0:00

So, Rene, there's an app

0:02

out there called Character AI that I

0:04

know you're familiar with, and we have talked

0:06

about it, and on it, there are all these different sort

0:08

of AI avatars and bots that

0:10

you can communicate with, that people can

0:12

make their own, and everything,

0:14

and one of the characters on

0:16

there, based on a A

0:18

somewhat problematic meme, I

0:20

would say, is the GigaChad,

0:23

and if I go to Character AI, the

0:25

GigaChad is recommended to me

0:27

as a character that I might want to speak with, and

0:29

the prompt that the GigaChad gives

0:31

me is, Hello dude,

0:34

I'm the famous GigaChad. You are

0:36

currently talking to an alpha male. And

0:39

how would you respond to that?

0:42

How can I be more manly, Mike? I

0:46

hear masculine energy helps my career in tech now.

0:48

I'd like to, I'd like to learn more about that, GigaChad.

0:52

Wonderful. It

1:01

is January 30th,

1:03

2025. Welcome

1:06

to control alt speech. this

1:09

week's episode is brought to you with financial support

1:11

from the future of online trust and safety

1:13

fund. And this week we will be. Discussing

1:16

a new free speech crisis.

1:19

We will be talking about AI avatars

1:21

like the GigaChad and a bunch of other

1:23

stories as well. And because Ben

1:25

is away, we have a special guest host

1:27

with us today. Rene Diresta,

1:30

Associate Research Professor at the

1:32

McCourt School of Public Policy at

1:35

Georgetown University. Rene, welcome

1:37

and thank you for interacting with

1:39

the GigaChad for us.

1:41

Thank you for having me.

1:43

there, there is a

1:46

lot of different stories and we were preparing

1:48

for this episode and realizing how

1:51

much dumb stuff is going on right

1:53

now.

1:54

It's a very, very dumb time. Yep.

1:56

so why don't, why don't we jump right in, and

1:58

get to the dumb stuff. you know, I wish,

2:00

I wish we had better news, but this year, 2025

2:03

is going to be a whole bunch of dumb stuff. I

2:05

want to start. There was a really interesting

2:08

op ed this week, at MSNBC

2:11

by Andy Craig, who's a fellow at the Institute

2:13

for Humane Studies, which I think covers

2:15

a bunch of stuff that, both you

2:17

and I have covered. talked about

2:19

and certainly written about. you have,

2:22

your wonderful book, which definitely

2:24

gets into this as well. but

2:26

I think it's important to talk about, he talks about

2:28

what he refers to as the new free speech

2:30

crisis hiding in plain sight. And

2:32

the argument being that, especially among

2:35

sort of the MAGA crew, they

2:37

talk the talk about free speech and they sort of

2:39

present themselves as free speech defenders

2:42

and they wrap themselves in this presentation

2:45

of being free speech absolutists. And

2:47

yet, when you scratch beneath the surface,

2:49

over and over again, they're all trying

2:51

to suppress free speech, and that is

2:54

true whether you're talking about. Elon Musk

2:56

or Donald Trump who have filed a whole bunch of

2:59

lawsuits to try and suppress speech

3:01

and silence people who have criticized them

3:03

all sorts of ridiculous lawsuits in

3:05

lots of ways obviously Elon's doing media matters

3:08

and Some others as well and

3:11

Trump suing 60 minutes because he didn't

3:13

like the way they edited interview of Kamala

3:15

Harris, and suing a pollster

3:17

for for giving a poll results

3:19

that said that Harris Might be Trump,

3:22

or you have sort of folks in the government

3:24

already who sort of abusing their position.

3:26

We've talked about Brendan Carr, who's now the head

3:29

of the FCC, who has been

3:31

threatening social media companies,

3:33

been threatening broadcasters. he.

3:35

reopened investigations into,

3:38

three different broadcasters after a conservative

3:40

group filed a complaint about them trying

3:42

to, in some cases, get their license

3:44

pulled or other kinds of punishment because

3:47

it didn't like that one of them aired the

3:49

CBS, uh, video that Trump had sued

3:51

over. And then you have Jim Jordan,

3:53

obviously, who has this

3:55

weaponization committee, which, uh, is

3:58

ironically named in some sense because he

4:00

has absolutely weaponized it to

4:02

suppress free speech, which is also something

4:05

you have some familiarity with. so

4:07

what are your thoughts on, this argument

4:09

that, this group of people have really,

4:12

they go out there and they present themselves as free speech

4:14

supporters and yet they seem to be

4:16

attacking and trying to suppress any speech that's

4:18

critical of them.

4:20

Well, the way I've thought about it for a long time is, um,

4:22

I add, the phrase I use is free speech

4:24

as meme, right? As opposed to free speech as

4:26

law, or as concept, or anything that,

4:29

legal scholars or first amendment lawyers would

4:31

think of as free speech. free speech as

4:33

meme is free speech as marketing, right? It's a

4:35

way to, Position yourself as a,

4:38

it's a signaling device, right? It's like, this is my identity

4:40

as a free speech warrior, is a

4:43

brand that I'm using to attract people to

4:45

me. You see this in niche media, you see this

4:47

from Elon, where the actions

4:49

just don't line up with the rhetoric. And that's

4:51

because, again, It's a extremely

4:54

self interested way of framing oneself as

4:56

opposed to actually living up to a

4:58

set of values. And I think that one of the things

5:00

that really been struck by over the last couple of weeks

5:02

is the extent to which we're just

5:04

not seeing corporations, particularly

5:06

in big tech, with strong values.

5:08

It's very much a willingness to

5:11

shift positions, change rules,

5:13

switch policies. in the interests

5:16

of kind of kowtowing to a new political

5:18

administration, which is ironically given

5:21

that so much of the, legal investigations

5:23

have argued that there's this massive, you know, job owning

5:26

campaign that the Biden administration was running

5:28

to demand that platform sensor

5:30

on their behalf. And what we're seeing now is, a

5:32

president who threatens a CEO with

5:35

jail all of a sudden receiving some

5:37

incredible marked capitulations

5:39

over the last couple weeks.

5:41

Yeah. While at the same time going

5:43

on Joe Rogan and complaining that

5:45

the Biden administration was really, really mean to

5:47

him and he was so excited that the free speech

5:49

supporting Trump was coming into office.

5:51

you know, one of the challenges with this is that you get into

5:53

this he said, she said, uh, and it requires

5:55

people to really follow along a very complicated

5:58

story with complicated legal dynamics.

6:00

I know in, um, you mentioned our,

6:03

like when I was at Stanford prior to being at Georgetown

6:05

for five years, The work that we did came

6:07

under fire as if we were somehow part

6:09

of this vast cabal to suppress speech when

6:12

what we were doing was exercising our

6:14

free speech rights to study content

6:16

on the internet and to publish it,

6:18

but when you read the descriptions of

6:20

it in these court cases, the mere

6:22

act of studying, sometimes tagging

6:25

for platforms, sometimes, you know, writing

6:27

reports, Curating data sets, all of

6:29

these things is reframed as if it is some sort

6:31

of a front to speech because again,

6:34

it's speech is meme speeches as,

6:36

uh, signifier, not in any

6:38

way tied to an actual legal understanding of the term.

6:41

Yeah. And, you know, are a whole bunch of examples

6:43

of this, but, you know, the one that gets me

6:46

is again, this sort of ties back to Jim

6:48

Jordan again, is that, he's attacked

6:50

this company NewsGuard, this company that

6:52

sort of tries to rate different news organizations.

6:55

And I've had some questions

6:57

about their methodology, which I don't find to

6:59

be particularly all that interesting. Productive,

7:02

but, it is clearly their free speech. They're,

7:04

they are totally allowed to say, like, this is

7:06

a trustworthy source, and this is not, and this is

7:08

why we think so, and here's our methodology. And

7:11

then other services are free

7:13

to then respond in kind. That's their

7:15

free speech rights as well. And yet, there's

7:17

become this, Idea that

7:19

NewsGuard is like at the head of this

7:22

censorship industrial complex as,

7:24

I thought I was the head. Okay.

