0:00

It's a UNIX

0:04

system. I

0:07

know this. It's

0:12

got all the files

0:14

of the whole park.

0:19

It tells you everything. Sir, he's uploading the virus. Eagle

0:21

One, the package is being delivered. What are you guys going to do for Thanksgiving? I am

0:23

going to be in the hospital. I'm going

0:24

to be in the hospital. I'm going to be in the hospital. I'm going

0:26

to be in the hospital. I'm going to be in the hospital. I'm going to be in

0:28

the hospital. I am already

0:30

at my in-laws. We're

0:33

splitting the cooking up three ways, so

0:35

I'll be doing that. Starting

0:37

as soon as I am done with work

0:39

today. What about you? We've

0:43

got a friend coming over. We

0:46

live pretty far away from... My

0:48

family's in Texas. I'm in South Carolina. Right.

0:51

We make the trip for Christmas,

0:53

but not for Thanksgiving. So we have friends

0:55

come over. Classic. We're

0:59

doing a big meal here. Cool. That's

1:01

great. What about you, Emily? I'm

1:04

making the trek out to Long Island tomorrow

1:06

morning because

1:08

I refuse to take the Long Island Railroad

1:10

the Wednesday night before Thanksgiving.

1:12

I took

1:16

it once when I was going to

1:18

my in-laws for Thanksgiving, and there were

1:21

kids that seemed like they were

1:23

on the precipice of throwing up from drinking

1:25

so much on the train. Hell yeah, there were. I

1:28

was like really close. I could not get farther.

1:31

I was so close to them, and I was like,

1:33

this is just going to end so horribly.

1:35

Look, I've seen it all on the

1:37

Long Island Railroad. It's a real service of magic, whatever that kind of magic

1:39

is, whether

1:43

it be like dark magic or, you know, good magic. Who's

1:46

to say?

1:46

I think that's

1:48

definitely right. Our mileage may vary. Yeah.

1:52

It's funny. I didn't think about the need to

1:54

get really messed up before

1:56

you go home for Thanksgiving when you're like 22.

2:00

Seems like it's a rite

2:02

of passage for certain people heading

2:04

east from Brooklyn. Well, I mean, I

2:06

think it's also like, you know, there's

2:10

the Thanksgiving tradition in many a suburban

2:12

hometown of, you know, going to the local

2:15

bar that you were not old enough to go to when you were in high

2:17

school, then see all the people.

2:19

Not something that I have ever partaken

2:22

in or have any desire to do, but... Oh,

2:25

that I've done. Some of the, like,

2:27

one version of that is also going to the city

2:30

from Long Island, getting super trashed

2:32

at whatever shitty bars that you and your friends would get

2:34

trashed at when you were in high school.

2:37

Part of the thing too is that you can drink

2:39

open container on the Long Island railroad,

2:42

but there are a couple of days every year that you can. That

2:44

is a big part of it, I would say.

2:47

Yeah,

2:47

they used to sell, like, they used to have, like, little, like,

2:49

wine carts on the platform so you could

2:51

get, like, a little can of wine. But

2:54

the big thing is you get beer

2:57

in a giant plastic cup with a straw.

3:00

Like, it's like a giant, like, soda cup that you would get at,

3:02

like, McDonald's or whatever.

3:04

Right. That is, like, a pinnacle. You

3:06

can't do that all the time? No, you

3:09

can't do that on, I think, the

3:11

day before Thanksgiving and

3:13

St. Patrick's Day. Well, there

3:15

we go. I forget exactly what

3:17

the deal

3:19

is, but needless to say, the way

3:22

that you get around that in high school is by buying a Gatorade

3:25

at the local, you know, train station

3:27

bodega and putting whatever amount

3:29

of shitty plastic, you know, jug vodka

3:32

you can find in there. Right.

3:33

Work around. And I'm not even,

3:36

you know, I'm personally not saying

3:37

this from experience, I'm not even thinking about

3:39

that, but I wouldn't get it

3:41

firsthand. Underage drinkers

3:44

are nothing if not innovative. Yeah, really. The

3:47

southern version of that

3:49

is go to Sonic and get a Route 44

3:52

sized drink, just

3:54

like this massive 44-ounce soda,

3:57

and then you pour out some of it and get it. and

4:00

put your preferred spirit

4:03

in there.

4:04

I've heard that Sonic's ice is very good. Oh,

4:06

it's amazing. It's like a very specific ice cube

4:08

shape. Yeah,

4:11

it's like these tiny little

4:14

balls. I don't know how to explain it, but it's

4:16

perfect. And everyone

4:18

should try Sonic once in their life. It's

4:20

one of America's finest foods. Sonic

4:23

popular. You know what, I think I'm

4:25

gonna have Sonic for lunch tomorrow pre-Thanksgiving.

4:28

God bless. It's

4:32

one of the only places here that's open 24-7 and

4:35

like every day, so. I

4:38

feel like it's legendary to people who

4:40

have not had it. I've never had it and

4:42

I've been like, it's on my bucket list.

4:44

I don't know if it's worthy of that, but it definitely

4:47

is. No, it's gonna be like, it'll

4:49

be like when In-N-Out Burger started coming into

4:51

Texas and all the people from

4:53

California loved it. Everyone that

4:55

was here that already had Sonic and Water Burger were

4:57

like, this is not, what is this with

4:59

this California shit? Get this out of here.

5:02

I went to a DQ for the first time

5:04

when I was in Washington State

5:06

over the summer. Surprisingly,

5:09

it was like the DQ in the Walmart parking

5:11

lot in Kurt Cobain's hometown. Hell

5:14

yeah. Which is just like various levels of like

5:16

suburban ennui that I'm like, wow.

5:18

You really understand them now.

5:21

Yes. The lyrics, they make more sense.

5:25

Dairy Queen is legitimate. Let's not

5:27

pretend it's not. It's great.

5:29

I mean, I had some very good chicken fingers.

5:33

Like literally no complaints.

5:35

Ice cream like 75 different ways. There's

5:37

nothing wrong with that. Exactly. So

5:41

we're actually here to talk about some

5:44

of your reporting Maxwell. Can

5:47

you introduce yourself

5:49

and then we'll get into it a little bit.

5:52

Sure, I am Maxwell Stron.

5:55

I am a reporter at Motherboard to cover

5:57

technology focused on sort of financial.

6:00

issues. So

6:02

we're having this this lovely Thanksgiving

6:04

conversation the day before Thanksgiving it is November

6:07

22nd at this moment

6:09

right now at 11 a.m. who

6:12

is in charge of open AI?

6:15

That is a

6:17

very good question and I

6:20

think the fact that I can't give you a

6:22

super hard answer

6:25

is maybe evidence of

6:27

evidence of how chaotic things have been.

6:29

I mean de facto the

6:31

answer is Sam Altman once again.

6:35

He's agreed to come back and join

6:37

the board. I don't know if the paperwork is signed

6:40

so you know it could technically for a

6:42

few more minutes there will be the Twitch CEO

6:44

Emmett Shear who was CEO for

6:47

I don't know a couple hours

6:49

or something like that. So

6:52

but yeah I mean for all intents and purposes

6:55

Sam is back in

6:57

charge after pretty crazy

6:59

five days or so.

