0:01

As regular control alt-speech listeners will know,

0:03

Mike, we start every week of the

0:05

podcast with a phrase or prompt that

0:07

you might see if you open an app

0:10

or a kind of web service. And we've

0:12

been doing this because we think these little

0:14

bits of text and taglines prompt us and

0:16

make us interact and engage. And this week

0:18

we could not do signal, right? Everyone has

0:20

been talking about signal. Even if we've done

0:23

it before, I can't remember. Signal

0:25

doesn't have a prompt or tagline

0:27

in the same way. So we're

0:29

stealing a bit of marketing material

0:32

from one of their online pages.

0:34

And so this week I'm prompting

0:36

you to speak freely. Well I

0:38

would just like to say that if any

0:41

random CEOs of tech companies want to

0:43

accidentally add me to their group chats

0:45

and tell me what they're thinking about

0:47

content moderation these days, I would not

0:49

be opposed to that. We're going to

0:52

talk about a few of the CEOs

0:54

today actually. I think so, I think

0:56

so. And what about you Ben? Can

0:58

you speak freely for me please? Well

1:00

I was speaking freely at the marked

1:02

insurgent event that we had in London

1:05

last night. I'm a little bit worse

1:07

for where this morning as a result.

1:09

It was a super fun evening,

1:11

but I managed to incorporate, you'll

1:13

be unsurprised Mike, my renovation. into

1:16

my presentation about content moderation. So I

1:18

spoke freely both about moderation and renovation

1:20

and yeah I got a few comments

1:22

and a few jokes about it but

1:25

it was fun. And you didn't you

1:27

didn't end up winning the Twitter sign

1:29

from last week's discussion for as part

1:31

of your home renovation? Not this time,

1:33

not this time unfortunately. Hello

1:43

and welcome to Control Alt Speech,

1:45

your weekly roundup of the major

1:47

stories about online speech, content moderation

1:49

and internet regulation. It's March 28th

1:51

2025 and this week's episode is brought

1:53

to you with financial support from the

1:56

Future of Online Trust and Safety Fund.

1:58

This week we're talking about Elon Musk.

2:00

courting of tough men, porn appearing on

2:02

platforms you might not expect, and rage

2:04

baiting. My name's Ben Whitelaw, I'm the

2:06

founder and editor of Everything in moderation,

2:09

and I'm with a man who's in

2:11

a lot of signal chats and may

2:13

be lurking in one that you're in

2:15

too, Mike Masnick. Please, please accidentally add

2:17

me to your signal chats. Yeah, I'm

2:19

presuming you've never had this done to

2:22

you before. I was going to ask

2:24

you. I have never had anything. I

2:26

am in, as you know, I am

2:28

in quite a lot of signal chats.

2:30

That is, it appears to be the

2:33

group chat app of choice for lots

2:35

of people, but I have never been

2:37

accidentally added to a group chat. What

2:39

a wild story of the nature of,

2:41

yes, the US government right now. Yeah,

2:44

such a shame, such a shame. So

2:46

yeah, if anyone's listening and wants that,

2:48

Mike, feel free to do so. I've

2:50

got some congratulations to give to you,

2:52

Mike. a special award this week. Well,

2:55

I think I think technically I don't

2:57

get the award until October. Oh, okay,

2:59

okay. But it's been announced. You don't

3:01

have an announcement. Okay, talk to us.

3:03

The protector of the Internet Award. Yeah,

3:06

yeah, I've won the protector of the

3:08

Internet Award. I've asked them to give

3:10

me a shield to go with the

3:12

award. It's from the Internet infrastructure coalition,

3:14

which is a really good group. I've

3:16

done various... projects with them in the

3:19

past. I've spoken at their conferences in

3:21

the past and they, I think only

3:23

recently, have started doing this. This is

3:25

not the first year that they've done

3:27

the Protector of the Internet Award, but

3:30

they've been doing it for a little

3:32

while. They have a big thing where

3:34

they fly a bunch of their members

3:36

into DC and they meet with Hill

3:38

and they meet with a award ceremony

3:41

where they give awards to a few

3:43

people. And this year, very... It's a

3:45

nice, nice, you know, little thing. Yeah.

3:47

And yeah, so that'll be in October.

3:49

So you get to dress up and

3:52

receive your shield. Yes, yes. Excellent. Well,

3:54

you might remember last week that we

3:56

had a call out for some podcast

3:58

reviews. And I'm glad to say, Mike,

4:00

we've got some very funny listeners. Because

4:03

the bar for submitting the review was

4:05

not hating the podcast. And this is,

4:07

let's be clear here, this is what

4:09

you told them. You told people, if

4:11

you don't hate the podcast, please review

4:14

it. And our listeners. took us up

4:16

on that with that specific prompt. Yeah

4:18

with glee with actual glee we had

4:20

three reviews which is three more than

4:22

we've had in the last six months

4:24

so thank you to those three listeners

4:27

and all of them said in their

4:29

review in some form I don't hate

4:31

this podcast which is which is and

4:33

I'm very glad to hear that so

4:35

thank you to those people who've left

4:38

the review. For those of you who

4:40

have left a review or who left

4:42

a review a long time ago and

4:44

want to update how you feel about

4:46

the podcast, please go and leave a

4:49

few words and a star rating on

4:51

your podcast platform of choice. When you

4:53

can, it really helps us really reach

4:55

our tentacles out into the wider world.