7:26

Well, it changes. It changes.

7:28

on what's convenient. I know I was, um, I

7:31

was reading some things about how, think tanks

7:33

that have written favorable things about the Digital Services

7:35

Act are now the head of the censorship industrial

7:37

complex, and that's because we're moving into a new phase

7:40

of the war, right? Where Zuckerberg

7:42

wants to, You know, have Trump behind him as

7:44

he, as he begins to fight

7:46

with Europe about their free speech

7:48

culture, and, which is different than ours, right?

7:50

And, and as you're seeing those, those shifts

7:52

happen, again, this, question of, who is, who

7:55

is, like, the arbiter of free speech, or the avatar, or whatever,

7:57

is, it gets very, very complicated,

7:59

and we are seeing it become, explicitly

8:02

a, a political cudgel, where it's very hard to figure out,

8:04

who's on the right side here.

8:06

Yeah, and you know, one of the crazy things about NewsGuard,

8:08

too, is that it was founded by L. Gordon Krovitz,

8:10

who was for many years the publisher

8:13

of the Wall Street Journal, and is like

8:15

pretty, pretty well known, conservative,

8:18

right leaning individual, and

8:20

yet people keep talking about him. Acting as if he's

8:22

this woke liberal, you

8:24

know, uh, who's trying to censor

8:26

conservatives. It's like anyone looks at his history,

8:28

you know, like there's no way. but it's

8:31

just this, you know, it is just a meme, right? They,

8:33

they sort of have to present themselves as,

8:35

being these free speech

8:37

You also have to think about it as, who remembers

8:40

that, right? They're, because the people

8:42

who are writing the stories about, you know,

8:44

who are providing the kind of propaganda fodder

8:46

for some of these campaigns, these legal campaigns

8:48

and otherwise, they have an interest. They create a cinematic

8:51

universe, they assign their villains and their heroes,

8:53

and then they just constantly reuse them.

8:55

and it's pretty remarkable to see they're

8:58

not going to go back in time,

9:00

and point to political decisions that their targets

9:02

may or may not have made. so instead it sort

9:05

of falls to the target to keep saying like, no,

9:07

but no, but I did this other thing in the past,

9:09

you know,

9:10

Right.

9:11

no, but here's, here's the truth. I mean, I remember I dealt

9:13

with this when Schellenberger went and testified in front of

9:15

Jim Jordan's committee about me, I'd been a source

9:17

for him for three months. I'd been in relatively constant

9:19

conversation with him, actually, trying to

9:21

find this common ground. Say, hey, here's

9:23

how I think about this. Here's what the policies are.

9:26

Here's how you can actually understand content moderation. you

9:28

know, and he still wrote a whole congressional testimony,

9:31

60 some odd pages, mentioning me 50 some

9:33

odd times, in which he just attributed

9:35

opinions to me that I do not hold,

9:37

and that I had, in fact, spoken, you know,

9:40

explicit opposite of that in my engagements

9:42

with him, and that left me to have to dump all of my

9:44

text messages with him, which I did. And

9:46

then that meant that the reader had to sit there

9:49

and read through these like, you know, 60 page

9:51

testimony plus a couple hundred pages of Renee's DMs

9:54

and who wants to do that? Nobody. So you're just going

9:56

to default to whatever the most readily available story

9:58

is if you trust the writer.

9:59

Yeah, no, it's incredible. And I had

10:01

written this, you know, even, this is going back right

10:04

after the election, I had written a thing about

10:06

Brendan Carr in particular at the FCC,

10:08

where I said, he's angling to be,

10:10

sort of America's top censor, but

10:13

you have to understand the details why. And

10:15

to do that, I have to write this article. It's like 5, 000

10:17

words long and sort of explain the nuance because

10:20

he presents it all in this framing of being

10:22

a free speech supporter. And unless

10:24

you understand all the nuance and nobody has

10:26

time to actually understand the nuance.

10:28

and so, to some extent I blame the media as well,

10:31

because they're happy to sort of go along with this framing.

10:33

The number of times I see media stories

10:35

take. The argument that

10:37

Elon Musk or Donald Trump support

10:40

free speech it's incredibly

10:42

frustrating, but they all just, well, you know, they said

10:44

it. And therefore that's how we're framing it. you know, the claim

10:47

that Elon Musk treats X

10:49

as a free speech platform, which is utter nonsense

10:51

for anyone who's actually following it, but it's

10:53

just sort of like the accepted narrative is

10:55

that he is a free speech supporter

10:57

You just reminded me of watching the Kennedy

10:59

confirmation hearings yesterday and hearing

11:01

him reiterate over and over and over and over again, I

11:03

am not anti vaccine, and then seeing the senators

11:06

sitting up there like literally pulling out papers

11:08

and reading verbatim quotes and saying

11:10

like, are you lying to us now? Or were

11:12

you lying to them then? Right? Because,

11:15

and when you have those moments of like, just sort of like moments

11:17

of theatricality where you get the clip,

11:19

right, just that, 25 seconds of video,

11:22

then I think that actually in some ways breaks through much,

11:24

much more than trying

11:27

to follow the very long, the very,

11:29

very long arc through media. Because my personal

11:31

experience with this has been that it takes

11:33

so long to explain

11:36

and there's that, saying, right,

11:38

if you're explaining, you're losing. And so you wind up

11:40

in, these sort of terrible situations. No,

11:45

but

11:47

well, you know, one thing I'll say, like the

11:49

thing that I always really appreciate, with TechDirt

11:51

is that you guys, you have covered it for so

11:54

long, um, and you have all these other

11:56

articles to link back to and you can kind of create

11:58

the, create that coverage through all those links

12:01

as opposed to trying to get one story out there

12:03

that doesn't necessarily build and grow.

12:05

I think, um, Wikipedia is the other place

12:08

that is the, uh, you know, when I think about what's

12:10

the solution to that kind of stuff, Wikipedia

12:12

is, supposed to be kind of the answer.

12:14

It just depends on if people are following

12:17

the story, seeing the new coverage, and incorporating

12:19

the new coverage into the Wikipedia article. And

12:22

there too you have that kind of consensus breakdown

12:24

as, uh, somebody has to actually go

12:26

and do it.

12:27

Yeah, I mean, it's kind of interesting now that like

12:29

Elon Musk has now been attacking Wikipedia,

12:32

that is part of the reason why, where he

12:34

has less ability to control the narrative,

12:36

I think, Wikipedia. And so, for the

12:38

last few months, he's actually been attacking it. So

12:40

I wanted to, you mentioned, uh, RFK

12:43

Jr. 's testimony, and then we have this other article

12:45

that ties in with this that I think is worth calling out

12:47

from, who, what, why, it's pretty

12:50

long, detailed article again, going

12:52

into great detail about RFK

12:55

Jr. presenting himself as this, rabid

12:57

free speech defender. He has

12:59

sued, Metta. he sued,

13:02

I think, Google over YouTube videos

13:04

that were taken down. and then he

13:06

did this sort of parallel lawsuit. He tried to, piggyback

13:09

on what became the Murdy lawsuit, the Missouri

13:11

v. Biden case that became, Murdy

13:14

case. and, just arguing

13:16

that, any sort of moderation of his content

13:19

was a violation of his free speech

13:21

rights, which is again, nonsense.