7:02

What the fuck is going on? What happened

7:05

and I know or something I've been thinking about

7:07

Emily I know that I have to stop cursing at least in the

7:10

first 10 minutes when

7:12

we move to YouTube. So I'm

7:14

bad at that too so we'll

7:16

have a swear jar, a virtual swear jar. We need

7:18

a cyber swear jar but what is

7:21

what happened five days ago like what is going

7:24

on over there? Sure

7:25

so the long

7:27

and short of it is that

7:31

the board which is

7:33

a very very particularly

7:35

small board that is

7:38

collection a strange collection of people

7:40

to be honest

7:42

was basically at odds with one another

7:44

for a long time over

7:46

some sort of philosophical differences

7:49

seems like maybe some personal differences

7:52

and some

7:55

one faction that's more academic

7:57

more do worry about AI sort

7:59

of formed on

8:01

one side and Sam

8:04

Altman and Greg Brockman who's

8:06

kind of you know his one of his

8:08

allies underlings I

8:10

guess you could say he was the president on

8:13

the other side and the

8:15

board made a move I think there were outside

8:18

of Sam and Greg just four people

8:21

that made this decision only

8:23

four people that caused all of this news

8:26

came together and decided

8:28

that they would push Sam out, demote

8:32

Greg and Greg later

8:34

quit so we can get into

8:36

the reasons of why they decided

8:38

to do that and what the philosophical differences

8:41

were but that's the thing that sort of precipitated

8:44

everything that's happened since then.

8:46

Let's back up a little bit I don't kind

8:48

of jumping around on our map here but I

8:52

want to put OpenAI the company

8:54

into like some sort of corporate context for

8:56

people before we start dissecting them. This

8:59

is kind of like if

9:01

you didn't know much about the company other than

9:04

like its financials it's kind of a fairy

9:06

tale startup right like

9:08

what's their evaluation at like how long have they been around

9:12

how much money are they making?

9:14

Sure

9:16

the I

9:18

mean yes in lots of ways they are a fairy

9:20

tale startup story

9:22

I mean they are valued at something like 80

9:24

billion dollars I think maybe

9:28

looking for had been looking for money that would

9:30

value them at 90 you know they were

9:33

I think valued at 30 billion a year

9:35

ago so that's you know the 3x

9:38

that you're looking for if you're a startup

9:40

in a single year they had this

9:42

charismatic founder Sam Altman

9:44

who's one of the most well-connected people in

9:46

Silicon Valley with the head of Y Combinator

9:49

which is you know really influential

9:52

startup accelerator in the

9:55

in the Bay Area and yeah

9:58

and most importantly They had

10:00

two products that had really,

10:02

I mean, inarguably taken

10:04

the world by storm in ChatGBT,

10:08

which you'd be hard pressed to find someone

10:10

who doesn't know what that is at this point,

10:13

and Dolly, which was the text

10:16

to image generator that took the

10:18

world by storm, I think, before

10:20

ChatGBT. So I mean, all

10:23

of those things combined. You

10:26

have the profile of a generational

10:30

startup, and a lot of people had talked about

10:32

Altman regardless

10:34

of what they thought about him as being

10:37

the Mark Zuckerberg of the 2020s, or the

10:39

Bill Gates, or the Jeff Bezos. I

10:44

mean, that was kind of the sense that people had,

10:46

this person was going to define

10:49

the world of AI and be the public face

10:51

of it in the same way that Zuckerberg

10:54

became the face of social media. Now,

10:58

there are some key differences between

11:00

open AI and your typical startup

11:04

that really paid a key role

11:06

in what we've seen happen over the last

11:08

five days. And

11:11

I can go into that history or not, but

11:14

the long and short of it is that it started as

11:16

a nonprofit. It was not

11:18

actually started as a sort of for-profit,

11:21

typical, you know, raised venture capital,

11:24

prepared an IPO company. It

11:26

was started as a nonprofit organization

11:28

that sort of transitioned and transformed

11:31

in a lot of ways into a more traditional

11:34

powerhouse startup. And that caused

11:36

a lot of the corporate weirdness

11:39

that we've seen

11:41

today.

11:42

Yeah, it's like I feel

11:44

very silly because I keep getting Sam

11:46

Altman and Sam Bankman freed mixed up. And

11:48

it's not just because their first names are both Sam,

11:51

but there's some effective altruism

11:53

kind of running throughout this whole thing. And that's kind

11:55

of part of the reason why

11:58

it seems to me. from what I've been

12:00

reading of your work and other people saying,

12:03

there's a lot of like the philosophical

12:05

differences between the effect

12:07

of altruism and then also effective

12:11

accelerationism.

12:13

Yeah,

12:14

effective altruism just keeps popping

12:16

its head into every major

12:21

catastrophe occurring in the technology

12:23

sector somehow. But

12:25

the way that effective altruism is popping

12:27

its head in here is very different from the way

12:29

that it popped its head into the FTX thing.

12:32

So at FTX, Sam

12:34

Bankman Freed was a hardline

12:36

believer in effective altruism. And

12:39

as part of their philosophy, it was

12:41

make as much money at all costs so

12:44

that you can then donate it or

12:46

use it to save the world

12:50

through X, Y, or Z. And so

12:53

in that philosophical school

12:56

of thought, it didn't

12:58

matter how the money was made. It didn't matter

13:00

if it was made unethically as long

13:02

as it was accumulated and then used to

13:05

save the world. Effective

13:07

altruism has this other aspect, though, which

13:09

is a sort of doomerism related to

13:12

artificial intelligence, a real fear

13:14

that AI could end the

13:18

world, the robots take over, the terminator

13:20

kills everyone, that sort of thing.

13:22

Now,

13:23

Sam Altman is not an effective altruist,

13:27

but two of the members of the board were,

13:31

or at least are heavily affiliated

13:33

with it just to cover our bases. And

13:36

those two people were

13:38

kind of it was surprising that they had come

13:41

to have such influence

13:44

at the company. They had

13:46

no real firm association outside

13:48

of that. One of them was an academic,

13:51

not really in the corporate

13:53

world at all. Maybe that

13:56

as a result of its nonprofit

13:58

origins. became really,

14:00

really worried that Sam Altman,

14:02

Greg, everyone that kind of stood with him, was

14:05

moving too fast, breaking things,

14:07

blah, blah, blah. And so those two

14:09

factions, Sam on one side and this

14:11

woman, Helen Toner on the other side, started

14:14

kind of go at each other. Helen

14:16

even wrote a paper in

14:19

which she criticized OpenAI. It was a very

14:21

academic paper and celebrated

14:24

another AI startup.

14:27

And Sam was like, what the hell? And

14:30

she was like, well, it's an academic paper.

