

other nodes. For example, if you have

47:16

four nodes as

47:18

a subset of a giant graph that has

47:20

a hundred nodes. And those

47:22

four nodes form a complete graph

47:25

they're densely connected if each of those four

47:27

is connected to the other three. You know, it doesn't have to be

47:29

a fully complete graph, but it's just a densely

47:31

connected graph that's much closer

47:33

to complete graph then a random group of ten thousand digital

47:35

nomads might have no connections between themselves. So

47:38

that's not what I define as a cloud formation.

47:40

This gets to our original

47:42

thing about people are used to

47:44

using the Internet individually

47:46

and informally collectively, and

47:48

this is now formally and

47:50

collective. If you have a densely connected

47:52

sub graph, amidst a bunch of loosely connected nodes.

47:54

That's a natural unit of

47:56

association where these folks

47:58

should be capable of

47:59

self government. Right? Because

48:02

they all kinda know each other. My my

48:04

friend, Jan Talin, who's, like,

48:06

early Skype engineer, you know, very

48:08

prominent investor, and so I'm he had a really

48:10

amazing take on this, which I'm gonna

48:12

call, this is this is Jan's phrase. He's like,

48:14

Balaji, you know, he heard paraphrasing me. He's like, you know,

48:16

I've been early to three things

48:18

in life. AI,

48:19

the i crypto,

48:20

crypto and

48:21

dance. I'm like, dance. Now,

48:23

so he's like

48:24

the investor in deep mind. Right. Right.

48:27

and, you know, we both in early on a bunch of these

48:29

steps. So say, Ankur, those I'm like, dance, what

48:31

do you mean? And so

48:33

what he meant was

48:36

basically that when you see you probably see this in movies, but maybe you see it

48:38

in real life, you know, let's say

48:40

there's a a couple who's really

48:42

good at dancing. Right? What

48:44

happens

48:44

is they're,

48:45

like, locked on, you know, like a

48:48

very choreographed kind of thing. The

48:50

entire dance floor kind of clears

48:52

back and admiration and looking at the coordination of these two

48:54

people. People like seeing

48:56

other people moving in formation.

48:58

Or put it in another

49:00

way,

49:02

when

49:03

you can get a group that is very highly aligned as

49:05

demonstrated by the fact that they've even if you're not constantly thinking

49:07

of, oh, they must have practiced a lot for that.

49:09

Right? Their motion is pleasing

49:11

to the eye. But they have

49:13

they are obviously self sacrificing, like, one person's elbow

49:16

is not where their person's faces at that time.

49:18

You know? They -- Right. --

49:20

they have coordinated

49:22

the whole thing. Yep.

49:24

k? That is now something that Gartner's

49:26

respect from

49:28

the outside because that's a unit. It's a coordinated

49:30

unit. You can visually see it's a coordinated

49:32

unit. These people are all

49:34

acting together And

49:36

so as such, because they respect each

49:40

other, they gain respect from the

49:42

outside world. So

49:44

to a non obvious

49:46

extent, this community, if it's

49:48

highly aligned, if it can do

49:50

a dance, Digitally.

49:52

Right? Where it's moving in

49:54

formation. That group, now you start

49:56

thinking of it as a unit. and

49:58

you both respect them and

49:59

you don't

49:59

wanna mess with them. If I have to rephrase another

50:01

way, it's AI, crypto, and

50:04

social. Okay. Okay? Lock me through

50:06

because social technologies allow

50:08

people to coordinate. Yes. Right?

50:11

Messaging apps, discords, you're allowing people to do this digital dance. One way of

50:13

kind of putting it all together, have you seen the movie

50:16

transcendence? I haven't. With Johnny Depp? No. No.

50:18

Okay. So I'm gonna spoil it for

50:20

the viewers. SPoilers. Okay.

50:22

Go watch it. Alright. Fine. Transcendence is a

50:24

awesome movie. It's underrated. I think I I think

50:26

it's a pretty good movie. There's a lot stuff like this

50:28

that I like like this. are a good sort of heavy. They're just

50:30

good sci fi explorations of, you know, potential future.

50:32

Even if they're a dystopian, which we can

50:34

correct, by the way. On that note,

50:37

Digression, digression. Many Hollywood

50:39

movies are dystopian because they steal

50:41

a base. They implicitly assume the

50:43

present is okay. and then some tech guy came in

50:45

and ruined it with their autonomous robots or

50:48

something. They messed it up. Oh my god. We had this

50:50

good ed and going, and they came to the garden ed and

50:52

spoiled it.

50:54

Right? And the alternative framing is that the present is

50:56

dystopian, and there's like a

50:58

few founders that just might

51:02

be able get us out of the situation if only they can get etcetera,

51:04

etcetera. Right? And that's a different framing

51:06

of, like, implicitly, is it is

51:08

a present badder than you ever. Okay. Point

51:12

is transcendance is, I don't think it's this opiant. Hence, it being this

51:14

opiant early on, but then it actually gets more

51:16

positive. One of the things in

51:18

it is you have Johnny

51:20

Depp is this AI

51:22

that coordinates all these human beings.

51:24

And I thought it thought provoking because

51:26

It extends the concept of like what you can do

51:28

with smartphones in terms of coordinating humans

51:31

to another level. Right? Well,

51:33

it's not just, you know, let let's say you're an

51:36

alien looking down at the earth and

51:38

you saw, like, somebody walking in New

51:40

York. Right? And then they

51:42

suddenly take a right angle turn and,

51:44

you know, they're just going in a totally different

51:46

direction. Why? Because somebody in

51:48

Hong Kong hit some

51:50

keys to send them a text

51:52

message saying, oh, no. Actually, the office is

51:54

down the street the

51:55

other way. Okay? that logic,

51:56

like, if you're some alien observing this, I mean,

51:58

there's an obvious logic to to from our

52:01

perspective. But, like, it seems like a

52:03

very subtle signal that moved from this person over

52:06

here. Like, backing that out is, like, really

52:08

difficult to figure that out.

52:10

Okay? But in the

52:12

network,

52:12

it's very visible. It's like, oh, this person

52:14

coordinated a certain person. They did a dance.

52:16

Okay?

52:17

And if you take that

52:19

up and it's I send a text

52:21

message every day or or hour or something

52:23

like that. We're sending packets.

52:26

Now you can have a bunch

52:28

of robots dancing in unison like this.

52:30

Right? Or a bunch of people dancing in unison

52:32

or some combination of people in robots, people

52:34

in servers.

52:36

Okay? and you can coordinate.

52:38

And so transcendent shows

52:40

like essentially many becoming one. Right?

52:42

This is actually something else that's going in v

52:44

two of the book. there's at least three different ways of

52:46

thinking about many becoming one. There is

52:49

let's call it There's democracy,

52:51

right, which is and

52:53

aggregation and election, many become one

52:56

because you all vote and then the decision

52:58

and it goes out.

53:00

There's markets where many

53:02

become one where you have an order book,

53:04

you have supply and demand, let the market

53:06

decide. Right? You get a price. And there's

53:08

actually a third version which

53:10

is harmony. Right? Many become one

53:12

because many actually become a single

53:14

organism. In the case of network

53:16

states

53:16

though,

53:18

you go from, yes, I understand this collective unison

53:21

online. What would be

53:23

the reason that

53:23

someone would want

53:26

to go from that online digital collective unison

53:28

to doing that in the physical

53:30

world. One point I make in

53:31

the book is, look, obviously, I love

53:33

digital space and so

53:35

and so forth. this is something by the way, which

53:37

I found, the v three is

53:40

hard to communicate because people will

53:42

hear something and they'll immediately bucket it into v

53:44

one or v two, but

53:46

not v Okay? So cloud first,

53:48

land last but not land never.

53:50

Okay? Why do I see this?

53:52

Whenever I

53:54

talked about the network state or something like that. People

53:56

would hear one of two things.

53:58

They'd either hear, oh, it's

54:00

totally digital. Yeah. That's great. You

54:04

the physical world doesn't matter. We're all online, cryptocurrency,

54:06

Internet, blah, blah, blah. Or

54:08

that here,

54:09

it's purely digital.

54:10

You more on,

54:12

like, you know, humans are still physical beings. We're still guns.

54:14

They're still buildings. They're gonna eat

54:16

online. What are you, you know, you're an idiot.

54:18

Like, another tech guy, stupid tech robot.

54:22

Right? Okay. So you'll basically hear that it's purely

54:24

digital, but it's not

54:25

purely digital. It is a

54:28

cloud formation

54:30

that has the ambition of projecting into the physical

54:32

world. Okay? Just like

54:35

Google has offices all around

54:37

the world and your login gets

54:39

you Google Office. Right? Why do

54:42

they have physical offices? Well, they do have

54:44

meet up points and so on. Right? You might

54:46

argue there's less need for physical offices than

54:48

there was. But there's

54:50

still a need for it. You still wanna have meetups.

54:52

You still wanna have you have data centers. You have

54:54

this kind of thing. Cloud first land last

54:56

but not land never. Right? Another version of this

54:58

are, like, the embassies of a

55:00

country around the world. Right? Another

55:02

version of this are like Starbucks

55:04

chains around the world. Right? Like

55:06

basically a chain that has restaurants or

55:08

storefronts around the world. And

55:10

the examples I've just given are those that are

55:12

basically commercial real estate or their

55:14

companies or their states like embassies. But

55:16

you could have it where it's not a

55:18

Starbucks or Google Office, but residential bill seat. You

55:21

could have essentially a network

55:23

of communities where it's not

55:25

a Google login enter and you see

55:27

a Google logo. It's not your workplace. It's your

55:30

residence. Right? You have essentially an

55:32

ethnic diaspora around the world

55:34

and you walk in and it's a little piece of

55:36

home everywhere you got. Right.

55:37

So I understand that. Why would you

55:39

want so

55:39

why would you want something like that? Well, even

55:41

more so than that, so I can imagine why a digital

55:44

community would want to have these physical locations around the world. Google is great

55:46

example for offices. I can also

55:48

imagine an example where

55:50

it's a bunch of people who want these got

55:52

these

55:54

coliving houses and they just operate

55:56

around them. But one

55:57

very important aspect

55:58

of the network state

56:00

is the need to

56:02

be diplomatically recognized.

56:04

So why that step? Why

56:06

not just operate within the

56:08

states that already exist? Where you

56:10

have these communities that are digital with

56:13

a physical element. Why that extra step to be diplomatically

56:15

recognized in your own state? Not

56:17

every business

56:18

needs to have

56:21

the ambition of becoming a public company, let

56:23

alone becoming Google or Facebook.

56:25

Right? It's totally fine to have

56:27

a sandwich store or a hardware store

56:29

or something like that. In the same way,

56:31

think that they're being a funnel. Okay? So in the book, you know, in the v

56:33

two, the way I'm defining it is network

56:36

society, network union, network archipelago,

56:38

network state. society

56:40

or start to say, that's just like one person

56:42

with a dream. Okay? It starts as

56:44

a one person with a dream and they assemble

56:46

a small group. Right? then stage

56:49

two is a network union where they're now able to have that union

56:51

being able to take collective action. Right?

56:53

Which is a huge step above,

56:55

like, ninety nine point nine set

56:57

of online groups do not have this capability.

57:00

In fact, messaging apps, Discord are not

57:02

built for tracking

57:04

collective action. this is kind of

57:06

the unification of cryptocurrency and Discord. You will get something that

57:08

actually tracks that and has leaderboards

57:10

for people and everybody can see

57:13

did what collective action and so on who participated, there's a

57:16

karma board, all the type of stuff. Right? So the short

57:18

answer to your question

57:20

is, there's things that are useful just as network unions as

57:22

network archipelagos that don't need to become full

57:24

network states just like there's things

57:26

that are useful as hardware

57:29

stores or small businesses that don't need to become

57:31

Google. Right? What's an example of a useful

57:34

network union that's that can do digital

57:36

collective action? An example I give in

57:38

the book is of, like, a professional guilt.

57:40

Okay? So it's a guilt of

57:42

designers. And why so those

57:44

thousand designers let's say

57:46

they're in network union?

57:48

Well, ninety nine percent of the time, they're just

57:50

chatting and they are exchanging

57:52

info on their latest designs.

57:54

Maybe each

57:56

day, somebody says, hey, you know, like, here's my work. I'd love a

57:58

boost. Right? And so, you

58:00

know, if you think about it, there's a thousand

58:02

people. So people

58:04

a day, each person gets, like, one booster

58:06

a year or something like that. Kaboom. So

58:08

they can ask for it. They're actually asking

58:10

for something from the community.

58:12

In return, they are expected to give exclusivity and then r t or fave

58:15

or whatever. And it's disclosed on their on

58:17

their thing. They're like, look, I'm part of this community. I'm

58:19

part of this network union. we're

58:22

all signal boosting each other. And then some

58:24

folks may not be able to put in time or they

58:26

may not have an account, they can put in capital

58:30

instead. they'd be like, look, I couldn't make it today, but I'm paying my union dues

58:32

in capital rather than time. And

58:34

that exchange rate is determined by, like, the union leader.

58:36

Right? So ninety percent of the time,

58:40

this Gild. It's boosting its members. It's helping to find

58:42

those members a job. Right? Somebody needs

58:44

a job. I'm out of work. Okay.

58:47

and then everybody just, okay, have another nine hundred and

58:49

ninety nine people, someone can find them a job right

58:51

away. Right? It starts to be a very powerful

58:53

support member, a professional guild. That's why

58:55

these things existed for many years. We're kind

58:57

of rebuilding them online. Okay? And that's useful. And so and

58:59

then one percent of the time, that person is

59:02

under attack online. And

59:04

then the

59:06

union leader looks at their bylaws

59:08

and, you know, it's not an unqualified

59:10

defense of them. Right? But most of

59:13

the online attacks or distortionary or, you know, they're distorting of of

59:15

some some kind. Say basically, say, okay. This

59:18

guy is being attacked or canceled

59:20

or something

59:22

unfairly. Here's their version of the story. We're gonna signal boost that to get them out.

59:24

Get get that up there because it is actually

59:26

in somebody's time of need when

59:29

their status is lowest. that

59:31

they need the boost the most. And if everybody in

59:33

the community acts together, the other nine

59:36

ninety nine people cannot be attacked for

59:38

supporting this person. Right? So you essentially join

59:40

the union ninety eight percent of the time. It is

59:42

for, like, kind of, upside.

59:44

One percent of the time, it's, like,

59:46

cancellation insurance protection against

59:48

downside. Right? that's an example of a structure that should exist, that will exist,

59:50

that we're gonna see lots of, that's a purely

59:52

digital structure, and you can do that for

59:54

designers, you can do that for

59:56

electrical engineers, basically, you'll

59:58

see all kinds of gills like this. Right?

59:59

Okay. And that's digital

1:00:00

only though. You don't have to have any physical

1:00:03

for that. The next step is like

1:00:05

a network archipelago, which takes that capacity for collective

1:00:07

action that the network union has demonstrated.

1:00:10

And now applies, as I mentioned, to crowdfunding

1:00:12

territory in the physical world, and having people

1:00:16

live together. Now when even two people from

1:00:18

an online community start living together, let

1:00:20

alone ten people, it completely

1:00:22

changes everything. Right?

1:00:24

You know, one concept that I mentioned is,

1:00:26

like, companies have logos,

1:00:28

but communities

1:00:30

have flags. And an important thing is if you go to the network c

1:00:32

dot com, if you see the logo in the

1:00:34

upper left, it is what we call the

1:00:36

plus flag. the plus flag

1:00:38

stands for new country.

1:00:40

Okay? Mhmm.

1:00:40

That's kinda clever.

1:00:41

I I thought it. Right. Yeah. Yeah. because,

1:00:43

you know, the plus every day

1:00:45

you're hitting that, in the upper right of Chrome. Right there, what

1:00:47

do you see the plus new tab? Right? Every single app,

1:00:50

the plus is new to this, new tab.

1:00:52

Right? So

1:00:54

It stands for the principle that

1:00:56

one can create a new country. Right? It also

1:00:58

stands for win and help win. Right?

1:01:00

Positive sum. So

1:01:02

you could

1:01:03

imagine each of these people, they're part of

1:01:05

this network union, and they want to

1:01:08

become part of the network

1:01:10

archipelago, as they're hanging the flag of their

1:01:12

community in their

1:01:13

room. They're seeing it every

1:01:15

day. They've got it on their profile.

1:01:17

And eventually, they save up the

1:01:19

money or the time they're able to

1:01:21

get to a remote work job or something like that. And they're able to find one of

1:01:23

the other nine hundred ninety nine people in the community and

1:01:25

be like, hey. Look. Why don't we go and get

1:01:27

a group house together? And

1:01:29

now, let's say, there's six people who are associated with this

1:01:31

in Boston. They start doing meetups and stuff. They

1:01:34

don't live together right away, but they do

1:01:36

meetups and other kinds of things. So I mentioned this, but

1:01:38

they start doing meetups built trust.

1:01:40

And you might like three of the people there,

1:01:42

but only wanna live with two of the people or

1:01:44

something like that. Okay? This council of

1:01:46

group house hacker houses room. All that stuff is a big thing among, you know, kind of

1:01:48

younger generation as you know. Or as your scale

1:01:50

increases, you buy an entire apartment building with like three

1:01:52

hundred people, or you buy like

1:01:54

essentially a

1:01:56

small town. and guess what? Now you've got self driving car down.

1:01:58

Because you have unitivity, you're the

1:01:59

self driving car community.

1:02:04

You put four

1:02:04

hundred people into middle of nowhere in Nebraska. Okay?

1:02:06

A thousand people, whatever. And now the

1:02:08

roads are all zoned for self driving cars. You can rip

1:02:11

up the roads. You can do whatever you

1:02:14

have route access to the physical world because you have

1:02:16

alignment. If you're building a network

1:02:18

archipelago, you want to build it on

1:02:20

like burning man style territory,

1:02:22

probably. Right?

1:02:24

or at least territory that nobody else wants. There's a website called

1:02:26

like land and farm dot com, but you

1:02:28

can see that there's like properties which on

1:02:30

a per square meter basis are like

1:02:34

one thousand cheaper than Palo Alto. You know, why is why is

1:02:36

Cornell? Why does Ithaca have higher property

1:02:38

values and other things nearby? It's

1:02:40

because lots of smart people

1:02:42

move there. and thereby increase property values by moving there. So

1:02:44

these thousand people can

1:02:46

make something out of nothing into

1:02:48

something in

1:02:50

the same you can take a domain name that doesn't have any value turn

1:02:52

to something because all these pointers are now looking

1:02:54

at, all these backlinks are looking at. So capacity

1:02:58

precollective action You have this thousand people,

1:03:00

they take a territory in the middle of nowhere, the

1:03:02

disimmunity thing, and they make it into something. And now

1:03:04

suddenly it's valuable.