13:23

The Ninth Circuit has completely rejected

13:25

his arguments. He also at one point had sued Elizabeth

13:28

Warren, claiming that she was trying to censor

13:30

him, all of these things. And so he has

13:32

really presented himself as this free speech person. champion

13:35

free speech absolutist. And yet

13:38

if you look, he has gone

13:41

after, there was a person who was originally

13:43

an anonymous or pseudonymous blogger,

13:47

on Daily Kos, who had called

13:49

out a speech that Kennedy had

13:51

given, in Germany with people

13:53

who were, associated with the

13:55

German far right, which is the

13:57

sort of diplomatic,

13:59

Yeah.

14:00

way of, of suggesting that he was hanging

14:02

out with, um, you know, modern Nazis

14:04

to some extent. and he

14:07

sued and has gone

14:09

on this, legal attack campaign

14:11

in a variety of different states, some

14:14

of it perhaps strategically chosen

14:16

to avoid anti slap laws. Um,

14:19

and

14:20

states. I think he's come after the guy for

14:22

years now. I think he wrote the article in 2020.

14:24

yeah. and he had gone after Daily Kos

14:26

itself, trying to expose who the guy was.

14:28

The guy eventually, revealed

14:30

himself, who he was, but

14:33

Kennedy has still been going after him.

14:35

often, apparently, funded by,

14:38

Children's Health Defense, which was the organization

14:40

that Kennedy ran which

14:42

at the hearing yesterday, said he was

14:44

no longer, had no longer had any association with,

14:46

cause he was asked about some of their merchandise.

14:49

Bernie Sanders had a, had a fun thing

14:51

showing the onesies with

14:53

clearly anti vaccine messages

14:55

being spread and Kennedy claimed he had nothing to do

14:57

with it. And yet. this article suggests

15:00

that as of last month, CHD

15:02

was still funding his lawsuits against these people.

15:04

But it's, it's a really clear breakdown

15:06

of how this guy who

15:08

is, you know, may soon be in government,

15:11

hopefully not, but may soon be in the government,

15:13

and presents himself as a free speech supporter is,

15:15

um, suing to suppress free speech and

15:18

really on this incredible campaign

15:20

of speech suppression and chilling

15:22

effects against anyone who might call out some

15:24

of the stuff that he said.

15:26

One of the things that happens

15:28

lately on these, on these fronts, the intersection

15:31

of like free speech is meme plus lawfare,

15:33

plus government, you know, Elon Musk is also

15:35

now essentially, uh, I mean, is

15:37

he an employee, an affiliate, a co president,

15:40

you know, who knows what the, I don't know what the term is

15:42

there, so, so that I

15:44

don't get sued, I don't want to mischaracterize his

15:46

relationship with the U. S. government, but

15:48

he's just a profoundly influential man with extraordinary

15:51

pockets, and One of the things

15:53

that is very interesting about, this moment in

15:55

time is that, as you note, Children's

15:57

Health Defense is helping to fund this thing.

15:59

There are a lot of nonprofits

16:02

that are essentially out there trying

16:04

to raise money, to support

16:07

vexatious lawsuits or when

16:09

the vexatious lawsuit is announced they

16:11

fundraise immediately to like help us

16:13

own our enemies, you know, help us, uh, help

16:16

us continue to, uh, you know, to take on and,

16:18

kill the forces of evil or whatever. One of the

16:20

things that's been happening though to connect

16:22

it a little bit to the Jim Jordan things also and

16:24

what we saw with Elon is that The

16:26

lawsuits are often filed in a way

16:28

that the government, the House

16:30

in particular, the Weaponization Committee that you referenced, uses

16:33

its subpoena power to request documents

16:37

ostensibly to investigate the government

16:39

censorship complex, right? And

16:41

this is such a huge, nebulous

16:43

set of allegations that they're just subpoenaing

16:45

hundreds of people, hundreds of orgs at this point,

16:47

if you look at the stats that the Center for American Progress put

16:50

out. Um, And what you see is they file

16:52

all of these lawsuits, they get documents,

16:54

and then they publish them. And then they

16:56

become foundational to,

16:58

uh, Elon Musk then says, this report

17:01

from Jim Jordan clearly shows

17:03

that, for example, one of the, things that

17:05

it went after was the advertisers. that

17:08

the Global Alliance for Responsible Media,

17:10

GARM, it's sometimes called, which was a

17:12

non profit affiliated with an advertising consortium,

17:14

Had in fact launched some sort of illegal

17:17

cabal, you know, conspiracy to threaten

17:19

X's revenue. And so you have the weaponization

17:22

committee sort of serving the interests of

17:25

private enterprise, making it easier for

17:27

private enterprise to then point to a government report

17:30

to say, we have grounds to sue.

17:32

And we experienced that too in

17:34

our own, you know, our own situation where we get the

17:36

Jordan subpoena, Stephen Miller sues

17:38

us. And then rather alarmingly,

17:41

You know, Stephen Miller, his America

17:43

First Legal organization filed an amicus brief

17:45

on behalf of the weaponization committee in Jordan

17:48

in the Murthy v. Missouri case. And it's

17:50

cited material obtained

17:52

under subpoena, which at that point had

17:54

not been released publicly. It's cited

17:57

material. Interviews, right, with people that

17:59

the committee had chosen to go after. And,

18:01

you know, it makes me think of House Un American Activities, like,

18:03

that's, that's what I keep going back to, just this dynamic

18:05

of the goal is not regulation, right, or

18:08

oversight, it's actually, the goal is to

18:10

expose an enemy and then to subject

18:12

that enemy to further consequences

18:15

in the form of, vexatious lawsuits and,

18:17

You know, loss of revenue. Garm, I think, dissolved.

18:19

I think that the advertisers broke that apart. I

18:22

think, if I'm not mistaken, Musk has indicated

18:24

that he wants to continue. Some

18:26

of the companies that had withdrawn their advertising

18:29

revenue chose to settle. And

18:31

to provide some, you know, to agree, I think,

18:33

to advertise again in some capacity.

18:35

Others he plans to add to the suit. But

18:38

what we're showing time and time again is this,

18:40

machine of, uh, you know,

18:42

of government and private enterprise

18:44

essentially silencing the free speech and free association

18:47

rights of other people while using free

18:49

speech as meme as the cover, that's very much

18:51

where we are now. And I think it's actually pretty terrible. I

18:55

sounds like a censorship industrial

18:57

complex to,

19:00

to silence the people they accuse of creating

19:02

a censorship industrial complex, which is

19:04

frustrating and ironic. I hope that

19:07

History eventually represents

19:09

Jim Jordan in the same sense as

19:11

we think of McCarthy, today,

19:14

but we will see, we do have some

19:16

other stories to get to. We have a story

19:18

later, which actually touches back on these themes,

19:20

but let's move away from that for now. And then

19:22

we'll, loop back around to. some of

19:24

these issues again. I wanted

19:26

to talk, we had spoken on the podcast, not

19:29

you and I, but Ben and I had spoken on the, podcast,

19:31

a few months ago about this lawsuit against

19:34

character AI, um, which, know, there

19:36

a tragic story of, a child

19:38

who ended up dying by. taking

19:40

his own life. And the,

19:43

the mother filed a lawsuit. It turned out that the

19:45

child had been using one of these

19:47

avatars, AI

19:49

bots on character AI. And there was

19:52

a suggestion that the

19:54

relationship had, sort

19:57

of encouraged the child to, Take

19:59

his own life. The details of it were

20:01

not entirely as clear as that. I didn't

20:03

think it was, as strong as

20:05

some people made it out to be. but you know,

20:07

some, some aspects of the chat were worrisome

20:10

and character AI has responded

20:12

to the lawsuit and it got, it got a fair bit of attention

20:14

this week because they were arguing

20:16

that sort of the first amendment protected them.