14:32

And he was like, everything you say matters

14:34

to this company. And so they were kind

14:36

of butting heads with each other. And increasingly,

14:39

someone in the middle was the

14:42

company's chief scientist, Eliot Settskever,

14:44

who worked with Sam,

14:47

but also was really a big

14:50

time doomer. And so he was kind

14:52

of found himself in the middle. And he consequently

14:55

also was the person who kind of invited

14:59

Sam onto the Zoom call or the Google Hangout

15:02

and said, we've got some news

15:04

to share with you. And you

15:06

know what happened next. They're not using Microsoft

15:08

Teams to do that. That feels very not

15:11

aligned with the brand. I

15:13

think when the reporting came out, I

15:16

cannot imagine Microsoft was super happy

15:18

about that. But yeah, Google Hangout. I'm

15:20

sorry, does anyone like Teams? No one

15:23

likes Teams. No.

15:24

No one likes Teams. Karen

15:27

uses it all the time. And it does not look good. I'm

15:33

sure even the Microsoft employees are using

15:35

it reluctantly, let's be honest. Microsoft

15:37

is important to this story. This is an aside before we

15:40

dig into some of the other wild things.

15:44

They injected, what, $13 billion? They

15:46

had the $13 billion investment?

15:48

$13 billion. So as I've

15:51

said, I know there's like a million different aspects

15:53

of this story. And that's why everyone's so confused.

15:56

But just to

15:58

try and put the Microsoft thing in context. So,

16:01

OpenAI was a nonprofit.

16:04

Elon Musk was associated. Elon

16:07

Musk decides he's

16:09

bailing.

16:10

They go, oh, we need money.

16:13

It requires a lot of money to build a complex

16:15

AI system. It ends up, of course. And

16:18

so they go, let's kind of reformat

16:20

this, create this for-profit arm. We

16:22

can use it to get investment.

16:24

Microsoft goes,

16:26

great. We would

16:29

love to give you, first, a billion dollars

16:31

and, ultimately, $13 billion. So

16:35

they own, technically, 49% of the company.

16:38

But unlike a typical

16:40

corporate structure,

16:42

an investor that owns 49% of generic Corporation X

16:48

has a lot of sway in the decision-making

16:50

of that company, because the board

16:53

is beholden to the shareholders so

16:55

the shareholders can exert a lot of influence.

16:58

OpenAI actually is not like

17:00

that.

17:01

The board is not,

17:05

its mandate is not to please

17:07

shareholders. It's to make

17:10

sure, effectively, that AI is developed

17:12

in a safe way and doesn't take over

17:14

the world and that Terminator doesn't come to pass.

17:17

And so, as a result, Microsoft

17:18

had invested all this money and

17:23

did not have any legal

17:25

effective power at the company. And

17:27

so when they found out that –

17:30

yeah, go ahead. So I just want to make sure that I understand.

17:33

So there's a nonprofit with

17:36

the board, and that nonprofit

17:38

is in control of the for-profit

17:40

company that Microsoft has invested in. Exactly.

17:45

And that board is very small. And that board is very

17:47

small. And those two entities are separate

17:49

enough that if you are Microsoft

17:52

and you invest the billions into

17:54

the for-profit thing, you don't

17:56

have much control over it.

17:59

Yeah. And I mean –

17:59

And until Friday, there had been

18:02

a pretty good argument that

18:04

they were creating a better version of a corporation,

18:07

one a corporate or better

18:10

version of a business at least,

18:12

one in which we can take all this money from

18:15

Microsoft so that we can build all these

18:17

fancy toys, but we're not

18:19

in the end of the day, just

18:21

like a shareholder capitalist function.

18:24

Well, our shareholder, we're

18:26

not just behold it to shareholders. We

18:29

in fact have a higher mission. We're going

18:31

to create a responsible AI,

18:33

a better AI. And as a result, we've

18:35

created a better system that looked great

18:38

until Friday, when obviously

18:41

shit hit the fan. It's funny

18:44

as when all of this originally went down, I

18:47

joked to myself in my own head, because I remember

18:50

maybe like a week ago, a

18:52

hot Sam Altman quote dropped where he said

18:55

that, AGI is

18:57

going to be a magic intelligence in the sky.

19:00

I was like, oh, it sounds like they're trying to build God or something.

19:03

I wonder if that's why they fired him.

19:05

It sounds like something

19:08

like that may actually be at the root

19:10

of this.

19:12

I think the root

19:14

conflict is very similar to that.

19:18

Yes.

19:19

Sam, well, one way to think about it would

19:21

be Sam Altman

19:24

and I'm generalizing here, but

19:26

if you'll let me, Sam Altman is

19:28

the Silicon Valley move fast and break

19:30

things guy at the company. And

19:32

then you have the more conservative

19:35

members of the board and this guy, Elliot

19:38

Sutskever, who was the chief scientist,

19:40

and they are the, we could destroy

19:43

the world. Everything could end

19:45

as a result of what we are building right

19:47

now. And so those two factions

19:50

were constantly kind of at each other's

19:52

throat. I don't know which

19:54

is right, but I could say

19:56

for sure that that was creating a lot of conflict

19:59

at the company.

19:59

companies the highest levels.

20:02

What? I okay.

20:05

Just the whole thing is so strange.

20:07

It's just like that was that was really what I learned

20:10

watching all because I was

20:13

not paying attention to the corporate structure of open

20:15

AI. You know, I was interested in chat GPT

20:18

and AI and their effects on society

20:20

and my job, obviously, but I

20:22

didn't realize how

20:24

weird it was at the top

20:27

of this thing. Will

20:29

you tell me so we know who the chief scientist

20:31

is. Who are the other people on the board? I'll

20:33

tell you who they were two days ago and I'll

20:36

tell you who they were now. How does that say?

20:38

This is I love this. This is great. Yeah,

20:42

because I assume that when you lose

20:44

a fight like this, you

20:46

maybe don't get to stick around, right?

20:49

Yes. So

20:52

let's say a year ago, there were nine

20:54

people on the board.

20:56

Three of them left.

20:57

Some of them were like Reed Hoffman,

20:59

but blah, blah, we don't have to get bogged

21:02

down and that

21:03

that then there were six

21:05

Sam Altman,

21:06

Greg Brockman move fast and

21:08

break things.

21:09

Then you had the four others that voted

21:12

against them in the end. One was

21:14

Elliott sets cover the chief scientist,

21:16

little bit do me but loves

21:19

AI. Then you had weird lay for

21:21

reasons I still can't quite understand. Adam

21:24

D'Angelo, who's the chorus

21:26

CEO. And then you had two other

21:28

people, Tasha McCauley

21:31

and Helen Toner. These are the effective altruist

21:33

types. One of them is weirdly married

21:36

to Joseph Gordon Levitt. Just

21:39

as in the side. Wait, wait, wait, wait, really?

21:42

That is true. Yeah. Okay, cool. Cool,

21:44

cool, cool. So he's just giving you a little bit

21:46

of flavor. So anyway, that

21:49

was the six you had Greg and Sam

21:51

on one side versus these four others.