4:57

And if you want to leave any

5:00

other kind of coded messages within your

5:02

reviews, please do. And we'll try and

5:04

discern them and filter out what it

5:06

is that the secret languages. really really

5:08

appreciate it and with that note we'll

5:11

crack on with today's stories. We're going

5:13

to start with a familiar figure Mike

5:15

but in a context that is slightly

5:17

unfamiliar to him and to us you

5:19

found a few stories about Elon Musk

5:22

and his courting of a couple of

5:24

strong men. Yeah, this has been interesting.

5:26

I think, you know, last week we

5:28

tried to avoid talking about Elon, but

5:30

you can only avoid talking about Elon

5:32

for so long before he must enter

5:35

the conversation. I think it's part of

5:37

the, you know, contract he has with

5:39

the world at this point. He must

5:41

be the main character. And so, yeah,

5:43

there are a couple interesting stories, and

5:46

it sort of started with the fact

5:48

that X decided to... file a lawsuit

5:50

in India basically complaining about the nature

5:52

of the content moderation demands from the

5:54

government. And this struck me as interesting

5:57

on a few levels. First being that

5:59

we had spoken about the fact that

6:01

old Twitter had also sued the Indian

6:03

government over its demands and in fact

6:05

that had been sort of an ongoing

6:08

fight and at one point the Indian

6:10

government had raided Twitter's offices in India.

6:12

Nobody was there because I think it

6:14

was in the middle of COVID. and

6:16

there's all this sort of back and

6:19

forth and pressure and when Elon Musk

6:21

had taken over Twitter and declared that

6:23

the old regime was all into censorship

6:25

and stuff and then almost immediately he

6:27

started obeying a bunch of the orders

6:30

coming from the Modi government to take

6:32

down speech of critics of that government

6:34

and we called out the fact that

6:36

hey this looks bad if you're going

6:38

around saying that you're against censorship and

6:40

then you're obeying the censorship commands from

6:43

the government, it raises questions about how

6:45

committed you are to that, especially when

6:47

the former administration that you claimed was

6:49

anti-speech was willing to fight them in

6:51

court. And then there were a bunch

6:54

of stories for the next two years,

6:56

basically, of X being willing to take

6:58

down critics of Modi over and over

7:00

again. There was a documentary. There was

7:02

a few other things along the ways.

7:05

So just the fact that he is

7:07

now stepping up and decided to sue

7:09

is interesting and different. because that's a

7:11

big shift. And as we've seen over

7:13

the last few years, the thing that

7:16

became clear was that mostly, Elon was

7:18

willing to fight when he disagreed publicly

7:20

with the government. So if it was

7:22

a more left-leaning government, then he would

7:24

go on Twitter or X and declare

7:27

himself to be a free speech martyr

7:29

and talk about how, oh, you know,

7:31

the awful Brazilian government is trying to

7:33

get me to take that stuff. But

7:35

when it was a more right leaning...

7:38

authoritarian government, Turkey and India. the sort

7:40

of classic examples, he seemed to be

7:42

willing to go along with it. So

7:44

it was really interesting to see him

7:46

shift in India at this moment too,

7:48

because this was also a moment where

7:51

his other companies are making headway in

7:53

India. Yeah. So before we get into

7:55

that, just took us through the kind

7:57

of actual case that he's bringing against

7:59

the Indian government and what it leads

8:02

into. So this is actually really kind

8:04

of interesting. So Indian internet law. has

8:06

been kind of back and forth over

8:08

the last like decade or so. And

8:10

there had been lawsuits on this and

8:13

I had written about this years ago

8:15

where they have these different IT acts

8:17

that sort of lay out the intermediary

8:19

liability questions around content moderation and for

8:21

a while it actually looked like the

8:24

Indian law was actually going to kind

8:26

of match. Section 230, but then that

8:28

upset some people in the government who

8:30

wanted content to be more easy to

8:32

take down. And so it shifted in

8:35

a pretty drastic way and made it

8:37

so that the government had a lot

8:39

more power to sort of order content

8:41

to be taken down. And so what

8:43

X is now doing is challenging the

8:46

sort of latest version of the law.

8:48

It's Section 69A of the IT Act.

8:50

And they're saying that that violates free

8:52

speech rights because it basically creates a

8:54

way for the government to send information

8:56

to the platforms that they say have

8:59

to be taken down. And the way

9:01

the law is currently being interpreted, and

9:03

again this is after a few different

9:05

court cases and a few different challenges

9:07

and changes, the way it is being

9:10

interpreted is that if the government sends

9:12

you... content that they believe should be

9:14

taken down, that you really will get

9:16

in trouble if you don't. And we

9:18

saw that again with the previous regime

9:21

and Twitter who did try to fight

9:23

this in court and eventually ended up

9:25

losing. Now the government has responded to

9:27

this and I thought this was really

9:29

interesting and this response came out just

9:32

before we started recording, de honest. And

9:34

they're presenting it as a very different

9:36

thing. So in the lawsuit... Elon is

9:38

referring to, or not Elon, but X

9:40

and their lawyers. I'd love it if

9:43

he wrote his own lawsuits. Yeah, that

9:45

would be, that would be quite something.