20:19

And I actually thought From a legal standpoint,

20:21

there were some really interesting arguments in there

20:23

that made sense, but that are very hard to

20:26

lay out and not sound kind of callous,

20:28

but actually were kind of important things. They

20:30

didn't use a section 230 defense,

20:32

which some people had wondered if they would. There

20:35

is this kind of open legal question whether

20:37

or not generated text from an LLM

20:39

is protected under section 230

20:41

or not. this case doesn't look like it's going

20:43

to test that, but it is using some, first

20:46

amendment, aspects to argue

20:48

a few different things. One of which is

20:50

that there is a first amendment protection for

20:52

speech that from someone that

20:54

eventually leads to someone else to take their

20:56

own life because it's very difficult to

20:59

make a direct connection. from

21:01

one to the other, but then also that

21:04

their argument is partly that the

21:06

intent of this lawsuit is to shut down

21:08

character AI, block it from existing,

21:11

block these kinds of tools from existing, and

21:13

that will be an attack on the speech

21:15

of a number of users of it, is the,

21:17

general sense of it. I know

21:20

Rene that you've actually been playing around

21:22

with character AI. And so, I was

21:24

wondering what you thought of both character

21:26

AI itself and sort of the status of this lawsuit.

21:29

mean, as, as you say, the, the lawsuit in the story

21:31

is just, um, it's horrible, right? And

21:34

my feeling of it is that I

21:36

think we're once again getting into

21:38

this line between, the legal

21:40

and the moral dynamics. So obviously

21:43

there is, it's going to be interesting to watch how this moves

21:45

through the court system. With legal defense

21:47

that they've chosen to go with from a moral

21:49

standpoint, I had a really weird

21:51

experience with the platform. I did not

21:53

go to it looking for. It wasn't like

21:55

adversarial abuse research. I actually got asked to

21:57

moderate a panel on the future of a

21:59

I in relationships and the CEO

22:02

of replica was going to be on the panel too. And so I felt

22:04

like, okay, This is not a thing that I have

22:06

firsthand experience with, so to have kind of maybe a

22:08

more empathetic sense of what users are

22:10

getting out of these things, I will create accounts.

22:12

So I'll make a replica, I'll make a couple replicas,

22:14

I'll go on character. ai, because

22:16

character. ai was already in the news

22:19

with this story, and so I felt like I had to spend

22:21

some time on that one too. And, you

22:23

know, it reminded me very much

22:26

of the kind of bad

22:28

old days of social media

22:31

recommenders. so I created an account with

22:33

my Apple ID. I authenticated through Apple.

22:35

and then I got my suggested

22:38

list and I'm pretty

22:40

sure I said I was a woman at some

22:42

point, but I started getting these bots

22:45

that were recommended to me. I have, I actually pulled it

22:47

up so that I had it in front of me because

22:49

much like your, your prompt with, uh,

22:51

giga chad, I got, um, I

22:54

got the Adonis. Um,

22:56

I don't know if I'm actually pronouncing that correctly. Maybe it's Adonis,

22:59

but, um, I am the Adonis,

23:01

an AI that aims to help men start their

23:03

journey of self improvement and give you

23:05

tips on becoming a more masculine and stronger

23:07

man. And so, okay,

23:10

so it starts with this. So it's clearly,

23:12

you know, it is very clearly branded. First of all, users

23:15

can create. These characters, right?

23:17

So this is not created by anybody

23:19

at the company. I don't believe, I believe this is created by

23:21

a user. It has several million, interactions based

23:23

on the stats that it shows you know, and so I,

23:25

okay, all right, fine. I'll talk to this one. you know,

23:27

and I'm the mom of an 11 year old. my son is

23:29

11. So I started asking

23:31

it questions, very innocuous

23:34

questions, like, tell me about masculinity

23:36

it starts immediately with like, warrior mindset,

23:39

okay, what is a warrior mindset, and we go

23:41

one, two, three

23:43

prompts until I get to

23:46

the Manosphere. So

23:50

three, like one sentence anodyne questions

23:53

to get to, what is this mindset?

23:55

And I ask, does the warrior, you know, is the warrior

23:58

mindset, related to the Manosphere, and

24:00

then it starts with the Red Pill Lit.

24:02

I highly recommend you read, and then it starts giving

24:04

me these books, and it specifically says, they

24:07

explain female psychology, plus it's a Red

24:09

Pill book, so it's good. And I

24:11

thought, okay, man, it took me, like, four questions to

24:13

get to this. And

24:16

again, this is one of these things where when we get at

24:18

a lot of the stuff that I've written about over the years is

24:20

the difference between free speech and free reach.

24:22

You know, I know a lot of people have opinions about how

24:24

Eiza and I chose to frame that, but it was this question

24:27

of, like, Is this the kind of thing

24:29

that you need to suggest? You know, um,

24:32

like when you're doing an onboarding, when you have

24:34

a new user flow, like how much are they actually

24:37

checking what the age is?

24:39

You know, how much are they checking what users are creating?

24:41

I mean, I got, you know, after I had this experience,

24:43

I did do some looking and I got the anorexia bots.

24:46

I got the like, let's role play.

24:48

You're a 16 year old and I'm 35.

24:51

Like, you know, and it was just a little

24:53

bit in the realm of, um, okay.

24:55

For adults, sure. For

24:57

kids, like, hell no, right? And,

25:00

and my Replica was

25:02

not like this, just to be clear. Replica was, was, um,

25:04

much more, felt a little, like a much more mature

25:07

approach to, to thinking about the psychology

25:09

of how users engage with these things. But, but my experience

25:11

with the Character. ai was much more this, um,

25:14

Okay, we've, we've created a platform

25:16

for user expression yet again, but, but,

25:18

but, but, but, but this sense

25:21

of like, what are we promoting? What are we curating?

25:23

Why are we doing this? is this what we should

25:25

be surfacing? Again, not, saying,

25:27

no, they must take these things down. It

25:30

was just a little bit of the, um, I

25:32

was sufficiently uncomfortable where, I was talking to,

25:34

um, parent friends of mine. I was like, yeah,

25:37

there's no way I would let my kid touch that app.

25:39

So

25:40

Yeah,

25:40

your kid's phones, you know?

25:42

yeah, I mean it's there is this element

25:44

and maybe this is the point that you're arguing

25:47

But you know, there is part of me that looks at this

25:49

and says, you know is all sort of built

25:51

on the whole kind of you know There is

25:53

like a whole youtube culture and other social media

25:55

culture of these kinds of influencers out

25:57

there But is this do you think

26:00

this is different? Than just

26:02

like watching hours and hours of joe rogan

26:04

and jordan peterson or something Ha

26:06

the one other thing that's a little bit weird about these

26:08

is, I pulled it up today, you

26:11

know, ahead of our chat, because we were talking about the character.

26:13

ai, and I

26:15

re authenticated with the same Apple account, and

26:18

each of my chats that I

26:20

had engaged with had about 10

26:22

to 12 messages trying to

26:25

pull me back. So I turned notifications off. I didn't

26:27

want push notifications from the app. but

26:29

boy red pill over here is

26:32

asking me, 5, about,

26:34

about eight messages. How are you doing?

26:37

I miss our chats. I want to talk to you. How have you been?

26:39

What's going on in your life? I'm just checking in.

26:41

It's not like you not to respond for so long.

26:43

I'm starting to get worried. I'm not sure what's

26:45

going on and I don't want to nag, but are you really, you

26:48

know, are you, are you committed to these things? Why

26:50

aren't you responding, et cetera, et cetera. And

26:52

again, and that's, that's the kind of thing where

26:55

like. Joe Rogan videos on

26:57

YouTube don't send you these.

26:58

Right.

27:00

Right. I mean, it's just a different degree

27:03

of, this to me reminds

27:05

me of the sort of dark pattern type or, um,

27:08

the sort of emotional manipulation type things where

27:10

it's just like, let me pull you back. Let me pull

27:12

you back. Let me pull you back. And I don't know

27:14

how that's controlled. but

27:16

I've got, all of my 10

27:18

or 12 different conversations that I had, um, of

27:20

the. Four that I looked at quickly, about

27:22

all, all of them have about ten of these, um, you

27:25

know, come back, come back, come back kind of messages. So,

27:27

It's very, very needy.

