RETURN TO CHANNEL

Search Captions & Ask AI

E115: The AI Search Wars: Google vs. Microsoft, Nordstream report, State of the Union

February 11, 2023 / 01:49:44

This episode discusses communication skills for children, the Chinese balloon incident, and the Nord Stream pipeline explosion. The hosts, including David Sacks and Chamath Palihapitiya, engage in debates about media reactions and government accountability.

The conversation begins with a humorous anecdote about a father trying to teach his son proper phone etiquette. The father expresses frustration over his son's lack of verbal communication skills, highlighting a generational gap in phone usage.

The discussion shifts to the Chinese balloon incident, where the hosts critique media reactions and the government's handling of the situation. They analyze the implications of the balloon's presence and the public's response to perceived threats.

The episode then transitions to the Nord Stream pipeline explosion, with the hosts debating the potential involvement of the U.S. government and the media's portrayal of the event. They reference investigative journalist Seymour Hersh's claims and discuss the geopolitical ramifications.

Throughout the episode, the hosts emphasize the importance of critical thinking and accountability in media and government, while also sharing personal anecdotes and insights into current events.

TL;DR

Hosts discuss phone etiquette for kids, the Chinese balloon incident, and the Nord Stream pipeline explosion, emphasizing media accountability and communication skills.

Video

00:00:00

I had this thing with my oldest son

00:00:01

where I don't know if it's all kids but

00:00:04

he's 13. he doesn't know how to answer

00:00:06

the phone not to save his life oh oh he

00:00:10

picks up the phone huh hello and so and

00:00:12

so I said listen from now on when you

00:00:15

call me

00:00:16

I expect a certain way that you pick up

00:00:18

the phone and that's going to be

00:00:19

practice for you how you interact with

00:00:21

anybody else and vice versa says well if

00:00:23

I call you you have to pick up the phone

00:00:25

and if you don't I'm hanging up right

00:00:26

away

00:00:27

so yesterday he calls me and and I'm

00:00:31

like Tremont speaking you know it's like

00:00:33

hello speaking

00:00:35

Dot

00:00:38

hi honey

00:00:40

he calls back hello speaking

00:00:44

dad

00:00:47

hung up again we did this three more

00:00:50

[ __ ] times and then finally I said

00:00:52

when will you get it through your head

00:00:53

can you please just answer like if I say

00:00:55

hello to my speaking Hey Dad it's your

00:00:58

son or Hey Dad it's

00:01:00

and it's unbelievable and he's like well

00:01:02

none of my friends pick up the phone

00:01:04

like this and I'm like oh my God like

00:01:06

don't they think they need to have

00:01:07

verbal communication skills it's

00:01:09

unbelievable

00:01:11

and then when I call him he picks up the

00:01:13

phone oh

00:01:15

what not even hello huh

00:01:20

it's like a grunt minimal efforts like

00:01:22

the minimum minimum number of syllables

00:01:24

it's it's embarrassing are you kids like

00:01:26

this or is it just my kid actually I

00:01:28

don't think I've heard him pick up at

00:01:29

the phone I need to test that

00:01:33

[Music]

00:01:37

Rain Man

00:01:38

[Music]

00:01:45

[Music]