9:47

But they're referring to it as a

9:49

censorship portal, basically saying the system that

9:51

is set up to have the government's

9:53

and request is a censorship portal, which

9:56

it probably is actually a fairly accurate

9:58

description of it. The government is pushing

10:00

back and saying, no, no, this is

10:02

not a censorship portal. This is just

10:04

a website that allows us to notify

10:07

you of harmful content. And this struck

10:09

me as really interesting because this has

10:11

been the debate that we've had in

10:13

other countries and in particular in the

10:15

US and the whole thing with the

10:18

Twitter files a few years ago when

10:20

Musk took over, which is that Twitter

10:22

has set up. various portals to allow

10:24

governments or government officials to alert them

10:26

to content that they believe might be

10:29

violating the terms of service. And this

10:31

is, there's like a very specific distinction

10:33

in here that is important, which is

10:35

that the US system and, you know,

10:37

the way it works is that certain

10:40

actors have the ability, they have access

10:42

to a portal where they can submit

10:44

stuff, but it still is up to

10:46

Twitter. or the company, you know, to

10:48

decide, does this actually violate our rules?

10:51

The point is, it's like flagging content,

10:53

and it is a more trusted flagger

10:55

because it's coming from the government, it

10:57

will be reviewed in order and determined

10:59

whether or not it actually violates. And

11:01

as we saw with the actual details

11:04

that came out later of the Twitter

11:06

situation, Twitter often would reject those and

11:08

say this doesn't actually violate our terms

11:10

of service and we will reject them.

11:12

And so the Indian government is sort

11:15

of presenting this as the same thing,

11:17

like this is just a way for

11:19

us to alert you. And I think

11:21

they are deliberately sort of mimicking the

11:23

language that was used in the US

11:26

to present it as not as threatening

11:28

or problematic, whereas what X is claiming

11:30

and what I actually think is probably

11:32

more accurate is that under Indian law,

11:34

unlike in the US, when you get

11:37

a request from the government through this

11:39

particular you feel very strongly compelled to

11:41

remove that content. Okay. And so I

11:43

find it interesting. that India is sort

11:45

of now using the language that was

11:48

used to describe the American situation where

11:50

it wasn't censorship, even though some people

11:52

claimed it was censorship, including Elon Musk,

11:54

to now reflect the situation in India,

11:56

where there is much more clear government

11:59

coercion as part of this process. And

12:01

now we'll see how it fights out

12:03

in the courts, though it is very

12:05

strange that at the time when Elon

12:07

is seen as fairly close with Modi

12:09

and has been doing deals for his

12:12

other businesses including both Tesla and Starlink

12:14

for him to suddenly decide that this

12:16

is a fight with fighting. Yeah that's

12:18

what's really interesting is that Elon Musk

12:20

and Modi met in Washington in February

12:23

and I remember the photos there's a

12:25

bunch of photos that were published of

12:27

him with Modi and some of his

12:29

kids as well. It was a very

12:31

kind of odd photo op, I don't

12:34

know what they were trying to achieve,

12:36

but clearly a kind of photo moment,

12:38

and then as you say, he's trying

12:40

to kind of expand his business interest

12:42

there. Do you think that this court

12:45

case is being used as leverage in

12:47

the push to expand into India with

12:49

Tesla and increase the market share of

12:51

starting? I have no idea. I mean,

12:53

it sounded like he was getting those

12:56

deals anyway, so I'm not entirely sure

12:58

that this is beneficial. So I'm honestly

13:00

a little confused by it. I'm wondering

13:02

if more information will come out at

13:04

some point, that something else came up

13:07

within this process. It strikes me as

13:09

a weird way to have leverage because...

13:11

I'm sure Modi is the stronger player

13:13

in this situation. You know, Musk needs

13:15

access to those markets for his other

13:17

companies. So it strikes me strange, just

13:20

also the fact that, you know, for

13:22

the last two years, he's been willing

13:24

to go along with Modi's demand. So

13:26

I'm not sure what pushed him over

13:28

the edge here. And in fact, that

13:31

leads into one of the other stories

13:33

that we did want to mention here,

13:35

which is that in Turkey, which is

13:37

the other place where Musk has shown

13:39

a willingness to roll over and do

13:42

what was demanded, he's been taking down

13:44

the accounts of various activists and opponents

13:46

to the Erdogan government. And so... Again,

13:48

we've seen it over and over again

13:50

when Moscow is in agreement and aligned

13:53

with the governments of these authoritarian countries

13:55

He seems to have no problem pulling

13:57

down Content that the government is criticizing

13:59

and he did so again this week

14:01

in Turkey But in India suddenly he's

14:04

challenging it. So it's it's a very

14:06

strange situation and I really have no

14:08

idea why yeah, it's funny in a

14:10

way that our main story this week

14:12

is the fact that Elon Musk doesn't

14:15

have a consistent approach to speech. It's

14:17

a story that isn't a story in

14:19

a way. But yes, there's enough there

14:21

I think to kind of bring it

14:23

to listeners and explain that Musk continues

14:25

to be very hard to predict when

14:28

it comes to understanding why he says

14:30

one thing and does something else. And

14:32

not only that by why that seems

14:34

to change from almost month to month.