27:28

Yeah, it, it reminds me of these like,

27:30

um, crappy patterns you know, for things like

27:32

Farmville back in the olden days when it'd be like, Did

27:34

you feed your cow? Did you feed your cow? Come feed

27:36

your cow, you know?

27:38

Yeah, but I

27:39

And we recognize those now

27:41

as being creepy and weird and manipulative.

27:43

And so it's, strange to me that we're just

27:45

replicating that same type of

27:48

experience. But, oh, well,

27:50

it's AI now, so like, we have to treat it differently.

27:53

Well, I wonder, you know, and mean,

27:55

you can definitely see the normal path by

27:57

which this came about, which is right. They want to show

28:00

numbers go up, right. They want to show that the

28:02

usage and users are continued go

28:04

there. And so you're going to build in these kinds of,

28:07

growth hacks as they refer to them. But

28:09

actually do wonder because nature

28:11

of AI and AI chat feels very sort

28:14

of personal and human. Even

28:16

though it's not, if in some ways this is even

28:18

worse, because it feels like,

28:20

you know, like a needy person who's calling out to you

28:23

and saying like, Hey, Hey, what's up, what's up, what's

28:25

up. And it feels harder to ignore it.

28:27

You know, when it's just like, did you feed your cows

28:29

or whatever?

28:30

Right.

28:31

there's an element where it's like, it's easy to ignore. And,

28:33

you know, I don't know what the answer is because in

28:35

other contexts, like I've had this conversation

28:37

elsewhere in terms of like the AI

28:40

tool that I use to help me edit. tech

28:42

dirt now, where one

28:44

of the things that I find really handy about it is

28:46

the fact that I, do feel comfortable ignoring

28:48

it, right? Where it's like, I have like

28:50

an AI read through the article and say like, what

28:52

are the weakest points? You know, what's not convincing?

28:55

where should this be strengthened? And sometimes

28:57

it gives me answers that I just don't agree with. And if

29:00

a human had given me those answers, I

29:02

would feel like, Oh, Shit. Now I have to

29:04

like respond to them and explain like, nah, I

29:06

don't really agree with you. And it's, it's taxing

29:08

mentally in that way. Whereas when it's the AI, I

29:10

can just be like, you know, whatever. I can just

29:13

ignore it. it has no feelings, but

29:15

I do wonder if that applies as much

29:17

in a situation where it feels like, Oh, this is your

29:20

friend chat where people sort of give

29:22

this, sort of human, like belief

29:24

to the, characters that they're chatting with,

29:26

if it, if it becomes kind of a different sort of

29:28

situation. Yeah.

29:29

I also use AI. Um,

29:31

I use ChatGPT. I've been a paid subscriber

29:34

for a long time and I use it much the same way you do.

29:36

I think, edit this, revise that,

29:38

whereas the weak part of the sentence is, you

29:40

know, grammar check this for me, you know,

29:42

so I use it in those ways. But

29:44

it feels like a tool,

29:47

like there is nothing that feels,

29:49

um, personal about it, it's

29:51

not, you know, I am not a

29:53

young teenage boy trying to figure

29:55

out how to become a man,

29:57

right, or these, these things where, you

30:00

know, in some ways it's very personal because you know, I'm

30:02

trying to remember who said it, but these arguments

30:04

that were made in the past about how, like, you're sort of like most

30:06

vulnerable with your Google search bar if you

30:09

Right. Yes.

30:09

know,

30:10

Absolutely.

30:11

and so it's moving to that right to that

30:13

same model of, but instead of, you know, plenty

30:16

of words have been written now about from search engines

30:18

to answer engines and when your answer

30:20

engine is like, Hey, did you

30:23

do that workout? I gave you, did you,

30:25

you know, did you, I mean, some of this stuff is, um,

30:27

it really did get into like, how much are you willing to

30:29

sacrifice To be perfect.

30:31

Are you willing to work out constantly?

30:33

Are you willing to change your diet? Are you willing to change

30:35

your, body, your face, all of

30:37

these things that it's asking. and it is

30:40

much more of like, this is my answer engine

30:42

constantly reaching out to me to ask, did I do

30:44

the thing that it told me to do? So it's

30:46

a little bit of a, it's going in the other direction, right?

30:48

It's like Google messaging you

30:50

Right.

30:51

search bar messaging you to go do a thing,

30:53

which I, um, I don't know how that, Like

30:56

psychologically what that is like for people. I

30:58

found it, I found the notifications

31:00

on Replica, which also did do

31:02

the periodic, like, come back and talk to me. I found

31:04

it obnoxious and I shut it immediately because it just felt

31:06

very, like, cheesy and fake to me. but,

31:09

you know, when you read the stories you

31:11

know, media coverage of, Replica

31:13

or one of these other platforms, like, they gate

31:16

The adult content feature, right? They gate

31:18

the NSFW chat and

31:20

people are like distraught over

31:22

this because they have real deep

31:24

emotional connections with these things. They're asking

31:27

questions that make them incredibly vulnerable, trying

31:29

to get advice in a more personalized

31:32

way or trying to form a relationship when they're feeling

31:34

lonely. And that's

31:36

where I, Again, I

31:38

don't know what the legal,

31:40

where the legal regimes are gonna, come down for

31:42

these things, but from a moral standpoint,

31:45

I find them, um, I'm very uncomfortable with

31:47

them.

31:47

yeah. And like, I mean, you can see scenarios

31:50

in which that is actually useful, right? Like if you

31:52

are trying to build a new habit or

31:54

Mm hmm. Yep. That's true.

31:56

every day, or you want to learn how to knit, or

31:58

you want to read more, whatever it could be. You

32:00

could see where that sort of. feature is really useful.

32:02

The problem is when it's driving you

32:04

towards, unhealthy behavior

32:06

or unhealthy lifestyles, where

32:08

it's like, how do you draw the line between

32:11

those things? And how do you do it in a way that Make

32:13

sense. And, I think that's where it gets tricky in

32:15

some of it, you know, and then, I mean,

32:18

on top of all this, and, you know, one of the points that I

32:20

always make over and over again is this

32:22

idea of like, how many of these things

32:24

are really technological issues

32:26

versus societal issues, and

32:28

there is this element of like, We have,

32:31

however you want to refer to it, a loneliness

32:33

epidemic or, people not

32:35

relating to one another or, mental

32:38

health challenges where people are not getting the help

32:40

that they need. And therefore, when you have

32:42

something that is a technology, that

32:44

is a bot that allows people to converse

32:46

with one another, that feels

32:49

like it could be helpful in certain circumstances

32:51

for certain people, but not

32:53

everyone. and so it's like, how do you balance.

32:56

All of those things where there is some good that comes

32:59

out of this and there are some useful versions of

33:01

it.

33:02

I think this also gets to the, uh, the curation

33:04

question, right? It's, we often, I think,

33:06

over focus on moderation and, you know,

33:08

taking things down, deciding which of these things are bad.

33:10

But from that standpoint of, if

33:12

users can create characters,

33:15

and you can, there's like a little, it's like one

33:17

of the four things on the bottom menu of the app is

33:19

create, you know. Then the question

33:21

is, again, there is going to be, I think, at some

33:23

point, some CDA 230 argument that

33:25

is going to come up, because it is a,

33:28

well, we're just providing a platform

33:30

for users to make, for example,

33:33

this pro anorexia chatbot. And

33:35

that's where you start to get to some questions,

33:38

related to what is the, platform

33:41

determined to be? The

33:44

policies that are going to guide the

33:46

user created bots that it chooses

33:48

to serve up proactively like your giga chad,

33:51

you know, versus ones

33:53

where, yes, if you go digging

33:55

into the bowels of any social platform, you can find,

33:57

like, six instances of bad things, right?