00:01:49

Jay Cal was on Twitter calling me out

00:01:52

for not denouncing the Chinese balloon a

00:01:54

strong enough language although I don't

00:01:55

think he understood the language I was

00:01:56

using

00:01:58

um did you understand what the word

00:01:59

errant means what did you think it meant

00:02:01

I you know I do understand

00:02:04

yeah it generally means Strang of course

00:02:08

yeah and traveling in search of

00:02:11

Adventure at least according to

00:02:13

Merriam-Webster so I just thought maybe

00:02:15

you're being a little but I know you're

00:02:17

a Duff so but it was a little dovish

00:02:19

thinking they were just off course I

00:02:21

think you'd think they were off course

00:02:22

or they were doing it deliberately you

00:02:24

buy that it was off course genuflect go

00:02:26

genuflect Jacob go go go so I knew Jacob

00:02:30

would have to

00:02:31

join in this Nationwide panic over this

00:02:34

balloon I have no panic over

00:02:36

yeah

00:02:42

Jacob I got some really bad news for you

00:02:45

there are these things called spy

00:02:47

satellites

00:02:49

everything oh I know it's obvious people

00:02:51

they've been sending these balloons over

00:02:53

here for a while I just thought your

00:02:54

framing as it was errant

00:03:02

it's such a hair brain scheme

00:03:05

to send a balloon flying over American

00:03:08

territory that the Occam's razor

00:03:11

explanation here is that they somehow

00:03:13

lost control over it and these things

00:03:15

are not steerable so like my guess is

00:03:18

that it probably wasn't deliberate just

00:03:21

because of how stupid a plan it would be

00:03:24

and how like obvious it would be

00:03:27

but it could be I don't really know what

00:03:29

I do know is that

00:03:31

the whole nation got in a lather and a

00:03:33

tizzy and started hyperventilating about

00:03:35

this balloon and it just shows how

00:03:37

reflexively hawkish the media is you

00:03:39

know it's like hey can we just wrap up

00:03:40

this war in Ukraine before we start

00:03:42

another war with an even more the media

00:03:46

superpower the balloon got more

00:03:48

intention than us blowing up north

00:03:49

stream exactly okay well hold on we're

00:03:52

ready to jump injured no what I would

00:03:54

say about that is uh the media Insight

00:03:56

is the valid one I was just responding

00:03:57

to the errand and I was curious if you

00:03:59

actually thought it was an accident I

00:04:00

don't think it's an accident I don't

00:04:01

think there are acts well I think it

00:04:02

clearly was off course and problematic

00:04:04

so you know I don't know about the

00:04:07

intentionality of it I I think I think

00:04:09

it very well could have been intentional

00:04:11

but I tend to think because it's such a

00:04:13

stupid hair brain scheme like I almost

00:04:15

given the benefit of the doubt not

00:04:16

because they're not capable of spying

00:04:18

I'm sure obviously sure obviously

00:04:20

they're spying on us just like we're

00:04:22

spying on them that's what yeah we both

00:04:24

do but I it seems like such a stupid way

00:04:26

because it's made for TV moment you have

00:04:30

to understand the nature of live

00:04:31

television why the media overreacted to

00:04:33

it it's ongoing so because it it's not

00:04:37

final it's like a live event occurring

00:04:39

like when kids are trapped in a mine

00:04:41

those are the best stories for CNN

00:04:44

because you can keep updating people and

00:04:46

people keep turning the TV on to check

00:04:48

on the status so this one was just made

00:04:51

for CNN because

00:04:53

obviously we're going to shoot the thing

00:04:54

down and obviously you can interpret

00:04:57

into it if it's an accident or not and

00:04:59

just gives them something to talk about

00:05:00

on a slow news week but it is

00:05:02

interesting I think your point that is

00:05:03

interesting is is the nordstream story

00:05:07

correct or not why is that not being

00:05:09

covered it's not being covered because

00:05:13

there it's not an ongoing story so that

00:05:16

just shows you you know I don't think

00:05:18

that's right I think it's not being

00:05:19

covered because the mainstream media

00:05:21

already took aside on this oh so

00:05:24

remember cynical yeah well it's not it's

00:05:27

not just cynical so when when Nordstrom

00:05:28

got blown up the administration came

00:05:31

racing out with the line that the

00:05:34

Russians did it that they this was

00:05:35

self-sabotage and by the way the media

00:05:37

repeated this endlessly this was the

00:05:39

media line and it never really made

00:05:42

sense to anyone who's paying attention

00:05:43

because first of all this was an

00:05:45

economically vital asset to Russia

00:05:47

second it was their main source of

00:05:49

Leverage over Europe was there control

00:05:52

over the gas Supply it's the idea that

00:05:55

they would shoot themselves in the foot

00:05:57

that way just to somehow what show how

00:06:00

crazy they are it never really made

00:06:02

sense anyone was paying attention and

00:06:04

the fact that the administration the

00:06:06

media so quickly raced to that

00:06:08

conclusion suggested that it was maybe a

00:06:11

cover story because if we had nothing to

00:06:13

do with it you would just be more

00:06:14

neutral and say yeah we don't know what

00:06:16

happened but they had to like promote

00:06:18

this line that well the Russians did it

00:06:21

which just never made any sense and now

00:06:23

cyhurst has come out with this story

00:06:25

so it laying out in great detail how we

00:06:27

did it it's not just saying we did it

00:06:29

it's laying out who did it and how and

00:06:32

the steps and all that kind of stuff and

00:06:33

just so you guys understand who this guy

00:06:35

is he's this legendary Pulitzer

00:06:37

prize-winning journalist he's like 86

00:06:39

years old or something like that now

00:06:41

but he broke during Vietnam he broke the

00:06:44

story of the milai massacre which the

00:06:47

military denied until he proved it uh he

00:06:49

broke the story during the Iraq War of

00:06:51

the Abu gray prison abuses which again

00:06:54

the media the military denied until it

00:06:58

was undeniable these Pulitzer winning is

00:07:00

all you really need yeah

00:07:02

basically another covert military action

00:07:05

and look we don't know for sure whether

00:07:07

it's true or not but it looks it looks

00:07:10

pretty bad do you think that Europe knew

00:07:12

do you think that the Europeans were

00:07:14

told that the Americans were going to go

00:07:15

and blow this thing up if that's true by

00:07:17

the way the Administration has said

00:07:19

clearly this is a false statement I

00:07:22

think the specific code was this is pure

00:07:24

fiction the Ukraine would be the most

00:07:26

likely uh person to do disability you

00:07:29

got to look for means motive and

00:07:30

opportunity they don't have the

00:07:31

opportunity to go down the story I don't

00:07:34

think I don't think the ukrainians have

00:07:35

the capability to do an undersea Mission

00:07:38

like that ah interesting the Norwegians

00:07:40

do and the story maintained that we did

00:07:43

it with them the British do

00:07:46

story didn't say whether they were

00:07:48

involved or not so no my guess is that

00:07:50

if this was a covert U.S activity it was

00:07:52

kept to a very small group

00:07:54

which is why it'll be hard to

00:07:56

prove more definitively than this the

00:07:58

other side of the argument is this is

00:08:00

such a provocation

00:08:02

and the administration saying it's

00:08:04

fiction so what is your response to that

00:08:07

sex like would the U.S do something so

00:08:09

provocative or would they have somebody

00:08:12

else do it and would Norway do something

00:08:13

so provocative it seems an extremely uh

00:08:17

it seems like an extremely offensive

00:08:19

technique as opposed to just backing the

00:08:21

Ukraine to defend itself against

00:08:23

Russia's Invasion here's the thing

00:08:24

before The Invasion Biden at a press

00:08:26

conference said that if the Russians

00:08:29

invade Nordstrom would be no more and

00:08:33

they asked him well how can you make

00:08:34

sure that that's a deal between Germany

00:08:37

and Russia we're not involved and he

00:08:38

said we have ways we have ways distrust

00:08:40

me it'll happen separately Victoria

00:08:42

Newland is our deputy secretary state

00:08:44

said something very similar about how we

00:08:47

would stop nordstream if the Russians

00:08:49

invaded and then after Nordstrom got

00:08:51

blown up

00:08:52

blinken at a press conference said that

00:08:55

this was a wonderful opportunity and was

00:08:57

extolling all the benefits of this and

00:08:58

then Victoria Newland had a

00:09:00

congressional hearing said that I'm sure

00:09:02

we're all very glad that it's a hunk of

00:09:03

metal at the bottom of the sea again

00:09:04

extolling all the benefits and we've

00:09:06

discussed the benefits on this program

00:09:07

of now we've shifted the European

00:09:09

dependence on Russian gas to dependence

00:09:12

on American gas so the fact the matter

00:09:15

is the administration kind of

00:09:16

telegraphed what they were going to do

00:09:18

they had the capability to do it and

00:09:21

they had the motive to do it so it's

00:09:23

certainly a plausible story you're right

00:09:25

that we can't know for sure it's a

00:09:27

single Source

00:09:30

depends on the credibility of Psy Hersh

00:09:33

but as it stands right now whose story

00:09:37

do I think makes more sense probably Psy

00:09:40

hershis you believe Sai harsh over the

00:09:42

spokesperson for the CIA who wrote the

00:09:44

claim is completely and utterly false

00:09:45

just to be clear well these these same

00:09:47

people denied Abu

00:09:49

Gravely you believe they deny everything

00:09:52

okay they deny that

00:09:55

um NATO expansion anything to do with

00:09:57

the you know breakout of this war yeah I

00:09:59

mean this is one of the situations where

00:10:01

we can't possibly know yeah you make a

00:10:03

good point jcal which is this if true

00:10:06

and look I'm not saying I don't know if

00:10:08

it's you're leaning towards believing it

00:10:10

I lean towards thinking it's more

00:10:11

plausible than not okay uh because also

00:10:14

like who else could have done it and who

00:10:15

again who had the motive means an

00:10:17

opportunity but you make a great point

00:10:19

which is if we did it it's an incredibly

00:10:22

provocative act it's basically

00:10:24

perpetrating an act of war against a

00:10:27

country that has thousands of nuclear

00:10:28

weapons Biden promised us at the

00:10:30

beginning of this war that he would keep

00:10:32

us out of being directly involved so

00:10:34

this would directly contradict what he

00:10:37

said at the beginning of this war

00:10:38

so I think it's a very

00:10:41

scary situation here yeah I know yeah

00:10:46

if the US didn't do it like why don't we

00:10:49

find the real killer I actually have

00:10:51

this I have the theory here's my theory

00:10:53

the CIA knew how to do it

00:10:55

Biden uh wants to do it the Republicans

00:10:58

want to do it obviously the the people

00:11:00

who are the most pro-stopping Nordstrom

00:11:01

have been the Republicans they've LED

00:11:03

this charge uh even more than the

00:11:04

Democrats so it's it's a bipartisan

00:11:06

issue stop Nordstrom bipartisan issue

00:11:08

hands down

00:11:10

I think the CIA probably knew how to do

00:11:12

it and just like we equipped the Ukraine

00:11:15

to do it we might have facilitated the

00:11:17

Ukraine and a collaborator UK

00:11:20

Norway some Freelancers you know we have

00:11:24

those freelance uh former Navy Seals

00:11:26

that operate perhaps the CIA just said

00:11:28

to the Ukraine here's a you know Black

00:11:31

Ops group

00:11:32

if you wanted to engage them

00:11:34

you could feel free to do so which would

00:11:36

then split the difference between what

00:11:38

Psy is saying what the CIA is saying

00:11:40

which is typically where the truth lies

00:11:42

is probably between what this

00:11:44

investigative journalist of note has

00:11:46

said and what the CIA is denying the CIA

00:11:50

probably has plausible deniability

00:11:51

that's my opinion well if you could find

00:11:53

a source for that story J Cal that lays

00:11:55

out in the same level of detail that

00:11:57

Hirsch has laid out in terms of the

00:12:00

meetings that occurred what was approved

00:12:02

when how they did it uh the you know

00:12:05

when they did it the explosives they use

00:12:08

the divers they used it goes into a fair

00:12:11

amount of detail incredible detail

00:12:16

right but you're you're trying to put

00:12:18

what you just said which is basically

00:12:19

you inventing a story on the same level

00:12:22

as Russia's story yeah I think yeah I

00:12:24

understand but he actually has a lot of

00:12:26

detail in his story yeah he had he had

00:12:28

umpteen sources there was a lot of

00:12:30

people that were willing to tell the

00:12:32

story

00:12:33

no no I don't think that's actually the

00:12:34

case I don't think it has umpteen

00:12:35

sources and there's no on the record

00:12:37

there's one main source there was one

00:12:39

main source yeah but Who provided a ton

00:12:42

of detail I don't know exactly how much

00:12:43

he was able to directly corroborate with

00:12:45

other sources I don't know that the

00:12:46

other thing I wonder sax is he

00:12:49

with this one-store story he has a

00:12:52

collaboration with the New Yorker

00:12:54

and anybody would if this was a really

00:12:57

well sourced and you could back it up

00:12:59

kind of story

00:13:00

they would love to have the ratings for

00:13:02

this story there's a blockbuster story

00:13:03

or as it happens a long time which we

00:13:06

learned was it through the Valerie plan

00:13:08

Affair which is that these major news

00:13:10

Publications will some sometimes have a

00:13:13

back Channel back to the National

00:13:14

Security apparatus when they have

00:13:16

something like this and the message was

00:13:18

you can't hit print on this yeah in

00:13:20

which case possibilities too he goes and

00:13:22

just self-publishes himself which wasn't

00:13:24

really even an option

00:13:25

a few years ago no yeah so I mean this

00:13:28

is this is the raw shock test of raw

00:13:30

shock tests you have the media you got

00:13:32

the CIA it's fodder for a great movie

00:13:34

when I go back to Jay Cal is I think you

00:13:35

can lay out some theories about let's

00:13:37

say you know the polls did it or maybe

00:13:41

ukrainians with the British or something

00:13:43

yes you could lay out those theories but

00:13:45

the media wasn't willing to entertain

00:13:48

any of those theories when this news

00:13:49

broke what did they do they blame the

00:13:51

Russians and that story made no sense

00:13:53

but they said it so definitively they

00:13:56

just for the source on that I'm looking

00:13:57

for Nick if you could pull up the source

00:13:59

of the administration blaming the

00:14:01

Russians I just want to make sure we're

00:14:02

active at a press conference yeah

00:14:05

by the Russians now they didn't push it

00:14:08

that hard what was interesting is the

00:14:10

bite Administration set up but they

00:14:11

didn't keep coming back to it but the

00:14:13

media really ran with it and when people

00:14:16

on both the left and the right like

00:14:18

people like Jeffrey sacks on the left

00:14:20

and Tucker cross on the right basically

00:14:21

started a question whether the U.S could

00:14:23

have done it they were accused of being

00:14:24

conspiracy theorists and you know Putin

00:14:27

Stooges and all the rest of it and now

00:14:29

all of a sudden Here Comes Seymour Hersh

00:14:31

with a pretty detailed story lending

00:14:34

Credence to that point of view

00:14:36

yeah it just may indicate we don't know

00:14:39

everything that's going on with this war

00:14:41

and um I think the longer this goes on

00:14:43

the more dangerous it is

00:14:46

Freeburg you have any thoughts on

00:14:47

geopolitical issues and who might have

00:14:49

blown up north stream uh what are your

00:14:51

sources saying Freeburg

00:14:54

what are your sources and science quote

00:14:56

are saying speculation is best left to

00:14:59

the future I don't know about the whole

00:15:01

speculating on what could have been done

00:15:03

in the past by someone those are

00:15:05

conspiracy theories and little things

00:15:07

they're not they're not actually

00:15:08

conspiracies

00:15:10

North stream was blown up you understand

00:15:12

that right like there's no doubt I'm

00:15:14

saying

00:15:15

Theory oh yeah Jay calistheory is this

00:15:18

you've got to bring data to the table

00:15:20

Yeah pure speculation you got to bring

00:15:23

data to the table to make it uh by the

00:15:25

way speaking of which Radix Sikorski who

00:15:27

is a Polish

00:15:29

Diplomat I think he was like their

00:15:30

foreign minister when Nordstrom blew up

00:15:33

he tweeted a photo of it saying thank

00:15:36

you USA oh

00:15:38

which was one of the reasons why people

00:15:40

thought that okay like yeah of course

00:15:43

the US did it you know who has the

00:15:44

capability to do it who has the motive

00:15:46

to do it who said they were going to do

00:15:48

it

00:15:49

and who benefits

00:15:51

who benefits who is conducting NATO

00:15:54

exercises in that region right three

00:15:56

months ago yeah exactly and Jeffrey

00:16:00

sacks pointed that out on I think it was

00:16:02

a CNBC interview before they basically

00:16:04

stopped him because he's not allowed to

00:16:06

talk about this on you know Network TV

00:16:08

apparently but he basically pointed out

00:16:10

they were like us radar signatures in

00:16:12

the area President Joe Biden declared

00:16:14

that this is from Bloomberg that a

00:16:15

massive League from the north stream gas

00:16:16

pipeline system in the Baltic Sea was an

00:16:19

intentional act and that the Russian

00:16:20

statements about the incident shouldn't

00:16:21

be trusted it was a deliberate Act of

00:16:22

sabotage and now the Russians are

00:16:24

pumping out disinformation lies Biden 12

00:16:26

reporters Friday at the White House so

00:16:27

he didn't exactly say the Russians did

00:16:29

it he just said it's an act of Russia

00:16:31

what's that the media did you could do

00:16:33

one of these montages where it was like

00:16:35

Russian sabotage in fact

00:16:40

he said it was an act of sabotage you

00:16:43

should actually look at what Biden said

00:16:44

everything he said is absolutely true if

00:16:47

the US also did it correct

00:16:49

yeah

00:16:51

every single sentence that Joe Biden

00:16:54

said is 100 true whether we did it or

00:16:57

whether Russia did it it was a

00:16:59

deliberate Act of sabotage by us think

00:17:03

about it if we did it we know they

00:17:05

didn't do it and then we have to be like

00:17:07

careful about

00:17:08

suggesting they did it what are the

00:17:10

Germans think because this is the

00:17:12

Germans pipeline this so if we blew it

00:17:14

up that's also an uh explain to me

00:17:17

you're thinking on the chessboard of our

00:17:19

relationship with Germany if we blew it

00:17:21

up would they not also see that as a

00:17:23

hostile act this is why I asked if

00:17:24

Europe knew because I think you have to

00:17:26

tell Germany that it's going to happen

00:17:28

and I think the quid pro quo

00:17:31

with Germany is some amount of

00:17:33

Guaranteed Supply that the U.S directs

00:17:35

into Europe

00:17:37

so that they know that their long-term

00:17:38

LNG Supply is intact so that they become

00:17:42

ambivalent right there's a point of

00:17:43

indifference where the Germans say okay

00:17:46

we don't know when this thing is going

00:17:48

to get turned back on and we don't know

00:17:49

about the implications of it are here's

00:17:51

what our demands are meaning our energy

00:17:53

Supply our energy needs are

00:17:55

and so as long as the United States can

00:17:58

say look worst case we have stuff in the

00:18:01

you know spr that we can give you

00:18:02

there's probably a point of indifference

00:18:04

where the Germans say okay we're just

00:18:06

going to turn around and not say

00:18:07

anything and I'm curious did the Germans

00:18:09

say anything when this happened