21:54

The four, you know, kick them out effectively

21:57

will kick Sam out Greg leaves and

21:59

blah, blah, blah. blah, blah, in fighting, in fighting, in

22:01

fighting. It's all blown

22:03

up through this deal that got cut basically

22:06

last night. And now you have only

22:08

three people on the board as of today.

22:11

The Quora guy remains. And

22:13

then you have two kind of weird characters

22:16

come in. One is Brett Taylor, who used

22:18

to run the Twitter

22:20

board and he shoved the deal down

22:22

Musk's throat. He's kind of

22:24

like your kind of classic old school

22:27

Silicon Valley type, former CTO

22:29

of Facebook. And then for reasons

22:31

I really, really, really cannot

22:33

understand. Larry Summers,

22:35

who is just like a strange

22:38

character that continues to pop up in

22:40

the American story, former

22:43

treasurer secretary, former Clinton

22:45

guy, former Obama guy, former president

22:48

of Harvard, somehow weirdly

22:50

involved in seemingly every major crisis

22:52

in American history. So you've only got

22:54

three people there now. Now, supposedly they're going

22:58

to grow it, but who knows? So I

23:00

mean, this is a huge blow

23:03

up overnight. The structure isn't changing.

23:05

Structure is not changing as far as I

23:07

know. You know, by the time you publish this podcast,

23:10

maybe it will. It is 1130 a.m. on November 22nd. Again, it is 1130 a.m. on

23:12

November 22nd.

23:19

It's

23:21

hard to say anything. You know, there's

23:23

lots that we don't know, but I do know that

23:25

the effective altruists, and

23:28

they are also, you know, I don't want to just boil

23:30

them down to that alone. They are, you

23:33

know, also entrepreneurs

23:35

and researchers in their own rights. And I do think there's

23:37

validity to being concerned about AI,

23:40

but they are gone.

23:41

They've decided they've made it cut a deal.

23:44

And as part of that deal, Sam can come

23:46

back, but he can't be on the board and

23:48

he has to undergo an investigation

23:51

into the vague claims that they had made about

23:53

his communication issues. So there's all

23:55

sorts of weird deal making happening. I

23:57

mean, personally, I think they really.

24:00

screwed it up and blew

24:02

their hand. But, you know, that's just me.

24:05

Yeah. Talk about the, because that was the,

24:08

the board issued a statement. Maybe

24:10

that's the wrong way to say it. The board said that

24:13

the reason they fired him was that he had weird

24:15

communication issues. Right. Yeah.

24:18

That's what they said. And then everyone was like,

24:20

well, what does that mean? Did he

24:22

pull a Sam Bankman free?

24:24

Did he lie to

24:26

the board? Did he, uh, I don't

24:28

know. You,

24:31

there was all sorts of speculation about it was what

24:34

it was publicly and at the company,

24:37

they were like, what did he do? What did

24:39

he do that caused you to do this? And,

24:41

you know, ironically, the,

24:44

they, you know, they're, they're, they're, the accusation

24:46

they leveled about him was that he wasn't good at

24:48

communicating. But ever since

24:50

they put out that statement, their communication

24:53

has been ridiculously bad, both

24:55

in

24:56

the company, just saying,

24:58

uh, what, what it wasn't, it wasn't

25:00

safe to you. It wasn't financial malfeasance,

25:02

but not ever saying what it was. And they also,

25:05

and this is my personal opinion, opinion,

25:07

absolutely screwed

25:09

it publicly. Because once they

25:11

put out that statement, they basically went

25:14

silent and Sam was able to drive

25:16

the conversation through the media. And

25:18

then like, we still don't know what their

25:20

argument was. And that was like a huge,

25:23

huge misstep. But to me, it seems like

25:25

they're kind of more academically focused

25:27

people who are a little bit out of their league in

25:29

sort of this sort of succession type

25:32

story. But

25:33

yeah, they never said anything. And they

25:35

should, and in my opinion, they should have. I

25:37

mean, even if they had to leak it through the press, because

25:39

like, one, I want to know.

25:42

And two, if that was it, that's,

25:45

you know, that's pretty weak to just say

25:47

that someone has communication issues and leave it at

25:49

that. And

25:53

I'm not really a Sam Altman defender.

25:55

Like, I don't believe that this guy is God.

25:58

I think he's like a deal making, you know. So

26:00

the deal maker is not even like

26:03

the genius behind all

26:05

of the technological breakthroughs, but

26:07

like they really screwed the

26:10

pooch is what I think personally.

26:13

And I mean, after, you

26:15

know, as this was all going on, there was that letter

26:17

from all of those open

26:19

AI employees who were basically saying,

26:22

bring Sam back, this is ridiculous. And

26:25

I remember, you know, this is this all

26:27

happened so fast. So please do correct me

26:29

if I am bungling anything.

26:31

If you get anything, yes.

26:35

And if you get anything wrong about this story,

26:37

it's completely fine. If I get anything wrong

26:39

about the story, maybe a little bit less fine,

26:41

but still, I think socially acceptable because

26:44

the amount of stuff that's come out about this over

26:46

the last five years, five days, it's five

26:48

years. Yeah, I feel like as

26:51

a

26:51

total aside, I feel like, you know, as

26:53

we were getting to record like, you know, getting

26:55

ready to record this morning, I feel

26:57

like I woke up, saw that Sam Altman

27:00

was once again named the CEO

27:02

of open AI or the head of the board or whatever.

27:04

And I was like, you know, I guess

27:08

we've done a lot of there's been a lot of stuff

27:10

happening, but none of it matters now.

27:12

And we literally just spent

27:15

a week going around and

27:18

getting back to the start. Yeah,

27:20

I mean, that there's a whole

27:22

moral somewhere in there, although I'm not sure what

27:24

it is. But to answer your earlier question,

27:26

yeah, I mean, the employees were

27:28

completely behind Sam and I think that was actually

27:31

a really important reason that

27:34

the members of the board ended up

27:36

cutting a deal to bring him back because,

27:38

you know,

27:39

they claimed though, they're all about safety. And

27:41

I believe that they really are the

27:43

members of the board that were really concerned about

27:45

it. But then what happened

27:48

was you had 700, I think it was 700 of 70, 770 employees saying, if

27:51

you don't bring them back

27:54

and you don't resign, we're just going to quit

27:56

and go elsewhere. And then you have a real

27:58

problem on your hand that I think

27:59

think actually is arguably an even

28:02

worse safety issue, which is that you have

28:05

this incredibly influential technology

28:08

with no one literally running

28:10

it that had created it. And that,

28:12

you know, we know is a problem. I mean, you can look

28:15

at like what happened to Twitter after they got

28:17

rid of everyone. It became

28:19

a disaster. The idea that ChatGBT,

28:22

something similar could happen to ChatGBT,

28:24

which people are just putting in like

28:26

untold amounts of information to every

28:28

day was really scary. So

28:31

I think that, you know, it's

28:33

not how I would exert my

28:36

influence as an employee, but you know, they

28:38

wanted to do it to bring back their boss and they were

28:40

successful.

28:42

All right, cyber listeners, we're gonna pause there for a break.