14:36

Yeah, I mean there's clearly no consistency

14:39

and so I'm sure there's some other

14:41

reason why X suddenly decided it was

14:43

worth fighting this. But this one doesn't

14:45

seem to fit the same pattern, where

14:47

you can sort of easily slide it

14:50

into, well, he likes this government, he

14:52

doesn't like that government, or he needs

14:54

some other thing here. This one is

14:56

just surprising. I almost wouldn't be surprised

14:58

if this lawsuit gets dropped very quickly,

15:01

if because of other back channel discussions,

15:03

Musk is like, hey guys, knock it

15:05

off. I mean, maybe is it that

15:07

he was distracted because he's running the

15:09

US government and wasn't the one to

15:12

make this decision, and then once this

15:14

gets back to him, he'll change his

15:16

mind. I don't know, but it is

15:18

a slightly surprising development. Yeah. It's not

15:20

the only story in which he appears

15:22

this week, Mike, and we both spotted,

15:25

as is always the case, we both

15:27

spotted a story in the verge this

15:29

week that told of Musk back channeling,

15:31

again, the CEO of Redit, Steve Huffman,

15:33

about some issues that he had about

15:36

moderators on Redit taking down and blocking

15:38

links to X. And so this is

15:40

a kind of story that picks up

15:42

from last year but has been recently

15:44

reported this week about how the essentially

15:47

these two CEOs were texting each other

15:49

probably sat back on the out, chilling

15:51

on a, after a long day of

15:53

thinking about how they were going to

15:55

take over the US government. And he

15:58

was kind of complaining about this development

16:00

and subsequently Huffman banned some of the

16:02

sub-readits, deleted all of the comments. And

16:04

so what did you make of this

16:06

as an indicator of musks? Again, inconsistent

16:09

approach to materials. it's yet another example

16:11

of his pure hypocrisy, right? And so,

16:13

you know, Huffman and Musk have been

16:15

friendly for a while, and Huffman in

16:17

the past has clearly been inspired by

16:20

some of Musk's actions and sort of,

16:22

you know, freeing up other tech CEOs

16:24

to be a little more aggressive in

16:26

their viewpoints. But this is just crazy,

16:28

right? Because we know that Elon and

16:30

X have blocked links to all sorts

16:33

of competitors for various reasons. Sometimes for

16:35

a while, right? They were slowing down

16:37

links to sub stack for a while.

16:39

They were blocking links to sub stack

16:41

for a while. They were blocking links

16:44

to Mastodon. They were closing accounts of

16:46

people. Overall, X has completely downranked links

16:48

because Musk wants to keep people on

16:50

the platform. And so he's admitted. He

16:52

finally admitted it. People sort of recognized

16:55

it. That post with links don't get.

16:57

rated as highly in the algorithm. You

16:59

know, so he's clearly done things to

17:01

try and keep people within his platform

17:03

to then go and complain to Huffman.

17:06

that a few subredits had decided as

17:08

a kind of protest that they were

17:10

no longer going to include X links

17:12

or links to X would no longer

17:14

be allowed in those subredits that this

17:17

was some sort of major breach that

17:19

needed often to step in. It suggests

17:21

a level of hypocrisy which is not

17:23

uncommon with Elon Musk, but it does

17:25

seem notable. The other element of it

17:28

was it wasn't just the blocking of

17:30

links to X that he was concerned

17:32

about. He was also concerned about people

17:34

calling out doge employees, the various kids

17:36

that Elon has brought into the federal

17:38

government or working havoc all over the

17:41

place. And he was upset that some

17:43

of those the people were being named

17:45

or talked about. There's a claim that

17:47

it was like advocating violence against them,

17:49

though I think that was somewhat exaggerated.

17:52

People are saying like even naming them

17:54

is advocating violence against them, which is

17:56

not accurate. And so he seemed to

17:58

be partly upset about that. And then

18:00

as part of that discussion also apparently

18:03

upset about some of the subredits and

18:05

the moderators within the subredits deciding that

18:07

they weren't going to allow links to

18:09

the former Twitter. Yeah, it does make

18:11

me think that I should be throwing

18:14

my weight around via text a lot

18:16

more. Apparently, that's the way. I never

18:18

really think of it as a weapon

18:20

in my armory, but it's just making

18:22

me think that there's maybe some people

18:25

that I can make me do something

18:27

for me via text. Well, see if

18:29

you can get into the chat with

18:31

the officials from the US government, you

18:33

start throwing your weight around there. that

18:36

would make for a good podcast next

18:38

week. Yeah, telling JD Vans to shut

18:40

up would be a start. Yeah, bring

18:42

me a coffee, JD. So we're going

18:44

to talk a bit more about CEOs

18:46

of platforms now because I've kind of

18:49

been listening to a couple of podcasts

18:51

with some CEOs of platforms this week

18:53

and there's a really interesting difference in

18:55

how they talk about content moderation. Let's

18:57

start with Evan Spiegel, the CEO of

19:00

Snapchat. who was on diary of a

19:02

CEO with Stephen Bartlett this week and

19:04

talked at length about a whole range

19:06

of different issues with a little segment

19:08

about both Snapchat's approach to content moderation

19:11

and also other kind of tangential issues

19:13

like how meta is thinking about content

19:15

moderation as well and We're going to

19:17

play a bit of a clip for

19:19

you. This is kind of technical wizardry

19:22

that we haven't tried on control or

19:24

speech before, but you're now going to

19:26

be able to hear a little bit

19:28

of that interview with Stephen Bartlett, because

19:30

I think it's a really good response

19:33

and a really good interview around some

19:35

of the issues we talk about here

19:37

all the time on control or speech.

19:39

And there's a great example, I think,

19:41

of a CEO who gets it to

19:44

a large degree, so have a listen.

19:46

really matters right and that's why we

19:48

have content guidelines because we want people

19:50

to feel like they're an environment where

19:52

they can express themselves and I think

19:54

some of the the conversation about different

19:57

content guidelines or having content guidelines or

19:59

not having them has been really interesting

20:01

because I think people are missing the

20:03

broader point if you have a platform

20:05

with no content guidelines and it's full

20:08

of people yelling at each other or

20:10

saying really mean or offensive things or

20:12

posting a lot of pornography that's a

20:14

really uncomfortable thing for most people. Right?