00:18:11

I'm looking for the jury I was literally

00:18:14

just typed to Nick what was the German's

00:18:15

position on this because that's

00:18:16

interesting too that's a tell well yeah

00:18:18

if you're breaking this down like a

00:18:19

poker hand we're trying to figure out if

00:18:22

you construct this hand right pre-flop

00:18:24

it's like you know where both of these

00:18:25

folks are

00:18:26

right okay there was a really

00:18:28

interesting video that a guy named uh

00:18:31

Matt orphelia who puts together these

00:18:33

really funny videos he put together

00:18:35

these montages of media reactions to

00:18:38

things and what he shows is that you

00:18:40

know you can have like 20 different

00:18:41

media outlets and they all use the exact

00:18:43

same words

00:18:44

and when he Clips it together you can

00:18:46

see that the reading from somebody's

00:18:48

talking points and it's not really clear

00:18:50

who and and basically if you look at his

00:18:52

video here on who blew up north stream

00:18:55

pipeline you see that there was like a

00:18:57

party line from the mainstream media on

00:18:59

the stuff the Nordstrom too we will

00:19:01

bring it into it what how would you how

00:19:05

will you do that

00:19:07

I promise you we'll be able to do it

00:19:11

the Nordstrom pipeline I mean

00:19:15

I said it's most likely Russia Russians

00:19:18

sabotage on its own infrastructure it's

00:19:21

a common sense matter I think it's

00:19:23

Putin's way of sending a message what

00:19:25

guten is saying To Us by blowing up this

00:19:27

pipeline is look I can blow up a

00:19:29

pipeline everyone knows that Putin did

00:19:30

this himself who are these Talking Heads

00:19:33

without the correct proof yeah I think

00:19:35

logic and Common Sense will tell you

00:19:37

that without the evidence Russia was

00:19:38

behind the incident we can say it for

00:19:40

sure who sabotaged the nordstream 2

00:19:43

pipeline I had enough it's nonsense come

00:19:46

on you know it's these Talking Heads who

00:19:48

have no first-hand experience

00:19:51

it's fantastic it's fantastic all his

00:19:54

videos are like that where like he has

00:19:57

one on the hunter buying laptop as well

00:19:58

where again he's got like 20 different

00:20:00

Talking Heads and and media Outlets all

00:20:03

portraying it in exactly the same

00:20:05

language

00:20:09

Sunday morning shows you hear the same

00:20:12

narrative from each side how does that

00:20:15

actually get coordinated each side

00:20:16

builds those bullet points emails there

00:20:18

what do they call them um surrogates

00:20:20

they email all the surrogates and say

00:20:22

just keep saying these things over and

00:20:24

over again to codify yeah so what's up

00:20:27

group

00:20:27

it's a WhatsApp group and they're just

00:20:29

like say this over and over again how

00:20:31

does it work sex I think it's partly

00:20:33

talking points memos I go out to chat

00:20:35

groups I think it's also just people

00:20:37

looking on Twitter and then there's like

00:20:39

certain Keynotes that they follow and

00:20:42

they know okay this is the party line

00:20:43

because such and such key person is

00:20:45

saying it and they take their cues it's

00:20:47

that uh memetic memic effect that you

00:20:51

guys always talk about people

00:20:52

yeah the memetic yeah the thing to

00:20:55

understand is that all the prestige

00:20:56

Outlets repeat the same party line and

00:20:58

have the same perspective yeah

00:21:01

you got you gotta do your own search for

00:21:03

information this is the beauty of sub

00:21:04

stack actually that's why sub stack is

00:21:06

so important is it actually gives you

00:21:07

disruptive yeah alternative because like

00:21:10

you've got 10 different

00:21:13

mainstream media networks or newspapers

00:21:15

and magazines but they all have exactly

00:21:16

the same talking points except you know

00:21:18

maybe Fox News is kind of the one

00:21:19

exception although even Fox on the whole

00:21:21

Nordstrom thing you saw that a fox can

00:21:24

be pretty militaristic and they had the

00:21:26

same generals basically blaming the

00:21:28

Russians for this on Fox yeah Germany's

00:21:30

position is just hey everybody uh this

00:21:33

is sabotage so that's it

00:21:36

sabotage I think you asked a really good

00:21:38

question there about the German interest

00:21:39

in this right now the German economic

00:21:42

interest and the German foreign policy

00:21:45

interests are not aligned what's clearly

00:21:47

best for the German economy is to have

00:21:49

cheap natural gas powering its

00:21:51

Industries even if it comes from Russian

00:21:54

Pipelines

00:21:55

and they no longer have that anymore in

00:21:58

fact they may never have that again so

00:22:00

they're going to pay a very high price

00:22:01

economically maybe forever and remember

00:22:03

their their whole economy is based on

00:22:05

industry they're a very industrial power

00:22:08

so if this war drags on for a long time

00:22:11

I think Schultz might be in some

00:22:12

political trouble precisely because he's

00:22:14

gone along with the Americans on this

00:22:17

and there is there is a growing

00:22:19

political opposition to this war inside

00:22:22

of Germany War fatigue is a real thing

00:22:24

and this thing's got to wrap up at some

00:22:26

point any uh final thoughts Freeburg I

00:22:28

don't you didn't get too involved in

00:22:30

that conversation but what is uh any any

00:22:32

game theory from you

00:22:37

no okay there you have it folks

00:22:40

Sultan of science checking in what's the

00:22:42

problem you don't want to criticize the

00:22:44

establishments

00:22:47

is it predicted why don't you have my

00:22:50

position on this is it because I'm

00:22:52

trying to understand because I don't

00:22:53

think anyone I'm not I'm not

00:22:55

pro-establishment I think you know that

00:22:57

I think I'm I'm just analytical around

00:22:58

the fact that I think there's a strong

00:23:00

orientation towards conflict and I think

00:23:02

there still is I don't think that

00:23:03

there's much of an incentive or a

00:23:06

motivation to back down because this

00:23:08

conflict creates a significant amount of

00:23:10

debt owed back to the US it creates a

00:23:12

potential future asset stream

00:23:15

creates a realignment of power

00:23:17

everyone's looking externally as

00:23:20

internal you know economic

00:23:22

conflict and wealth disparity issues

00:23:24

arise and economic growth is challenged

00:23:27

and inflation is soaring it's a great

00:23:29

place to you know address one's energy

00:23:32

so I think all of this stuff is detailed

00:23:34

analytical Shenanigans around who's

00:23:37

saying what or who's doing what I think

00:23:38

the underlying thesis and the underlying

00:23:40

River that's flowing

00:23:42

is one that's like looking for external

00:23:44

conflict I think the same is true with

00:23:46

the US and China

00:23:48

and you're talking about the

00:23:49

military-industrial complex is going to

00:23:51

benefit massive limits the longer this

00:23:53

drags out the more the conflict in

00:23:56

Taiwan heats up

00:23:58

the more we're going to invest in our

00:24:00

military we often talk about these

00:24:02

things as if they're like top-down like

00:24:04

master plan driven and as we all know

00:24:06

like there are more Ouija board driven

00:24:09

it's a bunch of guys that got their hand

00:24:10

on the Ouija board and they're all just

00:24:12

had a little too much caffeine

00:24:14

you know and in this case I think it's

00:24:16

just more about like everyone's a little

00:24:18

anxious and the anxiety's leading to

00:24:21

a desire for more conflict we're not

00:24:23

happy at home if you're happy at home

00:24:25

you're not looking externally for

00:24:26

conflict that's true in nearly every

00:24:28

developed Nation on Earth today

00:24:31

that's it I don't know it's pretty

00:24:33

simple I like your position I think it's

00:24:34

great take I think it's a great thing I

00:24:37

think there are a lot of interests who

00:24:39

benefit from war and I think the foreign

00:24:41

policy establishment is funded by those

00:24:44

interests and it's kind of wired for war

00:24:46

at least in terms of the reflex right

00:24:48

even something as relatively harmless as

00:24:52

a balloon I saw that becomes like a

00:24:54

casus bella it's like people are ready

00:24:56

to go to war against China over that I

00:24:58

saw a couple military leaders give a

00:25:00

talk a few months ago it's in a private

00:25:02

thing

00:25:03

so it wasn't on public record with the

00:25:06

establishment no yeah I was at the

00:25:08

establishment Gathering where the

00:25:09

establishments oh this is the Illuminati

00:25:11

did you genuflect twice going in

00:25:14

because I'm genuflect on each knee then

00:25:15

they give you the bag of capitalism

00:25:19

yeah what was striking to me in this

00:25:22

particular thing where these guys were

00:25:24

being interviewed on stage at like a

00:25:26

dinner thing and they were

00:25:29

so oriented around their

00:25:33

next steps in escalation and I think it

00:25:36

speaks to the point Zach so like none of

00:25:38

them were thinking about like where are

00:25:40

we at today how do we de-escalate what

00:25:42

is this going to get there was no

00:25:43

conversation at all from anyone about

00:25:46

resolution or de-escalation with every

00:25:49

single one of them it was all about like

00:25:50

my orientation for getting bigger going

00:25:53

deeper going harder going stronger

00:25:55

making this thing bigger and I think

00:25:57

that was really scary to me because I

00:25:59

didn't I didn't hear anyone having a

00:26:00

conversation around like how do we

00:26:05

everyone was thinking assumptively it

00:26:07

was going to get bigger I like the Ouija

00:26:10

board you have to follow the financial

00:26:11

incentives when the last time we looked

00:26:13

at this right Leon Panetta and all these

00:26:15

other guys who were screaming for war

00:26:16

they were getting paid by the

00:26:18

military-industrial complex I remember

00:26:20

when Lloyd Austin was nominated as

00:26:22

defense secretary he had some conflict

00:26:24

issues because he was just on the board

00:26:26

of Northrop Grumman or one of these big

00:26:28

military industrial companies and so of

00:26:31

course these generals have to push for

00:26:33

war because as long as they're girding

00:26:35

forward they're guaranteed to have

00:26:37

for them a very

00:26:40

lucrative job once they leave the

00:26:42

military the Ouija board though sacksaw

00:26:44

and I'll throw to you maybe you can keep

00:26:46

this metaphor going the media's got

00:26:48

their hands on it they want ratings you

00:26:50

have the energy industrial complex in

00:26:51

this German conflict who seeks to

00:26:53

benefit massively if people invest in

00:26:55

Renewables or you find other you know

00:26:58

oil you know off Norway Norway's oil is

00:27:02

one of the largest reserves that's

00:27:03

untapped so you have this Ouija board

00:27:05

media energy and the military industrial

00:27:08

complex all moving it at once maybe you

00:27:10

can speak to that everyone wants to move

00:27:11

it to the side of the Ouija board that

00:27:13

says escalate there are very few people

00:27:15

that are that have the energy to move it

00:27:16

to the other side that says de-escalate

00:27:18

the escalation means less energy more

00:27:21

less investment so yeah we're gonna go

00:27:23

escalate you know who warned us about

00:27:25

the military-industrial complex yes

00:27:27

Dwight D Eisenhower Supreme online

00:27:29

during World War II wins the war Patriot

00:27:32

war hero top General becomes president

00:27:35

Republican president and is his

00:27:36

departing address warns us that yes we

00:27:39

need to defense industry but they become

00:27:42

a vested interest in favor of foreign

00:27:45

interventions of war and in 1960s

00:27:49

where's the interest on the other side

00:27:52

of it I can tell you this the American

00:27:54

people don't want to be in a war with

00:27:55

Russia I don't even think most American

00:27:57

people want to send 100 billion over

00:28:00

there they want to send 100 billion to

00:28:02

their cities to fix crime and all the

00:28:05

other problems homelessness everything

00:28:07

yeah if you've never seen that farewell

00:28:08

address from 1961 it is well worth

00:28:10

watching you can find it on YouTube just

00:28:12

search for military industrial complex

00:28:14

Eisenhower and he and this is a person

00:28:16

who was part of the military industrial

00:28:19

complex saying watch out for this it was

00:28:21

a very impressive yes he was in the

00:28:24

machine he helped build the machine

00:28:25

welcome to the all-in podcast with us

00:28:28

again David the dove sacks chamath

00:28:30

polyhapatia and the Sultan of science

00:28:32

who is on his podcast or my God so many

00:28:35

podcasts are doing the Sultan of science

00:28:36

is in

00:28:37

hot demand what are all these podcasts

00:28:40

you're doing Freeburg holy science

00:28:41

podcast pulling you in I did a podcast

00:28:43

with Brian Keating last week who was

00:28:45

really kind enough to reach out he's had

00:28:46

some awesome guests he's a

00:28:49

cosmologist at UCSD Professor down there

00:28:53

and we were supposed to record that day

00:28:55

and then we canceled I think last minute

00:28:57

right yeah yeah I was only supposed to

00:28:58

be on with him for an hour and I'm like

00:29:00

oh well my next thing just got freed up

00:29:01

so

00:29:02

ended up doing like three hours it was I

00:29:05

was so I was like exhausted that day so

00:29:07

I look really hungover on the video and

00:29:09

probably par for the course stumble a

00:29:10

lot yeah

00:29:11

but no Lex Friedman for you so chamath

00:29:13

and I have done Friedman Alex Friedman

00:29:15

but you have not he's invited you yet

00:29:16

Lex has Lex invited you no no he's not

00:29:19

here's my invitation what's going on

00:29:22

here collect all four Lex what are you

00:29:23

doing

00:29:24

all right let's know Davos and no luxury

00:29:26

Moon

00:29:27

what's going on you remember

00:29:29

anti-establishment what's going on no

00:29:31

I'm too anti-establishment sucks that's

00:29:33

you know mistress nobody nobody's

00:29:36

inviting this quartet to anything

00:29:39

well full stop we're not doing it all in

00:29:42

life from Davos it's not happening folks

00:29:44

sorry they don't stop that heat hi Grace

00:29:47

I don't want to be part of any club that

00:29:49

would have be as a member and they also

00:29:52

don't want you there anyway so yeah it

00:29:54

works out for everybody it does fill me

00:29:56

with like a rage where I actually might

00:29:59

agree to doing the all in Summit again

00:30:02

by the way proposal coming your way this

00:30:05

weekend if you want to really really

00:30:06

really thumb your nose at The

00:30:08

Establishment yes let's do it set it

00:30:10

during the exact same dates and times as

00:30:13

an establishment conference oh the

00:30:16

altitude and then invite all the best

00:30:17

guests so that they are they come to

00:30:19

ours Justice

00:30:23

establishment conference you could cut

00:30:25

the anti-establishment oh that's a

00:30:28

tagline let's come up with a tagline

00:30:29

that just tweaks everybody

00:30:31

list we should create the anti-list you

00:30:34

know they have their like establishment

00:30:36

list we should have the anti-list that's

00:30:37

a good point I like the

00:30:39

anti-establishment start with them

00:30:41

I have a great Vanity Fair establishment

00:30:44

thing story

00:30:46

okay I snuck on that list they put me on

00:30:49

that list a decade ago let's pull it let

00:30:51

me find a link go on the most incredible

00:30:53

thing about it is that when you go to

00:30:55

the event which is kind of a cool event

00:30:57

we all had photos taken by Annie

00:31:00

Leibovitz oh and I have a montage of

00:31:04

some of the people that took photos that

00:31:07

day

00:31:08

that's pretty cool me Aaron Levy

00:31:10

Priscilla Chan Bezos bunch of people but

00:31:13

I look like a toy I was about to say is

00:31:15

just before you had just what you had

00:31:17

the dad bod and nose fashion sense yeah

00:31:19

it was like yeah oh look he looks great

00:31:22

here it is

00:31:23

yeah basically I had no stylist

00:31:27

it was like glasses it was like a year

00:31:30

or two post Facebook it was not a good

00:31:32

look for me not yeah yeah but that's not

00:31:34

that bad I'll show you the picture Tom

00:31:36

Ford that was your time

00:31:38

he's doing the jeans Blazer thing which

00:31:41

is like a really tired look for a

00:31:43

Silicon Valley three skinny jeans though

00:31:45

it's like a little three skinny jeans

00:31:46

yeah

00:31:50

that's not so bad just don't pull up the

00:31:53

chamoth pictures when he's wearing like

00:31:55

oh my God you know his Macy's shirt if

00:31:57

you do the Google search you put the

00:31:59

images before 2011 you will find some

00:32:02

photos

00:32:04

I wear the same thing every day for four

00:32:07

years five years is brutal

00:32:09

oh oh my God what is

00:32:13

that's awesome we gotta break this down

00:32:15

yeah see this is when the sweater game

00:32:17

was not tight this is like sweater 1.0

00:32:19

game it says I didn't know what I was

00:32:21

doing back then stop but take these

00:32:22

pictures off please look oh he's also

00:32:24

got the watch subtly peeking out this is

00:32:26

back when he was like oh I got a Rolex

00:32:27

ooh that's a picture but anyways yeah

00:32:31

this is what the watches were only five

00:32:34

years out of zero I don't know if that

00:32:36

watch guys or two but yeah same story

00:32:39

same same Jake out same same same all

00:32:42

right well listen all right listen I got

00:32:43

an Apple Watch and I'm gonna upload oh

00:32:45

geez

00:32:47

yeah there's the dad vibes

00:32:50

wow that must have been from yeah

00:32:53

Mayfield 13 14 years ago wow God you had

00:32:57

no style I really didn't it was rough it

00:33:01

really is rough I mean don't pull up

00:33:02

pictures of me I was fat yeah yeah

00:33:04

pictures of me eating egg sandwiches all

00:33:06

over the Internet oh yeah chubby Jacob

00:33:08

oh man look at that face there he is oh

00:33:10

that's plus 20 pounds let's keep this

00:33:12

going search Wars Microsoft versus

00:33:15

Google okay it's been a rough couple

00:33:16

days for Google we've all seen it Google

00:33:18

and Microsoft both did live demos of

00:33:21

their new generative AI yada yada yada

00:33:23

you guys all know about Chachi PT but

00:33:25

now Bing is integrating it into their

00:33:27

search engine getting there before

00:33:28

Google and Microsoft CEO Satya nadala he

00:33:33

is going ham he looks great he's fit

00:33:36

he's wearing a tight t-shirt and he is

00:33:38

saying he's going to make Google dance

00:33:39

dance Google dance he is getting up in

00:33:43

their business and uh listen he is in a

00:33:46

distant second place so it makes sense

00:33:48

on the other hand uh Google's AI demo

00:33:51

was frankly a bit of a disaster poorly

00:33:54

received stock that dropped 12 since

00:33:56

this event and

00:33:58

and their presentation did not include

00:34:01

the chatbot Bard because in search

00:34:04

because it wasn't working it seems like

00:34:06

there was an error in it when they said

00:34:07

what new discoveries from the James Webb

00:34:10

Space Telescope can I tell my

00:34:12

nine-year-old and Bard answered that it

00:34:14

took the first pictures of a planet

00:34:16

outside our solar system which is false

00:34:18

which of course we all know about chat

00:34:21

gbt it's only right half the time and

00:34:23

it's a little woke on the margins

00:34:26

so um anyway there was a screenshot

00:34:30

circulating today which is probably

00:34:32

false but it says the following me and a

00:34:35

bunch of co-workers were just laid off

00:34:37

from Google for our AI demo going wrong

00:34:38

it was a team of 168 people who prepared

00:34:41

the slides for the demo all of us are

00:34:42

out of jobs I can't imagine that's real

00:34:43

uh but if it was that would be a

00:34:46

hardcore moment for Google to fire a

00:34:47

bunch of people for screwing it up

00:34:50

listen you worked in the belly of the

00:34:51