28:44

We'll be right back after this. Green

28:47

Chef is the number one meal kit for eating

28:49

clean with dinners that work for you, not

28:51

the other way around. No matter

28:53

what's going on, my wife and I would like to come

28:55

together at the end of the day to cook a fresh meal. And

28:58

I love Green Chef because it handles all

29:00

the shopping and planning for us. There's

29:02

no last minute runs to the grocery store or fussing

29:05

over what we're in the mood for. Green

29:07

Chef, it has handled all of

29:09

that. All we need to do is follow the simple

29:11

instructions and a healthy meal is ready in 25 minutes

29:13

or less. Are

29:16

you looking to stock up on functional snacks

29:18

and clean beverages to energize you through the holidays?

29:22

Well, shop Green Chef's new green bundles,

29:24

which are available at Green Market. It's one-stop

29:26

shop for nutritious grab and go breakfasts.

29:29

They've got vegan options, brunch kits, wholesome

29:32

lunches, ready to eat snacks, veggie sides,

29:34

and more. You can easily add those to

29:36

your weekly order. So feel your best

29:39

this November with seasonal recipes featuring

29:41

certified organic fruits and vegetables, organic

29:43

cage-free eggs, and sustainably sourced

29:46

seafood. Cyber listeners

29:48

can get a great deal. For Green Chef's best deal

29:50

of the year, get $250 off

29:52

with code Cyber250 at GreenChef.com

29:55

slash Cyber250. For

29:58

Green Chef's best deal of the year. Get $250

30:01

off with code Cyber250 at

30:04

GreenChef.com slash Cyber250. Green

30:07

Chef, the number one meal kit for eating

30:10

well. Hey

30:12

there cyber listeners, Matthew here. This episode

30:15

was brought to you by Delete.me. A

30:17

few years ago I did some reporting that made

30:19

a few people pretty angry. I got a few

30:22

death threats and I started to worry about just

30:24

how much of my private information was available in

30:26

the public. We've got a safety team

30:28

here at Vice that helps journalists navigate these kinds

30:30

of situations. But the single best thing

30:32

that the security team did was sign

30:34

me up for Delete.me. Delete.me

30:37

is a service that will help you take control of all

30:39

the bits of information about you and your family that end

30:42

up online. Your name, your address,

30:44

the names of your family, even your shopping habits

30:46

are scraped and cataloged by data brokers.

30:49

Delete.me will help you find that information and,

30:52

when it can, remove it. Now,

30:54

not everyone is in danger of being doxxed, but

30:57

we're all getting more spam in our inbox

30:59

and on our phones. Everyone is in danger

31:01

from identity theft scams and phishing attacks. Everyone

31:04

has more private information squirreled away on the hard

31:06

drives of data brokers than they'd probably like.

31:09

Delete.me helps you fight back against the scammers, spammers,

31:12

and other malevolent forces online. After

31:14

you sign up for Delete.me and tell it what information

31:17

to look for, it scouts, go to work rooting

31:19

out your personal information from websites and

31:21

data brokers. Then it gets removed.

31:24

And if it can't get the information removed,

31:26

it lets you know who has it and what exactly

31:29

they have. Delete.me even

31:31

generates a custom report for you that lets

31:33

you see how exposed you are and

31:35

new tools like email masking reduce inbox

31:37

clutter better than a typical spam

31:39

filter. Using Delete.me

31:42

is like having a partner that's looking out for you. We're

31:44

all living on the internet more than ever, putting

31:46

more of ourselves out there, but that

31:48

exposure creates risk. Delete.me

31:51

is a partner that helps you navigate that risk. It

31:53

certainly helped me feel safe after

31:55

I started getting death threats and

31:58

been really excited to be signed up. up for them for the

32:01

next year. My new report

32:03

is coming in soon. They've already removed me from

32:05

like 70 odd lists in the past few months

32:08

and I'm excited to see how many more they can get

32:11

me off of. So, take control

32:13

of your data by signing up for Delete.me.

32:16

Now, get 20% off your Delete.me plan when

32:18

you go to joindelete.me.com slash

32:20

cyber and use promo code cyber.

32:22

The only way to get 20% off is to

32:24

go to joindelete.me.com

32:27

slash cyber and enter promo code

32:29

cyber at checkout. That's joindelete.me.com

32:32

slash cyber, promo code cyber.

32:35

AI might be the most important new computer technology

32:38

ever. It's storming every industry

32:40

and literally billions of dollars are being invested.

32:43

So, buckle up. The

32:45

problem is that AI needs a lot of speed and

32:47

processing power. So, how do you

32:49

compete without costs spiraling out

32:51

of control? It's time to upgrade

32:53

to the next generation of the cloud, Oracle

32:56

Infrastructure or OCI. OCI

32:59

is a single platform for your infrastructure, database,

33:01

application development and AI needs.

33:04

OCI has four to eight times the bandwidth

33:07

of other clouds, offers one consistent

33:09

price instead of variable regional

33:11

pricing and of course, nobody does

33:13

data better than Oracle. So,

33:15

now you can train your AI models at twice the speed

33:18

and less than half the cost of other clouds. If

33:21

you want to do more and spend less like Uber,

33:23

8x8 and Databricks Mosaic, take

33:25

a free test drive of OCI at oracle.com

33:29

slash cyber. That's oracle.com

33:31

slash cyber, oracle.com

33:33

slash cyber. All

33:35

right, cyber listeners, we are

33:38

once again on talking about OpenAI.

33:41

So, for a little while in the

33:45

liminal space that we all lived

33:47

in, it may still in fact be living in, Microsoft

33:51

was going to scoop up everybody?

33:55

I guess so, yeah. I mean, I

33:57

think, you know, like I said, Microsoft

33:59

had no legal power as a shareholder

34:02

in this situation, but they did try

34:04

as hard as they could to exert control over

34:07

it. So, you know, Microsoft

34:09

CEO really

34:11

pissed off that he didn't get any

34:14

notice about what was happening on Friday. And

34:18

then he was trying to get Ram Sam

34:20

back into the organization on Sunday.

34:22

And then when they went instead with the Twitch guy,

34:25

they, he was like, well, I'm just

34:27

gonna hire them all myself then. Basically,

34:31

he was like,

34:32

if you work at OpenAI and

34:34

you wanna come work for Microsoft, come

34:37

on over. And so I think that he

34:39

was also, I mean, in retrospect,

34:41

definitely, trying to exert pressure

34:43

on OpenAI to say, bring this

34:46

guy back, or we're literally gonna take

34:48

this thing down. I

34:51

wanna highlight some of the wilder characters

34:53

in this before we put

34:56

a bow on where we are at the moment. So

35:01

for a hot minute, Image Shear was

35:04

gonna be in charge. Who's Image

35:06

Shear?

35:07

Who's Image Shear? Who's

35:09

Image Shear?