1:03:06

Right? This this dance, right, that collective action,

1:03:08

turns something that any one person could

1:03:10

not have done in something that a thousand people

1:03:12

can do together when they're moving as a

1:03:14

multicellular organism. Just

1:03:15

one example, by the way of that is

1:03:17

is Tenguin, which I mentioned earlier. Back, and I think

1:03:19

it must have been around, like, twenty thirteen or

1:03:22

so. Tenguin

1:03:24

was very empty. You know, there's pictures of it back in the day where there's

1:03:26

like deus and one other building and and then a

1:03:28

bunch of rice fields. And it took one person

1:03:30

who ran a co working

1:03:32

or wanted to open one, and he worked with the locals. He built

1:03:34

a fiber line there. And now if you look at

1:03:36

Chengdu, it's I mean, it's it's completely full,

1:03:39

thousands of nomads. And to your

1:03:41

point about property values, I think I saw a tweet

1:03:43

the other day of someone, a friend of mine who lives there,

1:03:45

who basically said that his rent there was more

1:03:48

expensive than his rent in Singapore. where he used to And

1:03:50

so yes, that's a perfect example of

1:03:52

where property is you're saying that

1:03:54

no one cares about can be turned into

1:03:56

property that people care about. But

1:03:58

do a lot of people want to do this? How many

1:03:59

people do you need to do Google? What

1:04:01

percentage of the world moves

1:04:03

to the United States?

1:04:06

tiny percentage. Right? That completely

1:04:07

changed the world. Yeah. I mean, it strikes me

1:04:10

that that there are people who will wanna do

1:04:12

this, but will they want to move

1:04:14

to like rural, Utah, or some land that no one cares about. I

1:04:16

guess, you just need a few is what you're saying. A

1:04:18

tiny percentage of the world

1:04:19

moves to the US. And

1:04:21

the example of the United States though has

1:04:24

been because there was new land,

1:04:26

right, so to speak, at least from the perspective of the

1:04:28

Europeans, there were

1:04:30

social experiments in

1:04:32

democracy and capitalism in

1:04:34

particular that then propagated back to

1:04:36

Europe and to the rest of the world. Like, the French

1:04:38

revolution ball

1:04:40

is bad. was influenced by the American revolution. All the

1:04:42

democratization, lots of the market stuff, all

1:04:44

kinds of things happened in the new world and came

1:04:46

back to the

1:04:48

old world. Right? To an extent that I don't think people fully appreciate. You know, it's

1:04:50

like the Margaret Meade thing. Right? Never

1:04:52

doubt that a small group of highly

1:04:54

motivated people can change the world. Indeed, it's the

1:04:56

only thing that

1:04:58

ever has. Right? I mean, one thing I think that's kind of

1:05:00

interesting is just to hover on that point

1:05:02

for a second. Mass appeal

1:05:04

comes last.

1:05:06

Right? That's, like, after somebody is completely being

1:05:08

proved out and and so on and so

1:05:10

forth. And nothing ever goes from zero percent

1:05:12

to mass movement or I should

1:05:14

say nothing ever. It is hard to

1:05:16

get it to mass movement overnight. And even

1:05:18

if it does seem to do that, it is because

1:05:20

lots of prerequisites were installed.

1:05:22

Billing's of people with smartphones, and

1:05:24

they've got experience with

1:05:26

Instagram, and you've got five g

1:05:28

LTE, and you've got the app

1:05:30

store, and you've got this and you've got that.

1:05:32

That's why TikTok could ride behind

1:05:34

all those prerequisites because people knew

1:05:36

what social networks were, they knew what smartphones were, they

1:05:38

knew what app stores were, they knew

1:05:40

what a feed was,

1:05:42

blah blah. So sliding behind all of those things, I mean, TikTok

1:05:44

is, like, you know, a good company or whatever.

1:05:46

It's it's well executed. I'm sure people also

1:05:48

say it's Fireware, blah

1:05:50

blah, but leaving that aside. They were not just an overnight success, ten years

1:05:52

in the making. There were like ten companies

1:05:54

in the making. It's very easy to dismiss this

1:05:56

idea

1:05:56

of

1:05:58

startup societies and then eventually in network state because there are going to be

1:06:00

many failures, just like there are many failures with

1:06:02

startups. And then we look at the successes

1:06:05

today, but there's almost infinite numbers of people who

1:06:07

tried to build the next Google somewhere in

1:06:10

the last thirty years. Right?

1:06:12

And, of course, we see the

1:06:14

eventual success. And

1:06:16

I wonder how you think about that in terms of the

1:06:19

number of experiments that need

1:06:21

to happen in order to you

1:06:23

know, see this network state. How valuable

1:06:26

do

1:06:26

you think plane flight is? I don't know.

1:06:28

Very valuable. Yeah. Very valuable.

1:06:30

Right? Okay. So click this link.

1:06:32

and this is a century of aircraft accidents. And so you

1:06:34

can literally see in this graph, it's

1:06:36

like as of, you know, August twenty

1:06:40

twenty two, twenty eight

1:06:42

thousand eight hundred ninety six accidents

1:06:44

with a hundred and fifty nine thousand eight hundred and

1:06:46

fifty nine fatalities in basically the

1:06:48

last hundred something years

1:06:49

of aviation. Okay? Twenty thousand accidents, fatal accidents.

1:06:52

Right? And a lot of non fatal accidents.

1:06:54

Okay? So the point I mean, no

1:06:56

plane crashes,

1:06:58

no planes. No trains crashes. No

1:07:00

trains. Right? No explosions.

1:07:02

No internal combustion engines.

1:07:04

Right? Basically no risk,

1:07:06

no reward. And so, you know,

1:07:08

in a big way, like, we

1:07:10

have virtualized all risk, and that's

1:07:12

progress. Right? And certain of all risk, but a lot

1:07:14

of risk is being virtualized. where it's just

1:07:16

financial risk and so on and so forth. I I kinda

1:07:18

think people are foolhardy about

1:07:20

physical risk where people

1:07:22

will, you know, like, do

1:07:24

bungee jumping or skydiving

1:07:26

or things like that. They'll take

1:07:28

risks in my view with no reward beyond

1:07:30

just the thrill of it or whatever.

1:07:32

but they'll be averse to, okay, let me try experimental drug

1:07:35

that could save your life.

1:07:37

Right? You know, euthanasia is legal.

1:07:39

In fact, it's like six

1:07:41

leading cause of death in candid, like you can kill

1:07:43

yourself, but you can't take an experimental

1:07:46

drug. Mhmm. Okay? That's

1:07:48

that's actually a misallocation of risk in the

1:07:50

physical world. Instead, an

1:07:52

alternate society would

1:07:54

lionize those people who

1:07:57

took one for the team and

1:07:59

took a risk on

1:07:59

an experimental drug. This is actually at the beginning of the pandemic, by the way, we could have

1:08:02

had a vaccine in a week.

1:08:04

Why?

1:08:05

Because

1:08:06

because MRNA

1:08:07

vaccines could be printed out all

1:08:09

that time over those nine months was

1:08:11

not spent, I mean, some of it was spent

1:08:13

manufacturing or whatever. but a lot of

1:08:15

it was spent information gathering on whether the thing worked. There's a way

1:08:17

to fast forward all of that, and that would have been with

1:08:20

challenge trials. what's

1:08:22

a challenge trial? You have a bunch of healthy, brave

1:08:24

volunteers at the beginning of the

1:08:26

pandemic. Okay? And they get

1:08:29

the vaccine. and they volunteered to expose themselves to it. I mean,

1:08:31

you could've you could've drafted the military for this, by the way. If

1:08:34

you wanted to have

1:08:36

groups who said people who said, I will

1:08:38

risk my life for the country or whatever. Right? Soldiers, you could have asked for military volunteers. You would have

1:08:41

gotten soldiers who would

1:08:43

have volunteered. Okay? and you could have gotten

1:08:45

it from different demographics because some because some people are in the forties and fifties or whatever even in the

1:08:47

military. Right? You could

1:08:51

have you could have done extremely good data

1:08:53

very quickly and iterated on it. If you had people who were willing to risk their

1:08:55

lives for the

1:08:58

greater good, for the health of the population. And of course, you pay them and

1:09:00

so on just like you pay soldiers in a

1:09:02

in a battle and and you reward them

1:09:05

for taking this risk for the whole team. and they're not doing it

1:09:07

to kill other people who are doing save lives. Right? Look, I'm not saying

1:09:09

that you don't need a military at times.

1:09:12

Okay? But this

1:09:14

is actually the kind of thing, which is is like

1:09:16

a brave thing to do. Right? Now

1:09:18

that kind of society, which is

1:09:21

clear eyed about physical risk taking. Okay? And

1:09:23

that is considered the pros and

1:09:24

cons. But what it does is it

1:09:26

basically says, again, no plane crashes, no

1:09:29

planes. I mean, there's a book called the CRC handbook

1:09:31

of chemistry and physics. You're you're what this

1:09:33

is? I

1:09:34

have not read that yet. The CRC

1:09:36

handbook of chemistry and physics, is

1:09:38

the kind of thing that I would flip through in high school.

1:09:40

Why? It was basically like printed

1:09:42

out math world or printed

1:09:44

out the math portion of

1:09:47

Wikipedia before that all

1:09:48

the compounds, especially the older compounds,

1:09:50

you'd see, you know, smell

1:09:51

and taste like

1:09:54

cyanide, how did we know that cyanide smells like almonds? There's some

1:09:56

guy who took a hit for the

1:09:58

team. He was like,

1:09:59

almonds, you know,

1:10:02

croaks after that. Right? Because

1:10:04

like,

1:10:04

the old school chemists were

1:10:06

crazy enough to, like, smell and taste the compounds. That's why we

1:10:11

have you know, that that's why we know how do flavors in foods.

1:10:13

It is because somebody is taking

1:10:15

the head to take

1:10:17

that test at some point, it needs contact with this girl,

1:10:19

someone must take a risk at some point. Right? No risk. No reward.

1:10:22

No risk. No reward. Now

1:10:23

the thing is, this is being pathologized to

1:10:25

be like, oh, you're encouraging me to take

1:10:27

a risk to your ex. exploiting

1:10:29

them. You're exploiting them to take a risk and so on. Okay. Well, like, the issue

1:10:31

with this is it kind of assumes

1:10:33

that there's

1:10:34

no informed consent as possible.

1:10:37

you know, if you're if you're not

1:10:40

paying somebody to do something, you're exploiting them. If you

1:10:42

are paying them to do something, because you're paying them,

1:10:44

you're convincing them to do it, and they wouldn't have

1:10:46

ever seen it. So you're also exploiting them. Everything is exploiting other

1:10:48

than letting people go bunched up and

1:10:50

take purposeless risks and self destructive risks

1:10:53

and military risks, and that's all fine.

1:10:55

But but you can't have calculated risk. Right? Okay. So I disagree with

1:10:57

that philosophy. And I understand why people have

1:11:00

that philosophy that all

1:11:02

other people's physical risk is

1:11:04

expectation. But I think it can be

1:11:06

done in a proper way. And I think the way to do it is, you start with this online community, you start with the network union.

1:11:08

You develop lots of movies.

1:11:10

It's basically a content union. Okay?

1:11:14

You develop movies, you develop books, you

1:11:17

develop short films, all the stuff

1:11:19

is open source. Okay? Why?

1:11:21

why It's

1:11:22

almost like, have you heard the term content? like in content.

1:11:24

Lots of folks on Twitter are

1:11:26

cranking out what I consider devotional content.

1:11:31

for their hashtag. They're cranking on devotional content, which is, here's

1:11:33

why my cause is good. Here's why

1:11:35

you're bad. Right? Here's why I want

1:11:37

to abolish the police. Here's why

1:11:39

I want to like, social security or

1:11:41

something like that. Okay? That's Bosch plays a more, like, the kind of thing that gets you to

1:11:43

Zellis level. Right? Bitcoin gets you to

1:11:46

Zellis devotional content. Okay?

1:11:48

So what

1:11:50

you have is lots of

1:11:52

movies like Dallas Firestorm.

1:11:54

Okay? Not

1:11:55

exactly in that register, in

1:11:57

that tone, but like,

1:11:58

some Dallas virus club and some

1:11:59

limit still limit list. I've heard of it. I haven't sure.

1:12:02

I'll let you describe this one. Yeah. Please do. I've seen Dallas

1:12:04

virus club, but please for those who

1:12:06

haven't. So why don't you describe

1:12:08

Dallas virus? So the movie is

1:12:09

about a man who happens to get HIV. And at the

1:12:11

time, I think there are drugs that potentially

1:12:15

could cure him, but he cannot get access

1:12:17

to them. Am I remembering this correctly?

1:12:19

That's right. Yes.

1:12:20

Yeah. And and it's

1:12:22

his fight to to basically get that access

1:12:24

and also get that access for other people.

1:12:26

And I think he basically ends up breaking

1:12:28

the law, getting in trouble because he's

1:12:30

getting these drugs to himself and other people

1:12:32

ends up dying, I

1:12:33

think. But sorry, another spoiler. But am I remembering that correctly? That's right. But

1:12:35

basically, the primary agency preventing him from

1:12:37

getting the drugs is the

1:12:40

FDA. Right? And

1:12:42

this is based on the true story

1:12:44

of like ActUp, where people would do dyes, where the

1:12:46

FDA is preventing them from getting access to these

1:12:50

I mean, these people were, you know, had this life threatening

1:12:52

condition, obviously. Fundamentally, they're miscalibrated

1:12:54

on terms of risk. Right?

1:12:58

They

1:12:58

want to minimize PR risk for themselves,

1:13:00

the agency, as opposed to

1:13:02

allowing people to take a

1:13:04

risk on their own. Right? And

1:13:07

why is that? It's because media

1:13:09

only covers it when a drug doesn't work and doesn't

1:13:11

cover it when a drug

1:13:13

does work but has

1:13:16

been delayed. calculations of so called

1:13:18

drug lag. I mean, like Alex Tabrock and others have pulled things together on

1:13:20

this. Right? You know, Daniel

1:13:22

Heminger has written about this depending

1:13:26

on how you calculate it, the number of lives that's being caused by the FDA, basically if you have a drug and it was held up

1:13:31

by ten years, And then

1:13:34

once it's approved, it produced morbidity and mortality by x percentage. That means FDA's

1:13:36

delay of ten years multiplied by

1:13:38

that is how many lives it cost.

1:13:43

relative to just taking the risk at the beginning

1:13:45

and shipping it. Right? Now there

1:13:47

is risk. Okay? But

1:13:49

by once

1:13:50

you start having the ability to take

1:13:52

calculated risks, you can move at the speed of

1:13:54

software. Basically, do you know for example who

1:13:56

banking investor?

1:13:57

I they're familiar, but why don't you

1:13:59

share? we need

1:13:59

to get banking and best

1:13:59

zones. Okay? Banking and best, they won

1:14:02

the Noble Prize in the twenties, nineteen

1:14:06

twenties. Why? because they came with the

1:14:08

concept for insulin supplementation

1:14:11

to treat diabetes. They

1:14:13

had the hypothesis. they tested it

1:14:15

on dogs. Okay? Then they

1:14:18

tested it on themselves.

1:14:21

Then they tested on patient

1:14:23

volunteers and like, you know, it's like a miracle drug,

1:14:25

like people jumping out of bed type stuff.

1:14:27

Right? And then

1:14:29

and then Eli

1:14:30

Lilly had, like, scale production for the

1:14:32

entire North American continent in, like, two years, and

1:14:34

they won the Nobel Prize in, like, two

1:14:36

years after they start. Okay? That is

1:14:39

when pharma moves at the speed of software. That is what is possible in

1:14:43

the physical world once

1:14:46

we can get a zone outside the FDA, once we can exit the FDA. People don't understand how FDA

1:14:49

is. Your

1:14:52

entire life It's

1:14:54

like tens of trillions of dollars of values being held back by Feet. It's bad

1:14:57

they are as

1:15:00

an agency. and we

1:15:02

saw this during COVID where they were holding back. I mean, FDA is why we're flying blind in

1:15:07

February twenty twenty. because they

1:15:09

were holding back the tests. That's why the journals were reporting that, oh, there's

1:15:11

no problem. It's not in the US

1:15:14

yet. Actually, it was in the

1:15:16

US. Why

1:15:18

was it in the US? Because

1:15:20

it actually come in via China. And

1:15:22

people were not able to run tests

1:15:24

to confirm it. because they had to

1:15:26

get the so called emergency use authorization from FDA,

1:15:28

which was delayed. So what happened was actually some labs actually

1:15:30

did civil disobedience just like Dallas Byers Club and

1:15:35

they defied FDA, and they just went and ran the

1:15:37

test. And later, they got sort of a

1:15:39

blessing from journals, which was,

1:15:41

okay, Dave Russell conjugated

1:15:43

it as simple immediate versus earning an illegal trust blah

1:15:45

blah blah. Right? So they they basically blessed it

1:15:48

retroactively. So FDA did not go and crack

1:15:50

down on those labs that actually managed to

1:15:52

get the

1:15:54

testing data out there. Okay? Vaccine could

1:15:56

have been had in a week. Okay? The

1:15:58

tests could have been there

1:15:59

immediately. Fundamentally,

1:16:02

FDA

1:16:02

is -- the problem with

1:16:05

it is it's got controller of

1:16:07

the entire world. The single

1:16:09

most important the logical problem to solve

1:16:11

in the world is a problem of regulatory harmonization.

1:16:13

Regulatory

1:16:13

harmonization is a process by which a group of unelected bureaucrats in

1:16:16

the US write

1:16:19

the regulations for the entire world. You just look at

1:16:21

a small website, right, will outsource its

1:16:23

login to Facebook

1:16:26

login. Okay? because,

1:16:27

hey, Facebook's big, you know. Right? But in

1:16:29

so doing its beholden to Facebook, you know,

1:16:31

I I respect Chuck, but I

1:16:33

also don't necessarily want the login assume to be completely

1:16:35

at the discretion of some engineer there. Right? In the

1:16:37

same way, a small country, you know,

1:16:39

Czechoslovakia or, like, some

1:16:41

some small country will outsource their

1:16:44

regulation to the

1:16:46

FDA, the SEC,

1:16:49

FAA. Right? And they'll say, oh, well, look,

1:16:52

America's regulators should be good

1:16:54

enough for us. We're a

1:16:56

small country of a few million people. This

1:16:58

is a big country. It's got a big market.

1:17:00

Therefore, if a company can get through FDA regulation, then

1:17:02

we'll approve the product for sale in our country. If the plane can get through FAA,

1:17:06

we'll approve for sale. If it can get through SEC, then it seems

1:17:08

like a decent financial product we might list so in

1:17:10

our stock exchange or some or allow

1:17:13

our citizens to buy or something like that. This is

1:17:15

actually something that is sought by FAA

1:17:17

SEC FDA, respectively.