Beast Freeburg what are your thoughts on

00:34:53

being poking the tiger and telling

00:34:56

Google dance

00:34:58

you know Sundar dance you know what's

00:35:01

interesting is Google's had like an

00:35:03

incredible AI competency particularly

00:35:06

since they bought deepmind and it's been

00:35:08

predominantly oriented towards kind of

00:35:11

you know internal problems

00:35:13

you know they demonstrated last year

00:35:15

that they're AI improved data center

00:35:18

Energy Efficiency by 40 they've used it

00:35:21

for ad optimization ad copy optimization

00:35:24

uh the Youtube follow video algorithm so

00:35:27

what video is suggested to you as your

00:35:29

next video to watch which massively

00:35:31

increased YouTube hours watched uh per

00:35:35

user which massively increased YouTube

00:35:36

Revenue

00:35:38

you know what's the right time and place

00:35:39

to insert videos in YouTube or insert

00:35:41

ads and YouTube videos so you know

00:35:43

AutoFill in Gmail and docs so so much of

00:35:46

this competency has been oriented

00:35:49

specifically to avoid this primary

00:35:51

disruption in search

00:35:53

obviously now things have come to a bit

00:35:55

of a point because you know this

00:35:57

alternative for search has been revealed

00:35:59

in chat GPT

00:36:01

and you guys can kind of think about

00:36:03

search and you know we've used this term

00:36:05

in the past Larry and Sergey the

00:36:07

textbook that they read you know one of

00:36:09

the original textbooks that's used in um

00:36:12

internet search engine technology is

00:36:13

called information retrieval information

00:36:15

retrieval so information retrieval is

00:36:17

this idea that you you know how do you

00:36:19

pull data from a static data set and it

00:36:23

involves scanning that data set or

00:36:24

crawling it and then creating an index

00:36:26

against it and then a ranking model for

00:36:28

how do you pull stuff out of the index

00:36:29

to present the results from the data

00:36:32

that's available based on what it is

00:36:33

you're querying

00:36:35

for

00:36:37

you know and doing that all in a tenth

00:36:38

of a second so you know if you think

00:36:40

about the information retrieval problem

00:36:42

you type in the data or some rough

00:36:45

estimation of the data you want to pull

00:36:46

up and then a list is presented to you

00:36:48

and over time Google realized hey we

00:36:51

could show that data in smarter quicker

00:36:53

ways like if we can identify that you're

00:36:55

looking for a very specific answer we

00:36:57

can reveal that answer in the one box

00:36:58

which is the thing that sits above the

00:37:00

search results like if you said what

00:37:01

time is it you know what when does this

00:37:03

movie show at this theater so they can

00:37:05

pull out the structured data and give

00:37:06

you a very specific answer rather than a

00:37:09

list from the the database

00:37:11

and then over time there were other kind

00:37:13

of modalities for displaying data that

00:37:15

it turns out or even better than the

00:37:18

list like Maps or shopping where you can

00:37:20

kind of see a matrix of results or

00:37:22

YouTube where you can see a you know

00:37:24

longer form version of content and so

00:37:26

these different kind of um you know

00:37:28

information retrieval

00:37:31

you know media were presented to you and

00:37:33

it really kind of changed the game and

00:37:35

created much better user satisfaction in

00:37:37

terms of getting what they were looking

00:37:39

for the the challenge with

00:37:42

what this new modality is it's not

00:37:45

really fully encompassing so if you can

00:37:47

kind of think about the human computer

00:37:49

interaction problem you want to see

00:37:51

flight times and Airlines and the price

00:37:54

of flights in a matrix you don't

00:37:56

necessarily want a a text stream written

00:37:59

to you to give you the um you know the

00:38:02

the answer that you're looking for or

00:38:04

you want to see a visual display of

00:38:06

shopping results or you do want to see a

00:38:09

bunch of different people's commentary

00:38:11

because you're looking for different

00:38:12

points of view on a topic rather than

00:38:14

just get an answer but there are

00:38:16

certainly a bunch of answer solutions

00:38:17

for which chat DPT type you know natural

00:38:21

language responsiveness becomes a

00:38:23

fantastic and better mode to present

00:38:26

answers to you than the Matrix or the

00:38:29

list or the ranking and so on now the

00:38:31

one thing that I think is worth noting I

00:38:33

did a back of the envelope analysis on

00:38:35

the cost of doing this compared to chat

00:38:38

GPT so so Google makes about three bucks

00:38:40

per click you can back into what the

00:38:42

revenue per search is a bunch of

00:38:43

different ways one way is three bucks

00:38:45

per click about a three percent

00:38:46

click-through rate on ads some people

00:38:48

estimate this is about right about five

00:38:50

cents to ten cents Revenue per search

00:38:52

done on Google or anywhere from one cent

00:38:54

to 10 cents even if they don't click the

00:38:56

ads because one out of 100 people click

00:38:58

an ad and that's where the money comes

00:38:59

from so let's just call it five cents

00:39:01

right

00:39:02

and you can assume a roughly fifty

00:39:03

percent margin uh on that search which

00:39:05

means a fifty percent cogs or cost of

00:39:07

goods or a cost to run that search and

00:39:10

present Those ads so you know right now

00:39:13

Google search costs them about you know

00:39:16

call it two and a half cents per search

00:39:18

uh to present the results a recent

00:39:21

estimate on running the gpt3 model

00:39:24

for chat GPT is that each result takes

00:39:28

about 30 cents of compute so it's about

00:39:31

an order of magnitude higher cost to run

00:39:34

that search result

00:39:36

than it is to to do it through a

00:39:38

traditional search query today which

00:39:40

makes today today that's right and so so

00:39:43

that's the point like it has to come

00:39:44

down by about an order of magnitude

00:39:47

now this is a this then becomes a very

00:39:50

deep technical discussion that I'm

00:39:52

certainly not the expert but there are a

00:39:54

lot of great experts and there's great

00:39:55

blogs and sub stacks on this on what's

00:39:57

it going to take to get there to get a

00:39:58

10x reduction in cost on on running

00:40:01

these models and there's a lot related

00:40:03

to kind of optimization on how you run

00:40:05

them on a compute platform

00:40:07

the type of compute Hardware that's

00:40:09

being used all the way down to the chips

00:40:10

that are being used

00:40:11

so there's still quite a lot of work to

00:40:13

go before this becomes truly

00:40:16

economically competitive with Google and

00:40:18

that really matters because if you get

00:40:20

to the scale of Google you're talking

00:40:22

about spending eight to twenty billion

00:40:24

dollars a quarter just to run search

00:40:26

results and display them and so for chat

00:40:29

GPT type Solutions on Bing or elsewhere

00:40:31

to scale and to use that as the modality

00:40:33

you're talking about something that

00:40:35

today would cost 80 billion dollars a

00:40:37

quarter to run from a compute

00:40:39

perspective if you were to do this

00:40:40

across all search queries so it's

00:40:42

certainly going to be a total game

00:40:43

changer for a subset of search queries

00:40:45

but to make it economically uh work for

00:40:49

for these businesses whether it's Bing

00:40:51

or Google or others there's a lot of

00:40:53

work still to be done the great part

00:40:55

about this chamoth is that Bing gave 10

00:40:58

Billy to uh our friend Sam and Chad GPT

00:41:02

to invest in

00:41:03

Azure uh which now has the

00:41:06

infrastructure and we'll be providing

00:41:07

the chat GPT infrastructure to startups

00:41:09

or corporations big companies and small

00:41:11

alike so that 10 billion dollars should

00:41:14

do enough to grind it down between

00:41:15

software optimization data optimization

00:41:18

chip optimization and Cloud optimization

00:41:20

yes you would think so or no

00:41:23

the ability to run this at scale is

00:41:25

going to happen because

00:41:27

we're getting better and better at

00:41:30

creating silicon that specializes in

00:41:32

doing things in a massively parallelized

00:41:34

way

00:41:35

and the cost of energy at the same time

00:41:37

is getting cheaper and cheaper along

00:41:39

with it when you multiply these two

00:41:41

things together the effect of it is that

00:41:43

you'll be able to run these models the

00:41:45

same output today will cost 1 1 10. as

00:41:47

long as you ride the energy and compute

00:41:49

curve for the next few years so that's

00:41:51

just going to naturally happen I have

00:41:53

two interesting takeaways

00:41:55

and one is maybe a little bit of a

00:41:57

sidebar so the sidebar is if you guys

00:42:00

were sitting on top of something that

00:42:02

you thought was as foundational as

00:42:05

Google search back in 1999 would you

00:42:08

have sold 49 of it for 10 billion

00:42:10

dollars

00:42:11

hard though I think the answer is no I

00:42:13

think the answer is no

00:42:15

not in an environment where you have

00:42:17

unlimited ability to raise Capital this

00:42:19

is something that we've said before

00:42:20

which is that chat GPT is an incredibly

00:42:22

important innovation

00:42:25

but it's an element of a platform who

00:42:28

will get quickly commoditized because

00:42:29

everybody will compete over time and so

00:42:32

I think what Microsoft is doing is the

00:42:34

natural thing for somebody on the

00:42:37

outside looking in at an entity that has

00:42:39

93 share of a very valuable category

00:42:42

which is how can I Scorch the Earth and

00:42:44

so Microsoft effectively for 10 billion

00:42:47

bought almost 50 percent of a tool and

00:42:50

now we'll make that tool as pervasive as

00:42:53

possible so that consumer expectations

00:42:54

are such that Google is forced to Decay

00:42:58

the quality of their business model in

00:43:00

order to compete so that as Friedberg

00:43:02

said you have to invest in all kinds of

00:43:04

compute resources that today are still

00:43:06

somewhat expensive and that will flow

00:43:08

into the p l

00:43:09

and what you will see is that the

00:43:11

business quality degrades and this is

00:43:14

why when Google did the demo of Bard the

00:43:17

first thing that happened was the stock

00:43:18

went off 500 basis points we they lopped

00:43:21

off a hundred billion dollars of the

00:43:22

market cap mostly in reaction to oh my

00:43:25

God this is not good for the long-term

00:43:26

business

00:43:27

that's not good for the long-term

00:43:29

business on a mechanical basis when you

00:43:31

get an answer you don't have to click

00:43:32

the links no right now if you look at

00:43:34

Google's business they have the best

00:43:37

business model ever invented on Earth

00:43:40

ever for for-profit company it just

00:43:44

Reigns money this is a business that

00:43:47

this year will do almost 100 billion

00:43:49

dollars of free cash flow it's a

00:43:51

business that

00:43:52

has to find ways and we kind of joke but

00:43:54

they have to find ways to spend money

00:43:56

otherwise they'd be showing probably 50

00:43:58

or 60 percent ebitda margins and people

00:44:00

would wonder hey wait a minute you can't

00:44:02

let something like this go unattended so

00:44:05

they try to do a lot more things to make

00:44:07

that core treasure look not as

00:44:10

incredible as it is

00:44:12

they have 120 billion of cash

00:44:15

this is a business that's just an

00:44:16

absolute Juggernaut and they have 10

00:44:19

times as many employees as they need to

00:44:20

run the core business I don't know what

00:44:22

that is but my point is that it's an

00:44:24

incredible business so that business

00:44:26

will get worse if Microsoft takes a few

00:44:29

hundred basis points of share if meta

00:44:31

takes a few hundred basis points of

00:44:33

share if 10 cent does if a few startups

00:44:35

do

00:44:36

Cora by the way launched something

00:44:38

called Poe which I was experimenting and

00:44:40

playing around with last weekend if you

00:44:42

add it all up what Satya said is true

00:44:44

which is even if all we do collectively

00:44:46

as an industry is take 500 or 600 basis

00:44:49

points of share away from Google

00:44:52

it doesn't create that much incremental

00:44:55

cost for us but it does create enormous

00:44:58

headwinds and pressure for Google with

00:45:01

respect to how they are valued and how

00:45:03

they will have to get revalued and

00:45:04

that's what happens so the last thing

00:45:06

I'll say is the question that I've been

00:45:08

thinking about is what does Sundar do

00:45:09

right so what's the counter measure

00:45:12

yes this is what I was going to get I

00:45:14

think the counter measure here

00:45:16

if I was him is to go to the board and

00:45:19

say guys we're going to double tack

00:45:22

right so attack is the traffic

00:45:23

acquisition cost that Google pays their

00:45:25

Publishers it is effectively their way

00:45:28

of guaranteeing an exclusivity on search

00:45:31

traffic so for example if you guys have

00:45:33

an iPhone it's Google search that's the

00:45:35

default search in the iPhone Google pays

00:45:38

Apple

00:45:40

this year the screen negotiation for

00:45:42

that deal could mean that Apple gets

00:45:43

paid 25 billion dollars for giving away

00:45:45

that right to Google so if these Google

00:45:48

does all these kinds of deals last year

00:45:49

they spent I think 45 billion or so so

00:45:52

about 21 when you think about that

00:45:54

Google basically paid Apple which was

00:45:56

working on search technology they were

00:45:58

working on a search solution they paid

00:46:00

them to stand there and they're paying

00:46:02

everybody so I think the question for

00:46:03

Google is the following if you think

00:46:05

you're going to lose share

00:46:07

and let's say you go to 75 share

00:46:10

would you rather go there and actually

00:46:12

still maintain your core strangle hold

00:46:15

on search or do you actually want 75

00:46:19

share where now all of these other

00:46:21

competitors have been seated well you

00:46:24

can Decay business model quality and

00:46:26

still remain exclusive if you just

00:46:27

double the tack

00:46:29

and what you do is you put all these

00:46:31

other guys on their heels because as we

00:46:32

talked about if you're paying Publishers

00:46:34

two times more than what anybody else is

00:46:37

paying them you'll be able to get

00:46:39

published to say hey you know what don't

00:46:41

let those AI agents crawl your website

00:46:43

because I'm paying you all this money

00:46:44

remember that right so do not crawl in

00:46:47

robots.txt equivalent for these AI

00:46:50

agents and I think that that'll put

00:46:52

Microsoft and all these other folks on

00:46:54

their heels and then as as you have to

00:46:56

figure out all this derivative work

00:46:58

stuff all these lawsuits

00:47:00

Google will look pristine because they

00:47:02

can say I'm paying these guys double

00:47:03

because I acknowledge absolutely this is

00:47:05

a core part of the service so that's the

00:47:07

game theory I think that has to get

00:47:08

figured out but if I was soon

00:47:10

I love the second part because hold on

00:47:13

let me get saxophone yeah I love the

00:47:14

second part chamoth because in this clip

00:47:17

I'm about to show Nile Patel from The

00:47:19

Verge did an awesome interview with

00:47:22

Satya and he basically would not answer

00:47:26

this question at least to my

00:47:28

satisfaction which is

00:47:29

hey what do the Publishers get out of

00:47:31

this you've ingested our information how

00:47:33

do we get paid watch this clip it's very

00:47:36

telling the answer or even in the chat

00:47:38

session but if I ask the newbing what

00:47:40

are the 10 best gaming TVs it just makes

00:47:42

me a list why should I the user then

00:47:44

click on the the link to the verge which

00:47:48

has another list of the 10 best gaming

00:47:50

TVs well

00:47:51

a question but even there sort of say

00:47:53

hey who added these things come from uh

00:47:55

and would you want to go dig in like

00:47:57

that yeah even search today has that

00:47:59

like we have answers they may not be as

00:48:01

high quality answers they just are

00:48:03

getting better so I don't think of this

00:48:05

as a complete departure from what is

00:48:07

expected of a search engine today which

00:48:09

is supposed to really respond to your

00:48:11

query while giving them the links that

00:48:14

they can then click on like ads and such

00:48:17

works that way in my mind there's a

00:48:19

terrible answer he needs to address how

00:48:21

they get paid he punted the answer and

00:48:22

just said Hey listen search works this

00:48:23

way Saks will the rights to the data

00:48:27

will will Google just say to quora hey

00:48:29

we'll give you a billion dollars a year

00:48:31

for this data set if you don't give it

00:48:33

to anybody else they should maybe sax

00:48:35

the strategist let me hear your strategy

00:48:36

here you are now CEO of Google what do

00:48:38

you do I think there's maybe even a

00:48:40

bigger problem before that which is I

00:48:43

think the whole monetization model might

00:48:46

change so the reason why Google

00:48:47

monetizes so well is is perceived as

00:48:50

having the best search and then it gives

00:48:52

you a list of links and a bunch of those

00:48:55

links are paid and then people click on

00:48:57

them now I think when you search an AI

00:49:00

you're looking for a very different kind

00:49:02

of answer you're not looking for a list

00:49:03

of 10 or 20 links you're just looking

00:49:06

for the answer

00:49:07

and so where is the opportunity to

00:49:09

advertise against that I mean maybe you

00:49:11

can charge like an affiliate commission

00:49:14

if the answer contains a link in it or

00:49:17

something like that but then you have to

00:49:19

ask the question well does that distort

00:49:21

like best answer like am I really

00:49:23

getting the best answer or am I getting

00:49:25

the answer that someone's willing to pay

00:49:26

for this is your key Insight the fact is

00:49:29

if Google gives you an answer you don't

00:49:31

click on ads Google has had a very

00:49:33

finely tuned

00:49:35

balance between hey these first two or

00:49:38

three paid ads these might the paid

00:49:41

links might actually give you a better

00:49:43

answer than the content below them but

00:49:45

in this case if the chat GPT tells you

00:49:47

hey this is the top three televisions

00:49:48

these are the top three hotels these are

00:49:51

the top three ways to you know write a

00:49:53

better essay you don't need to click you

00:49:55

have to give an answer and the model is

00:49:58

gone the paid link is still

00:50:01

a subset in that case so at Google we

00:50:03

used to have a key metric was the bounce

00:50:06

back rate so when a user explains yeah

00:50:10

so when a user clicks on a result on the

00:50:12

search results page

00:50:14

we could see whether or not they came

00:50:16

back and searched again and so that

00:50:19

tells you the quality of the result that

00:50:21

they were given because if they don't

00:50:22

come back it means they ended up getting

00:50:24

what they were looking for and so ads

00:50:27

that um performed better than organic

00:50:29

search results which means someone

00:50:31

created the ad paid for it and the user

00:50:35

clicked on it and didn't come back and

00:50:37

came back with less frequency than if

00:50:39

they clicked on an organic result that

00:50:41

meant that the ad quality was higher

00:50:43

than organic quality and so the ad got

00:50:45

promoted to kind of sit at the top and

00:50:47

it became a really kind of important

00:50:48

part of the equation for Google's

00:50:50

business model which is how do we Source

00:50:52

how do we monetize more search results

00:50:54

where we can get advertisers to pay for

00:50:57

a better result than What organic search

00:51:00

might otherwise kind of show and so it's

00:51:02

actually better for the user in this

00:51:04

case than say just getting an answer for

00:51:07

example I'm looking for a Playstation 5.