35:11

What's weird about Adam Shear that's with the Twitch

35:13

guy? I

35:16

don't know. According

35:18

to Twitter, a lot of things. Right after

35:20

he got it down, he

35:22

got it on there, and a

35:25

lot of surfacing between, how can we

35:27

say, of a questionable matter

35:32

about some of his theories on questions

35:35

about sexuality. I'll leave

35:37

you the listeners to

35:39

look that up themselves. Probably

35:42

shows that they did not have

35:44

a huge amount of vetting in the two hours

35:47

before they named him. But yeah,

35:49

I think the main thing about

35:51

Emmett was

35:54

the key was a doomer. He

35:57

had referred to a or compared AI.

35:59

I'm going to do a fusion

36:02

bomb and we just said, you know,

36:04

opening arm is

36:06

moving at

36:07

level 10, the

36:13

E-causification is moving at level 1 or 2,

36:16

and so I think

36:19

that if you look at the board, it's really

36:21

interesting to me.

36:22

I'm not going to be nice in the morning, but I'm going to do

36:24

a move below. So we got Maraud in

36:26

and did a commendable job for

36:28

about 24 hours. It

36:34

seems to me that there was this kind of...

36:37

There seems to be one of the... Oh, there have been

36:39

one of them. What

36:41

is the big, going concern,

36:42

this is more fundamental? That

36:45

is the hot ticket.

36:47

That was run by people

36:50

who thought that they were building a

36:52

nuke and maybe shouldn't do it. It's

36:55

a very strange

36:57

situation. I still

36:59

have a little bit of trouble wrapping

37:02

my head around, and it's

37:04

made me think,

37:06

should the

37:07

people... It's a failure

37:09

of self-regulation, I think, if

37:12

nothing else, you know.

37:14

Should the people within the company

37:17

be responsible for putting

37:19

the brakes on a potentially

37:22

world-changing,

37:24

world-altering, world-destroying technology?

37:27

I think the evidence from this weekend

37:30

says we shouldn't at least on

37:32

our own trust them to be responsible

37:35

for that.

37:36

Anyone who appears as a character in a Reddit

37:40

atheist-aligned Harry Potter fanfic

37:43

should not be in charge of anything. Oh my

37:45

God, how could I forget the fanfic? Don't do this

37:47

to me. I'm sorry,

37:50

you know about this, right?

37:51

I very much know about this. This

37:54

is going to be a very, very brief overview,

37:57

which is because I am...

37:59

Harry Potter

38:01

and the methods of rationality is a fan fiction

38:04

that I know was basically in

38:06

name only and that is like this huge

38:08

Harry Potter fan fiction that Emmett Shear

38:10

is a named character in. 404 Media

38:14

did an article about it last night that I haven't read

38:16

yet because I try to do what

38:19

I can to preserve my sanity

38:21

as much as I can and

38:23

knowing things about the Harry Potter fan fiction

38:26

world drama, I've

38:28

sworn off that.

38:30

So it's written by a guy,

38:33

Harry Potter and the methods of rationality,

38:35

it's written by a guy who's like the founder

38:37

of Less Wrong and is one of these

38:40

big Reddit atheists extreme rationality

38:42

guys who

38:45

had an op-ed in time

38:48

maybe the middle of last year about

38:51

why AI is going to lead to the nuclear

38:53

apocalypse. He's

38:56

extremely doomerie. He's

38:59

one of the guys that thinks like we're

39:02

going to hook up AI to like

39:05

3D printers and the 3D printers are going to make

39:08

diseases that will kill all of humanity

39:10

because why would AI not do that? It's

39:13

the rational thing for AI to do. And

39:16

so like Harry Potter and the methods of rationality

39:18

is like this big fanfic opus

39:21

that he wrote that's

39:24

like a way to rational pill Harry

39:27

Potter fans. The whole thing is so

39:29

weird. Is this something I have to read or that

39:31

I can avoid for the rest of my life?

39:34

I don't think anybody in the year 2023

39:37

needs to read a Harry Potter fanfic.

39:38

I think we're good. Harry Potter's

39:41

time is well past us. No

39:43

offense to

39:44

my friends who still do that but let's

39:47

keep it 100 here.

39:48

Read

39:51

it, rational Harry Potter. Nobody

39:54

wants that. Go

39:56

read my immortal instead. Exactly.

39:58

Thank you.

40:01

So

40:03

the other

40:04

person I just want to have

40:06

some words said about is the chief scientist.

40:10

There's some stuff written in

40:13

the Atlantic about him

40:15

and how he's kind of, we've already said

40:17

he's a weirdo, but

40:19

he would do things like lead chance

40:21

at the company. What

40:24

is, what's, what? Elliot

40:27

Sitzkiver, probably the most interesting

40:29

person in this whole tale

40:32

of corporate intrigue. I

40:35

guess the main thing that I would say is

40:37

that he is well regarded as a

40:39

complete genius. He is like

40:42

a AI deep learning genius,

40:45

one of the most well respected people in

40:47

the field in the pure nuts and bolts of

40:49

deep learning. He came

40:51

from the academic world. He did a little

40:53

bit of Google stuff. Musk personally

40:56

recruited him over to open

40:58

AI,

40:59

but he also is super

41:01

into the sort of doomer

41:04

safety side. And so, you know, he,

41:06

he, I think, created

41:09

a sort of, I'm trying

41:11

to think about how to even describe this.

41:15

He basically burns a,

41:18

something in effigy that was supposed to

41:20

represent, you know,

41:22

a bad AI and he was supposed to

41:25

be burning away all of like these bad

41:27

AI. He

41:29

made a totem. Yeah, he made a totem to burn

41:31

effigy as like a religious ritual

41:33

to burn away the bad

41:36

AI. Exactly. And I think the other,

41:38

there was sort of almost a spiritual component

41:41

to his belief. So on the one

41:43

hand, you had this really hard line belief

41:46

in the power of deep learning and

41:48

AI. And on the other hand, he had this

41:50

sort of spiritual side that

41:52

said, you know, we have to remain

41:54

cognizant of the fact that this can destroy

41:57

the world. And like we were talking about earlier.

41:59

It's a confusing dichotomy

42:02

that he is so obsessed with building

42:04

it and simultaneously, you know,

42:07

like, absolutely terrified

42:09

of what he could be building. But

42:12

yeah, I mean, he, as a result of that,

42:14

became increasingly concerned about

42:17

Altman. He was eventually went

42:19

to the board with its concerns. He,

42:21

you know, was worried that Altman was, you

42:24

know, heading down a path that, you know,

42:26

was

42:28

the antithesis of what OpenAI should

42:31

represent. He was the one who called

42:33

Altman into the board meeting to

42:36

get him fired. If he

42:38

didn't do the firing himself, he very

42:40

well might have. And then, bizarrely,

42:44

he switched and called for Altman

42:46

to return, probably maybe

42:49

because of pressure from the other employees.

42:51

But he's just been all over the place. I

42:53

mean, philosophically, he's building

42:56

the thing and terrified of it. He,

42:58

you know, wanted Altman fired.