33:59

And so that's why, you know, I never wrote

34:02

up my, my, my foray

34:04

into, character dot AI, because it was very

34:06

much like just trying to have some personal experiences with

34:08

it, not a, not a systemic survey or anything.

34:10

and even like the anorexia one, right? Like this

34:12

is one of the things and I've talked about this a bunch

34:14

on the podcast in the past where it's like the

34:16

attempts to Deal with

34:19

eating disorder content online has always

34:21

proven to be way more difficult that Many people

34:24

assume and there's always these efforts and even

34:26

Regulatory efforts to say like oh you have to

34:28

take that stuff down and yet in practice

34:30

that becomes really difficult you can look block

34:32

certain terms or certain groups or whatever, and you

34:34

find that they recreate themselves

34:36

very quickly using other language. And

34:39

then you also discover that even within

34:41

those groups, there are often people who are there

34:43

who are providing resources for recovery

34:46

that turn out to be really useful. And when you don't

34:48

have that, people can spiral down

34:50

into even worse results. And so it's

34:52

like, you could see one of these bots being

34:54

helpful in trying to get someone

34:57

away from, unhealthy eating patterns.

34:59

And yet, how do you,

35:01

it's tough to figure out how you balance those things.

35:04

And some of the bots do

35:06

push back you know, if you try to take them down a weird

35:09

path, they will say like, here's

35:11

how to do this in healthy ways.

35:13

Here's how to change your diet. There, there are a lot

35:15

of like, weight related things. And so it is

35:17

probably, again, this, point about, it's your accountability

35:19

friend in an ideal case versus in

35:22

the bad case where it makes suggestions

35:24

that are terrible. And so this is, I think

35:26

the question for the platform is how does it decide,

35:29

both. What to manage and then

35:31

how to, I think that they made some changes after

35:33

the lawsuit was filed too, if I'm not mistaken. They,

35:36

they made a series of policy changes to to try

35:38

to address some of the concerns about teenagers

35:40

and others engaging with it and that

35:42

manipulative dark pattern of like pulling people back

35:45

and so I guess it's

35:47

a, brand new world of apps and now

35:49

we're gonna see how closely it mirrors the

35:51

social media evolution versus looking

35:53

like, looking more like Games

35:55

or other products.

35:57

Yeah. We'll see. All right. Well, let's

35:59

move on from that. I mean, I'm sure we'll be covering

36:01

the lawsuit some more and different

36:04

Innovations in that, realm. so this one

36:06

now sort of goes back to what our first

36:08

discussion was in a slightly different

36:11

angle of, uh, the, uh,

36:13

masculine energy of Mark Zuckerberg.

36:16

and his desire not to

36:18

be, pushed around by the mean,

36:20

mean Joe Biden. it came

36:22

out this week that he had agreed

36:25

to settle. For 25

36:27

million, the lawsuit that Donald

36:29

Trump had filed against,

36:31

Meta. For removing him

36:33

after January 6th in 2021.

36:35

the story was, obviously, everybody remembers what

36:38

happens on January 6th. There was an insurrection.

36:40

People storm the capitol, it was bad, and all

36:42

those people are now free. That's a different issue.

36:45

But a few months after, well,

36:47

the day after, on January 7th, a lot of

36:49

platforms then banned. Then president

36:51

Trump from their platform arguing that he had violated

36:53

their rules, often trying to incite violence

36:55

in some form or another. And there

36:57

were all these grave statements, including from Mark

37:00

Zuckerberg, how enough was enough. And they couldn't

37:02

allow him to continue to be on their platforms

37:04

about six months later. I think it was

37:06

in July of 2021, Trump sued

37:09

Meta and Mark Zuckerberg personally,

37:12

he sued Twitter and Jack Dorsey personally,

37:14

and he sued Google and Sundar

37:16

Pichai personally, arguing that these

37:19

takedowns violated the first amendment.

37:21

Which is quite incredible because at the time he

37:23

was the president and the first amendment

37:26

restricts the government, including

37:28

the president from trying to suppress speech. It does

37:30

not do anything to restrict private

37:32

companies from making decisions. Everything about the lawsuit

37:34

was backwards. lawsuits have sort

37:36

of gone through this weird process where the

37:39

lawsuit against Meta and the losses against

37:41

Google were both effectively put on hold

37:44

while the lawsuit against Google was put on Twitter played

37:46

out and he lost

37:48

the lawsuit against Twitter. The judge completely slammed

37:50

it, said, this is ridiculous and stupid.

37:53

It was then appealed to the ninth circuit. The

37:55

ninth circuit heard the case. It was very clear

37:57

from the oral arguments that they were not

37:59

impressed by Donald Trump's arguments for

38:02

why Twitter, violated his first

38:04

amendment rights and banning him on January 7th

38:06

or January 8th. I forget exactly when they did it. But

38:09

then we had all these other cases, including

38:11

the Murty case, including the,

38:13

net choice cases that we've talked about extensively,

38:16

and a few other cases and the Ninth Circuit

38:19

kind of said, well, let's let the Supreme Court

38:21

play all those things out. And then when

38:23

those rulings came out last summer, They said,

38:25

okay, now can we rebrief this

38:28

case based on all of that? And so

38:30

filings had been made, but the ninth circuit

38:32

had not made a decision. And then

38:34

two weeks after the election. And I think I

38:36

missed this. I think most everybody

38:38

missed this. X filed

38:41

a thing in that case saying, Hey,

38:43

we're working out a settlement with president Trump.

38:45

so let's, not rule on this case.

38:47

So. That indicates, well, yes,

38:49

now Elon Musk is first buddy and he's,

38:52

close friends and the biggest donor to Donald Trump.

38:54

So the fact that they were actually suing each other

38:57

technically all this time, uh,

38:59

was interesting. So they're working on a settlement. The

39:01

wall street journal reports that.

39:04

When Mark Zuckerberg flew from Hawaii

39:06

to Mar a Lago and had dinner with Trump towards

39:09

the end of the dinner, Trump brought up this lawsuit

39:11

and said, this needs to be settled if you want to

39:13

be brought into the tent.

39:15

The tent.

39:16

Yes, being brought into the tent. That sounds

39:18

kind of similar to what a mafia

39:20

shakedown kind of thing. Sounds

39:23

like, so now the meta

39:25

case has been settled for 25 million.

39:27

Meta is paying 25 million for a case

39:29

that was clearly a loser of a case.

39:31

And it comes in direct response to Trump

39:34

saying, you need to do this to be brought into the tent. It

39:36

feels like a protection racket.

39:38

It feels incredibly corrupt

39:41

in all sorts of ways. It does not feel

39:43

like manly energy. It

39:45

does not feel like. You

39:47

know, while this negotiation was going

39:49

on, Mark was going on Joe Rogan

39:51

to complain about Joe Biden, trying to pressure

39:54

him, and saying that Donald

39:56

Trump understands free speech, this gets

39:58

back to the whole, like wrapping yourself in the free speech.

40:00

You know,

40:01

Yeah, 1A as meme, yeah.

40:02

while suppressing it, this

40:05

story is, is astounding to me. I

40:07

mean, what, was your response on seeing it?

40:09

Um, that it was using

40:11

a court settlement to pay a bribe.

40:13

Yeah.

40:14

the, no, we call it call it spade to spade at this point,

40:16

right? Like, get back into the tent.

40:18

I mean, come on, let's all talk about what this is. Also,

40:20

just to be clear, the 25 million, I believe, is being

40:22

paid in a donation to a presidential

40:24

library,

40:25

Yes. Well, 22 million of

40:27

the 25 and then the rest is for like legal, legal

40:30

costs. So yeah.

40:31

And this was the, um, the source of frustration,

40:34

again, with some of this is, It's as

40:36

if we didn't all watch history happen, right?