20:16

That's uncomfortable. You say, maybe this platform

20:19

isn't for me. Maybe I don't feel

20:21

comfortable expressing myself here because all the

20:23

stuff I'm seeing isn't really appropriate or

20:25

aligned with my values. And so one

20:27

of the things we discovered really early

20:30

on is if you want to create

20:32

a platform where people feel comfortable expressing

20:34

themselves, feel comfortable communicating with their friends

20:36

and family, having content guidelines is really

20:38

helpful because it means that the content

20:41

experience is one that feels more comfortable.

20:43

But isn't that. People would say, well,

20:45

that censorship. I'm thinking now of the

20:47

video that Mark Zuckerberg released about matters

20:49

changed to their moderation systems, moving to

20:51

Texas, realizing that, I think he said

20:54

that they'd over-indexed with their moderators in

20:56

terms of left-leaning politics. So a lot

20:58

of the right leaning content had been

21:00

censored. What do you make of that

21:02

argument for content moderation? That we don't

21:05

want to censor people. I think it's

21:07

a misunderstanding of the First Amendment and

21:09

how it applies. If we look at

21:11

our country, the way, you know, at

21:13

least here in the United States, with

21:16

the First Amendment, that really focuses on

21:18

the way that the government interacts with

21:20

content creators or content publishers. And it

21:22

says, hey, it's not okay for the

21:24

government to interfere with individuals or publishers'

21:27

self-expression, right? That's not allowed. But one

21:29

of the things the First Amendment also

21:31

does is say... you know, platforms or

21:33

individuals can make choices about what sort

21:35

of content they want to promote or

21:38

want to have on their platform. That's

21:40

part of the First Amendment. You can't

21:42

force the Wall Street Journal to, you

21:44

know, put this article or that article

21:46

or accept any article from any author

21:49

all around the world. The Wall Street

21:51

Journal as a paper can decide what,

21:53

you know, what authors. it wants to

21:55

include on its pages, and that's part

21:57

of the protected First Amendment expression we

21:59

have here in this country. So this

22:02

whole notion of censorship doesn't apply to

22:04

companies that are private businesses that actually

22:06

have a First Amendment right to decide.

22:08

what content is on their platform. And

22:10

they may want to decide we're open

22:13

to literally anything. Anything goes, no problem.

22:15

And it seems like some platforms are

22:17

making that choice. But other platforms like

22:19

ours say, hey, in order to have

22:21

a healthy set of discourse across our

22:24

platform, in order to make sure people

22:26

feel comfortable when they're viewing content on

22:28

our platform, we don't want people to

22:30

come across pornography, for example, or violent

22:32

content, or hateful content. That's not something

22:35

that makes people feel good. want to

22:37

make sure that that content isn't on

22:39

our platform because it doesn't comply with

22:41

our our guidelines. And that may be

22:43

one of the reasons why in some

22:46

of these studies it shows that people

22:48

feel better when they use Snapchat because

22:50

they're not encountering you know really violent

22:52

content when they're using Snapchat. So listens

22:54

have heard Evan speak there about how

22:57

Snapchat approaches content moderation Mike. I mean

22:59

I think his response to Stephen Bartler

23:01

around the First Amendment was particularly interesting

23:03

and I wanted to note the fact

23:05

that Bartler has said a bunch of

23:07

things in the recent memory around content

23:10

moderation in relation to meta. So don't

23:12

if you remember, he posted on his

23:14

LinkedIn page, which has many, many millions

23:16

of followers, about the fact that meta's

23:18

move away from fact-checking and its changes

23:21

to its content moderation policy, represented one

23:23

of the most important videos that people

23:25

will see this year and a course

23:27

correction to... what he seemed to suggest

23:29

that it was slightly coded, but seems

23:32

to suggest was kind of overreach around

23:34

contour moderation. And I've listened to a

23:36

few kind of Stephen Ballet podcasts. I

23:38

don't love the guy, but he has

23:40

a slight tendency to kind of air

23:43

into the manosphere, I find. And so

23:45

it's interesting that he quizzes Spiekel quite

23:47

openly about his contour moderation. And the

23:49

speaker has a really good response, I

23:51

felt. Yeah. Were you surprised by how

23:54

well he handled that and the way

23:56

he seemed to understood it? Yeah, I

23:58

mean, partly, and partly not, I think.

24:00

Partly not because it was a great

24:02

answer. I mean, it's an absolutely fantastic

24:05

and very thoughtful and correct answer, understanding

24:07

the things, that this is not a

24:09

First Amendment issue, that values determine what

24:11

kind of community you want to build,

24:13

and that is what users appreciate as

24:15

well, and that there are reasons to

24:18

do this that have nothing to do

24:20

with censorship, but just what kind of

24:22

community you're trying to build. I think

24:24

it was a fantastic answer. Spiekel's been

24:26

the founder and CEO of Snapchat for

24:29

a while. He's gone through a bunch

24:31

of these fights and arguments and they've

24:33

been involved in some of them and

24:35

I feel like he has a really

24:37

deep grasp. So I'm not surprised in

24:40

that he gets it right. The only

24:42

thing I'm surprised in is that like

24:44

feels like every other CEO in the

24:46

deck space no longer does. Yeah. And

24:48

if anything like Spiekel had the reputation

24:51

historically and this is probably unfair that

24:53

he was you know, he was a

24:55

little bit more of like a frat

24:57

boy not really deep in the policy

24:59

weeds on these things. And yet this

25:02

answer suggests someone who's really thought deeply

25:04

about these things and actually has a

25:06

deeper understanding of it and is willing

25:08

to explain it clearly and not, you

25:10

know, what a lot of CEOs do

25:13

is kind of deflect and mislead and

25:15

sort of dance around it. And he

25:17

was just very direct. He's just like,

25:19

this is not a First Amendment issue.