1:17:20

If you go to their

1:17:22

website and look for harmonization, they

1:17:25

intentionally convene in working groups with all the rest of

1:17:27

the regulators in the world to say use US regulations, harmonize,

1:17:29

have everybody singing

1:17:32

in unison. Nothing is both big

1:17:34

business and big government like this. Obviously, the regulators like it because when you

1:17:36

become a

1:17:40

federal regulator, You're

1:17:40

not doing it for profit. You're doing

1:17:43

it for Ambit. Right? Ambit in, like, the the scope, like, you

1:17:45

know, Ambit Samroot

1:17:48

is ambition. Right? You know, it's

1:17:50

like a relatively uncommon word, but it's like the scope, extent, or bounds of something. Right?

1:17:55

so So If

1:17:56

a CEO wants maximum profit for their

1:17:58

company, a regulator wants maximum

1:17:59

ambit for

1:18:03

their regulatory agency. They just want scope. I'm

1:18:05

like the most baddest, you know, biggest regulator around. Right? FDA

1:18:08

brags, they

1:18:10

regulate something like thirty cents out of every dollar. They like brag about

1:18:12

this. How big their regulatory impetus,

1:18:14

the most powerful regulators were up. It's

1:18:16

a food and drugs administration. Literally everything

1:18:19

you put in your body every

1:18:21

bite you take every single day is regulated by FDA in the

1:18:23

U. S. and around the world. Right? So power is

1:18:26

absolutely immense. We have

1:18:30

Bionic Guys. They exist.

1:18:32

We have Super Soldier Serum. It's already real.

1:18:34

Do you see that poster mic? No, I

1:18:36

didn't. There's a little mouse on the

1:18:38

left hand side, and there's a myostatin null mouse on the right hand side. Okay. And second row,

1:18:43

that's the chest. of

1:18:46

the mouse on the left and the mouse

1:18:48

on the right. And that's literally like Captain America before and after. That paper

1:18:50

came out like fifteen years ago. People have known about Myastatin all literally

1:18:55

this one if you just inhibit myostatin, there's ways

1:18:57

of doing you can do knock down other

1:18:59

stuff. And there's other

1:19:01

things, full yourostatin blah blah blah. you could make

1:19:03

somebody basically naturally jacked. And you know

1:19:05

the side effect would be because

1:19:08

muscle is

1:19:10

metabolically expensive, they'd be able

1:19:11

to eat whatever they want. And of course, people say, oh my god,

1:19:13

you know, they'll have to be side effects.

1:19:15

It's like, pancreas, we flew

1:19:17

too close to the sun. actually, no. They don't have these side

1:19:19

effects. Yes. It might be plane crashes, but we actually

1:19:21

figured out how to get aviation basically without

1:19:23

plane crashes. In the same way, you could

1:19:26

figure out dosage and with enough volunteers, enough

1:19:28

people you could probably

1:19:30

get, like,

1:19:30

very effective, either natural or artificial steroids

1:19:36

or equivalents that would make somebody naturally

1:19:38

muscular and fit. Just like we take caffeine every day. It's a known drug

1:19:39

with known plus and minuses.

1:19:41

You have something that's like

1:19:44

that. Right? This

1:19:46

is just like one example. I can give

1:19:48

like fifty more. Okay? The point is the FDA is holding back so so

1:19:50

so much. So that is like why you want a network state.

1:19:55

Okay? So

1:19:55

network union, as I mentioned so, you know, the

1:19:57

progression. Network union just does things digitally and collectively. So

1:19:59

it's like a like

1:20:02

a guild working together. Right? a

1:20:04

network archipelago, now you're getting offline territory. But some

1:20:06

things do, yes, require changes in law. So that's the highest level

1:20:10

in network state. is something that has collective action, kept

1:20:13

capacity, and the content, which

1:20:15

aligns people, has

1:20:18

the physical footprint and it has like the square meters of physical

1:20:20

space. And then the last step

1:20:22

is it gets big enough and

1:20:24

strong enough and morally enlightened enough

1:20:27

it can do the dance Right?

1:20:29

That a

1:20:30

city or state or

1:20:31

country recognizes it and

1:20:34

does a deal with

1:20:36

it. Okay? And what are

1:20:38

the precedents for this? So, like, Wyoming has its Dow lock. It's literally built an interface effectively

1:20:41

to the Ethereum

1:20:44

network. Right? or El Salvador, Bitcoin's

1:20:46

national currency there. So it's built in interface, the Bitcoin network. Or a Nevada

1:20:48

has, you know, done a deal

1:20:50

with a lawn for the GEICO factory.

1:20:54

Right? So that's like there. Okay? Or

1:20:56

Amazon with HQ2. Google tried something like this with

1:20:58

sidewalk labs that got pushed back, but the

1:21:00

concept was there. Okay? Point is, that lots

1:21:02

of these sovereign entities, cities,

1:21:05

states, countries, or let's

1:21:07

say not lots, enough of

1:21:09

them, okay, have been doing

1:21:11

deals with companies and digital currencies that there

1:21:13

is feasibility there. And so

1:21:15

what you'd want is you want

1:21:17

a group of people which is

1:21:20

large enough motivated enough. Maybe it's a thousand

1:21:22

people, maybe it's ten thousand, maybe it's a hundred thousand, maybe you need

1:21:24

a million people. I don't know the

1:21:26

scale of it. That's an empirical question.

1:21:30

that they can go to

1:21:32

all of the different mayors and governors and

1:21:34

maybe presidents of the world because they're

1:21:36

they're it's a parallel process. That's

1:21:38

the awesome thing. you're no longer just negotiating with just the US

1:21:41

government, etcetera. Right? Forget them. Like,

1:21:43

the US federal government will

1:21:45

be the last mover

1:21:48

on everything Like, everybody's over indexed on. We

1:21:50

need to reform the US government. No. You need to exit the US government. I shouldn't say ignore,

1:21:52

but just assume they

1:21:55

can do nothing good. Okay?

1:21:58

Assume they can do nothing good basically just on the other six percent the

1:21:59

world, right, of

1:22:02

who are not American. and

1:22:05

the enormous part of the US, which is breaking away from the US government. Meaning, you

1:22:07

you wanna try to

1:22:10

get

1:22:11

a sanctuary city

1:22:14

for

1:22:14

for biomedicine biomedicine,

1:22:16

a sanctuary city for cryptocurrency,

1:22:17

a sanctuary

1:22:19

city for self driving

1:22:21

cars, where federal law

1:22:23

is not

1:22:23

enforced. Right? Officially. And

1:22:25

is it

1:22:26

not enforced

1:22:27

only along that

1:22:29

plane or is

1:22:31

it you are you are creating completely new sets of laws?

1:22:33

Well, it's both. So basically,

1:22:35

the thing is you can

1:22:37

do this both

1:22:38

within and outside the US. Right?

1:22:40

Outside the US, harmonization is rejected, inset,

1:22:42

so basically, the unelected bureaucrats of Silver Spring, Maryland

1:22:45

no longer have power. But why

1:22:47

do I see that? Because Those

1:22:49

bureaucrats I mean, for example, when Facebook is criticized, you know who runs Facebook at Zoc.

1:22:52

Right? You know he's

1:22:54

in control of it. He's

1:22:57

like attack by name constantly. People personalize it. It sucks company blah

1:22:59

blah blah. Can you name a single person who

1:23:01

works

1:23:02

at FDA? No.

1:23:04

Right. despite the fact

1:23:06

that every single bite you take every day is regulated by them. There's zero personal accountability. Okay?

1:23:09

If they're

1:23:12

ever criticized, It's this abstract thing

1:23:14

of the agency, which has billions of dollars in annual budget, and essentially infinite hit points can't

1:23:19

go bankrupt. Right? has appropriations from the

1:23:22

government. Right? You know, so no individual is ever accountable there. They're not accountable

1:23:24

to the electorate.

1:23:27

There's no elections Okay?

1:23:29

And they're not accountable to the market because they have career tenure and can't be fired. And they're anonymous. They're not

1:23:31

even accountable in terms of public criticism. So

1:23:35

this is totally unaccountable

1:23:38

agency that determines everything you

1:23:40

eat and drink, your vaccines, your drugs,

1:23:43

your this and that. And has

1:23:45

de legitimated itself in its performance over COVID, where it held back to vaccines. And also, you know, the Johnson

1:23:47

Johnson vaccine got held back by, like,

1:23:50

a few weeks over some stupid scare,

1:23:52

like, you

1:23:55

know, the problem, by the way, with the COVID conversation

1:23:57

around, like, vaccines is either

1:23:59

being, like, Yay, everybody, you

1:24:01

know, is forced to take

1:24:03

the vaccine or you know, oh my god,

1:24:06

vaccines are world economic forum, conspiracy, classroom, etcetera. And what there hasn't been

1:24:08

is option three, which is

1:24:10

we needed this in a week

1:24:13

and we could have, like, essentially, gotten them

1:24:15

to the vulnerable first. Even if needed some time to

1:24:18

scale, which it would have,

1:24:21

you get them to, like, the elderly rapidly.

1:24:23

Right? And you could have avoided like a million dead. That million

1:24:26

dead is on the FDA's

1:24:28

head. like because

1:24:30

challenge trials could have been done, you know? And so that's like a different interpretation of the problem

1:24:33

with this. Right? Now

1:24:35

the thing is, that

1:24:38

no country around the world has

1:24:40

this formulated in quite this way. That

1:24:42

has to say, despite the fact that

1:24:44

these are you

1:24:45

know ostensibly sovereign countries, they don't have the narrative that

1:24:47

a, FDA is illegitimate and or delegitimate.

1:24:52

And

1:24:52

b, also, and this is important. I

1:24:54

want to make this important point. I'm not saying zero regulations

1:24:56

and a total

1:24:58

free for all of all

1:25:00

kinds. I'm actually saying

1:25:02

a V3, right? So I'm not saying just end the FDA. I'm saying exit the FDA. didn't

1:25:04

just end the Fed

1:25:06

and replace it with nothing.

1:25:10

you're ending the Fed or exiting the Fed

1:25:12

and replacing the Bitcoin. The reason being, just

1:25:13

to talk about this point for a second, because I

1:25:15

think it's important. I've just spent all this time

1:25:17

tearing the thing down, allowing me give something to it to build

1:25:19

it back up again. People

1:25:21

want a regulated

1:25:23

marketplace. In the sense

1:25:25

of they want a regulator

1:25:27

that will do star ratings, like quality ratings

1:25:29

of, you know, most actors and bands of bad actors. Right? So you get

1:25:31

a star rating one to five stars on

1:25:34

something, you know, it's a product in the

1:25:36

market. and

1:25:38

the zero star bad actors and scammers are

1:25:40

just like kicked out of the market entirely.

1:25:42

One way of doing that is a

1:25:45

national state based regulator. like FDA, FAA, SEC,

1:25:48

etcetera. Another way of doing it, which is actually

1:25:50

not thought of as such, but it's really important,

1:25:52

is a cloud

1:25:55

based international regulator like

1:25:56

Amazon, eBay, actually even Gmail, Apple, etcetera. How

1:25:58

are those regulators? What do you

1:25:59

mean? Well, obviously,

1:26:02

Amazon

1:26:03

has star ratings. eBay

1:26:06

has star ratings, and they also kick bad

1:26:08

actors. Right? And they kick bad actors

1:26:10

on the merchant side, Uber and Airbnb

1:26:13

are also regulators. They're

1:26:15

international regulators. They're better than the taxi medallions

1:26:17

or the hotel regulators because Uber is tracking

1:26:19

every trip. Right? They're giving star

1:26:21

ratings on both the

1:26:23

driver and the passenger. they're

1:26:25

checking that payment can go through and that the passenger is good for it and so on.

1:26:27

And then they are decommissioning people who

1:26:29

have low ratings on either side

1:26:32

or abuse. This

1:26:35

is just, in a sense, a much more intrusive regulator.

1:26:37

No taxing medallion could possibly do

1:26:39

anything like that.

1:26:41

Right? They they do I

1:26:43

mean, a very cursory inspection every six months that

1:26:45

the windshield wiper still works. They're not

1:26:47

getting reviews from every rider, certainly not collecting

1:26:49

them in real time. It has to be

1:26:51

like an accident for them

1:26:53

to actually write up that taxi driver. So that's to be a

1:26:55

right tail kind of thing. So the state based paper based

1:26:59

regulators fundamentally less efficient than

1:27:01

the cloud based international regular, which also has data from around the world. Moreover,

1:27:03

it's inefficient in a different

1:27:05

way

1:27:08

where Typically, we think of the state based regulator

1:27:10

as adversarial to the industry. Right? Oh, the tax regulator is supposed to be adversarial to

1:27:13

the tax companies, keep them in check. What actually happens is they

1:27:16

form a duopoly against the

1:27:18

customer, whereas a cozy relationship

1:27:20

between the taxi regular and

1:27:22

the taxi drivers, since the taxi to know the regulator,

1:27:25

but the taxi riders just flip in and

1:27:27

out. You know? Like taxi riders

1:27:29

have nothing in common

1:27:31

besides driving taxi, the drivers keep meeting

1:27:34

with regulators eventually farm social bonds in the same way. The FDA is, in many ways,

1:27:36

adversarial to pharma

1:27:39

companies, but it's also cooperative

1:27:42

with them against startups and against patients who want novel treatments. Okay?

1:27:44

So Uber and Lyft actually

1:27:46

changed this, Airbnb

1:27:47

changed this because

1:27:48

Now

1:27:51

you have a regulator, customer provider

1:27:54

complex. It's Uber,

1:27:56

the regulator, the Uber

1:27:58

driver, and the

1:27:59

Uber passenger. versus

1:28:01

lift the regulator and the lift driver for

1:28:03

the passenger. Right? So the star

1:28:05

rating and the provider are

1:28:07

linked together and efficiency

1:28:10

comes from competition between regulators. It is is what's called polycentric law. Okay? So it is within

1:28:12

the same jurisdiction, you

1:28:15

have choice of law. Okay?

1:28:19

I pick the graph or the left app

1:28:21

or in another country go check or

1:28:23

graph. Right? You have choice

1:28:25

of regulator. Right? Now even two choices better than

1:28:27

one. Yeah. I get that it's too optimistic in

1:28:30

in some ways, but there's other, like, ride

1:28:32

sharing apps and so on. Right? that's

1:28:34

a model where you're not saying no regulation whatsoever. You're saying a choice of regulator. Regulators are

1:28:36

actually incredibly valuable.

1:28:39

Like, these are multibillion,

1:28:41

multi because the reason being, because people wanna pay one price to enter a

1:28:43

market, they diligence say a legit regular, that one

1:28:46

isn't. And then they don't wanna

1:28:50

go and test every product in the market. They

1:28:52

wanna know it's at a basal level of quality

1:28:54

and and look through it. Otherwise, you

1:28:56

have a situation where every coffee you get

1:28:59

at Starbucks, you need to put a dipstick in

1:29:01

it to see, you know, is this poisonous or not? So

1:29:03

the point is the v three combines

1:29:06

aspects of the v one and the v to get like third version. example, has aspects

1:29:08

of gold, which people talk about,

1:29:10

but it also has aspects of fiat.

1:29:15

can it borders. You can represent digitally. You can program

1:29:18

with it. And so it's about you

1:29:20

can't do any of those things

1:29:22

with an inert block of gold. So

1:29:24

it's digital gold. Right? It is not just

1:29:26

a dumb, you know, throwback to v one. It's

1:29:28

not competitive as

1:29:29

V1V2 beat v one

1:29:31

for a reason. V

1:29:34

three takes the good aspects of v one and the good aspects of v two to make something that's better than be

1:29:37

AV3

1:29:40

where

1:29:42

yeah,

1:29:42

you don't want the totally unregulated market

1:29:44

for patent medicines and scans and

1:29:46

so they had great upside and they

1:29:48

had great downside. We also don't want

1:29:50

this you

1:29:51

know, harmonized environment where unelected

1:29:53

bureaucrats just impose uniformity

1:29:55

and everything and no

1:29:57

one can take any risks

1:29:59

and

1:29:59

it's so risk

1:30:00

intolerant that it's the

1:30:01

riskiest of all. Right? Because if you don't take small

1:30:03

risks, you end up

1:30:04

taking the biggest risk which is not

1:30:06

taking risks and then you just never

1:30:10

exploring anything new and you can't adapt. Right? So

1:30:12

the v three says no regulation

1:30:14

is bad, but this regulation is

1:30:17

also bad, and we can do

1:30:19

better with cloud regulators. right, international regulators. And what would

1:30:21

that look like again just to go into the FDA example? Because the reason I go

1:30:23

into this in in great detail is this is one of

1:30:25

the most motivating things as to why we

1:30:27

need new states. Right?

1:30:30

There's a certain level of innovation you

1:30:32

can get to that's sub sovereign, and there's

1:30:34

a level that you actually need sovereignty

1:30:36

for,

1:30:36

of some kind. At a minimum, a

1:30:38

sanctuary city for biomedicine. some governor, right, could put

1:30:40

their state on the

1:30:41

map by saying, this is a zone where

1:30:44

let let's say it's Texas.

1:30:46

Right? They say, guess what, approvals

1:30:49

we'll now go through UT Austin because the thing is the

1:30:51

reason is you need some name brand. Right? That that people respect locally,

1:30:56

to be the regulars. Okay. We've got some

1:30:58

scientists there. We've got some physicians there and so on. So you have some reputable brand in this jurisdiction,

1:31:01

which is taking

1:31:04

over review. So you're really not changing that

1:31:06

much. What you are doing now is you're introducing the crucial thing of competition between regulators.

1:31:08

Right? And you might

1:31:11

say, oh, this is in

1:31:13

biomedicine. I don't know. I mean, like, you know, bio is, like, really

1:31:15

important. And the thing is people just don't know about this market. Okay?

1:31:20

But basically routes outside the FDA, like

1:31:22

rights to try laws, like CLIA Labs and

1:31:24

the LDT pathway,

1:31:27

like compounding pharmacies, off label

1:31:29

prescription by MDs and countries that aren't fully harmonized with FDA. For example, you go to Germany

1:31:31

for stem cells. Each

1:31:35

of these things has been attacked by

1:31:37

FDA. So off label people think it's bad, right? But there's a good article

1:31:40

called assessing the

1:31:42

FDA by the anomaly of

1:31:45

off label prescription. Right? I think this is by Taberach, which is worth reading the psalms twenty years ago.

1:31:47

And essentially his point is that

1:31:50

while off label sounds bad, oh,

1:31:52

right? it

1:31:55

actually means that a doctor can prescribe a drug for some purpose other than

1:31:57

what the FDA approves it for. And that's actually

1:31:59

the

1:31:59

way that,

1:32:02

like, all kinds of things seem to

1:32:03

work. Once you understand the degree to which FDA holds

1:32:05

things back, once you understand that they have

1:32:08

harmonized the whole

1:32:10

world, once you understand democracy doesn't work in this case because they're not

1:32:12

up for election. You've never run

1:32:14

an election on the FDA, okay?

1:32:17

Markets don't work. because they can't be fired. They're not

1:32:19

like a company. They don't go bankrupt. So whether

1:32:21

you believe in an electoral or market theory

1:32:24

of accountability, they are completely

1:32:26

unaccountable and anonymous. Right? The presses isn't holding them accountable either. Right?