00:51:09

I don't want to just be told hey go to

00:51:11

Best Buy and buy PlayStation 5. I want

00:51:13

to be taken to the check account page to

00:51:15

buy a PlayStation 5 and I am more likely

00:51:18

to be happy if I click on a result and

00:51:20

it immediately takes me to the checkout

00:51:21

page and Best Buy is really happy to pay

00:51:23

for you to get there because they don't

00:51:24

want you looking around the internet

00:51:25

looking for other places and a lot and

00:51:27

we we can't convolute all sorts of

00:51:29

queries not all search queries are hey

00:51:32

you know what's the best dog to get to

00:51:34

not pee on the floor or whatever kind of

00:51:36

arbitrary question you might have that

00:51:37

you're doing research on many search

00:51:39

queries are Commerce intention related I

00:51:43

want to buy a flight to go somewhere I

00:51:44

want to book a hotel to go somewhere I

00:51:46

want to buy a video game system

00:51:48

Etc that series of queries may have a

00:51:51

very different kind of modality in terms

00:51:52

of what's the right interface versus the

00:51:55

chat GPT interface where yeah there's a

00:51:57

lot of kind of organic results that

00:51:59

people sift on the internet for today

00:52:00

and and and and the question earlier

00:52:03

can be resolved by Google doing a simple

00:52:06

analytical exercise which is you know

00:52:08

what's it going to cost us and what's

00:52:09

going to give the user the best result

00:52:10

and that's ultimately what will kind of

00:52:12

resolve to the Better Business model

00:52:13

it's really measurable I think on

00:52:15

jamoff's point you know today Google

00:52:17

pays Apple 15 billion dollars a year

00:52:20

to be the the default

00:52:22

search engine on iPhones on the uh

00:52:25

Safari browser

00:52:26

that's only about a quarter of Google's

00:52:28

overall tack the majority of Google's

00:52:30

traffic acquisition cost is actually not

00:52:32

being paid for search a good chunk of

00:52:34

that is being paid to Publishers to do

00:52:35

AdSense display ads on their sites and

00:52:39

Google's rev share back to them for

00:52:40

putting ads on their sites so you know

00:52:42

the TAC number

00:52:44

I think maybe you kind of want to move

00:52:45

the needle but the majority of Google

00:52:46

searches don't come through the default

00:52:48

search engine that they pay and it might

00:52:50

be on it it might it might move the

00:52:52

needle a bit but I don't think it really

00:52:53

changes the the equation for them my

00:52:55

comment is more Tac has to become a

00:52:57

weapon on the forward foot number one so

00:53:01

if you're going to spend 21 of your

00:53:02

revenue on Tac you should be willing to

00:53:04

spend 30 to 40 percent to maintain the

00:53:06

93 market share I don't think what you

00:53:09

want to see is your profit dollars Decay

00:53:12

because you lose share it's rather

00:53:13

better for you to spend the money and

00:53:16

Decay your business model than have

00:53:17

someone decayed for you at this point

00:53:20

apple is really the only tack line item

00:53:23

I understand I'm not talking about today

00:53:25

I'm saying take that idea you have an

00:53:29

entire sales team whose job it is right

00:53:31

now to sell AdSense right you have an

00:53:33

entire group of people who know how to

00:53:34

account for tact and how to think about

00:53:36

it as a cost but if you're basically

00:53:38

willing to say out of the 100 billion

00:53:41

dollars of free cash flow I'm willing to

00:53:43

go to 80. or 70 billion of free cash

00:53:46

flow combined with the 100 billion of

00:53:49

short and long-term Investments I have

00:53:50

and I'm going to use it as a weapon and

00:53:53

I'm going to go and make sure that all

00:53:55

of these Publishers have a new kind of

00:53:57

agreement that they signed up for which

00:53:59

is I'll do my best to help you monetize

00:54:01

you do your best by being exclusive to

00:54:04

our AI agents right so you deprive other

00:54:07

models

00:54:09

of your content on your pages because

00:54:11

that will get litigated and there is no

00:54:13

way just like again if you say Do not

00:54:16

crawl you're not allowed to crawl if

00:54:17

you're Google or Microsoft researches so

00:54:19

I mean so this is going to happen for

00:54:21

these agents it's unrealistic to expect

00:54:22

that it won't so my point is Google

00:54:24

should do this and Define how it's done

00:54:27

before it's defined for them because

00:54:29

right now people are in this nascent

00:54:31

phase where everybody thinks everybody's

00:54:32

going to be open and get along and I

00:54:35

just think that that's unrealistic it's

00:54:37

a really important kind of philosophical

00:54:38

question first off like Google today

00:54:40

just so people know on AdSense is

00:54:42

typically paying out 70 cents on every

00:54:44

dollar to the publisher there's you know

00:54:46

it's a it's a it's a pretty you know

00:54:48

generous and it's it's the way they've

00:54:49

kind of kept the competitive mode

00:54:51

you know wide and kept folks out of

00:54:53

beating them on third-party ad Network

00:54:55

bids because they bid on everything and

00:54:57

they always win because they always

00:54:58

share the most Revenue back so they they

00:55:00

own that market with respect to

00:55:02

acquiring content you know the the

00:55:04

internet is open it's an open protocol

00:55:06

anyone can go to any website by typing

00:55:08

in the IP address and viewing the

00:55:09

content that a publisher chooses to make

00:55:11

available on that server to display to

00:55:13

the internet and there's a fair use

00:55:14

policy and there's you can you can type

00:55:17

in this IP address I think 15 or 20 or

00:55:19

30 of the pages on the internet right

00:55:21

now are apps that are closed Facebook's

00:55:23

close Instagram I'm talking about the

00:55:25

open internet right so like the content

00:55:26

on the internet but I'm saying the open

00:55:28

internet matters less and less

00:55:30

yeah I don't know I mean look you're

00:55:31

right maybe there's there's the

00:55:33

enhancement of the models but my point

00:55:34

being that if the internet is open and

00:55:36

you and I spent a billion lifetimes

00:55:39

reading the whole internet and getting

00:55:40

smart and then we were the chatbot and

00:55:42

someone came and asked us a question and

00:55:44

we could kind of answer their question

00:55:46

um because we've Now read the whole

00:55:47

internet do I owe licensing royalty

00:55:50

revenues to the knowledge that I gained

00:55:52

and then the synthesis that I did which

00:55:55

ultimately meant excluding some things

00:55:56

including some things combining certain

00:55:58

things and the problem with these llms

00:56:01

these these large language models is

00:56:03

that you end up with a hundred million a

00:56:05

billion plus parameters that are in

00:56:08

these models that are really impossible

00:56:09

to deconvolute you don't know we don't

00:56:12

really understand deeply how the neural

00:56:14

network is the the model is defined and

00:56:18

run based on the data that it is

00:56:19

constantly kind of aggregating and

00:56:21

learning from and so to go in and say

00:56:23

hey it learned a little more from this

00:56:24

website and a little less from that

00:56:25

website I'm not saying it's a practical

00:56:27

impossibility I'm saying when you look

00:56:29

at Transformer archive texture today

00:56:30

every llm that you write on the same

00:56:34

Corpus of underlying data for training

00:56:36

will get to the same answer so my point

00:56:38

is today if you're a company

00:56:41

the most important thing that you can do

00:56:43

especially you have a one trillion

00:56:45

dollar plus market cap

00:56:46

that could get competed away is to

00:56:49

figure out how to defend it and so all

00:56:51

I'm saying is from the perspective of a

00:56:53

shareholder of Google and also from the

00:56:55

perspective of the board of director or

00:56:57

senior executive or the CEO this should

00:57:00

be the number one thing that I'm

00:57:01

thinking about and my my framing of how

00:57:04

to answer that question is build a

00:57:06

competitive mode around two things one

00:57:10

is at the end which is how much money

00:57:13

and what kind of relationship do I have

00:57:15

with my customers including the

00:57:19

publishers

00:57:20

and can I give them more so that number

00:57:23

two is I can affect who they decide to

00:57:27

contribute their content to so you're

00:57:29

right let's assume that there are

00:57:31

five of these infinite libraries in the

00:57:33

world you mean non-public content how

00:57:36

important and also public how important

00:57:38

is it if quora says you know what guys

00:57:40

I've done a deal where my billions of

00:57:43

page views and all of that really rich

00:57:45

content core has incredible content

00:57:47

Google's paying me 2 billion a year and

00:57:49

so I've decided to only let Google's AI

00:57:51

agents crawl it and so maybe when there

00:57:54

are questions that quora is already

00:57:57

doing a phenomenal job of answering

00:58:00

I think it does make a difference that

00:58:02

Google Now has access to quora's content

00:58:04

and others don't right for hot minute

00:58:07

they did have access to the Twitter fire

00:58:08

hose and that was the premise was we

00:58:10

could get this Corpus of data that we

00:58:13

can have in a very limited restricted

00:58:14

way they paid Twitter a lot of money I

00:58:16

don't think that those deals exist

00:58:18

anymore Twitter I mean you guys might

00:58:19

know better than I do but I don't think

00:58:20

they exist anymore here's where chamont

00:58:22

I think you're right

00:58:23

um and maybe

00:58:25

Dave I think you're being too forgiven

00:58:27

these models know where they got the

00:58:29

data and they can easily cite the

00:58:32

sources and they could easily pay for it

00:58:34

and if you want

00:58:36

go ahead you're absolutely right the

00:58:39

video In The Wall Street Journal where

00:58:41

Satya was interviewed showed a demo and

00:58:43

you're exactly right they actually

00:58:45

showed Jason in the search results yes

00:58:47

but it made no sense because it's like

00:58:50

how do you know that those are the five

00:58:51

most cited places that resulted in my

00:58:54

page rank technology or the authority of

00:58:57

the website or the author but let's

00:58:59

pause for a second here there is a

00:59:00

company called neva.com I'm not an

00:59:03

investor none of us are it's a former

00:59:05

googler they have 78 employees I think

00:59:07

according to LinkedIn I just typed in

00:59:09

what are the best flat panel TVs here's

00:59:11

the result and as you see

00:59:13

sentence by sentence as it rewrites

00:59:16

another person's content it links with a

00:59:19

citation just like the Wikipedia does

00:59:21

and when you scroll to the bottom of it

00:59:22

it tells you hey this is from Rolling

00:59:24

Stone this is from Best Buy this is from

00:59:25

ratings and if that answer is good for

00:59:27

you and you trust those sources those

00:59:29

people should get a commission every

00:59:31

time there's a thousand searches and you

00:59:33

come up you should get a dollar every

00:59:34

time your data was used

00:59:36

these sites should sue the Daylights out

00:59:40

of Google and forget how to say they

00:59:41

can't do it is

00:59:44

not fair use because in fair use you

00:59:49

have the ability to create derivative

00:59:50

works on future platforms and you are

00:59:53

taking this person the original content

00:59:55

owner's ability to exploit that and you

00:59:58

are co-opting it and you're doing it at

00:59:59

scale and that is against fair use

01:00:01

you're not allowed to interfere with my

01:00:03

ability to make future products David

01:00:04

you know this as an attorney the problem

01:00:06

with that lawyer sorry all right well I

01:00:09

play one on TV in this podcast the

01:00:11

problem with that idea just from a

01:00:12

product perspective for a second is that

01:00:14

if you limit

01:00:17

how they can tokenize to just being all

01:00:20

entire sentences

01:00:22

the product will not be that good like

01:00:24

the whole idea of these llms is that

01:00:25

you're running you know so many

01:00:27

iterations to literally figure out what

01:00:29

is the next most best word that comes

01:00:31

after this other word and if you're all

01:00:34

of a sudden stuck with blocks of

01:00:35

sentences as inputs that can't be

01:00:37

violated because of copyright the

01:00:39

product will not be as good I don't just

01:00:40

don't think it'll be as useful

01:00:42

correct these are also not deterministic

01:00:44

models and they're not deterministic

01:00:45

outputs meaning that it's not a discrete

01:00:48

and specific answer that's going to be

01:00:50

repeated every time the model is run

01:00:51

these are statistical models so they

01:00:54

infer what the right answer could or

01:00:57

should be based on a corpus of data and

01:00:58

a symphysis of that data to generate a

01:01:00

response to a query that ref that

01:01:03

inference is going to be you know assign

01:01:06

some probability score and so the model

01:01:08

will resolve to something that it thinks

01:01:09

is high probability but it could also

01:01:11

kind of say there's a chance that this

01:01:12

is the better answer this is the better

01:01:13

answer and so on and so when you have

01:01:16

like you have in the internet competing

01:01:17

data competing points of view competing

01:01:20

opinions the model is synthesizing all

01:01:23

these different opinions and doing what

01:01:24

Google search engine historically has

01:01:26

done well which is trying to rank them

01:01:28

and figure out which ones are better

01:01:29

than others and that's a very Dynamic

01:01:31

process and so if as part of that ingest

01:01:33

process One is using some open openly

01:01:36

readable data set that doesn't

01:01:38

necessarily mean that that data set is

01:01:40

improving the quality of the output or

01:01:42

is necessarily the answer from the

01:01:45

output correct let me just give

01:01:47

everybody a quick four Factor education

01:01:50

on fair use and here it is from Google's

01:01:52

actual website because they deal with

01:01:54

this all the time and when you look at

01:01:56

that the nature and purpose and

01:01:58

character of the work including whether

01:01:59

such use is not profit or educational

01:02:02

purposes so that first test of fair use

01:02:05

is hey if you're using it educationally

01:02:06

and you want to make a video that is

01:02:09

criticism of Star Wars prequels or how

01:02:13

to shoot a shot like this Quentin

01:02:15

Tarantino if it's educational it's fine

01:02:17

and Court's typically Focus I'm reading

01:02:19

here from Google on whether the use is

01:02:20

transformative that is whether it adds

01:02:22

new expression or meaning to the

01:02:23

original or whether it merely copies the

01:02:25

original it's very obvious that this is

01:02:27

not transforming if they're just

01:02:28

rewriting it the nature of the copyright

01:02:30

is pretty transformative to me I don't

01:02:32

think so they're coming out with

01:02:35

entirely new content they're just

01:02:36

rewriting it they're not actually adding

01:02:39

anything to it would be fine if a human

01:02:42

does it it's not

01:02:44

because

01:02:46

it's pretty cool listen Jacob thank you

01:02:49

I think the rights issue is just like

01:02:51

the cost issue which is a problem today

01:02:54

maybe but it's going to get sorted out

01:02:56

but here let me finish new technology

01:02:58

waves that are this powerful don't get

01:03:00

stymied by either chip costs or legal

01:03:04

rights issues They Do by Leo it's 100

01:03:07

YouTube got stopped dead in their tracks

01:03:09

and the only way YouTube and Napster got

01:03:11

stopped in the tracks I predict this is

01:03:12

going to get stopped dead in its tracks

01:03:14

with YouTube level

01:03:17

this is different and Google was enabled

01:03:20

piracy and then they had to build tools

01:03:22

disagree with jcal I deeply disagree

01:03:24

with you I think both you guys you guys

01:03:26

I think they think that all that costume

01:03:28

let me read you number four well he's

01:03:29

gonna stay in the way of the AI no it's

01:03:32

going to stay away

01:03:37

happening AI is happening like let's

01:03:40

move the conversation for it I actually

01:03:41

I want to tell you that's my point I

01:03:43

would like to make my point we don't

01:03:44

your amateur lawyer opinions I am going

01:03:47

to give my point I don't give a [ __ ] if

01:03:48

you want it or not the effects 500 bucks

01:03:51

an hour for pretending to be alone

01:03:57

before from you here it is okay take it

01:04:00

easy Mr Sub better call

01:04:03

jaycalif time you've done this the

01:04:05

effect listen to this the effect of the

01:04:07

use upon the potential market for or

01:04:10

value of the copper work that harm the

01:04:13

copyright listen to this very important

01:04:14

yeah it takes over the world like Skynet

01:04:17

Jacob's gonna be like I thought we'd

01:04:19

stop this with rights listen to this AI

01:04:22

is not going to be stopped but companies

01:04:24

using AI to steal content will be

01:04:26

effective use

01:04:27

is that farm the original copyright

01:04:30

owner's ability to profit from his or

01:04:32

her original work by serving as a

01:04:34

replacement

01:04:35

for that works

01:04:38

is the one okay great okay

01:04:41

you made your point and you may be right

01:04:44

get to Marjorie Taylor green for you

01:04:48

let's chew up Marjorie Taylor green no

01:04:50

no hold on hold on how much did you

01:04:52

weigh I have something I have another

01:04:54

aspect of the AI thing I want to talk

01:04:55

about besides just this like internal

01:04:58

rights issue that you're going on all

01:05:00

right better call Jake I'll go

01:05:02

so I I had an interesting you know AI

01:05:04

experience this week and I I think we're

01:05:06

all going to start having these stories

01:05:08

to make a script of how to talk to your

01:05:10

kids every week there'll be some new

01:05:12

like use case that you see that you're

01:05:14

kind of Blown Away by the use case I saw

01:05:16

this past week in a product demo was

01:05:19

they were showing me like an Excel

01:05:21

spreadsheet like a very complicated it

01:05:24

could sell spreadsheet modeling and

01:05:25

financial asset and they had a plug-in

01:05:28

to a chat GPT type Ai and so they just

01:05:33

asked it they typed in what does this

01:05:35

spreadsheet do and it spit out like a

01:05:37

one paragraph explanation of what the

01:05:39

spreadsheet did and it was really good I

01:05:42

mean because me just eyeballing the

01:05:43

spreadsheet I could not have figured out

01:05:45

like instantly what that thing did it

01:05:47

would have to take me like a while to

01:05:49

figure it out it told me here are the

01:05:50

key inputs here the key outputs so that

01:05:52

was number one then they they did

01:05:54

something

01:05:55

I think even more interesting which is

01:05:57

they said

01:05:58

give me the formula

01:06:01

that tells me when the yield is above

01:06:04

two percent and this and that and that

01:06:06

and the chat GPT spat out a formula that

01:06:10

was like perfect Excel logic that was

01:06:12

something that you know URI could never