25:21

It's a values thing. We want to

25:23

build a community. This is what our

25:26

people expect. This is what our users

25:28

expect. And this is the kind of

25:30

thing that we've decided that this is

25:32

what our values are based on. And

25:34

it was, you know, it's fantastic and

25:37

clear and appreciable. And you can hear,

25:39

I really have, you know, I think

25:41

I've maybe heard of Bartlett, but I've

25:43

never seen any of his videos before.

25:45

I was not really familiar with him.

25:48

And I only watched this one really

25:50

familiar with him. And I only watched

25:52

this one little section, you know, a

25:54

little bit longer than the clip that

25:56

we played of him talking about the

25:59

content moderation stuff. were taking down too

26:01

much conservative speech and they had to

26:03

move to Texas for it, which is

26:05

like we know is not true, but

26:07

he seemed really bought into the narrative

26:10

of what happened and so it was

26:12

really nice to see Spiegel just kind

26:14

of push back on him. Yeah, and

26:16

Bala has a kind of Zuckerberg aesthetic

26:18

doesn't he? He's got the kind of

26:21

black t-shirt, he doesn't quite have the

26:23

gold chain, but you know he needs

26:25

the Latin phrase on the t-shirt. Yeah,

26:27

exactly, exactly. And Bala has been kind

26:29

of had some focus on focus on

26:31

him focus on him focus on him

26:34

for... showcasing and highlighting kind of health

26:36

misinformation on the podcast as well. So

26:38

he's incredibly, in the UK at least

26:40

he's incredibly well known, he's incredibly well

26:42

listened to, he's got various books out,

26:45

and he does seem to kind of,

26:47

I would say, spotlight, some slightly odd

26:49

health experts, inverted commerce. And so, again,

26:51

you might expect Spiegel or anybody on

26:53

the podcast to somewhat side with him.

26:56

and actually I thought Speaker did a

26:58

great job of kind of standing his

27:00

ground. One guy who didn't do a

27:02

very good job of that was our

27:04

next CEO who was on a podcast

27:07

this week, Neil Mohan of YouTube. He

27:09

was on the Semaphore podcast Mix Signals

27:11

with Ben Smith and Max Tani and

27:13

we're going to play a bit of

27:15

a clip now of how he was

27:18

responded to some questions about YouTubes. quantum

27:20

moderation policy and some of the tensions

27:22

around the US administration and the kind

27:24

of clash in ideals, you'll notice I

27:26

think in this clip a very different

27:28

sound, a very different tone and he's

27:31

not only defensive I'd say but also

27:33

quite evasive in his answers. You've said

27:35

that the number one priority for YouTube

27:37

is the safety of YouTube's ecosystem and

27:39

we're in a moment when that's actually

27:42

like a slightly unusual thing to say

27:44

and a lot of platforms are really

27:46

backing off. anything like content moderation probably

27:48

because of pressure from this this White

27:50

House and this administration and I wonder

27:53

if you feel like your there's tension

27:55

between you and the in the administration

27:57

on I guess particularly issues around public

27:59

health. I'll say a

28:01

few things. First, and probably most important

28:04

and kind of really at the top

28:06

is everything that we talk about, everything

28:08

we just talked about in terms of

28:10

the business, how content works on our

28:13

platform, etc. is back to our mission

28:15

statement, which is to give everyone a

28:17

voice and show them the world. And

28:19

the first half of that mission statement

28:22

is really about free expression and freedom

28:24

of speech and I can say for

28:26

myself and I know for many of

28:28

the colleagues that I work with every

28:31

single day that's why we come to

28:33

work like that's the power of YouTube

28:35

right like that if you have an

28:37

idea you have a thought and you

28:39

want to share it with the world

28:42

then YouTube is a place where you

28:44

can go and share it without somebody

28:46

telling you that you don't sound the

28:48

right way or you don't look the

28:51

right way or you're saying the wrong

28:53

thing or what have you and that

28:55

is core to our mission and everything

28:57

that we do is ultimately, frankly, in

29:00

service of that. And so it's the

29:02

reason why actually I think we've had

29:04

community guidelines from the very early days.

29:06

And in order to allow creators to

29:09

be able to share their ideas have

29:11

this free sort of voice freedom of

29:13

expression and to earn a sustainable living

29:15

from it we also have rules of

29:18

the road in terms of how our

29:20

platform works right like no porn or

29:22

adult content or financial scams or what

29:24

have you right like back to the

29:27

question around like when you turn on

29:29

the TV like that's not what consumers

29:31

are looking for when they turn it

29:33

on and our advertisers right the brands

29:36

that support that content aren't looking for.

29:38

And so our approach to responsibility is

29:40

with all of that in mind, right?

29:42

But ultimately towards his goal of freedom

29:45

of expression, that's how I've always looked

29:47

at it. You know, and even in

29:49

years past when... You and I've talked

29:51

about it, hopefully I've been consistent in

29:54

terms of that sort of core thesis.