1:32:28

So you could only exit them and you

1:32:30

need a new technique to exit them.

1:32:32

That can be outside the US, you get

1:32:35

places like Germany and so on to allow

1:32:37

for stem cells. You get places

1:32:39

that allow for biomedical treatments

1:32:41

that the US is holding back.

1:32:43

Right? Or it's inside the US and you're having sanctuary cities. Just like

1:32:45

sanctuary city, they won't enforce federal immigration

1:32:47

law. Right? So

1:32:49

precedent is being set. In fact, actually, states are diverging from

1:32:51

the federal government on education,

1:32:54

on gun laws, on

1:32:56

abortion laws,

1:32:58

on marijuana laws, like, if marijuana be legal,

1:33:00

why can't you know, like

1:33:02

every

1:33:02

other kind of

1:33:03

experimental drug that

1:33:05

you might want to take be potentially legal.

1:33:07

Right? And obviously, you you have some framework

1:33:10

around it. It's not no framework, but

1:33:12

finally, you

1:33:14

have choice. you

1:33:14

have something that's outside its monopoly of the

1:33:16

state regulator. Right? Again, unelected and unfavorable.

1:33:19

I mean, there's so

1:33:20

many other reasons I think that we

1:33:22

need innovation on sovereignty. But that alone is how we get to life

1:33:24

extension. That alone is how we

1:33:26

get to transhumanism. We get

1:33:29

to brain machine interface. We get to

1:33:31

liver generation and super soldier serum and all this stuff. There's

1:33:34

so much stuff that's being held back.

1:33:36

Okay? Like, by I mean,

1:33:38

that's why you see articles constantly

1:33:40

like, scientists have

1:33:42

discovered necks. Right? You're like, wow, that's amazing. You're like reversing this guy's aging by like

1:33:44

ten

1:33:44

years. Their hair

1:33:47

is repigmenting from from

1:33:51

gray to black. Right? Here's an

1:33:52

example of that. because people

1:33:53

have to see this stuff because they don't

1:33:55

believe it otherwise. Right? They don't have

1:33:57

the same exposure. If if

1:33:59

you

1:33:59

don't see the examples outside

1:34:02

of

1:34:02

your reality of FDA

1:34:04

or non self driving

1:34:05

cars or whatever society that you

1:34:07

live in, it's impossible to

1:34:10

reimagine that there are these other realities. Exactly.

1:34:12

It's Bostiad's point

1:34:13

seen and unseen. Are you familiar

1:34:15

with that? I think so. I

1:34:17

I didn't realize that was him. But

1:34:19

yes, exactly. Like, it's it's impossible to imagine those realities until

1:34:21

you see it. And then it's like this

1:34:23

spiral of, wait a

1:34:25

minute. That's right. It's basic it's like you know what

1:34:27

I mean? It's basically something where it's like, I show people

1:34:30

so for example, take a look at this thing that I

1:34:32

just sent you. One, it says patient

1:34:34

image pretreatment and patient image post treatment. And yes, exactly like you it's like

1:34:36

gray hair. The man looks like

1:34:38

he's in his eighties to a

1:34:42

full head of brown, black hair. People don't understand

1:34:44

that we're not like within ten

1:34:46

percent of optimal or something like that.

1:34:49

In many areas, we

1:34:51

are a thousand x, ten

1:34:53

thousand x, a hundred thousand x off

1:34:55

from what is possible. Right? people

1:34:58

said the pandemic was not going to be something, and

1:35:00

it's like millions of people were affected by it.

1:35:02

Like, it takes years to build a train

1:35:04

station in the Bay Area, and it takes less

1:35:06

than a day in China. That is not a

1:35:07

oh, they're ten percent faster.

1:35:09

They are literally a hundred

1:35:11

x faster.

1:35:13

Maybe a thousand

1:35:14

x. If if you're talking one like, say, hours,

1:35:16

right, versus a year.

1:35:18

Okay? That is

1:35:21

about a thousand x faster. it's just

1:35:23

a totally different thing. You know, your cost, by the way, the

1:35:25

cost of everything comes way down because

1:35:27

it's now a sub routine.

1:35:29

You can just invoke that. Boom. Boom. Okay. Transition.

1:35:31

Transition. Let's just clone stamp like

1:35:33

this. Right? Versus the completely antiquated

1:35:36

kind of thing in the physical

1:35:38

world in the US. and all of that is the state.

1:35:40

All of that is the state

1:35:42

holding this back. Right? And so

1:35:44

ways

1:35:44

to exit the

1:35:47

state and gain sovereignty are how we reinvent

1:35:49

the school. And in a sense, by the way, we talk about the v three,

1:35:51

you know, TL has talked about how we can innovate

1:35:53

in bits, but not in

1:35:55

atoms. Right? Mhmm. And what

1:35:57

the network state is, is among other things, many other things, but it is a recipe

1:35:59

for using bits to reopen

1:36:03

innovation in apps. Well,

1:36:04

yeah, I think I think

1:36:07

that's a really important point because it's easy for people to see exponential difference

1:36:12

between physical and digital.

1:36:14

A simple example is physical mail can take days, weeks, months, digital mail take

1:36:19

seconds. Right? So So they can see

1:36:21

the delta there, but I do think it's hard for people to imagine that exponential

1:36:23

delta within the physical world

1:36:26

because it feels slower inherently.

1:36:30

I think the train station example is showcasing

1:36:32

that, yeah,

1:36:33

it's not gonna happen in seconds, but

1:36:35

there is a massive delta between what we

1:36:37

have in certain arenas and what

1:36:39

we could have. Yes. I showed on

1:36:41

those videos, he's like, okay,

1:36:43

you reset my

1:36:44

belief in

1:36:45

what is

1:36:46

possible. Right? This is

1:36:49

senior executive at a trillion dollar company didn't

1:36:51

know you know, he's like, this

1:36:53

is why we're doing bits and

1:36:55

so forth. I'm like, atoms

1:36:58

are actually possible in other

1:37:00

places and times. Right? So other

1:37:02

places, I'm showing China here

1:37:05

where there's a thousand x,

1:37:07

There's another link that's worth looking at, which is patrick

1:37:09

Carlson Fast. patrick Carlson

1:37:11

dot com forward slash fast.

1:37:13

That's also worth putting up

1:37:16

on screen. where he goes to other times and

1:37:18

shows that the US used to be much faster in building things. And by the way,

1:37:22

people will say, oh, China's authoritarian. That's why I can build fast. We're a

1:37:24

democracy. We you know, they're basically making

1:37:26

a virtue out of incompetence. Right?

1:37:29

That's actually not

1:37:31

the case because quote unquote, the US was a

1:37:33

democracy. I mean, if you think the US was democracy mid century, he was able to build fast then.

1:37:35

I actually do argue that's part of the

1:37:37

nature of the political system, but it's

1:37:40

not necessarily see

1:37:42

versus authoritarian. And in fact, actually, the democracy of the mid twentieth century America was quite authoritarian.

1:37:45

Right? Democracy like

1:37:48

capitalism or Crescendy

1:37:50

or communism is so capacious

1:37:52

a term that can mean both x

1:37:55

and its opposite.Communism, for example, meant kill all

1:37:57

the capitalists, and then it means capitalists can

1:37:59

be in the communist party. Right? Christianity meant tear

1:38:01

down the Roman Empire, and it also eventually meant have Christian kings

1:38:03

and build a holy Roman Empire. Like,

1:38:05

democracy means tons of different

1:38:08

things over time

1:38:10

periods, like the ancient Greeks thought democracy,

1:38:12

there's a great book called against

1:38:14

elections, the case for democracy. It's

1:38:16

a hilarious title. Okay? and it

1:38:19

basically says that the ancient Greeks,

1:38:21

they use a mechanism called sortition rather than

1:38:23

election. Do you know what that is?

1:38:25

It's like -- No. --

1:38:27

random selection from the population and any they drew lots. And so your

1:38:29

president could be anybody

1:38:31

from the community. What

1:38:34

did that do? It eliminated the whole that

1:38:36

you had to maintain a high

1:38:38

level of virtue in your community

1:38:43

and actually a high level of quality because anybody could be the leader.

1:38:45

And, you know, at first, it seems

1:38:47

like a crazy

1:38:50

system. But the

1:38:51

awesome thing about history is, you know,

1:38:53

people estimate there's a hundred billion people

1:38:55

who've ever lived. Right? You have

1:38:57

one life. And so just

1:38:59

like we look at other places

1:39:01

in the galaxy and we know what our star system looks like, but there's like

1:39:03

dual star systems and black holes, all

1:39:06

these crazy things in the huge

1:39:08

expanse in

1:39:10

in space. There's this enormous

1:39:12

expanse in time. There's this

1:39:14

other configurations of humans that

1:39:16

are totally counterintuitive to our

1:39:19

current orientation. We think we know this current state,

1:39:21

well, it doesn't work anymore. But let's say, this current state of affairs kind of

1:39:23

works, or at least the ones we've already always

1:39:25

known over our tiny

1:39:28

window of twenty or

1:39:30

thirty or forty years on the world. But when you take the hundred billion human lifetimes and look at all these other configurations, like other

1:39:33

star systems,

1:39:36

you're like, oh, that's how election just work.

1:39:38

That's how regulation just work. That's how it worked here. And you can pick a little somber teams from there

1:39:40

and say, this was compatible

1:39:42

with human nature back then.

1:39:46

And

1:39:46

maybe we can update it and make it work in

1:39:48

the present day. One thing

1:39:49

that you brought up, which is important, is this

1:39:52

idea that there there has

1:39:53

to be the potential for a challenger. So, like, if you take the taxi example before, it's like, yeah, I

1:39:56

could choose

1:40:00

person a's taxi or person b's taxi, but

1:40:02

I didn't have an alternative outside of the taxi system. And something that's coming to mind

1:40:05

is as we talked

1:40:07

about the line drawn around

1:40:09

the world or at least within our

1:40:11

lifetime or let's say my lifetime, they've been mostly fixed, obviously outliers, but mostly fixed. And

1:40:13

to me,

1:40:13

that reminds me of this idea

1:40:16

that they're there

1:40:19

are many people around the world who just assume that

1:40:21

there is no challenger. And we've

1:40:23

seen that competition drives innovation

1:40:25

in many domains. And

1:40:27

so it's almost fascinating

1:40:29

just imagine that we can get

1:40:31

to

1:40:31

potentially exponentially better societies

1:40:33

potentially exponentially better society if

1:40:35

we do have But there's almost like this underlying

1:40:37

assumption that there aren't. There are

1:40:40

challengers within the societies

1:40:42

that we have, but not challenges

1:40:44

to the societies. I mean, just to talk

1:40:46

about this for a second. Like, the US establishment is

1:40:50

currently fighting

1:40:51

fighting conflicts with tech, with

1:40:54

half its population, with

1:40:56

Russia, with China, with

1:40:58

to a lesser extent, with

1:41:01

Israel, India, Hungary, France, Brazil, with the

1:41:03

Brexiteers in

1:41:04

Britain, with web three

1:41:06

now, and with crypto, like,

1:41:11

It's just finding on so many different fronts at the same time.

1:41:13

At the same point, it

1:41:15

is printing tons

1:41:18

of money And in many ways, state capacity has

1:41:20

fallen through the floor where SF takes twenty

1:41:22

years to reopen a bathroom. Or, you

1:41:25

know, I had this thread where it's

1:41:27

like, a trillion dollar disaster

1:41:27

for the thirty five. And, like, the zumbol, this is,

1:41:30

you know, like, this navy ship, and that's

1:41:32

a disaster. And the Ford

1:41:34

class carrier, another multi ability, like,

1:41:36

incredibly expensive things that we just sort of renumb to

1:41:38

it. because you see another hundred billion dollars here, a hundred billion dollars there, a

1:41:41

trillion dollars here, a stimulus

1:41:43

here. It's just money

1:41:46

being thrown down a raffle. It's

1:41:48

like a kid who I mean, one of

1:41:50

my points, you know, this is the article

1:41:52

I wrote called founding verse inheriting. one really

1:41:54

good way of thinking about the current

1:41:56

US government and the US establishment really because

1:41:58

more than just the US

1:41:59

government. The reason that

1:42:02

it

1:42:02

it can't execute partly because

1:42:04

it hasn't had competition, but partly

1:42:06

because it's essentially inherited monopoly

1:42:08

from better men. If you think about

1:42:10

the difference, think about a founder who sets up

1:42:13

a factory. Okay? They pass it down

1:42:15

to their air, who passes it down, and it's

1:42:17

just cranking out widgets and churning out money, and it's

1:42:19

a great grand or whatever of the founders narrowed

1:42:22

the factory. Okay? One day. And so it seems like the thing is working. It's,

1:42:24

you know, DuPont brand widgets

1:42:26

or something like that. Right? and

1:42:30

the great the great grandson famous looks

1:42:32

at him legally

1:42:35

as legitimate heir all

1:42:38

of that founder because, hey, look at here's all

1:42:40

the documents to chain a custody rep. Right?

1:42:42

Then

1:42:43

one find a that factory has

1:42:45

to switch from making widgets to making

1:42:47

masks or like servers or something

1:42:49

new. It has to

1:42:51

do something new. This great

1:42:54

grandson does not have the skill of the founder. They cannot change the assembly line. They don't know what they're neither

1:42:56

do the career

1:42:59

managers who have been hired

1:43:02

over the time. Like, the ability to do something from scratch, to do something new, that was the demand of the founder.

1:43:04

So all these people

1:43:06

are just running systems that

1:43:11

men better than

1:43:11

themselves set up years and years ago.

1:43:13

Right? They're not the

1:43:16

founders of

1:43:18

heirs. And so we understand this within the context of

1:43:20

a factory that the, like, great,

1:43:22

great grandson of Rockefeller or

1:43:25

DuPont or whatever is not Rockefeller

1:43:27

DuPont. I mean, in one sense, like, know, half your genes go to your child

1:43:29

and then half again and half more

1:43:32

generations you go down, it's not

1:43:34

like the same person anymore. It's

1:43:36

like, one one half to

1:43:38

the end's power in generation sound. Right? So not the same person,

1:43:41

that the same person but

1:43:42

they are they're legitimate.

1:43:44

They're just not competent. They were selected

1:43:46

for legitimacy but not

1:43:47

competence. Right? Commerce,

1:43:50

if

1:43:50

you select for confidence, but not

1:43:53

legitimacy. That's like someone taking over the factory, blab, blab,

1:43:55

blab, hostile takeover, or not even hostile takeover.

1:43:57

That's the least legal thing. It's

1:43:59

like seizing the factory. Maybe

1:44:01

they're actually good at operating it, but they're considered illegitimate the studies.

1:44:03

Like, you just stole it from them, blah blah. Okay? That's, like,

1:44:07

for example, NASH additionalization of oil fields and

1:44:09

stuff like that, you know, with various independence movements and this kind of thing. So you

1:44:12

want

1:44:13

is

1:44:15

both legitimate and and simply just

1:44:17

identifying that as an axis. Right? Legitimate and competent. Well, that's

1:44:19

why you need to have

1:44:23

re foundings. because nobody would have elected

1:44:26

Elon to run Tesla

1:44:28

or elected Mark Zuckerberg

1:44:29

to go in at thirty

1:44:31

something years old. run

1:44:34

a three point six billion communication network. He could only have proved it by doing it, by founding from scratch.

1:44:36

So the from scratch

1:44:38

aspect is so ridiculously important.

1:44:43

That relates to

1:44:43

all of our current institutions because there are decades, in

1:44:46

some

1:44:46

cases, hundreds of years old, and

1:44:48

they're just built for a different

1:44:51

time and by better men. the people today,

1:44:53

you know, like George Washington organized, like, the armies of the

1:44:55

United States from scratch. Right? The NYPD was at

1:44:57

one point organized from

1:45:00

scratch. Right? Even FDA remembers

1:45:02

organized from scratch at one point. Right? These organizations are now so many generations down

1:45:04

that they're just

1:45:07

inherited by heirs. Okay?

1:45:09

And sometimes they're literal errors in the and I talk about this in the book, in the founding versus inheriting

1:45:11

chapter, they just paste it in here. But sometimes

1:45:14

they're literal errors in the sense of,

1:45:16

like, a

1:45:19

DuPont or a Rockefeller or, you know, like

1:45:21

Sellsburger who's inherited the New York Times from

1:45:23

his father's father's father's father's

1:45:25

father. Right? Looks like five generations. And by the

1:45:27

way, just on that, I do talk about that in the book. It's like there's this website you should

1:45:30

check out. It's called

1:45:32

Tech journalism is less

1:45:35

diverse than tech dot com. And the point being that actually,

1:45:36

the journals

1:45:37

who endlessly talk

1:45:40

about tech diversity

1:45:42

and so it's our far less diverse than

1:45:44

tech itself. Right? A lot of

1:45:46

these tech journalists are basically employees

1:45:48

of some East Coast nepotist who's inherited

1:45:50

millions of dollars and or a newspaper

1:45:52

and is funding their whole operation. So

1:45:55

they're meritless nepotus and or

1:45:57

employees thereof. who are attacking the self

1:45:59

made. Right? And once you see that, you're like, oh, these these journals are basically,

1:46:01

like, dogs on the lease are hitmen

1:46:03

for old money. Right? assassance

1:46:07

for the establishment. They have no legitimacy whatsoever. It's literally old

1:46:10

money attacking new money. The

1:46:12

meritless.

1:46:13

Attacking the merit you know, the meritocracy or the more

1:46:16

meritocratic as privileged. It's like

1:46:18

actually this amazing inversion when

1:46:20

you when you apply that

1:46:22

lens strip. The thing is this entire establishment has

1:46:24

you know, it's declining legitimacy, but it

1:46:26

has legitimacy. It doesn't have confidence. And

1:46:29

so it can't innovate because it's

1:46:31

not selected for founders. like, if you're

1:46:33

a founder, you but you go to the US government or the US establishment. No.

1:46:35

That's a whole point as you can't found anything. You get like,

1:46:38

your career ambition is to

1:46:42

try to paste some language

1:46:44

into an eleven hundred page bill tonight

1:46:46

before it gets approved in a

1:46:48

vote where nobody reads it. Right?

1:46:50

Like, and it's deployed to the entire country. It's

1:46:53

imagine if that's how Google was coded.

1:46:55

Some giant political meeting,

1:46:57

blah blah, and someone, like, pays sense of JavaScript.

1:47:00

Obviously, that could be malware. Obviously, it

1:47:02

could be self interested. Obviously, that's not

1:47:04

the way to test something, to roll it

1:47:06

out, to deploy it. Like, Future historians will look at our current time

1:47:08

and these institutions as

1:47:11

so insane in

1:47:12

many ways. Like, just one example is

1:47:14

the thing I just mentioned where you're deploying

1:47:16

code to three hundred million people that you haven't

1:47:18

read. You haven't tested in the sandbox. You haven't got

1:47:21

any, you

1:47:22

know, room for iteration. the

1:47:25

the people who are responsible for writing it

1:47:27

aren't responsible for implementing it. It's like totally broken on that

1:47:30

level. On another level,

1:47:32

you're voting for somebody, and the contract isn't binding. They

1:47:34

give some campaign promise. They say they're gonna do something.