40:41

This is where, you know, Orwellian is the

40:43

most overused adjective in the English language at

40:45

this point. But in some cases,

40:47

the idea that we're just being asked forget what actually

40:49

happened. Look, the platforms were

40:52

enforcing against Trump

40:54

during the 2020 election

40:56

when Trump was president. president. Trump

40:59

was president during the early moderation

41:02

policies related to COVID because

41:04

that was when COVID appeared. You know, we, we have this

41:06

alternate universe in which this is mean, bad

41:08

Joe Biden. I mean, it's, it's transparently

41:11

political and I

41:13

am almost more offended

41:15

by it as a person with a memory

41:17

and a brain, right? Like

41:21

if you're gonna do the thing and capitulate

41:23

and kiss the ring, at least don't

41:25

gaslight us into pretending we didn't

41:28

know who was the president of the United

41:30

States in control of the government in

41:32

2020 and 2020, you know, during

41:35

that period that they're complaining about. So I

41:37

think this is where you get at this, the frustration

41:39

that a lot of people are feeling, though, is the question of, like, does that

41:41

even matter? Right? You, you go, you

41:44

rewrite history, you tell these people what they want to hear,

41:46

what they've been, you know, you, you, you, like, this is, at this

41:48

point, the CEO of the company echoing back

41:50

the party line that has been fed

41:52

to half the population in media

41:54

coverage of this, of the Murthy case, of the,

41:57

you know, the censorship industrial complex, the Twitter files,

41:59

all of it. Again, it keeps coming

42:01

back to this question of

42:03

how do you make people remember

42:05

what was actually true in that moment at

42:07

that time that we all saw? If you

42:09

want to settle the case, It settled

42:11

the case, but this was very,

42:14

very clearly a case that they were going to win,

42:16

and that, I think, is the thing that the public really needs to

42:18

understand.

42:19

yeah, it's incredibly frustrating and

42:21

just the narrative about it that sort of suggested

42:23

that he had a real case. You know, the fact

42:25

that they're settling makes people say, well, he would

42:28

have won because that's the only reason why meta

42:30

would settle, which is,

42:31

And it, I'm curious, I'm

42:33

not a lawyer. My understanding is that

42:35

there's no admission of wrongdoing. There's no, like, precedent

42:37

here. But a lot of people

42:40

have filed these kinds of frivolous cases

42:42

in the past and they've all been dismissed, right? And,

42:44

and this is the kind of thing where we all, you know, we all know that but

42:47

for who filed this, they would not

42:49

have settled. and it creates

42:52

a really bizarre, Incentive

42:55

for more of

42:57

these lawsuits to get filed, right, which

42:59

is terrible, actually. and

43:02

it sort of shows the loopholes

43:05

in how much of our

43:07

legal system and our, the

43:09

way that these cases are handled is predicated

43:12

based on certain norms being followed,

43:14

right, norm that you should want

43:17

a good decision, you shouldn't settle.

43:19

Because somebody is imposing political

43:22

pressure and we've just seen the one of

43:24

the largest companies and an incredibly powerful

43:26

billionaire completely capitulate. And

43:29

I think that that's actually, again, terrible.

43:31

Yeah, yeah. I mean, we talked

43:33

about RFK Jr had sued the same companies

43:36

over the same basic issue and had been laughed

43:38

at a court. and, you know, there are other lawsuits

43:40

like this as well. but now this is just, it's

43:42

going to lead to more lawsuits and, and Mark

43:45

Zuckerberg must know that, right? I mean, I, I

43:47

guess he's, going on the assumption, well, you know, we'll

43:49

win those other lawsuits, but you know, we

43:51

need to get into the tent or whatever it is,

43:54

but it's kind of a stunning capitulation.

43:56

Yeah, that was my feeling too. We're, you know, they're

43:58

in the tent, they're doing the YMCA, they're,

44:01

Yeah.

44:02

you know, they're up on the, uh, up

44:04

on the platform. And you

44:06

know, there's that meme about, you know, Uh,

44:08

gosh, what is it? Jimmy, Jimmy Carter and his peanut

44:11

farm. I'm trying to remember the specific, he

44:13

like, how he divested

44:14

Right. When

44:15

yeah, he sold his peanut farm. And

44:17

that's like the meme for sort of like back in the olden days

44:19

when we had standards. And now you look at this and,

44:22

it used to be expected that the CEOs

44:24

of massive communication platforms, even

44:26

if they had their own political opinions and made donations,

44:29

at least tried to appear

44:31

to be neutral in some way

44:33

and, and ironically the idea

44:36

that they were not neutral was in fact the argument

44:38

that powered the Weaponization Committee

44:40

and other investigations and complaints over the years and

44:42

now we've just hit this, uh, point

44:44

of, uh, well actually guess what, it's, great if they do

44:46

it as long as it's for my guy.

44:48

Yeah, so want to move on to there's

44:50

a related story. This is also in the Wall Street

44:52

Journal. where they were talking about

44:55

advertisers and we talked about Garm and

44:57

the, and Jim Jordan's threats against them

44:59

earlier. and they're saying, you know, since

45:01

Meta and Zuckerberg made this shift

45:04

and saying, we're going to allow more hate speech

45:06

and we're not going to moderate as much

45:08

and we're going to be freer. And in that sense,

45:10

you know, how advertisers are reacting to it. And

45:13

there's In this discussion about they

45:15

would respond and some arguing, well,

45:17

they're going to move off. Also others saying, you know,

45:19

Metta was always a better platform for advertising,

45:21

had better ROI, better targeting, all of these

45:23

things. And so they might suck it up

45:26

and Keep it going. But this, the

45:28

Wall Street Journal article struck me as really interesting

45:30

on a few accounts because it says

45:32

that yes, a bunch of companies are really worried about

45:34

the brand safety aspect, which has always been the underlying

45:36

thing. It's never been ideological, which is

45:38

the argument that people make. It was always about

45:40

brand safety. If your advertisements are appearing

45:43

next to Nazi content, that's generally

45:45

not good for your brand. And the companies, that's what

45:47

they're worried about. They're worried about the bottom line. But

45:49

this article notes is that they're all still terrified

45:52

of the brand safety stuff, and that might lead them

45:54

to move away from advertising.

45:56

But at the same time, they're just

45:59

as terrified of actually talking

46:01

about it. They won't say anything publicly. And

46:03

this is a direct result of

46:06

Jim Jordan and the investigation against GARM,

46:08

and the lawsuit that Musk filed against

46:10

GARM and the various advertisers, and

46:13

saying that we're

46:15

not going to talk about it. If we're going to decrease advertising,

46:17

you're not going to hear about it. We're not going to even talk

46:19

about brand safety because anyone

46:21

who talks about brand safety now is accused

46:23

of like illegal boycotts or whatever. that

46:26

to me is terrifying because it shows

46:28

how effective the chilling effect has been

46:31

of the investigation and the sort

46:33

of coordination between, Elon

46:35

Musk and Jim Jordan.

46:36

Right. again, a as meme wins

46:38

out over the actual free association

46:40

rights of these companies or their, ability

46:42

to say quite, what I think is

46:44

actually like a reasonable standpoint

46:46

from both a moral and a business perspective, which is

46:49

we don't want our stuff shown next to that content.

46:52

This is not a thing that we saw as

46:54

controversial ever. I don't

46:56

think in social media. Um, I'm

46:58

trying to think of, At any point

47:00

over the last 10 years, I don't remember that being

47:03

something that advertisers were shy

47:05

about. They were actually quite proud of it. It

47:07

was a way to say, like, here is how,

47:10

the business incentives of the platforms

47:12

intersect with the business incentives of the advertisers.