01:06:14

figure out right you need like a super

01:06:16

pro user of excel to basically know how

01:06:19

to do this stuff so it spit it out and

01:06:21

like boom it worked instantly they copy

01:06:23

and pasted into the spreadsheet and you

01:06:25

could basically like the spreadsheet was

01:06:27

much more advanced now so what it got me

01:06:30

thinking about is that we're going to

01:06:31

have these little assistants everywhere

01:06:33

you combine that power with Say speech

01:06:38

to text right because we could have just

01:06:40

talked to it it would the speech to text

01:06:42

would transcribe the instruction spit it

01:06:44

back out and you're gonna have these

01:06:46

like little personal digital assistants

01:06:48

in applications I think you know it's

01:06:50

pretty obvious to see how AI could

01:06:53

replace call centers with you know

01:06:55

having the front line call center

01:06:57

operator be instead of being a human it

01:06:59

could be like an AI but this is actually

01:07:01

even before that like you could actually

01:07:03

I think in every single application that

01:07:05

we use there's going to be an AI

01:07:07

interface and like a it is probably

01:07:09

gonna be voice based where you can just

01:07:11

say to it hey I'm trying to accomplish

01:07:13

this

01:07:14

like how do I do it can you just make it

01:07:16

happen totally and it's going to be

01:07:18

really powerful actually I have an idea

01:07:20

I was hanging out with Andre karpathy

01:07:22

and I gave him this following

01:07:24

challenge so there I was so there I was

01:07:27

I said if you had to build stripe

01:07:30

I said how many Engineers do you think

01:07:32

it would take you and how long would it

01:07:34

take you to build a competitor I was

01:07:35

just just a thought exercise you know it

01:07:36

would take hundreds of millions of

01:07:38

dollars and years

01:07:40

now imagine you could

01:07:43

you were feeling threatened by stripe

01:07:45

imagine you're a large company

01:07:48

Visa Mastercard just as an example

01:07:51

you can now actually get one or two

01:07:53

really smart people like him

01:07:56

to lead an effort

01:07:57

where you would say here's a couple

01:08:00

hundred million dollars to compete with

01:08:01

stripe but here are the boundary

01:08:03

conditions number one is you can only

01:08:04

hire five or ten engineers

01:08:08

and so what you would do is you would

01:08:09

actually use tools like this to write

01:08:11

the code for you and the ability to

01:08:13

write code is going to be the first

01:08:15

thing that these guys that these things

01:08:17

do incredibly well with absolute

01:08:19

Precision you can already do unit

01:08:21

testing incredibly well but it's going

01:08:22

to go from unit testing to basically

01:08:24

end-to-end testing and you'll be able to

01:08:26

build a version of stripe extremely

01:08:29

quickly and in a very lean way so then

01:08:32

the question is well what would you do

01:08:33

with the two or 300 million you raised

01:08:35

and my thought is you use it again as

01:08:38

tack you go acquire customers you go to

01:08:41

customers and you're like well listen if

01:08:43

Braintree is gonna charge you one basis

01:08:45

point over Visa Mastercard and or sorry

01:08:48

100 basis points and stripel 50 you know

01:08:52

what I'll charge 10. margin destruction

01:08:54

margin destruction and this is going to

01:08:57

make everything what's so interesting

01:08:58

you can take any business that's a

01:09:00

middleman business I think this is the

01:09:02

point any middleman business right now

01:09:04

that doesn't have its own competitive

01:09:05

mode can be competed against because now

01:09:08

you can take all of those input costs

01:09:10

that go into human capital you can defer

01:09:13

that have a much smaller human capital

01:09:15

pool and push all of that extra money

01:09:17

into traffic Acquisitions

01:09:19

Silicon Valley of being more efficient

01:09:22

this is going to lead to that efficient

01:09:23

benefit of all of this is economic

01:09:25

productivity because the End customer

01:09:27

that's using that tool that you just

01:09:28

mentioned they now have a lower cost to

01:09:30

run their business and their total net

01:09:32

profits go up and this is what happens

01:09:34

with every technology cycle it always

01:09:36

yields greater economic productivity and

01:09:39

that's why the economy me growth and

01:09:40

that's why I just want to see it so

01:09:42

important that's why technology is so

01:09:44

important to drive economic growth not

01:09:46

debt we've historically used Financial

01:09:48

engineering to drive economic growth we

01:09:51

need immigration energy and innovation

01:09:52

drive this what about China yeah what

01:09:54

about China wait what are you talking

01:09:55

about what we're just immigration is is

01:09:58

AI making it harder and harder to make

01:10:01

humans productive

01:10:03

so you want to like bring in no millions

01:10:07

of Labor yeah

01:10:10

one human can can do the coding that 20

01:10:12

humans would do before assuming they're

01:10:14

skilled enough to use the AI guys go

01:10:16

back to go back to like traditional

01:10:17

Capital easier than programming go back

01:10:20

to traditional capitalism for a second

01:10:21

so please go back because the striped

01:10:23

example is another good one so if you

01:10:25

have this business model right how does

01:10:28

the ecosystem get efficient right how do

01:10:31

we create more opportunity to use

01:10:32

freeburg's language well the only way

01:10:34

that it really happens how does cost go

01:10:36

down is that certain entities become big

01:10:39

enough that they can drive the prices

01:10:40

down right an Uber a doordash or

01:10:43

whomever says yes I need payments

01:10:45

capability so Braintree give me your

01:10:47

best bid checkout.com give me your best

01:10:49

bid at the end give me your best bid

01:10:51

strike give me your best bid you compete

01:10:52

it Down Right Amazon comes in Walmart

01:10:55

comes in they do it in physical cpg

01:10:57

Goods they do it online they do it for

01:10:59

all kinds of Technologies

01:11:01

but you've never had an internal form

01:11:03

that can create hyper efficiency and and

01:11:05

basically create customer value

01:11:08

like this thing can because this thing

01:11:10

can allow let me show you an example

01:11:12

Julian 10 companies to get created that

01:11:15

can do the work of 10 000 people hold on

01:11:18

let me show you two things these are two

01:11:19

aha moments I had this week this first

01:11:20

one is called Galileo AI it was just a

01:11:22

tweet you describe the design of it here

01:11:25

they say an onboarding screen of a dog

01:11:27

walking app incredible and you type that

01:11:30

in and it gives you a welcome screen

01:11:32

that's like a seven out of ten then it

01:11:34

says oh a way for people to change their

01:11:36

name phone number and password you know

01:11:37

that classic screen on any app I need to

01:11:39

change my thing boom it gives you that

01:11:40

well then at the same time that people

01:11:43

are making text to ux user interface

01:11:47

beautiful here this stuff will get

01:11:48

dumped into figma but then

01:11:51

there's a GitHub co-pilot which if you

01:11:54

haven't seen we all know it here

01:11:56

this is allowing you GitHub and GitHub

01:11:58

co-pilot GitHub is bought by Microsoft

01:12:01

another one of Satya and nadella's

01:12:03

incredible Acquisitions this guy's like

01:12:05

the new Zuckerberg I mean what a what an

01:12:07

incredible

01:12:09

person to come after bomber who's just

01:12:11

so effective at what he's doing

01:12:13

as you're writing your code it fills in

01:12:16

your code it knows what you're writing

01:12:17

just like in an email and it's a smaller

01:12:19

subset of information than email email

01:12:21

you could write anything you could be

01:12:23

you know talking to a lover or a

01:12:25

business person whatever here when

01:12:27

you're doing programming it's a much

01:12:28

finer data set

01:12:30

these two things are going to come

01:12:31

together where you're going to be able

01:12:33

to build your MVP for your startup by

01:12:35

typing in text and then publish it

01:12:37

you're not going to need a developer for

01:12:39

your startup that is transformative in

01:12:41

the world let me ask you guys a question

01:12:42

do you think that this leverage and I I

01:12:45

argue this is all about like Leverage

01:12:47

it's one person can generate X units of

01:12:49

output more Yep do you guys think that

01:12:52

this commoditizes and puts at risk sex

01:12:55

in particular like all of Enterprise SAS

01:12:58

because it becomes such a commodity to

01:13:00

basically build a business that does

01:13:02

something now or does it create much

01:13:04

higher returns for investors because you

01:13:05

invest so much less Capital to get to a

01:13:08

point of productivity with that business

01:13:10

or revenue of that business than was

01:13:12

needed before

01:13:13

I'm not sure I tend to think that if it

01:13:16

gets easier then everything becomes more

01:13:18

competitive

01:13:19

so right

01:13:21

I don't know but so value gets competed

01:13:23

by the way I would slightly disagree

01:13:25

with the characterization does it scare

01:13:27

you as an investor and I I don't know I

01:13:29

mean like for example in that demo we

01:13:31

just saw what does the AI do it exports

01:13:34

the new design assets to figma format

01:13:37

because that's just making it faster so

01:13:39

if you can create a SAS product that

01:13:41

becomes a standard everyone's still

01:13:43

going to want to use it I mean there's

01:13:45

like really good reasons for that so I

01:13:47

don't know I don't think I don't think

01:13:48

business software is going away I also

01:13:50

don't think that like you're not going

01:13:52

to need to hire Engineers because

01:13:53

co-pilots is going to do it for you I

01:13:54

think what co-pilot will do is make your

01:13:56

typical engineer more productive right

01:13:58

yes they had some graphs around how like

01:14:02

copilot reduced coding Time by 50 right

01:14:05

so I think you'll be able to get a lot

01:14:07

more out of your developers I think

01:14:08

that's sort of the key is a lot of the

01:14:11

drudgery work gets taken care of the

01:14:13

answer Freeburg to your question is I

01:14:15

think more startups more Niche startups

01:14:17

will make better products and then

01:14:19

you'll just have many more folks

01:14:28

entry price matters it'll be poorer

01:14:30

returns well entry price matters too if

01:14:32

you're investing at 5 million like I do

01:14:34

in companies when they're just on

01:14:36

napkins and you know back of envelope

01:14:38

there's plenty of room if you have a 200

01:14:40

million dollar exit that's a 40x but

01:14:42

there could be a lot of new categories

01:14:43

too like a lot of new categories

01:14:47

the electrical industries that get

01:14:49

disrupted like we're thinking about just

01:14:51

software displacing software it could be

01:14:53

software displacing like industries that

01:14:56

aren't even soft video games I think the

01:14:58

entire video game industry is going to

01:14:59

get completely Rewritten with AI because

01:15:00

you're not going to have a publisher

01:15:01

anymore that makes one game that

01:15:03

everyone consumes you're going to have

01:15:04

tools that everyone creates and

01:15:06

consumes their own game you guys want to

01:15:08

watch this cool clip it's 58 seconds

01:15:10

long go ahead this is a music clip day

01:15:13

here on all in but I thought it was

01:15:15

really cool is this AI turning you into

01:15:17

Eminem this is what the world's waiting

01:15:19

for

01:15:20

I hope it's you doing I'm naturally

01:15:22

Eminem like but this is a only in your

01:15:24

anger this is the future

01:15:29

the future a sound I'm getting awesome

01:15:31

and on the ground What's That Eminem

01:15:35

at David Guetta playing Eminem a track

01:15:38

with Eminem's voice right yes

01:15:42

something that I made as a joke and it

01:15:45

works so good I could not believe it I

01:15:48

discovered those websites that are about

01:15:50

AI basically you can write lyrics in the

01:15:55

style of any artist you like so I typed

01:15:59

write a verse in the style of Eminem

01:16:02

about future Rave and I went to another

01:16:05

AI website that can recreate the the

01:16:10

voice I put the text in that and I

01:16:13

played the record and people went nuts

01:16:15

all right it's that is nuts it's nuts

01:16:19

so here it is folks whoever makes the

01:16:22

best Friedberg Eminem hybrid rap with

01:16:25

David sacks as the hype man's getting a

01:16:27

free VIP ticket to all in Summit 22

01:16:37

there are going to be a lot of

01:16:38

interesting mashups that get created

01:16:40

like for example you'd be able to create

01:16:42

a movie where let's say you want to make

01:16:44

a western you want John Wayne to star in

01:16:46

it I mean you obviously get the rights

01:16:48

but let's say you win a state but no

01:16:50

actor ever goes away you could there's

01:16:52

gonna be a database of all of them but

01:16:53

if you wanted to make it it's going to

01:16:54

be better you're going to be writing the

01:16:56

script as you write the script the AI is

01:16:58

going to be showing you that scene in

01:17:00

real time and you don't have to publish

01:17:02

it right that's the rainy day it'll just

01:17:04

change in real time

01:17:09

yeah just like Instagram and Tick Tock

01:17:12

basically democratize like everyone's

01:17:14

ability to create and publish content

01:17:15

this takes it to a whole nother level

01:17:17

where the Monopoly that big um

01:17:20

production houses have because they have

01:17:22

the big budget so they can afford to

01:17:23

make a big movie if that cost of making

01:17:25

a 10 million movie goes to ten thousand

01:17:27

or a thousand dollars of compute time

01:17:28

anyway one sitting in their studio in a

01:17:31

basement can start to make a movie and

01:17:33

it really changes the landscape for all

01:17:36

media not just movies music video games

01:17:39

and ultimately the consumers themselves

01:17:41

can create stuff for their own enjoyment

01:17:43

and maybe the best of those products

01:17:45

wins but but having each individual

01:17:47

here's what's going to happen this

01:17:49

become a lot more like the music

01:17:50

industry I mean remember anyone can

01:17:52

really create a song now and they do and

01:17:54

people do go viral on Twitter upload it

01:17:56

on distro kit and it's on Spotify did

01:17:59

you guys see this article in the New

01:18:00

York Times that was kind of throwing

01:18:02

some shade at the CEO of Goldman Sachs

01:18:05

David Solomon yes

01:18:08

look great in the picture but I didn't

01:18:09

read it because I figured it's hate they

01:18:11

were talking about his

01:18:12

his side gig as being a DJ but

01:18:15

specifically they called out a potential

01:18:17

conflict of interest and it's related to

01:18:19

this because I guess he had gotten a

01:18:22

license to a Whitney Houston song

01:18:24

and he remixed it and released it on

01:18:27

Spotify

01:18:28

and her biopic is about to come out and

01:18:31

they thought that there could be this

01:18:33

perceived conflict because Goldman works

01:18:35

on behalf of the publishing company and

01:18:37

my thought was along the lines of what

01:18:39

you guys said like why is this a story

01:18:41

meaning David Solomon should be able to

01:18:43

go to any website license the song Make

01:18:46

it and then submit it back to them for

01:18:48

them to approve because the quote in

01:18:49

here that that matters is the company

01:18:52

that licensed it said we are in the

01:18:54

business of making sure this body of

01:18:56

Music stays relevant so obviously you

01:18:58

want Whitney Houston songs Michael

01:19:00

Jackson songs You Want John Wayne you

01:19:03

want these people to live on in culture

01:19:04

because it's part of our culture look at

01:19:07

this subhead David Solomon brushes off

01:19:09

DJing the better way to maximize it will

01:19:12

be like this go and use it create a

01:19:13

derivative work let us see it if we like

01:19:16

it so like Guetta should be able to just

01:19:18

give that back to Eminem If Eminem's

01:19:19

cool with it he should be able to ship

01:19:21

it and just be done what a non-story

01:19:22

like the New York Times is just so

01:19:24

anti-billionaire so David Solomon

01:19:26

brushes off DJing as a minor hot hobby

01:19:29

that has little to do with his work at

01:19:31

the bank but his activities May pose

01:19:33

protect

01:19:35

it's like what are they writing about is

01:19:37

there not something more important than

01:19:38

this do we think that David Solomon or

01:19:40

any CEO of any major Bank on Wall Street

01:19:42

would put their job at risk running one

01:19:45

of the 20 or 30 most important

01:19:47

institutions in the financial

01:19:48

architecture of the world to license a

01:19:50

Whitney Houston song that they can play

01:19:52

at Coachella I mean does that come pass

01:19:55

the smell test no no and and just for

01:19:58

the record

01:20:00

iconoclastic David Solomon you're going

01:20:02

to be doing the opening night DJ set for

01:20:04

all in Summit 2023. the Griff the grift

01:20:07

is on let's get him

01:20:10

these bombs hates them you got anybody

01:20:11

that New York Times hates on you got a

01:20:13

slot you got a slot that's it that'll be

01:20:15

our lens here Dave Chappelle Dave

01:20:17

Chappelle for sure one thousand percent

01:20:19

I mean he looks good he looks like he's

01:20:20

living his best life it's a finance news

01:20:24

somehow it just like offends them that a

01:20:26

corporate CEO could well that anybody's

01:20:29

happy birthday

01:20:32

I think it's kind of cool this guy Punk

01:20:34

PJs yeah yeah [ __ ] yeah let me do his

01:20:36

thing I've always wanted to be a DJ yeah

01:20:39

oh yeah

01:20:40

spin the one and twos

01:20:45

after party and we'll use AI to help you

01:20:47

out yeah when I should have been

01:20:49

learning how to DJ I was playing the

01:20:51

violin

01:20:52

you should get up there and earlier

01:20:54

violin under a DJ said I played violent

01:20:57

for 10 years we can play a duet together

01:21:01

I played in an orchestra oh my God I

01:21:03

never knew this never again this is all

01:21:06

kind of Revelations Happening Here by

01:21:08

the way here is my most uh pull up that

01:21:10

first chat GPT I gave you this was an

01:21:12

aha moment me and sax were doing our

01:21:15

weekly Mastermind group sex and I get

01:21:17

together we kind of like co-mentor each

01:21:19

other here was uh something that blew us

01:21:21

away

01:21:22

um during our Mastermind group this was

01:21:23

a chat GPT I did this one because I was

01:21:25

trying to figure this out how do I make

01:21:27

my spouse and kids feel heard and then

01:21:29

judge if you did give us a great one

01:21:30

give them your full attention number two

01:21:32

empathize number three validate their

01:21:33

feelings there you have it Zacks just

01:21:36

put that into yourself let me use this

01:21:37

as an example what public so so that's

01:21:40

obviously the aggregation and synthesis

01:21:41

of lots of different self-help websites

01:21:43

how do you describe attribution of that

01:21:46

answer to a particular content publisher