29:56

So yeah, Mike, what did you think

29:58

about how Mohan responded to Ben Smith's

30:00

kind of very pointed questions? Yeah, I

30:03

mean this is the more typical unfortunately

30:05

the more typical CEO's response that we

30:07

hear on these kinds of questions Where

30:09

someone doesn't want to come out and

30:12

say anything that then will be taken

30:14

up by you know Probably a bunch

30:16

of idiots on various social media platforms

30:18

taken out of context and presented as

30:21

something to rally around and so he

30:23

says a lot of nothing and he

30:25

does so in a very sort of

30:27

defensive way and doesn't directly address the

30:30

issue and he could have right I

30:32

mean it would have been great to

30:34

have him come out and respond the

30:36

same way that does. And so, you

30:39

know, I think this, the contrast between

30:41

Mohan and Spiegel is really, really notable.

30:43

And it's, you know, it's unfortunate for

30:45

Mohan that they both came out around

30:48

the same time, but it's just such

30:50

a different answer to effectively the same

30:52

question. Yeah. And also I noted the

30:54

fact that Mohan's responses have changed significantly

30:57

in the last two to three years

30:59

as well. It was only, I think,

31:01

probably at that time made sense. You

31:03

can tell from the clip that he,

31:06

that idea is dead. You know, is

31:08

very much in a platformer in which

31:10

he talked about working with other partners

31:12

in the space, civil society organizations, non-profits,

31:15

kind of experts to help remove borderline

31:17

content. And that, probably at that time,

31:19

made sense. You can tell from the

31:21

clip that he, that idea is dead.

31:24

You know, is very much just doing

31:26

the bare minimum and no more. Well,

31:28

I think it's in some sense it's

31:30

even worse than that, right? Because this

31:33

kind of answer, it's a not trying

31:35

to say anything answer, it's an answer

31:37

that's designed to try not to get

31:39

anyone upset by not actually saying anything.

31:42

And it's a lost opportunity. It's an

31:44

opportunity where he could come out and

31:46

say the same things that Spiegel said,

31:48

which is that it's our place and

31:51

we get to determine how things are.

31:53

And like, yes, we're trying to enable

31:55

free speech. But one of the best

31:57

ways to do that is to have

32:00

a set up that reflects values and

32:02

that people don't feel harassed or feel

32:04

that there's misinformation flowing there and he

32:06

didn't say that. And so I think

32:09

it's a lost opportunity. But maybe an

32:11

opportunity gained in the sense that you

32:13

can now send this to folks in

32:15

the Republican Party in the US and

32:18

justify, and justify like. I've said publicly,

32:20

something that you agree with, and therefore...

32:22

But it's not, right? I mean, it's

32:24

not even, it's not even saying that,

32:27

right? It's not even saying what the

32:29

Republicans want them to say. It's saying

32:31

nothing. That's the problem, right? You know,

32:33

I mean, with Zuckerberg at least, I

32:36

mean, okay, so yes, you could say

32:38

that like this isn't like the complete

32:40

capitulation that Zuckerberg going on Rogan and

32:42

saying a bunch of nonsense was like

32:44

just obviously... completely ridiculous untrue fantasy land

32:47

stuff that the Republicans got excited about.

32:49

Nobody's going to get excited about this.

32:51

What he said was nothing. It was

32:53

empty. Yeah, but sometimes the bare minimum

32:56

is all these guys want to do,

32:58

isn't it? Yeah, I mean, sure, he

33:00

can point to it. But I don't

33:02

think this would satisfy anyone on any

33:05

side of this debate because it's not

33:07

the full-throated endorsement of any particular position.

33:09

It is clearly like trying to tiptoe

33:11

around landmines. Do you think that... the

33:14

US administration will have the potential to

33:16

go after Snapchat or put pressure on

33:18

YouTube because they're not saying what they

33:20

would like them to you. They're not

33:23

kind of towing the line or... We'll

33:25

see. I mean you never know who

33:27

the next target is going to be.

33:29

Right. Obviously there are investigations going on

33:32

now with the FTC that we've talked

33:34

about where they're, you know, they want

33:36

to go after tech but we don't

33:38

know who it is they're going after

33:41

because most of the tech companies have

33:43

sort of towed the But I don't

33:45

know, they haven't really been, you could

33:47

see where it's like, it's been really

33:50

easy because Snapchat is often considered one

33:52

of the ones that kids use. And

33:54

so whenever they're kids safety discussions, Snapchat

33:56

will often come up. So I could

33:59

totally see a kind of moral panic.

34:01

around Snapchat and they'll say that they're

34:03

not handling kids well and that they'll

34:05

do stuff around that but I don't

34:08

know it's impossible to predict with this

34:10

administration. Yeah nonetheless I think it's interesting

34:12

to see two CEOs of two major

34:14

platforms talking about content moderation in the

34:17

same world and it does have the

34:19

sense of these platforms trying to shape

34:21

the narrative and be on the front

34:23

foot with how they talk about this

34:26

topic. Speagle talks about proactively scanning for

34:28

pornography. in a way that Spotify on

34:30

our next story might have something to

34:32

learn from. You found this story about

34:35

Spotify being unprepared, let's say, for some

34:37

graphic content on the platform. Yeah, this

34:39

is almost hilarious, right? So anyone who's

34:41

been in this space for anyone at

34:44

the time knows that if you do

34:46

any kind of user-generated content, like at

34:48

some point you're going to have to

34:50

deal with pornographic content and have a

34:53

clear policy and a way to enforce

34:55

it. That is even true with text,

34:57

but as soon as you get to

34:59

video or imagery, you know, so Spotify

35:02

has always been audio for the most

35:04

part, and music, and so that was

35:06

less of an issue. They've gotten bigger

35:08

into podcasts, and there's been some controversy

35:11

there. But now, sort of realizing how

35:13

much stuff is video and how many

35:15

podcasts are now video, they've sort of

35:17

moved into the video space. And apparently,

35:20

even though their policies are... that they

35:22

will not allow pornography. They were unprepared

35:24

for sexually explicit material suddenly showing up

35:26

and getting very very popular. And so

35:29

they had the top business podcast. their

35:31

listing of top business podcasts apparently included

35:33

some fairly pornographic material, which is not

35:35

normally what I associate with business content.