1:47:38

They're not gonna do it. It's literally like buying labeled milk

1:47:40

and getting orange juice. That's

1:47:42

fraud. In the commercial,

1:47:44

you know, in the

1:47:46

commercial setting that's fraud. Right? So electoral

1:47:48

fraud is when I mean, the most routineized version

1:47:50

of it is when a politician says they're gonna

1:47:53

do x, gets your vote on

1:47:55

that basis and does y. then

1:47:57

people say, oh, Bob, that's representative obnoxious, they're like, well,

1:47:59

your vote literally didn't count.

1:47:59

And your vote doesn't count. There's literally

1:48:02

no you have

1:48:03

no recourse. They've got, like, whereas,

1:48:06

you know, some form of immunity. Right? Basically,

1:48:08

you cannot sue them for not voting your

1:48:10

way. Right? So it's all completely symbolic.

1:48:13

I mean, those are just like two

1:48:15

examples, current you cannot

1:48:18

so

1:48:19

fix that system. you have

1:48:21

to figure out a way to get outside it. Why do

1:48:23

you want to get outside it? Well, I mentioned the FDA has a huge motivation, but there's of others.

1:48:26

You want self driving? you're

1:48:28

gonna need sovereignty. Do you want to draw on

1:48:30

delivery? You're gonna need some form of sovereignty. Do you want, like, you know, to be able to have

1:48:35

nuclear power? probably gonna need some form of sovereignty. Like Wyoming is pushing some of this. There's good

1:48:37

things happening in some places that are like accelerating it.

1:48:39

You want it in the physical world, you're gonna need

1:48:41

some form of sovereignty. One of the big recipes of

1:48:43

the network state is not

1:48:46

just say, Till's correct observation, you can innovate in bits, you can't

1:48:47

innovate in atoms, you can build a billion dollar business online, you need

1:48:49

a billion promise to build a shed

1:48:51

in San Francisco. That

1:48:55

is true. We've reconciled attention by using bits

1:48:57

to innovate on atoms, by building

1:48:59

this aligned community online. And the alignment

1:49:01

is ridiculously important. They need to do

1:49:04

the dance They need to coordinate and

1:49:06

choreograph together. You need to be able to if they can't all hit like on something, they literally can't do anything

1:49:08

else. But if

1:49:11

they can do that, they might be

1:49:13

able to do a lot of other things. You turn them

1:49:15

into a multicellular organism. Right? The content is also really important. It's not just a code. It's community in the

1:49:17

content as well as

1:49:19

the code. Right? So all the messaging

1:49:22

they're putting out. The declaration of independence, it actually had that thing at the beginning, which says, a decent

1:49:24

respect opinions

1:49:27

of mankind requires that should declare the causes which impel

1:49:30

them to the separation. Okay?

1:49:32

That's a really critical

1:49:34

thing. Essentially, something that tech

1:49:36

people underestimate

1:49:38

and undervalued, is that something like ten or twenty or thirty

1:49:41

percent of

1:49:44

your actions you need

1:49:46

to allocate, like, at the header, just like HP headers. Right? You need to allocate space for the moral justification for

1:49:48

what you're doing and

1:49:51

almost everything you're doing. basically,

1:49:54

let's just say first order everything. Right? It's almost like, I'm doing this because and then you do For example,

1:49:56

if you're fighting

1:49:59

back on Twitter,

1:50:00

fighting back on twitter you

1:50:03

might

1:50:03

think that everybody has seen the

1:50:05

full context of the whatever number

1:50:07

of tweets and attacks and so

1:50:09

on back and forth. You basically

1:50:11

have to assume that most people are coming in with absolutely

1:50:13

zero context. And in your limited space,

1:50:15

you need to include

1:50:18

the reason that this person

1:50:19

you know, like,

1:50:20

this person attacked you in the first

1:50:22

place, you're defending yourself, and

1:50:25

then the attack. So you're literally this

1:50:27

very spare space, spare pass you have there, you

1:50:29

have to put your justification for the

1:50:31

attack or the counterattack or the

1:50:33

defense before you actually do the

1:50:36

defense. Right? in the same way for

1:50:38

every product that moral justification we are doing this because x is good and y is bad.

1:50:40

Right? That is there

1:50:43

and that is like just

1:50:45

baked in whether it's a logo or it's a slogan or it's a mission statement

1:50:47

or it's a catechism, or

1:50:52

whatever you wanna call, I call

1:50:55

it the one commandment. Right? Which is

1:50:57

your moral differentiation from society at large.

1:50:59

That tells you,

1:51:00

okay, Seizar has

1:51:02

gotten something wrong. We're gonna get it right. And to get it right, we're gonna need to shape physical

1:51:04

world. So you

1:51:05

use bits to come

1:51:08

to consensus. and

1:51:10

you say, this group is now aligned, we

1:51:12

need self driving. Why? Because

1:51:15

the current NHSTA is an

1:51:17

abomination. The current thing of tens of

1:51:19

thousands of automobile deaths here as an abomination.

1:51:21

We have the technology to fix it. We

1:51:23

can't just retrofit

1:51:26

the roads necessarily. Why? Because It's like trying to send maybe electrical

1:51:28

power over a waterline. Okay? The current

1:51:30

roads were not built for self driving.

1:51:32

If you built a town for

1:51:34

self driving, it would look totally different.

1:51:37

you would have the roads would have

1:51:39

sensors instrumented there. might be farther away from wouldn't have, like,

1:51:42

human crossings of the

1:51:44

roads. all kinds of

1:51:46

things, you can engineer your way out of the problem where you just assume the road is completely self driving.

1:51:48

Everything changes in terms of an

1:51:50

engineering thing. You're not trying to send

1:51:54

electrical power over waterline. That's one of the reason self driving is

1:51:56

so hard is you have to assume

1:51:58

all of this legacy baggage there. If

1:52:00

you could take that away, it could be easy.

1:52:03

And in fact, We know this to be the case because guys like Rio Tinto, this is

1:52:05

a this is a huge mining conglomerate. Even as

1:52:07

far back as like twenty thirteen,

1:52:11

on closed roads, on

1:52:12

private mines, they're running. When they're the only ones who

1:52:15

are running the miners and trucks, they just run them

1:52:17

all from Perth in Australia. They can

1:52:19

run the mine remotely. okay,

1:52:22

and move the things around because there's no one

1:52:24

else on the road or because it's a private road where

1:52:26

they know all the cars on. Right? This is a

1:52:28

great example of self driving car zone. get this moral

1:52:30

commitment, you align people with their bits,

1:52:32

you get that

1:52:33

alignment, and then they go and

1:52:35

take physical territory and they

1:52:37

just relentlessly negotiate with enough governments

1:52:40

until they get a sanctuary city or they get a

1:52:42

deharmonized zone. And they don't just replace it with nothing.

1:52:44

They replace it with a new regulatory power. one

1:52:46

clarifying question. So when you say that

1:52:48

they develop their own regulatory framework,

1:52:50

is that just using the example

1:52:52

of self driving? Is that just

1:52:54

around self driving, or is the intention that

1:52:56

they regulate or reregulate everything

1:52:59

from education to taxes,

1:53:01

etcetera, or is it

1:53:02

embedded Right? So the one commandant concept is I think it's

1:53:04

a it's a useful thing in the book. It's actually one of

1:53:06

the more important things in the book. So

1:53:09

I'll get the short version and the long version. the short

1:53:11

version is, you cannot change everything society at once, but you

1:53:13

can also not give up completely on

1:53:15

societal change. So just like a

1:53:18

startup sets out to fix one

1:53:20

thing, right,

1:53:20

in the market. A

1:53:22

starved

1:53:22

society sets out to fix

1:53:24

society one

1:53:25

moral failing

1:53:28

of society. Okay? So

1:53:29

you're not saying, okay, profitable, unprofitable. You're saying

1:53:31

good, bad. You're starting with a

1:53:34

moral premise. It's not

1:53:35

a market premise. this

1:53:39

is really important. Like, a lot of dows, for example, will

1:53:41

start with the market premise. And they'll be

1:53:43

like, hey, you can all

1:53:45

make money come here. Right? And, you know, what it reminds

1:53:47

me of is that scene for mediocrity. You

1:53:49

know, the guy's like, if you'll

1:53:51

like socks and Monty

1:53:53

too, wow, we too. Right? Like, you

1:53:56

know, it's like, these are these are like

1:53:58

the human universals where there's absolutely no

1:53:59

community there. People are just They're for the

1:54:02

money. They're they're when a drop. Right? When token. It's like joking, but it's like, you know, dogs, like, waiting

1:54:04

for some scraps and those grab a

1:54:06

scrap and then run off or whatever. Right?

1:54:10

And so there's nothing in that's not a real community.

1:54:12

That's just a bunch of people who are just there

1:54:14

for the air drop or something like that. Right?

1:54:17

So join a bunch of discords for the

1:54:19

air drop or whatever. Right? So starting

1:54:21

money first is wrong.

1:54:23

Right? Money comes, it's a superstructure on top of a

1:54:27

community. Okay? It's a way for the community to represent its debits

1:54:29

and credits between each other. And that's

1:54:31

why you see a

1:54:33

bunch of these coins drop to zero because there's no

1:54:35

true community. There's speculation on whether they build a community

1:54:37

and that drops to zero because they don't build

1:54:39

a community. One

1:54:40

way of thinking about it is So one command what's

1:54:43

what are examples of one command? So I

1:54:45

mentioned the the digital sabbath example.

1:54:47

Right? Where you're saying,

1:54:49

look, technology is technology is good. but you can have too much

1:54:51

of a good Just like, you know, highways, I think the

1:54:54

internal combustion engine is good, cars are

1:54:56

good, generally. I mean, electric

1:54:58

engines are even better or whatever.

1:55:00

but you could overdo it where San

1:55:02

Francisco, for example, had a highway that blocked access to the waterfront. And

1:55:05

so you had

1:55:07

overdone it. See, You

1:55:09

say, I'm not against cars. I want AV3 that takes

1:55:11

away that highway, so we have a walkable waterfront. And so in the

1:55:13

same way, you're like, look, I'm pro Internet,

1:55:15

but I'm not pro all

1:55:18

the time with all the notifications. So we

1:55:20

have community support and we say we're offline twelve

1:55:22

hours a day. And just Internet has cut

1:55:24

twelve hours a day. And everybody's now just like

1:55:26

the weekend synchronizes people. And they're like, okay, we have

1:55:28

Saturday and Sunday off. Right? The

1:55:30

Southern Union actually tried making it

1:55:32

so that they could work seven days

1:55:34

a week. So you had Mondays and and Tuesdays

1:55:37

off and somebody else said Tuesdays and

1:55:39

Wednesdays and somebody else said Wednesdays

1:55:41

and Thursdays. They tried these kinds

1:55:43

of experiments. Okay? And the

1:55:45

problem is that there's a utility of everybody knowing that Saturday and Sunday are off generally

1:55:47

means you don't have the coordination. Imagine

1:55:49

you have to schedule your

1:55:52

weekend like and

1:55:54

you could never overlap with somebody

1:55:56

else. Okay? So coordination and everybody

1:55:58

picking the same defaults totally

1:55:59

changes things. once you have

1:56:02

everybody except in this community that I'm offline nine PM to nine AM. Guess what?

1:56:06

that's what

1:56:07

That changes the community. Okay. Hey, guys.

1:56:09

Let's go and, you know, we'll go for a a run or or, you

1:56:11

know, we're all gonna have outdoors dinner

1:56:13

or something like that. I

1:56:16

don't know. you could

1:56:18

figure out the exact timing. And maybe on

1:56:20

the weekends, it's actually like you're offline twelve PM, but like

1:56:22

an entire day you're off or something like that. Right? And

1:56:26

this

1:56:26

is offline day, and we're gonna all do things.

1:56:28

Right? And so once you have that coordination

1:56:30

just like weekend, new things arise out of that

1:56:32

when you can assume everybody's point to the same

1:56:34

thing. literally new societal adaptations arise. And the

1:56:37

thing about this

1:56:38

is money runs

1:56:40

out,

1:56:40

but a

1:56:42

moral premise doesn't. being part of that community, if

1:56:44

digital sabbath is important to you, being

1:56:46

offline, some of the day is important to

1:56:50

you. If, you know, like, just okay, offline,

1:56:52

and it's okay that I'm offline.

1:56:54

And as this community expands, like more

1:56:56

and more style conventions, people

1:56:58

know not to email you then.

1:57:00

you're offline, by the way, nine PM to

1:57:03

nine AM, ninety five percent of jobs will be fine. Right? Like, some jobs or whatever,

1:57:05

you need to be site

1:57:07

reliability engineer or something like

1:57:09

that. It's more than, like, more than ninety five

1:57:11

percent. Like, most don't expect you. know, that's something

1:57:12

like that relative to

1:57:14

where normally working times are.

1:57:17

people expect you to be asleep eight hours,

1:57:19

something like that. So, nine PM to ten AM, you should be pretty much

1:57:22

okay even outside your site. But as it grows, it becomes a million

1:57:24

people. It's

1:57:27

a totally different world. You know? Like, all kinds

1:57:29

of things get scheduled for offline time,

1:57:31

for offline time, the coordination thing.

1:57:33

This is a resource that doesn't run out.

1:57:35

Money runs out.

1:57:36

Morality and moral premise doesn't. This

1:57:37

is differentiated from the outside world in like

1:57:40

a permanent and persistent and

1:57:42

interesting way. It's like a wellspring.

1:57:44

that doesn't run out. Right? Another example,

1:57:46

keto Kosher. Okay? So you've got a network archipelago by my

1:57:48

definition.

1:57:49

So you don't just

1:57:51

have an online community you've

1:57:53

crowds into territory and you're treating sugar like

1:57:55

cocaine. Okay? You're interditing

1:57:58

it at

1:57:59

the border. you're

1:58:01

literally not allowing so every store in every restaurant

1:58:03

is not filled with high fructose

1:58:07

corn syrup and chocolate. because

1:58:10

we have a you have these really sophisticated corporations that have it

1:58:14

have sherlock up so

1:58:16

that when

1:58:16

you're at the checkout line for

1:58:19

something, you're hitting the face for five minutes with some, you know, chewy chocolate

1:58:21

like

1:58:21

sugary kind of

1:58:24

thing. Right? So

1:58:26

the point is, like, to wear down your resistance. You just wanna be on a diet or something and but like,

1:58:32

literally, this experts from Madison

1:58:34

Avenue have set up something to be maximally tempting to you and put it in the checkout line so you're looking

1:58:36

at it the whole time. Right? So

1:58:38

you're trying to whittle it down. Right?

1:58:42

against that giant corporation, your community

1:58:45

collectively can provide

1:58:46

some resistance. By setting the

1:58:48

defaults in a different way, Right?

1:58:50

Now, you know, what I'm describing there is, like,

1:58:52

sort of outside of the normal, like, left right

1:58:54

kind of spectrum. Right? Because you're talking about

1:58:56

a community that's an interest of the community. It's

1:58:59

neither it's not a left or right. It's like it's basically something which is just for

1:59:01

what the community's values are. Right? And

1:59:03

what they would do is they'd say,

1:59:05

okay, no sugar at the border,

1:59:07

just not there. You have healthy food, you

1:59:09

have lettuce, you have fruits, you have tomatoes, you have other kinds of stuff.

1:59:12

Every meal

1:59:15

is healthy. You know, you don't have to ask what are

1:59:15

your dietary restriction surveys now? Snap to grid. It's

1:59:18

not I'm offline nine PM to

1:59:20

nine AM. It is I am Snap

1:59:22

to grid on a different thing, which is

1:59:24

I'm either full keto or

1:59:26

I'm just very low sugar or some something like that. You can imagine variance of this. There's keto kosher, there's papillae

1:59:32

people, there's a carnivore community. You

1:59:34

could also have the Beacon village, which is as different from the carnivore community

1:59:36

as both are

1:59:39

from mainstream society but I

1:59:41

would bet that both the vegan village and the

1:59:43

carnivore community would both be far healthier

1:59:48

than you know, McDonald's eating mainstream America. You know

1:59:50

what I mean? Like, they're quite different from each other,

1:59:51

but those are both bacon

1:59:54

village. They're eating lettuce and

1:59:56

tomatoes and stuff, carnivore. They're eating

1:59:58

at least. They're eating, like, real meat and stuff, but they're not eating this

1:59:59

processed, stuff they're nodding this process

2:00:03

horrible stuff. Right?

2:00:04

as

2:00:05

they're probably twenty pounds lighter, thirty pounds lighter. Okay?

2:00:07

And this, again, this is a resource that doesn't run out. When you

2:00:09

join that community, that

2:00:12

moral premise where

2:00:14

you've

2:00:14

inverted something that society said was

2:00:17

good. You're saying it's bad. Sunny said it's

2:00:19

good or at least acceptable to have sugar.

2:00:21

You're saying it's actually morally bad. and everybody in

2:00:23

your community agrees with that. So you get social support for that. The

2:00:25

defaults are set for that. You literally need to

2:00:27

travel outside of your day, outside of the

2:00:29

town, to go and pick it up.

2:00:31

It's literally like getting drugs. Right? So you've now got several levels of defense. And

2:00:35

and this

2:00:37

is, again, it's morality focused. Right? And, you know, what's interesting

2:00:39

about this, by the way, just to put it on to

2:00:42

relate to something else. Mike

2:00:44

Maritz, famous investor at Sequoia. He's like, quasi

2:00:46

retired now, famous guy. And he has this saying,

2:00:49

which I think

2:00:51

is very clever

2:00:53

that all

2:00:53

of Sequoia's best

2:00:54

starts the Samsung one seven deadly cents. Okay.

2:00:59

So it's like,

2:01:01

sloth

2:01:01

or gluttony, pride, lust, etcetera. Right? And

2:01:03

what he means by that is

2:01:05

it's really hard to

2:01:07

build a company And

2:01:10

so you need to satisfy some sort

2:01:12

of visceral drive. Right? The more intellectual, the more academic,

2:01:14

the more idealistic one's view of human nature like

2:01:18

we will all gather to deliberate on x. Right?

2:01:21

The less likely that is going

2:01:23

to be a thing. Right?

2:01:25

Not impossible, but less likely. And so once you

2:01:27

start doing this, you can actually take a lot of startups

2:01:29

and you can map them. Tinder is

2:01:32

lost. Twitter is

2:01:34

wrath. Instagram is pride and

2:01:36

so on and so forth. Right? Uber Eats is

2:01:38

like gluttony. Right? And many things approach of

2:01:40

the apps are arguably slots or whatever.