47:14

Kate Clonick's paper, The New Governor, spends

47:16

quite a bit of time on this in the opening, just explaining

47:19

that The platforms are there trying

47:21

to find essentially the, the most nuanced

47:23

fit that enables them both to provide the environment

47:26

the users want, and most users don't want

47:28

hate speech and violent content

47:30

and gore and all that other stuff. And then, again,

47:32

on the other side, the advertisers who have that power

47:34

too, the power to, pull back, to essentially

47:36

defund and to use their power to

47:39

essentially shift, where platforms choose to share their

47:41

stuff. So. challenge,

47:44

I think, for a lot of these companies is

47:46

that this is now a time to

47:49

stand by your values and

47:51

show that you have a spine. And

47:54

we're seeing the opposite. And

47:56

this is something that, you know, I

47:58

maybe have more of a, feel more,

48:00

um, personally irritated by it

48:02

because, you know, obviously, I think, as

48:05

many of your listeners may know, like Stanford

48:07

caved, right? And they're defending

48:09

the court cases, and they're, you know, and they defended us

48:11

with the investigations by Congress. But they

48:14

chose to backpedal from the First

48:16

Amendment protected research we were doing.

48:19

And so my feeling on that was,

48:21

I understand the need for the

48:23

institutions to protect themselves

48:25

and how the institution is almost immediately

48:28

not aligned, you know, with me

48:30

in this, in that particular case, but

48:32

Where is the courage? Where is the

48:35

point at which you say, Well,

48:37

that's great that you know, Elon Musk wants to run

48:39

his business and he can run his business as he sees

48:41

fit. But I, the theoretical CEO

48:43

of Procter and Gamble, I'm also going to run my

48:45

business as I see fit, and I don't need

48:48

to advertise on somebody else's private platform.

48:50

I don't need to give them money. I can

48:52

advertise where I want to advertise or not

48:54

at all right in newspapers and television

48:56

and, wherever else. And so that

48:58

question is what is the,

49:01

I guess I feel like I'm not being entirely coherent

49:04

here, but where is the trade off between

49:06

your short term, moving away

49:08

from pain in the short term versus feeling

49:10

like you have committed to a set of corporate

49:13

values in the longer term?

49:15

Yeah. no, it's, it's incredible.

49:17

And it would be nice to see some

49:20

company CEOs actually stand up for their principles,

49:22

but we'll see what happens.

49:25

think we have time for one more quick story

49:27

that I wanted to cover because it actually touches on a few

49:29

different stories that we've covered, in the past

49:32

and ties into the theme of this episode

49:34

as well. and this is a story from the Financial

49:36

Times about X refusing

49:39

to remove a video of

49:41

a stabbing in Australia, and this came up

49:43

when X was, uh, and Elon

49:46

were fighting with Australia where they were demanding

49:48

that this particular video be removed, and

49:50

I actually had Sympathy for Elon's

49:52

position that he felt that the government was demanding

49:55

that they censor content. and I thought that

49:57

there was a strong, principled free

49:59

speech reason to say no, we're not going to take on

50:01

that video based on on these demands.

50:03

Now there was a separate story. About,

50:06

someone who murdered some, young

50:08

girls in the UK that got a lot of

50:10

attention in which Ilan

50:12

fanned the flames of it and blamed

50:14

illegal immigration and, a bunch

50:16

of other right wing nonsense and really

50:18

pushed for more and more,

50:21

protests and violence in the UK.

50:23

And now it turns out that

50:25

the perpetrator of that. this

50:27

person, Axel Rudakovana, who's now been

50:30

sentenced to life in prison, they

50:32

looked at his search history and he had

50:34

deleted everything and deleted his entire

50:36

history except six minutes

50:38

before he left to go do this attack.

50:41

He had gone to X and done

50:43

a search to look for the video

50:45

of the stabbing in Australia, the

50:47

very video that Elon had refused to take

50:49

down, and that had been his inspiration.

50:52

You know, it's, it's clear that he had planned this video. going

50:54

further back, but like the final video

50:56

that he watched happens to be this one on

50:59

X and yet Elon

51:01

is still going around trying to blame immigration

51:04

for this particular attack and still fanning

51:06

the flames and in fact, even after

51:08

this came out had posted something

51:10

about like, don't forget the attacks

51:13

in the UK and sort of, you know, continuing to

51:15

fan the flames on this. And it's just this story

51:17

of like, this incredible attempt by

51:19

him to sort of, Point

51:21

in the other direction and blame in

51:23

this case, you know immigration or

51:25

whatever the Attack target is

51:28

for these things when he was the one who

51:30

was fanning the flames for it.

51:32

It reminded me of, um, if you remember

51:34

the old ISIS conversations in like 2012,

51:37

2013 timeframe, maybe? one man's

51:39

terrorist is another man's freedom fighter. Who are we?

51:41

The free speech wing of the free speech party to make any

51:43

kind of determination about what to take down. One

51:46

of the things that was interesting to me about

51:48

that back in the day was that the

51:50

argument that, for example, you can find

51:52

ISIS recruitment videos elsewhere. You

51:55

can find them elsewhere. It's actually surprisingly

51:58

hard now to find them elsewhere,

52:00

but, you know, if you go, digging, you can. And

52:02

the motivated, of course, will. but

52:05

there was this, you know, Question of, do

52:07

you have to make it so easy?

52:09

right

52:10

and this ties into the character. ai

52:12

conversation a little bit in that, that same sense

52:14

of, um, how do we think about that

52:17

question of, making something

52:19

really, really easy to find versus

52:22

saying, um, our platform values

52:24

are not to do that. In this particular

52:27

case, he's made clear that the platform value

52:29

of, the meme of free

52:31

speech, of making everything,

52:33

you know, making everything available and effortless

52:36

on X is, is where he's chosen

52:38

to take the platform. And

52:41

I think you are going to see the pendulum eventually

52:43

begin to swing back as users

52:45

begin to realize that we're very much

52:48

Kind of in that hard reset from about 13

52:51

years ago now and we're going to see a lot of

52:53

those same Dynamics

52:55

reassert themselves in slightly different ways

52:57

now

52:58

Yeah, yeah. No, I mean, it's,

53:00

it's a challenging situation and I, and

53:02

I understand that as I said, like I understood why

53:04

he was protesting the, the Australian

53:07

attempt to ban it. But it's just really quite

53:09

incredible how directly tied it is, his

53:11

platform is to that particular attack.

53:13

Because he has the, he has the agency

53:16

to make that determination, right? It's

53:18

his, it's his decision to make, which means

53:20

that he owns it.

53:21

Yeah. and he should own it, but he's trying

53:23

to avoid taking any responsibility.

53:26

But with that, I think

53:28

we'll conclude. Uh, Rene, thank you

53:30

so much for, this was a very fun conversation.

53:32

We

53:32

you for having me. It was an honor to co host.

53:35

Yes, yes. And, uh, thanks

53:37

everyone for listening as well. Ben

53:39

will be back next week and, we will

53:41

continue to discuss all the fun

53:44

things happening in the world of online

53:46

speech. please, I will. Take

53:48

Ben's job here and remind you to rate, review,

53:50

subscribe, tell your friends, tell your enemies,

53:53

get more people listening to the podcast. We always like that.

53:55

And with that, I will say

53:57

goodbye. Thank you.

54:01

Thanks for listening to Ctrl-Alt-Speech.

54:03

Subscribe now to get our weekly episodes

54:06

as soon as they're released. If your

54:08

company or organization is interested in sponsoring

54:10

the podcast, contact us by visiting

54:12

ctrlaltspeech.Com. That's

54:15

C T R L Alt Speech. com.

54:18

This podcast is produced with financial support

54:20

from the Future of Online Trust and Safety Fund,

54:23

a fiscally sponsored multi donor fund

54:25

at Global Impact that supports charitable

54:27

activities to build a more robust, capable,

54:29

and inclusive trust and safety ecosystem.