01:21:48

honestly serious question fair use

01:21:50

derivative work I think it is

01:21:55

an opportunity there there's no

01:21:57

monetization yeah you say that because

01:21:59

you hate content provides but here I

01:22:00

tell you this is publishing I think the

01:22:03

GPT the chat on this one no I they know

01:22:08

who they got it from

01:22:10

consists of lots of websites it's a very

01:22:12

smart synthesis

01:22:14

it's not pulling a result from someone's

01:22:16

website it's like read hundreds of

01:22:18

websites and it's like average them

01:22:19

bulldogs for the average it's that's

01:22:22

complete [ __ ] they could say as we

01:22:24

ingest this stuff

01:22:25

this is how it's being done and then you

01:22:28

could publish it when you publish it you

01:22:29

say hey where do we start is that what

01:22:32

you're saying no I'm sorry it's not

01:22:33

[ __ ] that is what you said is

01:22:34

exactly how it's being right and and

01:22:36

Jake out if let's say that the AI is

01:22:39

using 100 different websites and

01:22:41

synthesizing 100 websites what's the

01:22:43

incentive for the marginal 100th website

01:22:45

to say well opt me out unless you pay me

01:22:47

totally right

01:22:50

open AI will just be like okay finals

01:22:53

we'll just work with the other night and

01:22:54

this is why content providers is my best

01:22:56

piece of advice you ask a question I'll

01:22:57

give you the answer content providers as

01:22:59

a group need to get together and fight

01:23:01

for their rights New York Times Madison

01:23:04

right to party no fight for the right to

01:23:06

get paid and to survive

01:23:08

and say as a group either give us these

01:23:12

terms or don't index us all different

01:23:15

content creators absolutely you unionize

01:23:17

all of them they should be a united

01:23:18

front like the music industry why do you

01:23:20

think the music industry gets paid by

01:23:21

Peloton anybody because because there's

01:23:24

five of them get together and they fight

01:23:26

for them and you could organize but

01:23:27

there's millions and millions of

01:23:29

Publishers I think the point here is

01:23:31

that technology is fundamentally

01:23:32

deflationary here's the next great

01:23:35

example correct where the minute you

01:23:37

make something incredible costs go down

01:23:39

but also frankly revenue and profit

01:23:41

dollars go down in the aggregate doesn't

01:23:43

mean that one company can't husband a

01:23:45

lot of it and do incredibly well like

01:23:47

Google has done but it's just going to

01:23:49

fundamentally put pressure on all these

01:23:50

business models which is why I think

01:23:52

it's important Google

01:23:55

should go and they should cannibalize

01:23:58

their own business before it is

01:23:59

cannibalized for them freebar final word

01:24:02

here's another way to think about it I

01:24:03

think that if this goes as we all

01:24:05

predict and everyone's saying it's gonna

01:24:07

go it is more likely than not that many

01:24:09

of these quote content Publishers that

01:24:11

aren't adding very much marginal value

01:24:13

are going to go away then you could see

01:24:15

the number of content sites offering

01:24:16

self-help advice and how to do this and

01:24:18

how to do that 95 of them go away

01:24:21

because all of that work gets aggregated

01:24:23

and synthesized and presented in a

01:24:25

really simple easy user interface that

01:24:27

makes them completely oblivious and I'm

01:24:29

not discrediting the value that many

01:24:30

content Publishers provide but the um

01:24:33

you know the requisite at that point to

01:24:36

be valued as a quote novel content

01:24:38

producer is going to go way up like the

01:24:41

offset to that though is it's so much

01:24:43

easier to create content because of the

01:24:44

AI we have this company copy AI where

01:24:48

even before this chatgpt stuff you would

01:24:50

just go there and say I want to write a

01:24:51

blog about X Y and Z you just give it a

01:24:54

title and it spits out a post and then

01:24:57

they'll actually give you 10 different

01:24:59

blog posts and then you just select the

01:25:01

one that is the direction you want to go

01:25:03

and you keep doing you know human

01:25:04

selection

01:25:05

how does new intelligence get put back

01:25:08

into the system that's based existing

01:25:10

purpose system no you have like a

01:25:11

corporate blog so you publish it to your

01:25:13

corporate my point is if there is now

01:25:15

some new information in the world who is

01:25:17

going to add that to the Corpus if

01:25:19

everybody is just stealing content and

01:25:21

rewriting

01:25:24

its have always had a desire to create

01:25:26

most people create for free there's a

01:25:28

head of the long tail that actually gets

01:25:29

compensated the rest of the long tails

01:25:31

traditionally gotten nothing and they do

01:25:33

it because they want a free day it's

01:25:34

kind of like now the creation is gonna

01:25:37

explode because it's so easy Beethoven

01:25:38

listened to Haydn and then Beethoven

01:25:40

novel Symphonies and his Symphonies were

01:25:42

incredible and he built on the

01:25:44

experience of listening to Haydn the

01:25:46

same is true of how content is going to

01:25:47

evolve and it's going to evolve in a

01:25:49

faster way because of AI and this

01:25:51

content is not just being retrieved and

01:25:53

reproduced it's being you know

01:25:55

synthesized and aggregated and

01:25:57

represented in a novel way

01:25:59

I want to answer that please chat if

01:26:02

chat TPT takes a Yelp review and a you

01:26:05

know a Conde Nast Traveler review and

01:26:07

they represent it based on the best

01:26:10

content that's out there that they've

01:26:12

already ranked because they have that

01:26:13

algorithm with pagerank or Bing's

01:26:14

ranking engine and then they republish

01:26:16

it and then that jeopardizes those

01:26:18

businesses that is profoundly unfair and

01:26:20

not what we want for society and they

01:26:22

are interfering with their ability to

01:26:24

leverage their own content is profoundly

01:26:26

unfair and those magazines and

01:26:29

newspapers

01:26:31

it's possible YouTube is a great example

01:26:34

YouTube was going to get shut down

01:26:36

Sequoia and the YouTube founder sold it

01:26:39

to Google because they were so scared of

01:26:41

the Viacom lawsuit and how well it was

01:26:44

working against them they thought this

01:26:46

is this business will never fly if we

01:26:48

don't have a big partner like YouTube

01:26:50

like Google to support the lawsuit they

01:26:53

they won the lawsuit or they settled it

01:26:55

because they were able to do content ID

01:26:57

and allow content creators the only

01:27:00

reason YouTube exists is hold on let me

01:27:01

finish it's because they let content

01:27:03

creators Watermark and find their stolen

01:27:06

content and then claim it and when they

01:27:08

claim the stolen content they were able

01:27:10

to monetize it that's what's going to

01:27:12

happen here there'll be a settlement

01:27:13

where they are going to be able to claim

01:27:16

their content I will bet any amount

01:27:23

propose a bet

01:27:25

you've seen these AIS that generate

01:27:27

images right like stable diffusion and

01:27:29

like Wally or whatever you literally

01:27:31

just tell it I want this image in this

01:27:33

style and boom it's done and it would

01:27:36

take an artist weeks to produce that and

01:27:38

you can do it in five seconds and you

01:27:40

can tell the AI give me 20 of those and

01:27:42

then you just keep iterating and in five

01:27:44

minutes you've got something

01:27:45

mind-blowing so the fact that it's so

01:27:48

much easier to create content you can do

01:27:49

the same thing with the written word the

01:27:50

people who need to be compensated jcal

01:27:52

if they don't get what they want they

01:27:53

may just go away but they'll be 10 times

01:27:55

or 100 times more people here's how you

01:27:58

are thank you for bringing up this

01:27:59

example so that I can prove how wrong

01:28:01

you are Getty Images is suing stable

01:28:04

diffusion at the moment here is what the

01:28:06

[ __ ] at stable diffusion did they

01:28:08

train their AI on Getty Images with the

01:28:11

watermarks on them and they've been

01:28:12

busted and they are dead to rights now

01:28:14

and they're going to pay a hundred

01:28:16

million dollars or more to Getty Images

01:28:19

for stealing their content and allowing

01:28:22

it to be republished in a commercial

01:28:23

setting those images don't look too good

01:28:25

to me

01:28:29

stable diffusion

01:28:30

copied the Getty image

01:28:34

work okay I think I think stable

01:28:36

diffusion is bigger problem is they

01:28:38

can't do noses and ears and eyelids uh

01:28:43

that looks like a bigger problem anyway

01:28:44

shout out to stable diffusion for

01:28:46

stealing Getty image content so funny

01:28:50

Mozart influenced by hide it not

01:28:52

Beethoven Sorry by the way there's a

01:28:54

real nobody cares interesting topic

01:28:55

about AI that we don't have time to get

01:28:57

to this week but I think we should put

01:28:59

it on the docket for next week which is

01:29:00

should AIS be trained to lie super

01:29:03

important because what's happening right

01:29:05

now yeah

01:29:08

the last thing I'll say on this from my

01:29:11

perspective maybe we can jump on after

01:29:12

this is this is the best thing that

01:29:14

could happen for all of the monopolists

01:29:17

in technology because

01:29:19

Microsoft taking five or six hundred

01:29:21

basis points of share is the best way to

01:29:22

ensure that the FTC has zero credibility

01:29:25

in going afterwards or anybody else in

01:29:27

Texas right those those all of those

01:29:29

things I think are DOA so in some ways

01:29:31

actually Google leaking five or six

01:29:33

percent of the market share is a really

01:29:35

good thing because the FTC is rendered

01:29:37

Toothless in making any certainly

01:29:39

understand that that's such a good point

01:29:40

I mean it's it's kind of a good news bad

01:29:43

news scenario with this whole thing the

01:29:45

good news is that the Google monopoly's

01:29:46

finally been cracked the bad news is

01:29:48

that it's Microsoft and even bigger

01:29:50

Monopoly that's the one that's done it

01:29:52

but it just shows like how vulnerable

01:29:54

all these big tech companies are TBD and

01:29:57

they may all end up competing with each

01:29:58

other

01:29:59

everyone everyone's got a tactical

01:30:01

Nuclear weap now and we don't know where

01:30:03

it's going to get pointed and who's

01:30:04

going to set it off and where and like

01:30:05

the Weaponry has completely changed yeah

01:30:07

the Weaponry totally changed and to

01:30:10

prove how wrong you guys are here is the

01:30:12

Verge

01:30:14

here's the other lawsuit open source

01:30:20

I've been tracking this you people

01:30:22

haven't you guys need to watch what's

01:30:25

happening right now co-pilot GitHub chat

01:30:28

GPT and Microsoft are being sued by

01:30:30

developers because co-pilot was built

01:30:33

off of stolen content these lawsuits are

01:30:36

just beginning and

01:30:38

they'll make it awesome licensing fees

01:30:40

this will be a transitory effect and it

01:30:42

won't it won't change the Dynamics of

01:30:44

where this is going over the long time

01:30:45

change the Dynamics of YouTube in the

01:30:46

long term so let's keep going I don't

01:30:48

know they're doing pretty good we have a

01:30:50

portfolio company called Source graph

01:30:51

which is building and um why don't you

01:30:53

just put your logo page up if you're

01:30:54

going to go through the whole point no

01:30:56

they're I mean they're building

01:30:57

something similar but it's it's opt-in

01:30:59

you just opt into the you know you just

01:31:01

get all your customers to opt into it

01:31:02

okay

01:31:03

let's move on you want to move on you

01:31:06

want to do the Chromebook

01:31:10

would be a great chat let's talk about

01:31:12

it next week though I think yeah I think

01:31:13

there's more time I don't know about you

01:31:16

guys but I found it one of the more

01:31:18

profoundly disappointed saddening states

01:31:20

of the Union I've ever seen

01:31:22

why I think it was you know we often

01:31:26

kind of focus on the one-year cycle of

01:31:27

what the state of the union says but I

01:31:29

think what's more important is how much

01:31:30

the data that's coming through in the

01:31:32

State of the Union supports the more

01:31:34

scary long-term cycle I've talked about

01:31:36

this a lot on how scared I am about uh

01:31:39

kind of where uh where we're headed with

01:31:41

respect to the the US's ability

01:31:44

to fund its financial obligations and

01:31:47

the the scary moment at the State of the

01:31:49

Union besides Biden's inability to kind

01:31:51

of articulate much very well which was

01:31:54

honestly a really

01:31:55

discouraging sight to see was

01:31:58

you know when he talked about what you

01:32:00

know the Republicans are trying to cut

01:32:01

Social Security and Medicare

01:32:03

the U.S treasury put out a projection uh

01:32:06

which I tweeted last week

01:32:09

uh originally shared on Twitter by Lynn

01:32:11

Alden this is the U.S treasury's

01:32:12

forecast of uh debt held by the United

01:32:15

States over time and the assumptions in

01:32:17

this forecast are we've got a certain

01:32:19

amount of debt today and we're running

01:32:21

Social Security and Medicare forward

01:32:23

without cuts and so what happens as we

01:32:26

make these Social Security Medicare

01:32:27

payments and we accrue and pay interest

01:32:29

on the debt that we hold today and we

01:32:32

don't change the tax rates in this

01:32:34

country and this is what happens so it's

01:32:36

a runaway kind of debt scenario in the

01:32:39

U.S by definition has to default at some

01:32:41

point because you cannot tax every

01:32:43

dollar of the economy at 100 at some

01:32:46

point

01:32:47

and so you know there are two ways this

01:32:49

can go the first way is you have to cut

01:32:52

back on these major kind of you know

01:32:54

expense commitments that naturally

01:32:56

balloon over time and that is Social

01:32:58

Security and Medicare and the other one

01:33:00

is that you just tax a lot more and when

01:33:03

you tax a lot more economic growth gets

01:33:05

affected and it makes it really hard to

01:33:07

eventually pay off that debt and the

01:33:08

debt continues to spiral so

01:33:11

I think what we saw was number one

01:33:14

the announcement by Biden hey

01:33:16

Republicans are the ones who want to cut

01:33:18

Social Security Medicare and they all

01:33:19

screamed and they said no way no way

01:33:21

we'll never do that and a lot of them

01:33:22

did interviews afterwards and said it's

01:33:24

total BS and Biden would say that which

01:33:26

I think supports what the polls have

01:33:28

shown which is on both sides of the

01:33:29

aisle

01:33:30

people do not want to see Social

01:33:32

Security and Medicare cut in any way

01:33:33

right now that means you can't and you

01:33:35

guys saw what happened in France where

01:33:37

they pushed back the retirement age by

01:33:38

two years and there was effectively

01:33:40

riots across the country I don't know if

01:33:42

you guys saw this a few weeks ago we

01:33:43

didn't talk about it but it was pretty

01:33:44

brutal pretty ugly

01:33:46

and so this is a real cost that's coming

01:33:48

bare it's coming bear in the United

01:33:50

States not just with the publicly funded

01:33:52

Social Security and Medicare programs

01:33:53

but also with a lot of the private

01:33:55

pensions that are going to need to get

01:33:56

bailed out with the same Federal money

01:33:57

because they're not going to let those

01:33:58

things go bankrupt and that's another

01:34:00

trillion plus of liabilities so you know

01:34:03

that that cost is going to balloon and

01:34:05

the only solution at that point is to

01:34:07

introduce massive tactics and so they

01:34:09

proposed this billionaire tax this tax

01:34:10

on unrecognized capital gains it is

01:34:12

literally if you keep Social Security

01:34:13

Medicare where they are and you don't

01:34:15

pay down the debt and you don't grow the

01:34:17

economy fast enough you have to

01:34:18

introduce significant tax hikes across

01:34:21

corporate

01:34:22

um uh and um the individual taxpayer

01:34:24

base

01:34:25

um and so you know it really again if

01:34:27

you zoom out it really indicates this

01:34:29

steepening curve that the U.S has to

01:34:31

climb its way out of and as you tax more

01:34:33

there's less to invest in the economic

01:34:35

growth the government is a far worse

01:34:37

investor in economic growth than the

01:34:39

free market and that means that we can't

01:34:41

grow our way out and grow GDP enough to

01:34:43

ultimately cover our debt obligations

01:34:45

and this is is what dalio dalio's book

01:34:48

that I mentioned in 2021 was so kind of

01:34:50

importantly sharing

01:34:52

this is a multi-hundred year cycle and

01:34:55

the last couple decades get really nasty

01:34:57

and this chart which is a forecast from

01:34:59

the actual U.S treasury highlights the

01:35:01

problem and the comments made at the

01:35:03

Congress that's in front of the Congress

01:35:04

this week by the president of the United

01:35:05

States indicates how serious of a

01:35:07

problem this is going to be because no

01:35:09

one wants to cut these major cost

01:35:11

obligations that we have coming to you

01:35:15

aren't you and the Republicans you want

01:35:17

to cut Medicare and get rid of it is

01:35:19

that what Biden said you guys want to

01:35:20

get rid of it what was that kerfuffle

01:35:22

about with your uh who's the person on

01:35:25

your squad who's yelling and screaming

01:35:26

out at the president of the United

01:35:27

States no that doesn't matter that was

01:35:29

just sort of but who is

01:35:33

is that Biden was basically trying to

01:35:37

take a page out of Bill Clinton's

01:35:38

Playbook when Bill Clinton lost the

01:35:41

midterms in 94 he basically triangulated

01:35:44

to the center and he did two things he

01:35:46

started going for kind of um small ball

01:35:49

he started playing small ball politics

01:35:51

it was like school uniforms and things

01:35:53

like that that were relatively

01:35:55

unobjectionable and that regular middle

01:35:57

class people could get behind and then

01:35:59

he basically posed as the defender of

01:36:01

entitlement programs back then he um in

01:36:04

96 he ran against Dole by portraying

01:36:06

Dole he went all the way back to Dole's

01:36:08

vote against Medicare this is what the

01:36:09

Biden team is teeing up for the re-elect

01:36:13

in 24 is they're talking about things

01:36:16

like curbing Ticketmaster fees and

01:36:19

fixing right turn red lights I mean

01:36:23

seriously like total small ball okay

01:36:25

they're gonna try and pretend like he

01:36:27

wasn't the most radical tax and spend

01:36:29

Progressive over the last two years

01:36:31

we've really ever had in American

01:36:33

history they're going to try and make

01:36:34

everyone forget that and just talk about

01:36:36

the small objectable stuff and then he's

01:36:38

also going to again pose as the stalwart

01:36:40

defender of entitlement programs which

01:36:42