2:01:42

Right? because that it is hard to build a company and you do wanna

2:01:44

align it to a visceral drive. The problem

2:01:46

is that the seven deadly sins or

2:01:49

sins for a reason because you can overdo

2:01:51

it. all relate to lack of

2:01:54

self control. And the and

2:01:56

when you

2:01:57

have lack of

2:01:59

self that's bad in various ways. You're fat. You're not on

2:02:01

a diet or something like that. You spend too much money. You know, you

2:02:03

get mad when

2:02:06

you shouldn't get mad. like not being in control of yourself often means if

2:02:08

you don't have self control, other control is imposed

2:02:10

on you. Others will control you. Right?

2:02:13

If you don't control yourself, others will control you in

2:02:15

some which is that. The other issue is that once companies get really big, they're

2:02:17

no longer simply like fighting for

2:02:20

survival and they have the special drive.

2:02:22

Instead, they're kind of actually creating the

2:02:24

deadly in. Netflix,

2:02:26

you know, there is this is an off

2:02:28

end comment. I'm not like holding it against them or whatever, but they did

2:02:30

say something like, Netflix is competing with, like, having wine with your wife.

2:02:35

Right. They're competing for your time. Yeah. But they said

2:02:37

attention that you have. Yeah. Yeah. It's it's

2:02:39

like this famous quote. It's like competing

2:02:41

for all of your time, including, like,

2:02:43

wine with your spouse or something

2:02:45

like that. It's meant to be tongue in

2:02:48

cheek, but it's also true. It's like people would

2:02:50

rather have that than Netflix most of the time.

2:02:52

Right? So the issue is

2:02:54

that these companies, once they get to a certain level, are exacerbating genuine light sins. And

2:02:57

in many ways,

2:03:00

America's become Dutch America, the west

2:03:02

where our lots of world has become like a simple society in that sense. Right? Where people are being out

2:03:07

of their desired self control by these

2:03:10

companies that are pulling them in these directions for private profit. Right? And now

2:03:14

again, I'm like, you're like, wow, this is a really different tone from you. I thought you're like

2:03:16

a super capitalist pro pro profit VC

2:03:18

kind of person. I am, but

2:03:23

I'm also really fundamentally a community person with, you know,

2:03:25

both capitalism and democracy, by the way, are

2:03:27

tools or conflict resolution tools. You

2:03:29

know, democracy is elections, capitalism's

2:03:31

markets. But, like, within

2:03:33

a family or even within a

2:03:35

company a community, not is an election an auction. Many things

2:03:38

are just a decision that

2:03:42

is arrived at organically where people

2:03:44

are just aligned. They're harmonious. You know? It doesn't have

2:03:46

to be Like, it's it's a conflict where

2:03:48

you've got multiple competing options being in the

2:03:51

election. you'd have an auction because people really want to stay on a bit

2:03:53

on it. Lots of things can be allocated by community if

2:03:55

you have a real community. What's

2:03:57

the opposite of the seven deadly sense? while keto Kosher is the

2:03:59

opposite of

2:03:59

gluttony. Right? It's satiety.

2:04:01

Right? Digital Sabbath is

2:04:04

like, I mean, it's opposite of

2:04:06

several things it's probably the opposite of, like, wrath. You're not getting mad

2:04:08

on Twitter. It's opposite of sloth. You're getting out

2:04:10

of the house, right, rather than being online

2:04:13

all the time. And so now the opposite,

2:04:15

in some ways, the start company is saying, look, it's hard to build a start up so

2:04:17

we have this visceral

2:04:19

drive they're satisfying. here

2:04:22

what we do is we actually flip it and probably

2:04:24

you could take the seven cardinal virtues

2:04:27

and build the start society on

2:04:29

this

2:04:29

each of the seven cardinal

2:04:32

bushes, you invert the merits thing. Right? And

2:04:34

again, I'm not critiquing Mike here, like, his

2:04:36

framework is a good framework

2:04:38

in the sense of satisfying those visceral desires, but any good thing can be overdone. Okay? And I think we've overdone in

2:04:40

some ways. How do

2:04:43

you bring it back? these

2:04:45

startup societies as opposed to companies have a moral innovation that may

2:04:47

start with the seven cardinal virtues or these are

2:04:49

things that are like this

2:04:51

in other religion seven parishes

2:04:53

from Catholicism, but there's other religious kind of things. You take one of those and you're like, keto kosher is

2:04:56

attacking the gluttony problem. How are we

2:04:58

doing this? The community is giving you

2:05:00

support. our

2:05:02

one commitment is sugar is bad. Right? And

2:05:05

then that one

2:05:06

commitment, by the

2:05:07

way, leads to other things. If

2:05:09

sugar is bad, continuous glucose monitoring

2:05:11

might be good. So everybody gets CGM

2:05:13

meters. And now you're snapped and grid in a different way. You can assume everybody

2:05:15

in society, a thousand people, ten

2:05:18

thousand people, a hundred thousand people

2:05:20

have continuous glucose

2:05:22

responders. And maybe they all opt in in some privacy preserving data sharing way to say, here is

2:05:24

the glucose response of, you know, when

2:05:26

I eat lettuce, I get this result.

2:05:31

then you go further down the tech tree, hey, I'm not just monitoring glucose, I'm monitoring my vitamin

2:05:33

D levels, and this level, and that level,

2:05:35

so you get

2:05:38

a quantified self community out of that. Right? So once you have that the

2:05:40

moral innovation actually enables the

2:05:42

technological innovation. This is something

2:05:45

that we've sort

2:05:47

of forgotten because what's happened in modern society

2:05:49

is the technological

2:05:50

progressives and

2:05:51

the political progressives have separated and the

2:05:53

moral innovators who are coming up

2:05:55

with new slogans online are

2:05:57

different than the people who are making

2:05:59

new technologies. In the past, that was actually one

2:06:01

movement. For example, public health, right? In the

2:06:03

early nineteen

2:06:04

hundreds, the

2:06:06

moral innovation was like cleanliness is next to godliness. Let's not all be filthy and so on. The technical innovation were like

2:06:08

sewers, right, handwashing, and all

2:06:10

that type of stuff. Right? And

2:06:15

those two things went hand in hand. That moral innovation doesn't

2:06:17

cost you anything. But it

2:06:19

does mean you actually have

2:06:21

to be concrete about what you

2:06:24

believe in that is different than sight at large.

2:06:26

And I'm not saying, come up with your own ten commandments and your whole new religion. I am saying, however,

2:06:31

one commandment just like a focused has one thing. If you have one commitment like sugar is bad and then

2:06:33

you can take that, you build a community out of that, and

2:06:35

then you'll get your second and

2:06:38

third and fourth derivation for

2:06:40

that, the CGM stuff and so on that

2:06:42

comes out of it. Right? And that's just one example. But in this fashion, everybody has

2:06:45

one thing that they

2:06:47

think there may be a few things, but often they can infect one thing they think it was, like,

2:06:49

deep problem with US society. Like, they might say,

2:06:51

oh, people are gambling

2:06:53

too much. on coins and

2:06:56

assets and so on and so forth. They might say people

2:06:58

are eating too much or people are too mad online. They're

2:07:00

canceling each other or they're too lazy and

2:07:02

they're not fit. or or something. A lot of these things line up with the cardinal virtues thing

2:07:05

or something like that. Right? Or

2:07:07

Tinder encourages, like, bad personal

2:07:10

relationships and we need something that's you know, like much higher commitment level and

2:07:12

much more courting beforehand. Right? Everybody has

2:07:14

some critique of society like this. And

2:07:16

now, you can be present of

2:07:19

your own start up society. Okay?

2:07:21

You can start solving the problem because you set up your shingle, your one commitment, you have

2:07:23

your whole set of graphs and charts,

2:07:26

all your historical arguments. that

2:07:30

say, why you actually think this is bad? You're willing

2:07:32

to put this at the top of your tour

2:07:34

profile and defend this in public. Okay? Because

2:07:36

people will the thing is your arguing society

2:07:38

is wrong. Right? You're arguing, you are right. Society is wrong. And then

2:07:40

you've recruited community of like minded people that

2:07:42

believe in this. And then, either it's

2:07:45

online or it's offline or it's some combination or you

2:07:47

start getting territory. And then maybe

2:07:49

you eventually need, like, a sovereign

2:07:52

recognition to become a full network state. but even

2:07:54

a network union or a network archipelago can get very far. You are solving the problem for yourself. You are remember the thing we talking about,

2:07:56

the Boschiat, the scene

2:07:58

and unseen, which you mentioned,

2:08:02

you are now making the unseen scene.

2:08:04

You are showing in the sugar

2:08:06

free zone what the impact of

2:08:08

sugar is on the sugar full

2:08:10

zone. Everybody's thirty pounds lighter here. Right? Diabetes has

2:08:12

dropped

2:08:12

x percent. Everybody's like much better looking.

2:08:14

Oh

2:08:15

my god. Right? Whatever. Right?

2:08:17

You know, like, they have just all these other

2:08:19

conditions maybe go away when they're not overdosing in sugar all

2:08:21

the time. And so that thousand

2:08:23

people reforms a million people, ten million

2:08:25

people because they make movies about what

2:08:28

they're doing. I

2:08:28

like the analogy or almost

2:08:30

the framing of evidence over confidence. Right? In the in the

2:08:32

societies that we have

2:08:35

today, they tell us what

2:08:37

should be done and perhaps the best way to to

2:08:39

actually show an alternative is through evidence. Right? So through these small

2:08:41

societies that show not say

2:08:43

what is better.

2:08:46

That's exactly right. And basically, the key

2:08:48

thing is so I mentioned the term

2:08:50

president of a startup society, president of a

2:08:52

network society. Okay? The number of people

2:08:54

in the world who could feasibly become

2:08:56

the United States is actually so you

2:08:59

have to be like thirty five years old, natural born American citizen, some other requirements. Okay? So only

2:09:01

four percent of the world

2:09:03

is American. And so apply

2:09:07

those requirements, it probably chops in half again. Right? Let's say

2:09:09

it's maybe on the other two percent of

2:09:11

people could become president. And we're told this

2:09:13

thing in school that anybody can become

2:09:15

president and so on. but it actually

2:09:17

means ninety eight percent of the world cannot become president of United States. And even

2:09:19

the president of United States cannot actually fire these regulators who

2:09:21

have career tenure and so

2:09:23

on and so there's

2:09:25

a whole to do over that. Right?

2:09:27

Schedule f. And practically speaking, first, ninety eight percent of world actually become

2:09:32

president. and that president doesn't even empower

2:09:34

these regulatory agencies. So the current pathway cannot fix those regulatory agencies. There is no path

2:09:36

to reform within the

2:09:39

system. It's literally like making

2:09:42

blockbuster into Netflix. Right? Making Barnes and Noble, try and become Amazon, making BlackBerry

2:09:48

into Apple, you couldn't do it. You

2:09:50

just disrupted. You had to build the alternative system. Right? You could not have turned England into

2:09:52

America or the UK

2:09:53

into America. You had to

2:09:55

build America. Right? So

2:09:58

now

2:09:59

though, you declare yourself president of a

2:10:01

Starb Society. And guess what? The vast majority

2:10:03

of the world, they'll

2:10:04

all laugh at you. Law would look

2:10:06

at this idiot. like, oh, your president is sort of

2:10:08

old. Right? You know, it's like, you

2:10:10

know, a big issue, your imaginary president,

2:10:12

and you want that. You know

2:10:14

why

2:10:14

you want them to do that? because they

2:10:16

won't

2:10:16

stop you. Exactly. If you do have enough power, if

2:10:18

you do have enough attention, then you are actually putting

2:10:21

friction between you and your goal for

2:10:23

That's right time.

2:10:24

That's right. So so

2:10:26

like the whole gondi line which is overused, but it's also, you know, first day laughing. You want them to laugh at you. Right?

2:10:28

They're not taking me seriously.

2:10:31

You know what? Because here's

2:10:34

the thing. You don't need to win an election. You do

2:10:36

not need to

2:10:37

get fifty one percent support. That thing you're

2:10:39

mentioning at the beginning of, will

2:10:41

the majority go this? everything doesn't have to be

2:10:43

calibrated to whether the low attention

2:10:45

voter can be convinced of

2:10:47

this premise. And actually, the way

2:10:49

they're convinced, here's another deficit of

2:10:51

the current system. It's all acting. You're increasingly

2:10:53

seeing people who are selected just for being influencers or actors in

2:10:55

a literal sense, you

2:10:59

know, whether it's Al Franklin or it's Trump

2:11:01

or it's Reagan or it's you have folks who are literally actors and

2:11:03

are selected for

2:11:07

their acting ability. to

2:11:08

be politicians. But actors are liars.

2:11:09

Like, we don't think of it that because actors

2:11:11

are procedures and

2:11:12

liars, not procedures, but an

2:11:14

actor is a very skilled liar.

2:11:17

They're

2:11:17

they're incredibly convincing on camera. You believe they have that emotion. They've blended to

2:11:20

the character.

2:11:24

And then cut and they stop

2:11:26

crying. Just get a tissue. They're like, alright, what's my next scene? You know? Like that. Right? And so

2:11:28

you're selecting for

2:11:31

politicians that are liars.

2:11:35

Because, you know, there's a

2:11:37

saying like actions speak louder than words.

2:11:39

They actually don't Words speak louder

2:11:40

than actions. You know why? you

2:11:42

see the Twitter feed of, I don't know, AOC or,

2:11:45

like, you know, any other politician,

2:11:47

whatever. You see their

2:11:48

Twitter feed. But most people

2:11:50

are not looking at anywhere near us constantly is

2:11:52

not just their voting record,

2:11:54

but their actual actions. Their

2:11:57

actions are not legible in the

2:11:59

same way. one of the things that crypto by the way, is we're gonna be

2:12:01

turning the LinkedIn resume of self

2:12:03

declared things into a

2:12:06

feed of crypto credentials that is awarded

2:12:08

to people by you. You close the sale, you get a cryptic

2:12:10

credential. It's an on chain proof from your CEO at the time that you

2:12:12

close the sale for this

2:12:15

amount at this time. you solve the

2:12:17

math problem, you get a cryptic credential from your professor.

2:12:19

Right? So you now have a machine readable, on

2:12:22

chain resume that's provable,

2:12:25

It's not simply assertions, it's timestamped, it happened at

2:12:27

the time, it's digitally signed by that person. Now actions

2:12:29

start speaking on the same

2:12:31

scale as words. And

2:12:35

now you can start a word because there's a feed of them. It's visible in

2:12:37

the same way as the words online, it's if

2:12:39

it's on chain. And now

2:12:42

you can actually start filtering your society by those people who actually are doing actions

2:12:44

and not just words or actions plus

2:12:46

words. Right? Actions plus words is valuable,

2:12:49

actions alone is valuable, words loan is usually not as valuable.

2:12:51

And so, you know, you put this together

2:12:54

and what you've got is something where

2:12:56

your concept of evidence over

2:12:58

confidence is really good because confidence

2:13:01

is the confidence of the actor. It's

2:13:03

also the confidence of the confidence man. Okay? Evidence evidence of

2:13:07

one's own eyes. and here's the thing. You don't have

2:13:09

to go and move to the society. You can tour the society. You can tour it in VR.

2:13:11

You can tour it

2:13:14

online. In fact, a good start of society, as I mentioned, you know, twenty, thirty

2:13:16

percent of your case is making the moral

2:13:18

case for what you're doing. They're constantly

2:13:21

preaching their case

2:13:24

from scratch over and over

2:13:26

again. Just like you're constantly pitching, you're startup. Right? That tagline that you've heard ten thousand

2:13:28

times, that person on their side

2:13:30

is hearing once the first time. Mhmm.

2:13:34

Right? Yep. So you're constantly making them moral case. You're making in videos.

2:13:37

You're making in content because you're recruiting people to

2:13:39

your community. But even more than that,

2:13:41

you're trying to set an example for the rest of

2:13:43

the world. and they may not adopt the entire Okay? They may not say,

2:13:45

oh, you know, well, look, digital sabbath,

2:13:47

we can't

2:13:48

be offline

2:13:50

twelve hour day. Maybe we'll do eight or some silly iteration thereof.

2:13:52

Right? But that's good. It's

2:13:54

just like Google's example of

2:13:58

using Linux eventually force reform on

2:13:59

Microsoft. And now they own GitHub or whatever. Right?

2:14:02

That would never have been done within the organization

2:14:05

the reform had to come from without. but it

2:14:07

was reform. It was but it was practical reform. It wasn't saying,

2:14:09

hey, let's rely on the same to change. You

2:14:11

have the 1000x. You have the

2:14:13

people exiting to that jurisdiction.

2:14:15

You have the loss of faith in

2:14:17

the old, but you also have the building up of faith in the new and that forces reform. Yeah.

2:14:19

I think the

2:14:21

parallel to

2:14:24

startups is really illuminating

2:14:25

in the sense that every startup, as you said, has

2:14:27

some sort of

2:14:27

tagline, some sort of minor thing, minor

2:14:30

at the time that they're

2:14:32

founding, that

2:14:34

can extend. It has inertia. In some

2:14:36

cases, all the way to Google. In some cases,

2:14:38

it ends up just being a bootstrap startup.

2:14:41

In some cases,

2:14:41

it's somewhere in between. And they similar

2:14:43

to the idea of these startup societies

2:14:45

are often ignored, and then as

2:14:47

they get bigger and bigger and

2:14:49

bigger, they become too big to be ignored. And then there's

2:14:51

other routes at that juncture too. Right? Some

2:14:54

of them are absorbed and bought. Some

2:14:57

of them continue to ascend and actually

2:14:59

overthrow the existing company in that space. And so, yeah, I

2:15:01

think it's really interesting to consider that many

2:15:04

of these startup

2:15:06

society ideas will sound

2:15:08

silly. too many people at

2:15:10

first, but they do become something or they evolve over

2:15:12

time

2:15:12

or they can evolve.

2:15:14

That's right. And basically, like startups

2:15:16

Start

2:15:18

societies are like start companies in this way, where start companies,

2:15:20

as I mentioned in the book, it's like, you

2:15:23

think about Twitter or SpaceX. Right? tour

2:15:25

sounded insanely trivial. SpaceX sounded insanely ambitious. Okay? Oh,

2:15:28

hundred forty characters, breakfast, tweet,

2:15:29

and

2:15:30

what

2:15:31

is this? You know? And

2:15:35

SpaceX sounded, okay. Yeah. Sure. You're gonna be

2:15:37

NASA. Alright. Good luck. Right? So

2:15:39

essentially, they're at opposite ends

2:15:41

of the spectrum. But they're both now, like,

2:15:43

ten billion dollar companies have changed the world. One

2:15:45

of them was considered attributable, but it was

2:15:47

in feasible. Lots of start

2:15:49

up things are like that where they're just outside the room to window and

2:15:51

all of it is in the execution. And so, you know, guy, Adan Levine, had

2:15:54

this great tweet where he's

2:15:56

like, what

2:15:58

is, like, the craziest thing and he's,

2:16:00

like, a fan of of this. So he's, like,

2:16:02

what is the craziest thing that would work as

2:16:04

a one commandant? And that's a What is

2:16:06

your answer to that? I mean, things couldn't

2:16:08

even predict. Right? I mean,

2:16:10

the obvious stuff is like language

2:16:13

communities, you know. like religious communities.