43:01

And then three days later was saying what a

43:03

horrible mistake he was. He might

43:06

well represent OpenAI better than

43:08

anyone else because he has

43:10

the tools, but he also seems

43:13

to not quite understand

43:16

the corporate world and how to make

43:18

moves inside large, complex

43:20

organizations. I love

43:23

watching dudes struggle

43:26

with the idea of God and

43:28

how they're unable to resolve it. And then like what happens

43:31

after? Like so

43:33

many people, like, get

43:36

real into atheism, and then things

43:38

can get really strange afterwards. Right.

43:42

It's like we've got the tools to build what I would

43:44

think of as a classical image

43:47

of like God is this thing in this thing

43:49

in the sky that watches over us. We should

43:52

do that. That sounds like a great idea. Right.

43:55

Says some of the AI scientist.

43:58

I guess the thinking is like someone's

44:00

gonna build it, we should be the ones who

44:02

it since we're like the

44:04

responsible ones but

44:07

I think that might be a good

44:09

theory that might not have realistic

44:12

applications. Yeah,

44:14

you're spending too much time talking to your word calculator

44:16

buddy. Right, exactly. You're

44:20

doing some projection there. Yeah, right.

44:23

Let's

44:25

expand this out away from OpenAI if we

44:27

can. You also wrote

44:29

this great piece before this kind

44:32

of about Wall Street's love affair

44:34

with the idea of AI in general. You know,

44:36

we're at this era where we're post crypto crash

44:39

where there's kind of this big I would say

44:42

like drive to find

44:44

what the next tech bubble is

44:46

going to be and I would say we're firmly in the middle

44:48

of the AI tech bubble, right?

44:51

Definitely. I think you know when

44:53

Wall Street sees new they see

44:56

opportunity and

44:58

you know I think that they search for

45:00

a technological breakthrough

45:03

might mean you know some innovation

45:05

that leads to a productivity increase

45:07

that leads to more profits that means

45:09

that they should invest in the company. That's

45:12

one side of it. What I was interested

45:14

in is how they're using

45:16

planning to use AI within the banks

45:19

themselves. So I mean Jamie

45:21

Dimon,

45:22

CEO of the biggest

45:24

bank buy assets in the US, you know

45:26

he said that it's going to be critical

45:28

to his company's future success

45:31

and you know there's already

45:33

toying with ways to you know summarize

45:37

legal documents, sift through

45:41

earnings reports, you know sift

45:44

through news reports. That

45:47

stuff I think is really interesting. There's cyber

45:49

security, cyber

45:51

security elements, blah blah blah. What I'm

45:53

really interested in is this

45:55

kind of fringe that is testing

45:58

using AI to make investment

46:00

decisions, you know, like, let's

46:03

entrust the AI to

46:05

tell us to invest in company X,

46:07

Y, and Z, and how much

46:09

to invest and then to switch over to company

46:12

B. That is a really

46:14

big change that I think is pretty interesting.

46:17

And while it's on the fringe, I

46:19

suspect it's on the fringe in the way that 15

46:22

years from now, it will be part of the

46:24

mainstream. Yes. I mean, Wall

46:26

Street embraces this kind of thing

46:28

pretty readily most of the time, right? Like they were

46:31

a big reason that

46:33

we've got some

46:35

of the internet infrastructure we

46:37

do in the country is so that they can

46:39

make lightning fast trades from coast to coast.

46:42

Yeah, I mean, Wall

46:45

Street is always looking for

46:47

an edge. I mean, that's the whole

46:50

game. Financial

46:53

innovation essentially is looking

46:55

for a way to repackage new

46:57

things, old things to make

46:59

them look new, or, you know,

47:01

to figure out a way that they can be slightly

47:04

better than the trading firm across the

47:06

street. So, yeah,

47:08

they're always looking for an edge and computers

47:11

have increasingly been a part of that.

47:12

That's why we had high frequency

47:15

trading become so popular, you

47:17

know, over the last

47:19

over the century, it allowed

47:21

them to kind of

47:23

trade faster, get an edge, and

47:26

occasionally create a flash crash that

47:28

would take away billions in shareholder

47:30

value in a couple seconds. Is

47:32

it possible that so like,

47:35

what can you get in kind of the nuts and bolts of what they're

47:37

talking about actually doing with the AI? Yeah,

47:39

so there's a lot of different stuff.

47:41

I mean, you know, on

47:43

a more basic level of what's already

47:46

happening, Vanguard, big

47:48

financial institution is

47:51

using AI to generate retirement

47:53

portfolios. Morgan Stanley

47:56

has sort of a chat GBT assistant

47:58

for financial advisors.

47:59

users, help them kind of interact

48:02

with clients.

48:04

JPMorgan Chase has

48:07

created a pattern for something called index

48:09

GBT. Well, supposedly that's

48:11

going to help traders decide

48:13

where to invest. And then

48:15

there's people on, like I said, you know, there's

48:18

people whose entire investment

48:20

thesis is based on the development

48:23

of an AI that tells them where to invest.

48:26

And then there's also people who are trying to use large

48:28

language models, chat GBT

48:30

type things to help

48:33

traders say, like, help

48:36

me follow 50 stocks 24-7, help me

48:38

understand the

48:41

rule, financial rule through

48:44

which I can make bet X, Y, and Z. So

48:46

they're trying all sorts of things and

48:51

to varying degrees of success, I

48:53

think. So far, I think, you

48:55

know, AI has not been

48:57

able to beat the market, let's say.

48:59

Is this any different from, say,

49:02

bots that have already been deployed? I

49:04

mean, we've had those for years, right? Sam Bankman

49:07

Freed was using one, I think,

49:09

both at FTX

49:12

and Jane Street. Yeah, it's

49:14

a good question. I think it is

49:16

different. Definitely the use of large language

49:18

models to try and make decisions, financial

49:21

decisions, as a sort of assistant,

49:24

is a different sort of innovation

49:26

that we haven't seen before. It comes with concerns

49:29

because a

49:29

lot of large language models tend to

49:31

be wrong a lot of the time.

49:34

So that could be a concern.

49:38

I think also, you know,

49:40

the idea of using complex, you know,

49:43

generative AI to

49:45

create a portfolio is a

49:48

different sort of thing than just using a bot.

49:51

I mean, Betterment is like a robo-advisor

49:53

and they kind of have algorithms

49:56

or models that they create and then allow

49:58

it to kind of do its thing.

50:00

That is different from having a

50:02

sort of quote unquote smart, generative

50:05

AI that's sort of making the decisions on

50:08

its own learning, you

50:10

know, and developing a thesis as we

50:12

go.

50:13

That is different. It's,

50:15

you know, potentially creates

50:18

more autonomy by the AI,

50:21

which could be of concern

50:24

if it makes a incorrect

50:26

decision and invests, you know, $10

50:29

trillion into, I don't

50:31

know, bookstores

50:33

or something like that, you know? And then,

50:36

you know, everyone's like, wait, why have

50:38

we over invested in bookstores?

50:40

Creates a bubble, crash. I think that's

50:43

the concern.

50:44

I'm sure Michael Lewis will write a wonderful

50:46

book. I'm sure that he will. Hopefully

50:48

he

50:51

finds someone who isn't committing a major

50:54

massive fraud. The

50:57

profile this time. Well, he would profile

51:00

the AI and really humanize it. Sure,

51:02

sure. In a very touching

51:04

manner, for sure.