are very popular and it's and and partly

01:36:45

they're doing this um they're ready I

01:36:47

think getting ready for DeSantis on this

01:36:49

because if you read

01:36:50

so in the political analysis on this and

01:36:53

uh Josh Barrow had a good column and

01:36:55

Andrew Sullivan had a good column

01:36:56

talking about this that way back when

01:36:58

DeSantis was in Congress and he was like

01:37:01

a backbencher he voted for you know some

01:37:04

Republican budgets a Paul Ryan budget

01:37:06

that had some of this entitlement reform

01:37:08

in it so they're going to try and

01:37:10

portray him as against Atomic reform now

01:37:12

I don't think it's going to work because

01:37:13

all he had to say is like listen that

01:37:15

was a long time ago I wasn't voting for

01:37:17

cutting entitlements I just voted for my

01:37:19

party's budget

01:37:21

um that's irrelevant I can tell you I

01:37:23

will not cut entitlements so any smart

01:37:25

Republican is going to take entitlement

01:37:27

reform off the table because it is a

01:37:29

total third rail and they will lose and

01:37:32

I think Trump had the right instincts on

01:37:34

this and I'm sure that any major

01:37:36

Republican will have the right instincts

01:37:38

on this you see them why they're

01:37:40

throwing Rick Scott under the bus Rick

01:37:42

Scott had this proposal about

01:37:45

having

01:37:47

entitlements go from being sort of

01:37:51

permanently entitled to being something

01:37:52

that gets voted on every year and the

01:37:55

rest of the Republican caucus is like

01:37:56

hell no we're not touching that and they

01:37:59

can't run away fast enough from Rick

01:38:00

Scott so free Burger is right there is

01:38:02

no appetite for entombment reform and I

01:38:04

would tell any Republican if you want to

01:38:07

do entitlement reform it's got to be

01:38:08

bipartisan you've got any thoughts you

01:38:11

got to do what Ronald Reagan did which

01:38:13

is join hands with Tip O'Neill and

01:38:16

Moynihan and you jump off the cliff

01:38:17

together do not stick your neck out on

01:38:19

this turn off any thoughts on the State

01:38:21

of the Union writ large

01:38:24

no

01:38:26

Marjorie Taylor green yelling liar at

01:38:27

the present

01:38:30

um no okay well that's the most boring

01:38:32

statement I mean like both sides uh

01:38:36

engaged in a lot of like weirdness I

01:38:38

mean Biden was like bellowing at various

01:38:41

points in his speech it was quite

01:38:42

bizarre and you're right there were some

01:38:44

Republicans in the audience who were

01:38:45

bringing like jackasses and it's

01:38:48

unfortunate because I think utterly

01:38:50

utterly disappointed I think if the

01:38:51

Republicans had just calmed down I think

01:38:54

Biden's sort of weird mannerisms where

01:38:57

he was like practically yelling it was

01:38:59

like Abe Simpson you know old man

01:39:01

yelling at the cloud and yeah they they

01:39:03

do a good job of like right at the

01:39:05

moment of self-immolation they let him

01:39:07

off the hook totally it's really

01:39:09

incredible Republicans just they have no

01:39:11

impulse control

01:39:12

right when they could just be quiet sit

01:39:15

there calm quiet and let Biden do the

01:39:18

damage to himself they just cannot I

01:39:20

like McCarthy I like McCarthy telling

01:39:22

Margie telegrange and Mary let me tell

01:39:25

him that other glider guy to get the

01:39:27

hell out of here listen it's always been

01:39:29

hard to control backbenchers that's just

01:39:31

the reality that does not speak for the

01:39:32

entire party

01:39:34

how do you uh how do you guys feel about

01:39:36

being taxed on your change in net worth

01:39:40

from year to year

01:39:41

you mean a wealth tax wealth tax yeah

01:39:43

the uh if it's a if it's for

01:39:44

billionaires totally cool if it's for

01:39:46

centimillionaires or deck of

01:39:47

millionaires absolutely off the table

01:39:50

we want those people to grow their

01:39:52

wealth and invest the billionaires yeah

01:39:53

sure is that what you think is going

01:39:55

Freeburg yeah so hard to execute on what

01:39:58

are you supposed to do like I would love

01:40:00

to see a website where you have a piece

01:40:01

of yarn you know those those sites yeah

01:40:03

and then you have like a little uh pin

01:40:05

and you kind of push the pin out and the

01:40:06

other one has to go in you can't have it

01:40:08

all you can't have low taxes and have

01:40:11

these entitlement programs and have this

01:40:13

level of spending it's impossible you

01:40:15

have to tax

01:40:17

that's the only way or you have to cut

01:40:19

the entitlement programs or you have to

01:40:21

cut the spending well you're you're

01:40:22

right about that free bird but let me

01:40:24

tell you why the American people

01:40:26

think it would be ridiculous to cut

01:40:28

their entitlements they've watched as

01:40:30

Washington spent eight trillion dollars

01:40:33

correct on forever wars in the Middle

01:40:35

East they just watched as Biden spent

01:40:37

trillions enriching the Pharma companies

01:40:40

on a fugazi that didn't work

01:40:42

they've just watched as hundreds of

01:40:44

billions went to climate special

01:40:47

interests the Democrats

01:40:50

we talked about that let me think about

01:40:52

it

01:40:55

brilliant hold on they've watched as

01:40:58

trillions of dollars have gone to the

01:40:59

donor class and they are going to rise

01:41:02

up and say the hell with you if you cut

01:41:05

off our social security that we paid

01:41:07

into for decades while enriching all

01:41:09

these special interests yes

01:41:10

I don't advocate for any of these points

01:41:13

and the bailouts of corporations in the

01:41:15

great financial crisis yeah right the

01:41:16

cons I think

01:41:18

I think I think it's just an analytical

01:41:20

certainty that you know you gotta zoom

01:41:23

out and stop thinking about the yearly

01:41:24

cycle and the election cycle on this

01:41:26

stuff and just look at where we're

01:41:27

headed over a multi-decade cycle okay

01:41:29

and there's just no resolution to this

01:41:31

problem what happens the way we're all

01:41:33

oriented right now huh if there is a

01:41:35

wealth tax let's just say on billions

01:41:38

let me just say one thing sorry Jacob

01:41:40

let me just say what do they do they

01:41:41

leave the country well that did happen

01:41:43

in France by the way I think 40 of the

01:41:45

people that were taxed left and then

01:41:47

they came back they would violate our

01:41:49

constitution

01:41:51

that's that'll be it'll be um litigated

01:41:53

for sure yeah

01:41:54

if you can get the cost of energy in

01:41:57

this country

01:41:58

to drop by 50 to 75 and you can increase

01:42:02

energy

01:42:03

capacity by 10 to 20 fold then you have

01:42:06

a Fighting Chance because you can

01:42:08

actually grow the economy out of the

01:42:09

problem and that's really where I am

01:42:11

optimistic and excited about

01:42:12

opportunities like fusion and you guys

01:42:14

can make fun of me all you want but if

01:42:16

we can get to a point where we can

01:42:17

increment energy capacity by an order of

01:42:20

magnitude there is economic growth that

01:42:22

will arise from that from all these new

01:42:24

Industries and these production systems

01:42:25

and that's how we can grow our way out

01:42:27

of the debt entitlement tax problem

01:42:30

where one of those three things has to

01:42:31

give in the absence of that

01:42:34

you know what the cheapest source of

01:42:36

energy is today

01:42:39

Fusion three cents a kilowatt hour

01:42:41

Vision you mean no it's Fusion

01:42:46

it's the sun using solar panels there it

01:42:48

is yeah the problem tomorrow honestly

01:42:50

let me just say this one thing the

01:42:51

problem is can you scale energy capacity

01:42:54

by 10x and can you do it fast enough and

01:42:56

that's the real testicle while creating

01:42:59

jobs that's the real techno economic

01:43:01

question right like I posted a link in

01:43:03

my reading list today this week you guys

01:43:04

can go look at it the most prolific

01:43:06

distribution of fusion technology is

01:43:09

China actually deploying solar on every

01:43:11

single Rooftop in China the United

01:43:14

States could do it too and you will 10x

01:43:16

the power available well yeah I mean

01:43:18

they are entering it'll be zero they're

01:43:21

increasing their I mean it'll happen in

01:43:23

the next years it's just sometimes we

01:43:25

want to create intellectual complexity I

01:43:27

I love these different forms of fusion I

01:43:30

just think it's a 50-year traj to get it

01:43:32

because even I'm not betting I'm not

01:43:34

betting on it I'm just saying it's there

01:43:35

is a there's a I was agreeing with you

01:43:37

I'm just building on your point to say

01:43:39

it's actually happening Fusion is what

01:43:41

is actually creating abundant zero cost

01:43:43

energy today

01:43:44

yeah and so look if we can increase

01:43:46

energy capacity in this country by 10x

01:43:49

energy production capacity by 10x and we

01:43:51

can do it in the next 20 to 30 years

01:43:54

on every rooftop if we can we have a

01:43:57

path out of the entitlement tax debt

01:44:01

problem otherwise one of those three

01:44:03

things is going to give and it's going

01:44:04

to be ugly the thing that we are lacking

01:44:06

right now is not actually the generation

01:44:08

capability

01:44:09

which is incredibly cheap but it's

01:44:11

really scalable storage and once we

01:44:14

figure that out which is actually the

01:44:15

real technical bottleneck to Abundant

01:44:17

zero cost energy we'll have your

01:44:19

boundary condition met and we'll have it

01:44:21

well before different forms of fusion

01:44:23

are commercializable

01:44:25

France had an exodus of an estimated 42

01:44:30

000 millionaires between 2000 and 2012.

01:44:33

uh and really the and before yeah for 12

01:44:37

years and then they were uh reversed

01:44:39

they were just losing the tax base so

01:44:40

violently they had no choice what's

01:44:43

going to happen in California

01:44:44

it's happening in California already

01:44:46

with San Francisco right it's happening

01:44:49

already yeah they're stupidly telling

01:44:51

people that this wealth tax is going to

01:44:53

have a 10-year look forward so everybody

01:44:55

I know is at least talking about hey

01:44:57

could this happen in the next 10 years

01:44:58

because if it might happen five years

01:45:01

from now we got to leave now because if

01:45:03

I wait they're going to chase you

01:45:05

they're going to chase me yeah

01:45:06

specifically you

01:45:09

they hate you no I mean it's it's uh

01:45:12

it's a problem it is a serious let me

01:45:14

tell you I mean it's a freeberg's point

01:45:15

uh it's not just Ray dalio uh the great

01:45:17

uh political satirist Peter O'Rourke who

01:45:19

died last year he wrote that American

01:45:22

politics is defined by the formula x

01:45:25

minus y equals a big stink where X is

01:45:29

what people want from government and Y

01:45:32

equals what they're willing to pay for

01:45:34

government and that difference is

01:45:36

basically the big stink and the Java

01:45:37

politicians is to manage that stink and

01:45:40

the problem is the politicians have not

01:45:42

been doing a good job managing it and

01:45:44

they increasingly do a worse job

01:45:45

managing it so yeah at some point it's

01:45:48

going to blow up yeah you guys want to

01:45:49

hear a crazy statistic I was pulling

01:45:51

this out the other day you know what the

01:45:53

budget per capita is of the city of San

01:45:55

Francisco so how much the city spends

01:45:58

per year divided by the number of people

01:45:59

that live in the city well we know it's

01:46:00

a couple of billion dollar budget we

01:46:02

know only a couple of hundred thousand

01:46:03

billion dollar budget yeah and there's

01:46:05

800 000 residents it's eighteen thousand

01:46:07

dollars

01:46:08

per citizen per year that's how much the

01:46:10

city of San Francisco spends a third of

01:46:13

that of money by the way 30 of it goes

01:46:15

to or 25 goes to Public Health Care Now

01:46:18

when you look at that eighteen thousand

01:46:19

dollar uh budget per capita it is more

01:46:22

than every single state of the union on

01:46:25

a per capita basis except Oregon and

01:46:27

North Dakota which have very weird uh

01:46:29

budget so San Francisco spends more than

01:46:33

every other State per capita it spends

01:46:35

more in aggregate than 16 U.S states and

01:46:38

the federal budget per capita is fifteen

01:46:41

thousand dollars the federal

01:46:42

government's budget per capita is

01:46:43

fifteen thousand dollars so San

01:46:45

Francisco spends more than 15 states and

01:46:47

spends more per capita than every other

01:46:49

state except two and spends more than

01:46:50

the federal government per capita and I

01:46:53

think that really highlights and and you

01:46:55

need to tax the base to do that and now

01:46:57

we are seeing San Francisco is the

01:46:59

largest population of Exodus and

01:47:01

population access of any City and

01:47:03

business Exodus of any city in the

01:47:04

United States that is your predictor

01:47:06

that is your predictor of where this

01:47:08

goes right in Miami Florida peaked by

01:47:11

contrast has no state income tax they

01:47:14

just rely on property taxes sales tax

01:47:16

which California has as well there is a

01:47:18

a income tax and cap gains tax of zero

01:47:21

in Florida and they seem to make it work

01:47:24

880

01:47:26

000 people uh was the peak in San

01:47:28

Francisco 2018 2019 2020 and then in uh

01:47:32

2021 815 so

01:47:35

it's gone even further some people think

01:47:38

it's down to 650 now I think it is

01:47:40

pretty hard to track but yeah well you

01:47:42

have a lot of people who owned homes

01:47:44

there and maybe own second homes because

01:47:45

let's face it it was a well-heeled group

01:47:47

of individuals living there and a lot of

01:47:49

them just still have their places but

01:47:51

they've left and then they're in the

01:47:52

process of selling the point the point

01:47:54

is increasing the tax rate is a great

01:47:57

short-term solution but over the long

01:47:59

term if that budget per citizen isn't

01:48:02

brought in line there is no way to tax

01:48:04

the base enough without causing the tax

01:48:06

base to leave forget about what my

01:48:08

anyone's personal opinions are that's

01:48:09

just the economic reality of what

01:48:11

happened in France it's what's happening

01:48:12

in San Francisco there's a lot of great

01:48:14

predictors in history of where this has

01:48:15

happened

01:48:16

and so something has to give or you have

01:48:19

to have a miracle like an energy Miracle

01:48:21

but you know we'll all keep investing

01:48:23

for that all right everybody this has

01:48:25

been an amazing amazing episode of the

01:48:28

all in podcast 115 episodes for the

01:48:32

dictator Sultan of science

01:48:35

and for the pacifist David the dove

01:48:39

sacks I am the world's greatest model

01:48:41

Creator Jay Cal and we'll see you next

01:48:43

time bye bye love you Besties

01:48:45

we'll let your winners ride

01:48:48

Rain Man

01:48:49

[Music]

01:49:02

besties

01:49:04

[Music]

01:49:10

somehow

01:49:12

[Music]

01:49:34

[Music]

Episode Highlights

Chinese Balloon Controversy
A discussion on the media's overreaction to a Chinese balloon incident, questioning its implications.
“The media got in a lather and a tizzy!”
@ 03m 33s
February 11, 2023
Nord Stream Sabotage Speculation
Debate over the Nord Stream incident raises questions about U.S. involvement and media narratives.
“It's basically perpetrating an act of war!”
@ 10m 24s
February 11, 2023
Germany's Economic Dilemma
Exploring the conflict between German economic interests and foreign policy.
“Their whole economy is based on industry; they're a very industrial power.”
@ 22m 05s
February 11, 2023
Military-Industrial Complex Warnings
Reflecting on Eisenhower's caution against the military-industrial complex.
“He warned us that yes we need the defense industry, but they become a vested interest in favor of foreign interventions.”
@ 27m 29s
February 11, 2023
Microsoft's Bold Move
Microsoft's $10 billion investment in Chat GPT aims to revolutionize search and AI integration.
“Microsoft effectively for 10 billion bought almost 50 percent of a tool.”
@ 42m 47s
February 11, 2023
The Importance of Competitive Mode
Companies must build a competitive mode to defend their market share, especially in AI.
“The most important thing that you can do is to figure out how to defend it.”
@ 56m 41s
February 11, 2023
Transformative AI Tools
AI tools like GitHub Co-Pilot will enable startups to build without needing developers.
“You can build a version of Stripe extremely quickly and in a very lean way.”
@ 01h 08m 29s
February 11, 2023
AI's Impact on Productivity
AI tools are set to revolutionize productivity, allowing fewer engineers to achieve more.
“One human can do the coding that 20 humans would do before.”
@ 01h 10m 10s
February 11, 2023
Democratizing Content Creation
AI is changing how media is produced, allowing anyone to create their own content.
“It's going to change the landscape for all media.”
@ 01h 17m 36s
February 11, 2023
David Solomon's DJing Controversy
The media questions the Goldman Sachs CEO's DJing side gig, raising eyebrows about conflicts of interest.
“What a non-story!”
@ 01h 19m 22s
February 11, 2023
The Cost of Entitlements
There's a growing concern about the sustainability of Social Security and Medicare as costs rise.
“No one wants to cut these major cost obligations.”
@ 01h 35m 11s
February 11, 2023
Energy Solutions for Economic Growth
Increasing energy production capacity could provide a solution to the debt and entitlement crisis.
“If we can increase energy production capacity by 10x, we have a path out.”
@ 01h 43m 51s
February 11, 2023

Episode Quotes

Key Moments

Media Overreaction03:33
Nord Stream Debate10:24
Economic Interests21:45
AI Cost Analysis39:21
Business Model Shift48:46
Media Critique1:19:22
Content Creation Explosion1:25:34
Social Security Debate1:33:30

Words per Minute Over Time

Vibes Breakdown

Podcast thumbnail

01:19:20

E128: Google enters AI wars, Druck’s warning, Trump crushes CNN & more

Podcast thumbnail

01:42:51

E132: SEC goes after crypto giants, Sequoia splits, LIV/PGA, Messi's deal + LIVE Q&A!

Podcast thumbnail

February 02, 2024

01:33:50

E164: Zuck’s Senate apology, Elon's comp package voided, crony capitalism, Reddit IPO, drone attack

Podcast thumbnail

January 13, 2024

01:38:39

E161: US strikes Houthis, market instability, Q1 rate cuts in doubt, Carta's major mishap, DEI

Podcast thumbnail

September 08, 2023

02:08:15

In conversation with Chris Christie

Podcast thumbnail

02:01:47

In conversation with Robert F. Kennedy Jr.

Podcast thumbnail

01:42:28

Tucker Carlson: ICE Raids, LA Riots, Strong Economic Data, Politicized Fed, War with Iran?

Podcast thumbnail

01:20:35

In conversation with President Trump

Podcast thumbnail

September 15, 2025

41:54

Tulsi Gabbard on Russiagate Hoax Evidence and How She’s Reforming Politicized Intelligence Agencies

Podcast thumbnail

October 29, 2022

01:36:01

E102: Elon closes Twitter deal, $META uncertainty, Zuck's historic bet, big tech decline & more

Podcast thumbnail

August 16, 2024

01:46:02

Break up Google, Starbucks CEO out, Kamala’s price controls, Boeing disaster, Kursk offensive

Podcast thumbnail

01:47:41

Trump verdict, COVID Cover-up, Crypto Corner, Salesforce drops 20%, AI correction?