2:16:15

You know, you're gonna have for example, like,

2:16:17

Rob Dreyher has this book called a

2:16:19

Benjekt option. Right? If

2:16:21

you want to be a religious Christian,

2:16:23

Now actually that. And

2:16:24

you don't have to impose your

2:16:26

values

2:16:26

on the rest of society. So

2:16:29

it's at d three. Right? You're not imposing Christianity on people who don't agree with

2:16:31

it. Nor are

2:16:35

you seeing your Christian culture eroded or

2:16:37

what have you. you go and gather with others and you actually go in craft and terror and

2:16:39

you practice and you actually

2:16:43

live that holy life. And that's always harder

2:16:45

to do in practice than in deer. people will have your if you heard the

2:16:47

term tran. Right? People will lark being tran. I

2:16:50

know that's two interested

2:16:53

abbreviations in one. Okay?

2:16:55

But they will sort of pretend that they

2:16:58

are like super traditional or whatever

2:17:00

online because they inhabit this sort

2:17:02

of fantasy world.

2:17:03

But the actuality of how the

2:17:05

trad stuff interacts with modernity, you know, for

2:17:07

example, in Judaism, there's something called like modern

2:17:09

orthodox. Right? I don't know if you heard

2:17:12

the term. Right? Modern

2:17:13

orthodox. It's basically it's

2:17:15

a choice to synthesize, like,

2:17:17

traditional

2:17:17

orthodox criticism with the

2:17:20

modern world. and has to make interesting

2:17:22

kind of trade offs for that. You know,

2:17:24

when you actually build that society, maybe

2:17:26

they'll be pleasing, maybe it won't be. And the

2:17:28

way you can test it is whether it attracts

2:17:30

immigrants from the rest of the world. Right? In fact,

2:17:32

there's this really here's something. This is going in v

2:17:34

two of the books. This is new content. Okay?

2:17:37

Ready for new content? Take a look

2:17:39

at this pretty cool video. Okay? And then, here's

2:17:41

a visual version of it. So you can see it in

2:17:43

a graph. Okay? click

2:17:46

the second link first. The population rank of every

2:17:48

US state over a hundred years. Okay? And actually, there's a better version of this that

2:17:50

goes all the way back to seventy seven. This is only over ten years.

2:17:55

You can watch that YouTube link as well,

2:17:57

which kinda has an animation of

2:17:59

this. The thing is that the US has had something where, especially, over two hundred

2:17:59

something years, different

2:18:04

states have been like number one and the

2:18:06

ones that are pulling in all the

2:18:09

internal immigrants from other places. Right? people

2:18:11

talk about the US as a nation

2:18:14

of immigrants, and that is talked about as coming from other countries. But there's two wrinkles

2:18:16

on First

2:18:19

is, it's not just a nation of

2:18:21

immigrants, it's a nation of immigrants. Right? Todd, logically, every immigrant, emigrated somewhere. So they why leave

2:18:23

Poland or India or

2:18:29

Iran -- Right. -- or China. Sometimes it's leaving

2:18:31

like communism, like China,

2:18:34

or nazism, leave fleeing Germany, or

2:18:36

you're leaving an economic basket case like

2:18:38

India was. or you're leaving, you know, just for search for a better life like many people Right? But

2:18:41

fundamentally, the people

2:18:44

who leave are

2:18:47

not the wealthy necessarily what they

2:18:49

are or the politically powerless. This

2:18:51

is really important, by the way,

2:18:53

because people say, oh, apology. All

2:18:56

your exits stuff. It's all about like rich guys

2:18:58

leaving and leaving us holding the bag blah

2:19:00

blah blah. And I'm like, have you

2:19:02

looked at the profile of immigrants globally? Like,

2:19:04

they are not on balance

2:19:07

the wealthy. Right? In fact,

2:19:09

the wealthy and powerful have big

2:19:11

houses and political power in places like San Francisco, and

2:19:14

they control the

2:19:17

government, whereas the

2:19:20

new money Right? The

2:19:21

immigrants are the ones who do not

2:19:23

control the government and thus had to

2:19:25

leave San Francisco because they could not

2:19:27

get emailed through. because they could not fix the

2:19:29

streets or whatever, because they got attacked, because they got

2:19:32

priced out and so on and so forth. So those who leave are

2:19:34

not those who are financially wealthy, those who leave are those

2:19:36

who politically

2:19:38

powerless. Right? And once you look at

2:19:40

it, they are often those who are in a sense

2:19:42

fleeing. Right? I mean, one way of putting it in

2:19:46

this is kind of production out of

2:19:48

zurnum, and it may seem like a

2:19:50

digression, but entertaining. Was Stalin rich? Howard

2:19:51

Bauchner: Depends. In in what part

2:19:53

Well, society. I think it it

2:19:55

wasn't money. Right? In fact, it's like quote unquote

2:19:57

or something else. I'm not sure if this is true, but it's

2:19:59

it's very

2:19:59

plausible. Like, you

2:20:02

know, Stalin didn't even have change in his

2:20:04

pocket. Right? But he could walk around and it's like,

2:20:06

Grand Theft Auto, you just car, check a car. Right? that's

2:20:10

like how Stalin and, like, the nomenclaturea

2:20:12

walked around, like, the Soviet Union. They just had, like,

2:20:14

root access to everything. They could come into your your house, sometimes your wife,

2:20:19

you know, like, you're you're farm or whatever. That's

2:20:21

what collectivization meant. It was the people. It was the

2:20:23

people. It's the communist party and it's a nomenclature that controls everything. Right? So

2:20:27

was he rich? No, he

2:20:29

was powerful. Right? Meaning a wealthy person must

2:20:32

still persuade Right?

2:20:34

They're giving you money for something and

2:20:36

you can reject that offer or not.

2:20:39

But the powerful can compel. So persuasion versus compulsion, right, convincing versus coercing, Those

2:20:43

are different things. So was Talend Rich? No.

2:20:45

He wasn't Rich. He was powerful. And once we acknowledge that as a limit case, right? You can pull it back and be like,

2:20:48

okay. Actually, power

2:20:52

is a different axis from wealth. And there can be

2:20:54

people who are powerful but not wealthy.

2:20:57

There are people who can compel, who cannot

2:20:59

convince and vice versa. those who control territory or

2:21:01

those who are powerful, and those who set the laws that

2:21:03

are powerful. If you could set

2:21:06

the laws, why would you leave? Right. You gotta

2:21:08

Right? There's no incentive. There's no incentive. You're

2:21:10

setting the laws. This is all bespoke made

2:21:12

to you. Blah blah blah. Why would you

2:21:14

leave? Right? This is you've you've carved

2:21:16

out some stupid, you know, bail outs or something like

2:21:18

that, you know, you're you're you're not seeing the folks

2:21:23

who own housing in SF and who

2:21:25

have been blocking housing. those are not the folks who are leaving. It's

2:21:27

the new people who are leaving. Right? It's politically powerless

2:21:31

for leaving. And so the reason this is

2:21:33

important, by the way, is often things are framed as masses

2:21:35

versus leads or like, whose original lead

2:21:39

are leaving. It's actually typically it's

2:21:41

at least three groups. There's masses,

2:21:43

there's elite, and what Peter Churchill calls the counter elite. That

2:21:47

term is a glorious term because

2:21:49

it clarifies so much. Right? basically, you don't get

2:21:51

into this silly kind of situation of saying that someone

2:21:56

who VC, you know, millionaire or

2:21:58

like a tech founder or, like, biometric. You don't get into the silly game

2:21:59

like, are they elite

2:22:02

or are they massive?

2:22:05

Instead, they're the country.

2:22:07

The tech folks are

2:22:08

a global meritocratic group

2:22:11

relative to like basically

2:22:13

the white epictus of

2:22:16

the American I'm not the kind of person

2:22:18

by there who thinks white is an insult. Okay?

2:22:20

But the US establishment does. So once you kind of

2:22:22

apply that lens, you see how many of their businesses

2:22:25

many of their media corporations, how many of their

2:22:27

East Coast kinds of things are either literally inherited in

2:22:29

the sense of, like, passed by father to son, like the New York Times company, like

2:22:31

SALTzberger to SALTzberger five generations. or

2:22:35

informally inherited, like the Kennedy name, the Plus name.

2:22:37

It's not East Coast. I mean, she's on the West Coast, but, like, you know, these

2:22:39

are like famous family dynasties. Right?

2:22:44

A very large fraction if you go

2:22:46

to that article founding versus inheriting. There's a paragraph there which is like I say,

2:22:51

Now, of course, someone who

2:22:53

attends political office isn't always

2:22:55

a familial heir, though it's more common than you

2:22:59

might think. And if you click

2:23:01

those seven links, you'll see an article in so called

2:23:04

widow succession hereditary

2:23:06

politicians, history of wives replacing their

2:23:08

dead husbands in congress, the Kandi family, the Bush

2:23:10

family, the Clinton family, the Roose family, and

2:23:14

the list of the United States political families, and

2:23:16

there's actually quite a lot of them. Right? This

2:23:18

is also there's this movie, the distinguished gentleman, which is about like name recognition, you know. So America has an ability

2:23:24

It has de facto, like hereditary

2:23:26

titles. That's a huge advantage in

2:23:28

running for offices, that hereditary name

2:23:30

of being a Kennedy or something

2:23:32

like that. Right? And so against

2:23:35

that elite, that

2:23:35

hereditary elite, sometimes in

2:23:38

the literal sense of the inheriting

2:23:40

a media corporation, what I call the meritless

2:23:42

nepotists. Right? And just calling them that, by the

2:23:46

way, scales fall from the eyes.

2:23:48

So much becomes clear. old money against new

2:23:50

is an eternal story. It's literally the old money calling the

2:23:54

new money rich. That's like literally what

2:23:56

it is. Okay?

2:23:57

Once you kinda see that, You're like, okay, it's a

2:23:59

leap versus counter leap. And what that

2:23:59

will elite forces counter probably

2:24:05

mean is

2:24:05

the US establishment wins within

2:24:07

the territories of the US that they control, but they lose

2:24:11

outside there's parts from the

2:24:13

US that I think become more sanctuary city like, you know, diverging in different ways

2:24:15

to the left and

2:24:19

to the right of the federal government. And

2:24:21

there's countries outside that diverge away from the

2:24:23

federal government. And so happens is the have power fewer people. Well, something

2:24:25

that comes to mind is,

2:24:27

I think I

2:24:28

heard you say something along

2:24:30

the lines of Google News, basically

2:24:34

made

2:24:34

every magazine compete with every other

2:24:36

magazine -- Yes. -- or every other

2:24:38

blog, compete with every other blog. And

2:24:41

I I wanna hear your take on how

2:24:43

certain technological shifts like remote work. You know, we

2:24:45

can relate this back to the network state, but even

2:24:47

separate from that. change

2:24:49

the game in terms of every state

2:24:51

or every city competing with every other city

2:24:54

out there for talent. The US is

2:24:55

a nation of immigrant, but

2:24:57

it's also a niche of immigrants where the people

2:24:59

who left were those who are politically powerless as we just

2:25:01

talked about from other countries. But it's also a niche and it's not the financial that's a

2:25:04

political power but

2:25:07

it's also a nation

2:25:09

of internal immigrants. Less appreciated

2:25:11

is the degree to which The

2:25:14

number 123 and four states in

2:25:16

the US have been in flux over time. They

2:25:18

have attracted people and repelled people and so on and

2:25:22

looking at that over time actually tells

2:25:24

a really interesting story. Again, you know, we

2:25:26

live twenty, thirty, forty, fifty six. Like, if

2:25:27

you're very old, you've seen some of these ups

2:25:31

and downs. Right? But once you take the window that's

2:25:33

longer than human life, because we've got data going back. Remember, I think about a billion hundred billion people have lived. If you go back

2:25:35

and get more data, you only have one life. Right? let's

2:25:41

find you get some leverage out of that. Right? That's what history

2:25:43

is. It's leverage. Right? Financial leverage

2:25:45

and sense. It is alpha. Okay? Other configurations of

2:25:47

humans work. So you look at these charts and

2:25:49

you see, oh, wait a second. in the eighteen hundreds, Ohio was a really big deal.

2:25:51

It was, like, one of the top seats, attracting lots of

2:25:53

people. Like, penciled in your

2:25:55

your, like, oh,

2:25:56

yeah. That was, like, when

2:25:58

it wasn't the

2:25:59

rust belt, like making stuff. Right?

2:26:02

Like the higher river, all the stuff

2:26:04

is like really important. And then you come

2:26:06

forward and you see New York was like,

2:26:08

for

2:26:08

example, the very beginning of the

2:26:11

US, Virginia was big. And now

2:26:13

you kind of vaguely remember the

2:26:15

songwriter, early American history, you'll hear people

2:26:17

talk about, like, and as a

2:26:19

Virginian, you know, like state

2:26:21

identity was actually a big deal

2:26:23

then.

2:26:23

Why? Because we have

2:26:25

to work back. But in

2:26:27

the, like, the seventeen hundreds, individual

2:26:29

are now US

2:26:32

Pennsylvania was settled by, like, William

2:26:34

Penn and his followers and, like,

2:26:36

Massachusetts Bay Colony is, like, the Northeast,

2:26:38

and Virginia is, like, the cavaliers and

2:26:41

tie art, all these different districts.

2:26:43

Right? None of districts like subcolonies.

2:26:45

They were as different as tech companies were

2:26:46

in, like, the year two two thousand and eight. Right?

2:26:51

So you have Larry Page and Google.

2:26:53

You have Zuckerberg and Facebook. You have Steve John Snapple. You have Bill Gates and Microsoft. You have Jeff Bezos

2:26:56

and Amazon. These

2:27:00

are founder run companies with different

2:27:02

cultures and different personalities and so

2:27:04

on. Now today, now that four

2:27:06

out of those five companies have

2:27:08

had their founders leave, Right? Now

2:27:11

today, Microsoft, Google, Amazon, and

2:27:13

Apple are much more similar

2:27:15

than they were a ten

2:27:17

year a lot of people have the resume of I did three years here and two

2:27:19

years here and three years here, but, like,

2:27:22

big tech is its own thing. Facebook is the outlier

2:27:24

because Zuck is still a founder and so he can actually do

2:27:26

big things like rebranding to meta that can take

2:27:30

risks. Okay? Point being that putting those states

2:27:32

together, like Virginia was an identity, of its own,

2:27:34

like, Virginia's the whole thing. After the civil war, it went

2:27:38

from the United States r to the

2:27:40

United States

2:27:41

is. Like, these states lost

2:27:43

more of their individual

2:27:45

identity and became like a

2:27:47

federal government like like a union

2:27:49

where they were just administrative subdivisions. Right? And even more

2:27:51

of that was knocked down

2:27:53

over the early twentieth century and the Great

2:27:56

Centralization, you know, FDR is another big step

2:27:58

with the tenth amendment basically being repealed. Everything that was supposed to be left to the states was

2:27:59

now

2:28:00

something

2:28:03

where you could set up these regulatory agencies like

2:28:05

the FDA and so on. They just said, hey,

2:28:07

everything is interested commerce therefore the federal government regulates anything. Right? So all the centralization happened,

2:28:09

but even during this period,

2:28:11

even during the high centralization,

2:28:14

still have people moving between states.

2:28:16

And now, you have something like Florida

2:28:18

that's been coming up over the last twenty

2:28:20

years, and it's really with the Miami thing and

2:28:22

so on, it's really giving California competition. Right?

2:28:25

But that's not something that just came out

2:28:28

of nowhere. If you look at this graph,

2:28:30

the visual capital graph, you see, like, Florida's been gaining traction for

2:28:32

decades. And so it's

2:28:34

an overnight success like, you know, fifty years

2:28:36

in

2:28:36

the making. What you see with this is

2:28:38

that internal competition of this kind between administrative subunits is

2:28:41

very important for keeping things healthy. And this

2:28:43

is the remember that we're talking about,

2:28:45

like, the different leviathons, this is the polystatist model, as opposed to the monostatist model. Right? There's always a tension

2:28:48

between these. Okay?

2:28:51

There's this whole Chinese saying, which

2:28:54

is the empire long divided must divide. Long divided must unite.

2:28:56

Why? because

2:28:59

when it's Lone United, you

2:29:01

have stultifying, bureaucracy, harmonizing, like I was

2:29:03

talking about, it's not responsive, etcetera. then

2:29:06

you decentralize, you divide. Then what happens? You

2:29:08

have a period of, oh, it's all chaotic. Oh

2:29:10

my god. I wish we had common sand, etcetera. So then you reunify. Oh, wow. We've reunified.

2:29:15

And then there's a honeymoon period. And so each

2:29:17

time there's like a honeymoon period of, wow, we've got our independence. Oh my god. It's chaotic. Oh, it's chaotic. Oh my god.

2:29:20

It's chaotic. oh,

2:29:24

let's divide. Right? So it goes in cycles like

2:29:26

this. And that's not to say that you come back in the same

2:29:27

place. One of the points I make in the book is the helical theory of history

2:29:28

where From

2:29:33

one axis, it looks like progress, from another axis,

2:29:35

another projection, it looks like

2:29:37

going in a circle. But if you look

2:29:39

at it from outside, it's like, you know, maybe

2:29:41

it's a x of t equals cos t, y of t equals

2:29:44

sine t,

2:29:46

z of t equals t is the param

2:29:48

equation for Helix and, like, three space. So

2:29:50

it's cyclical on some axes and it's,

2:29:52

you know, linear on

2:29:53

others. Right? So you are still making

2:29:56

progress sometimes even if it seems like you're going

2:29:58

in a circle uniting dividing. Bring that

2:29:59

to the present day. What we

2:30:02

are going through and what we are going

2:30:04

to see is much more poly sadism. and

2:30:06

the network state is way of not just using the existing administrative units like

2:30:11

Florida and what have you but setting

2:30:13

up new ones. Imagine the United

2:30:15

States where you couldn't just found a company but you could found the California. Let's say the

2:30:17

first order,

2:30:19

the thirty

2:30:21

four million

2:30:23

Californians own California. Okay? Like this huge territory

2:30:25

around. Right? The five hundred thousand, let's

2:30:28

take a smaller state, Wyoming. Like about five hundred

2:30:30

thousand Wyomingans or Wyomingers actually don't know what to

2:30:33

I'm sure someone will be like it's I

2:30:36

don't either. Yeah. It's like, you know, Wyoming guy. I

2:30:38

don't know. It's like something like that. Right? Okay. Five hundred and eighty thousand. Yeah. So the smallest unit, I

2:30:40

think. So

2:30:44

the five hundred eighty thousand Wyomingans

2:30:46

basically own Wyoming. Right? They could, in a sense, crowdsource

2:30:47

on Wyoming.