51:06

I just want a social network of

51:09

this entire year's, you know, tech

51:11

rise and downfall.

51:12

Oh my God. When are we getting that

51:15

is my question. I promise

51:17

you that there are 45

51:19

agents in Hollywood right

51:21

now fighting for, you know, 10 scripts

51:23

or something like that. Yeah. Yeah.

51:26

Whichever one can sign Trent Reznor to

51:28

do the score, just let me know when that's there.

51:30

Exactly. Exactly. I would love Joseph

51:32

Gordon Levitt as Sam Altman.

51:36

It needs to be an inside job.

51:37

Yeah. It's the, they

51:40

hire Aaron Sorkin to write their

51:42

version of what happened. The

51:46

movie I want to see is

51:49

Sam Bankman's read versus the

51:51

Binance CEO, you know, like

51:54

kind of like they were rivals

51:56

and it seemed like the Binance CEO

51:58

like basically took down.

51:59

or started Sam Bankman's

52:02

free fall and then he fell this

52:04

week.

52:04

He fell as well. I

52:07

know we made fun of the Michael

52:09

Lewis book, but I actually do think there's

52:12

a lot of good stuff in there and a lot of the finance,

52:14

a lot of the stuff between them, he

52:17

captures their interpersonal rivalry

52:19

pretty well and it's pretty fascinating.

52:22

Also, nice segue. I

52:25

know how to segue. What can I say? Yeah,

52:29

what's going on here? We're a year

52:31

out from the crypto crash. We're

52:34

deep in the aftermath. CZ

52:36

was one of the last guys standing, right? Yeah,

52:41

CZ Chai Ping Zhao

52:44

is the CEO of Binance, which

52:46

is the world's largest crypto exchange.

52:50

I think Sam Bankman freed

52:52

in him. They were the big rivals,

52:54

the big two of the industry. I

52:58

think, ironically,

53:00

Sam fell first. I think there were always

53:02

a few more questions about CZ.

53:05

He lives in the UAE. They

53:08

really had no company headquarters.

53:11

There were all sorts of sketchy seeming

53:13

things that were happening. But

53:15

after Sam Bankman freed, fell, he

53:18

was like kind of the last man standing in a

53:20

lot of ways.

53:22

But

53:22

as these things go, the

53:26

US federal agencies were

53:30

well underway in a long

53:32

investigation into all sorts of stuff

53:36

that, well, I guess we can say now, was

53:38

completely illegal. This

53:41

week, he kind of surprisingly

53:44

said that he would step down, plead

53:46

guilty to a criminal charge. And

53:48

so he's out at Binance. The

53:51

company is going to pay a $4 billion

53:54

fine

53:55

related to charges like

53:57

allowing America.

53:59

Americans to trade with Iranians and Russians

54:02

and things like that, which is against

54:04

the law for all sorts of reasons. And

54:07

yeah, but Binance will continue under

54:10

a new CEO. I

54:12

think that was a big part of the deal. But

54:15

yeah, I mean,

54:17

it's a fall from grace and it's almost so

54:19

much faster and less dramatic that

54:22

I think it's going to go a little bit more under the radar

54:24

than Sam's. CZ

54:27

was never as big like a personality

54:29

or a character as Sam was, right? Sam

54:32

was hanging out with football

54:34

stars and was very seen and

54:37

was trying to be the face of

54:39

this thing in a way that CZ wasn't. I

54:41

think Sam was trying to be the face for like the

54:44

general person who didn't know who crypto was.

54:47

Your mom's face of crypto is

54:49

what Sam was. CZ,

54:53

I think, instead has sort of like,

54:55

I mean, in the crypto space,

54:58

he was like a god, you know, they were

55:00

like bowed down to him. They just thought he

55:02

was an absolute genius, smarter

55:05

than everyone else, you know, always seemed

55:10

to be a step ahead of the US, kind of laughing

55:13

at the federal agencies, pretending

55:15

to be in compliance and then saying,

55:17

you know, no KYC behind

55:19

the scenes. I mean, he

55:22

was, you know, he kind of represented

55:24

a little bit of like the thumbing of the nose at,

55:27

you know, people who thought that, you know, the

55:29

rules should apply to them. And

55:32

you know, like their whole mantra is

55:34

to

55:35

create a freer

55:36

flow of money. And it seems like

55:38

they definitely did that because the

55:41

money was flowing pretty freely between

55:45

or at least through them and some terrorist organizations

55:48

that came out.

55:49

Yeah, it turns out you can't just invent securities

55:51

a lot out of whole cloth and then say they're not

55:53

securities and shouldn't be regulated like

55:55

securities. Yeah, I mean, that I think

55:58

plays more into.

55:59

SEC investigation that's ongoing,

56:02

so we have not seen the last of it. But

56:04

it seems like he's going

56:07

to go into the background. He's saying

56:09

that he's going to

56:10

take some time off, maybe start investing

56:12

a little. Maybe I know the funniest

56:15

thing was that he was like, and I might make

56:17

myself available for mentoring if

56:19

people are interested.

56:21

I don't know if that's a good idea

56:23

or

56:23

not. I don't know. I

56:25

mean, it's a pretty crazy thing to say like,

56:27

hey, I just stepped down from my

56:30

organization and pled guilty. My

56:33

company had to pay a $4 billion fine. I'm

56:36

basically a disgrace now. And

56:39

I'm available if you want to chat

56:41

about how to start your startup. Just

56:43

put time on my calendar. Yeah, exactly.

56:47

You learn more from failure than you do from success.

56:51

That's definitely true. And in

56:53

that

56:53

way, I guess he has a lot to tell people. Would

56:56

you rather talk to Napoleon right at

56:58

the height of his power, or do you want to interview

57:02

Napoleon when he's in exile on the island? Yeah,

57:05

I will say I'm joking

57:07

a little bit. I'm obviously saying these things in jest.

57:10

I don't really feel the need to be super kind

57:12

to CZ, but he did

57:14

say, at a minimum, I'll tell

57:17

you what not to do. If

57:20

you're my mentee, I'll tell you what mistakes

57:23

to avoid.

57:24

Good times. All

57:26

right, Maxwell, I think that about covers

57:28

the speaks wild news. Thank you

57:31

so much for coming on to cyber. Everyone

57:33

enjoy your Thanksgiving. Have a good Thanksgiving,

57:35

guys.

57:37

Bye, everyone. Bye. Bye.

58:00

A A

58:09

A A

58:14

A A

58:21

A A

58:27

A A

58:35

A A

58:44

A A

58:52

A A

59:00

A A

59:09

A

59:11

A

59:16

A A

59:22

A A

59:32

A A

59:40

A A

59:48

A A

59:56

A

59:59

A

59:59

actors, authors, chefs,

1:00:02

musicians, and more about how the food

1:00:05

and the culinary traditions of their youth

1:00:07

shape their lives in interesting and

1:00:10

sometimes surprising ways.