2:30:50

let's say,

2:30:50

I don't know the exact ratio, but

2:30:53

let's say on the order of eighty percent

2:30:55

or ninety percent of Wyoming territory Wyoming residents. it's

2:30:57

just some out of state holding, some

2:31:00

Chinese holdings and some, but let's just

2:31:02

say, okay, first order. That's interesting. because what that

2:31:03

suggests is, five

2:31:05

hundred thousand people that

2:31:08

you unify in

2:31:09

an

2:31:10

online network society

2:31:10

network union, network archipelago, could

2:31:14

have a territory that amounts to

2:31:16

the territory of Wyoming just spread

2:31:18

out. I could actually found an administrative unit that competes

2:31:20

that competes with these

2:31:22

other

2:31:22

administrative units. And the key thing is that

2:31:24

because I'm being able to buy a piece of

2:31:26

tariffs that aren't connected to each other, I

2:31:30

don't have to expand it in one place. Right?

2:31:32

This is a key insight. Like, I talked about this in the book that

2:31:34

the big thing about the nation states, see, you know, there's city states, nation states and network this

2:31:40

again, the v three. Right? The city state was

2:31:42

like this independent unit. It was innovative and so on, but it got beaten by the nation state. You know why? because the nation state had

2:31:44

scale. you

2:31:48

know, what we now think of is Italy. And like the

2:31:50

language Italian, most people in the Italian peninsula,

2:31:53

people argue that most people didn't speak Italian. That's

2:31:55

the time that Garibaldi is unifying Italy. Eric Hobbesbaum

2:31:57

argues that at the time of the French revolution, most people in

2:31:59

France didn't speak what

2:32:02

we now know as French. They spoke other

2:32:04

dialects, other languages. Right? Germany, you know, Bismarck UNIFY Germany.

2:32:06

And there's a whole thing in the Eastern, it's called the German

2:32:10

question about, you know, whether, like, Hanover

2:32:12

and Austria, like, how would how they actually attract it's

2:32:14

sort of like, you know, who would run the American Republic? Would

2:32:18

it be Virginia or would it be the capital b

2:32:20

and d c? Right? Garryball the United Lee,

2:32:22

after the French revolution, France got unified

2:32:25

and homogenized, the state actually shaped

2:32:27

the nation. because now all these people have

2:32:29

no memory of a time when their ancestors

2:32:31

spoke a language out of the French. Right?

2:32:34

The nation and state of an interplay where

2:32:36

the state has all these power so it can

2:32:38

actually reshape the nation and rewrite people's brands like men in black,

2:32:42

you know, like, flash, you know, now

2:32:44

the children are educated by the state.

2:32:46

And so state shapes nation, there is a feedback effect. Right? It's sort of similar to

2:32:50

the whole currency argument. What can's first

2:32:52

barter? You know? Or is it like

2:32:54

debt? Right? You know, various arguments on this. And I do think it's barter that comes but it

2:32:59

is true that you can have a feedback effect

2:33:01

here. Okay. Why did I

2:33:02

bring this up? The city state lost the nation state because nation state

2:33:03

scale. Okay? And so

2:33:06

it rolled

2:33:07

up all of these

2:33:09

smaller units into this

2:33:12

big thing and had the scale

2:33:14

to beat others. And in fact, once

2:33:16

one group had that scale, then the nearby neighbors need

2:33:18

to also get together. Like, so the French revolution and the Napoleonic Wars

2:33:23

are what help lead to German unification.

2:33:25

All these guys are like, hey, we

2:33:27

got a gang up together. Everyone's France is gonna beat us. Right? There's actually a relic of the past like like

2:33:33

a duck build platypus. That, it's like a

2:33:35

missing link that lives into the present. Have

2:33:37

you ever been to you know what

2:33:39

San

2:33:39

Marino is? I I recommend it, but

2:33:41

I actually don't know. Why don't

2:33:43

you explain what it

2:33:45

is? San Marino is this tiny

2:33:47

little patch of it it's in what we

2:33:49

think of as Italy. That is like a sovereign

2:33:51

thing in the UN. It's got a flag. It's

2:33:54

like thirty thousand people. It's like this weird thing.

2:33:56

What what a country of thirty thousand people, it's an

2:33:58

enclave also. It's like in the middle of

2:33:59

Italy. ill

2:34:01

And so what it is, is basically,

2:34:03

before the unification of Italy, it's a

2:34:05

whole complicated story about how it got unified. But Garibaldi was able to

2:34:07

seek refuge in San Marino at one And in gratitude, He's

2:34:13

like, okay, we'll just do a contract with you guys who won't fold you

2:34:15

into the full

2:34:19

Italy. Okay? And so San Reno, by

2:34:21

hooking by crook managed to preserve its sovereignty to the present

2:34:23

day, and they're like

2:34:25

a relic of the past when the whole

2:34:28

thing looked like that. lots of little San Marino's all

2:34:30

over the map. Right? If you look at a map of Germany

2:34:34

before Bismarck, right? That's a map of Germany

2:34:36

before Bismarck. Right? Now here's another one, you

2:34:38

know, a map of the princely the

2:34:41

princely states okay states. Okay?

2:34:43

When India got independence, It

2:34:45

wasn't

2:34:46

just independence from the

2:34:46

British. Oh, that wasn't the

2:34:49

only problem they had to solve. They

2:34:51

also had solved the problem of the

2:34:53

Princely States. Take a

2:34:54

look at that. Basically, there are five hundred and sixty two princely states that think

2:34:59

that's the exact number about that constituted

2:35:01

what we now know of as the Indian Union. If

2:35:03

you look at that map, look at how complicated that looks. It's

2:35:07

widely

2:35:09

complicated. Wildly

2:35:12

complicated. Right? This is

2:35:12

the physical world used

2:35:14

to be complicated like this, but

2:35:17

it meant it

2:35:18

was very

2:35:18

simple within those communities because people

2:35:22

basically spoke the same language that things in common

2:35:24

and so on and so forth. You're trading off like

2:35:26

one kind of complex. I don't think it's like all roses or whatever. Right?

2:35:29

but you often can trade off one

2:35:31

kind of complexity for another.

2:35:33

Right? Clean lines on a map

2:35:34

in Africa, you know, have put people

2:35:37

who don't speak the same language, don't

2:35:39

share the same culture together, and has

2:35:41

led to a lot of the African civil wars, is these artificial nations

2:35:43

didn't reflect actual ethnic groups. Okay? This

2:35:47

is actually, you know, the Sykes Picot lines in

2:35:49

the Middle East. It's actually something that, like, ISIS and others also hate,

2:35:51

but basically, whenever you see these straight lines in the Middle East, these

2:35:55

are, like, artificial countries that don't

2:35:57

reflect ethnic boundaries. And so it just gives rise to, like, endless

2:35:59

conflict because people need to share culture before

2:36:04

they can agree on law. You

2:36:06

know, if your religion says

2:36:07

this must be done and disguised

2:36:09

religion says this must not be done,

2:36:11

you just guys gonna slug it out endlessly. You

2:36:13

can't just agree. And it might be something really stupid. It might be some dietary thing

2:36:15

or something that doesn't really matter in the grand scheme of things. or

2:36:19

at least not to not to people who don't share

2:36:22

either religion, but just slug it out in the sleep if the boundaries aren't set in the right way. This how the world used to work. If you look

2:36:24

at San

2:36:28

Marino, you get a piece of it. You look at the map

2:36:30

of Germany and before Bismarck, you look at Indian,

2:36:33

Prince States. This is how actually the

2:36:36

world also kind of works today. You know

2:36:38

why? because a lot of countries are small

2:36:39

countries. Howard Bauchner: Yes. So you know that part.

2:36:40

The

2:36:43

other thing is, if you've seen the

2:36:45

county map of the US and who's Democrat and

2:36:47

who's Republican. That also looks fractal in the same that also

2:36:49

looks fractal and this same way

2:36:51

way. It's complicated. Right.

2:36:53

Right. Yep. And The lines aren't

2:36:55

representative necessarily of

2:36:55

the nation behind

2:36:58

those lines. That's

2:36:59

right. And so in fact, you

2:37:01

know, they're saying like one nation

2:37:03

under god indivisible. Right? So there's this really important the

2:37:06

unfortunate one liner

2:37:09

is it's now

2:37:12

two nations that don't believe in

2:37:14

God, highly visible. Right? Why? Because,

2:37:16

you know, you can take

2:37:18

a

2:37:18

look at, like, this graph that's

2:37:20

at the level of

2:37:22

Congress. Okay? But

2:37:23

you can literally see it

2:37:25

going from people basically voting together

2:37:27

on things to essentially,

2:37:30

all Republicans vote in with all Republicans,

2:37:32

all Democrats vote in with all

2:37:34

Democrats. Right? And there's the rare bipartisan things, like where there's, like, gray lines that connect them are,

2:37:36

like, nine

2:37:40

eleven and, like, the financial crisis. So

2:37:42

for, like, the, you know, bombings

2:37:45

or bailouts type stuff, that's, like,

2:37:47

that's a bipartisan stuff. But otherwise, they just never

2:37:49

vote together on anything. You know? Then you you look at that. That's at the level of Congress, and

2:37:51

then you look at this graphic they just paced in. this

2:37:56

is at the level of Twitter or

2:37:58

Facebook. Okay? There's a article CGR

2:38:02

study twenty seventeen basically, both Twitter

2:38:04

and Facebook look the same, where

2:38:06

it looks like this blue and red separated thing. Right? So

2:38:07

this is literally

2:38:11

two nations. Going back to the very

2:38:13

first thing in our chat, it is not one nation. It

2:38:15

is two nations with different values.

2:38:18

What am I thinking about it

2:38:20

is something like ninety six percent

2:38:22

of Democrats marry other Democrats, only four percent marry Republicans. Okay? So Democrats will not marry Republicans.

2:38:28

And what that means is

2:38:30

ideology becomes biology in one

2:38:32

generation. Like, this is

2:38:33

becoming or

2:38:34

already is in some

2:38:37

ways, like Sunyoshiya. It

2:38:38

is not a political conflict in

2:38:40

the traditional sense. It is

2:38:42

a tribal conflict between groups that

2:38:45

do not marry each

2:38:46

other. Right? If you look at this graph over here, right? Marriages between at this graph

2:38:49

or be here Democrats

2:38:51

and Republicans are rare. So when

2:38:53

you have something like that, they are becoming different ethnic groups as different as, you

2:38:55

know, what do you call racial groups,

2:39:00

like Suni and Cheyenne, you know, or Prossen

2:39:02

and Catholic, you know, for a long time. That was a huge thing in Europe. Right? Or, you

2:39:05

know, Hudu and Tutsi

2:39:07

or whatever. Right? And

2:39:09

this relates

2:39:09

to your idea

2:39:11

of primary identity. Right? Because these people

2:39:13

are Democrats or Republicans. They're also

2:39:15

American. They're also many other things

2:39:18

left handed, right handed, as you said.

2:39:20

But there's one identity that people are

2:39:22

now trending towards, right, as their primary

2:39:24

in certain cases. And that's why you're saying basically

2:39:26

the nation, which previously would have been primary

2:39:30

identity, American. Exactly. Or in

2:39:32

the past, primary identity, Catholic. Mhmm. That's right.

2:39:34

It's like Yugoslavia, where

2:39:35

that primary identity was national. But

2:39:39

now it's a subnational entity. It's kind of

2:39:41

actually there's this book by I think by Barbara Okay? It's called, how civil start? She

2:39:43

talks how in a rock she

2:39:49

interviewed, like, a girl there. She's like, what was different about, you know,

2:39:51

after the US occupation? And so

2:39:53

she's like, well, people started asking me a question they

2:39:55

never asked me before, which is, are you sooner or

2:39:57

she had? So, Saddam, for all of his many faults, basically kept the

2:39:59

lid

2:40:00

on Iraq. If

2:40:02

you were in Iraq before Saddam, there

2:40:05

there are bad things about it. but it

2:40:07

was not in civil war and ISIS and the, like, insanity that

2:40:11

followed, and the satarian violence because in

2:40:13

the absence of the seat, once the state

2:40:15

was knocked out and the Americans were considered legitimate, you had You had you

2:40:18

people fell back on

2:40:20

that Leviathan, and you

2:40:23

had these sectarian disputes. which, of

2:40:25

course, were fueled by Iran, all these

2:40:27

terrorists. But it became sunni versus

2:40:30

the the national identity went away. Right? And

2:40:32

so what you're mentioning here is

2:40:34

exactly like the American flag is

2:40:36

not actually That's the state.

2:40:38

Right? but people don't identify it. You'll see all these articles, NYC, other places don't criticize, oh my

2:40:41

god, I don't like the

2:40:43

American flag. Instead, the way

2:40:45

of thinking about it, it's

2:40:47

not just Republicans, internationalists, Its Democrats are

2:40:49

also nationalist. They're for Democrats. And Republicans

2:40:51

are for Republicans. Both are for their own

2:40:53

tribe. Their their nation is not the American

2:40:55

nation. It is Republicans for

2:40:57

the Republican nation, Democrats for the Democrat nation.

2:40:59

And both of them actually have their own flags. They're

2:41:01

not called the Democrat flag and the Republican flag, but you can immediately

2:41:04

think of you

2:41:06

know, where it's a blue line, like the

2:41:08

the blue lives matter flag or the thin

2:41:10

blue line flag, the don't trade on me flag for the Republicans or the progress flag for the Democrats, etcetera.

2:41:15

Like, you they have actually their own

2:41:17

flags that are actually the flags of their nations. We don't recognize them as such today, but that's why I was saying, like, if in

2:41:19

the twenty tens, question

2:41:24

is what is the currency? The twenty twenty is the question

2:41:27

is what is a nation. Once there

2:41:29

is no longer an American nation, there's a Democrat

2:41:31

nation, the Republican nation. You have these fractals

2:41:33

that look just like the Princely States of India in nineteen forty seven or

2:41:35

Bismarck before Germany. We

2:41:39

already have the Fractal Democrat Republican America.

2:41:41

That complicated thing that I showed you

2:41:43

that seem like distant past is actually our present. How does this relate? The Internet has actually made

2:41:45

this possible

2:41:46

to deal with.

2:41:47

I mean, I'm not saying

2:41:49

this all gonna be roses or

2:41:51

what have you. But one of the

2:41:53

things the Internet has done as you

2:41:55

progress and connect it to the previous

2:41:57

thing, from

2:41:58

city state to nation state's network state,

2:42:00

Those city states like San Marino, there were

2:42:02

little tiny things that couldn't survive on their

2:42:04

own versus a big nation state, can now, with the internet group together

2:42:07

around the world and have a collective scale

2:42:10

that is again bigger than any local

2:42:12

nation state. A nation state can no

2:42:14

longer expand. Right? Because the boundaries there are fixed. the

2:42:17

network state can expand at the speed

2:42:19

of a TikTok or a Facebook. Right? You can

2:42:22

grab all these people globally. It's constantly competing for people. And so it can get

2:42:24

greater scale than

2:42:27

any nation state or not any.

2:42:29

Let's say many nation states. Right? And I mean, it'll be hard to

2:42:31

be bigger than China or India, but Facebook is. Right?

2:42:35

Now, Facebook is not in network state because

2:42:37

it was not set up as such. The Facebook employees report to Zuck. They are in the leadership hierarchy. The

2:42:39

Facebook users are just using a tool and

2:42:45

they're not taking orders from Mark on this or

2:42:47

that. But you could imagine a

2:42:49

network that did have that kind of

2:42:51

social contract. But when you

2:42:52

came in, you were like, look,

2:42:54

here's the thing.

2:42:55

This is the president. Here's the folks. Here's what we're doing. Here's

2:42:56

the one commandment. Here it'd

2:42:58

be like your slot in

2:43:00

the hierarchy if we have

2:43:02

to break up tasks and so like

2:43:04

when joining a company. Right? Here's your title

2:43:06

and your position. It doesn't have to be

2:43:08

quite like that as hierarchical, but it could be. Okay?

2:43:10

Different network side and select different varieties. and

2:43:13

then you are signing that social smart

2:43:15

contract upon entering it. And now you've

2:43:17

got something that has it's like

2:43:19

AV3 It has aspects of the

2:43:21

city state where it's like everybody agrees

2:43:24

on something. Like, it's innovative. It's

2:43:26

agile. It's nimble. It's like founder led. It

2:43:28

is aspects of the nation state, which beat

2:43:30

the city state because it's got scale. You

2:43:32

know? It has uniformity. Like, it is useful to have

2:43:34

people in a large group all speak Italian

2:43:37

or French. It loves laws and all the types

2:43:39

of this. There's a reason the v two happened.

2:43:41

And this v three combines aspects of both, just like Bitcoin combined aspects of both

2:43:44

digital gold. and

2:43:46

theatroncy. Right? And that is

2:43:48

maybe what we're going to

2:43:50

need as we move into this fractalized America where you do have

2:43:53

where you do have

2:43:55

two nations, not under god, because people

2:43:58

don't believe in god anymore. Like, that's dropped off the cliff as well, divisible. You're

2:43:59

some new

2:44:01

it is about you're going to need some new

2:44:04

thesis

2:44:04

of how to govern. And I

2:44:06

think network states may be a piece

2:44:08

of that. But even if you take

2:44:10

America aside for now, the rest of

2:44:13

the world, all of these other countries, all

2:44:15

of these other groups within those places, the ninety

2:44:17

percent of the world is South American, They

2:44:19

can also now have new ways of

2:44:21

doing self government that don't require winning a war or an election or have you. Anybody can declare themselves

2:44:23

president to start society set

2:44:28

up a one command and fix the problems they see in the world

2:44:31

and be an example for others. Well,

2:44:32

I think that's a

2:44:34

good place to end it. This was really, really

2:44:37

informative. I think I'm glad we covered certain topics like, is

2:44:39

this just for the wealthy? Or can

2:44:42

this actually scale to the point that

2:44:44

we consider states to be today? So, Palaji,

2:44:46

thank you for going through all this. If people are interested, they can go to what?

2:44:50

The network state dot com

2:44:52

and also Twitter dot com from such page.

2:44:54

Yes. So free online, and you

2:44:55

can also get it on Kindle. Awesome. And if

2:44:56

and you can also get around kenosha

2:44:58

people are curious, the network

2:45:00

state dot com slash dashboard, looks like you

2:45:02

have twenty six startup societies that you're starting to track. So also a very interesting place see

2:45:07

who's building, what they're building, and there's a

2:45:09

lot of

2:45:10

diversity there too. That's right. So the space is the things are happening here. Awesome. Balaji.

2:45:12

Thanks for listening

2:45:14

to the a sixty

2:45:17

z podcast. If you

2:45:19

like this episode, Don't forget to subscribe, leave

2:45:22

a review, or tell a friend. We

2:45:24

also recently launched on YouTube at youtube

2:45:26

dot com slash sixteen z underscore video, where you'll find exclusive video content.

2:45:28

We'll

2:45:33

see

2:45:35

you next time.

Unlock more with Podchaser Pro

  • Audience Insights
  • Contact Information
  • Demographics
  • Charts
  • Sponsor History
  • and More!
Pro Features