35:38

And they were sort of taken by

35:40

surprise and had to respond and said,

35:42

oh, of course, you know, it wasn't,

35:44

we didn't intend for that and that

35:47

violated the rules and they eventually took

35:49

it down once it was called out.

35:51

But it suggests that they may have

35:53

moved into the sort of. video market

35:56

without preparing their trust and safety folks

35:58

for the level of pornography and the

36:00

ways that people are going to attack

36:02

things like the trending lists and the

36:05

top lists. Yeah, indeed. And it's surprising

36:07

because it wasn't that long ago that

36:09

they were embroiled in the kind of

36:11

Joe Rogan scandal and they faced a

36:14

whole bunch of heat for their trust

36:16

and safety approach, I would say. They

36:18

then bought a... company called Kinsen, which

36:20

actually did some smart work to identify

36:23

misinformation and kind of false narratives deep

36:25

within podcasts and you know kind of

36:27

often hidden in far-reaching corners of the

36:29

platform. Full disclosure I did some work

36:32

for Kinsen at some point before they

36:34

were acquired by Spotify. That's how I

36:36

know about them. tooling up or maybe

36:38

skilling up around. Well it's been something

36:41

that has been known for a long

36:43

time. Everybody knows. You don't want to

36:45

make too big of a deal of

36:47

it, right? Because like, again, content moderation

36:49

scale is impossible, right? There's always going

36:52

to be something that slips through, people

36:54

are going to make mistakes, things are

36:56

going to get missed. So I don't

36:58

want to make too big of a

37:01

deal of it, but it is noteworthy

37:03

that like, you would think any platform

37:05

as they're expanding into video and pushing

37:07

video heavily, they have to realize. And

37:10

in fact, the video that sort of

37:12

made it, sort of made it, somebody

37:14

just ripped it, put it in there,

37:16

and was able to get a bunch

37:19

of downloads. But the fact that it

37:21

made it into the top business podcast

37:23

lists, you would feel that there would

37:25

be a little bit of extra review

37:28

before something gets to that level, and

37:30

it's just sort of noteworthy that it

37:32

appears that at least this particular attempt

37:34

to get pornography witnessed on Spotify, made

37:37

it past the guards. Did Hector ever

37:39

have a porn problem? In the comments

37:41

Not a bad one. I mean, we

37:43

certainly had the issue of spam, right?

37:46

And so, and a lot of spam

37:48

is sort of like linking to pornographic

37:50

content. And so we definitely have had

37:52

that. We did have, at one point,

37:55

we had this weird thing where somebody

37:57

was showing up and basically writing a

37:59

novel in the comments to a very

38:01

old tector post. The comments are probably

38:04

still there and it was like, I

38:06

mean, just reams and reams of text.

38:08

Well, I had nothing to do. Like

38:10

an erotic novel novel. Yeah, yeah, it

38:13

may still be somewhere in the textured

38:15

archives Because I don't think we pulled

38:17

it down. It was like it was

38:19

on like a really old post so

38:22

like nobody was reading it. It wasn't

38:24

interfering with anyone. Yeah, and so I

38:26

can't remember if we left it up

38:28

or if we pulled it down. It

38:31

was, and they were like coming back

38:33

every show often and just like adding

38:35

another chapter. It was massive. It was

38:37

massive. We've had some weird things happen

38:40

in the comments over time, especially like

38:42

older comments. Yeah. That just go back

38:44

ages and people put in all sorts

38:46

of weird stuff. Maybe we'll do a

38:49

special recording of the podcast and if

38:51

we read out the erotic novel written

38:53

on it in the comments under the

38:55

detector post decades ago. I don't even

38:58

know if I can find it find

39:00

it find it again. We're looking recently,

39:02

we're at somewhere at 82, 83,000 articles

39:04

on Tecta at this point. Wow. And

39:07

so, yeah, even finding that. And over

39:09

two million comments, so it's, we've got

39:11

a fair bit of content. Yeah, good

39:13

luck finding that. It's worth saying that

39:16

Mike and I have been thinking about

39:18

accompanying the audio version of controller speech

39:20

with a video version. And... We have

39:22

been arming and arming about it. We

39:25

would love to hear- We need to

39:27

get to the top of the business

39:29

list on Spotify. Seems like we have

39:31

a way to do that. But yeah,

39:34

if listeners, we'll keep it clean, I

39:36

promise. If listeners have a view on

39:38

whether they would listen and watch a

39:40

control or speech podcast on the platform

39:43

of their choice, drop us a no,

39:45

podcast at control speech.com, CTRL, AOT speech.

39:47

And give us your thoughts. It might

39:49

spur us on to do a version

39:52

where we have both of our big

39:54

heads on a screen together. Talking of

39:56

adult sites, the kind of other story

39:58

that I noted this week was one

40:01

about only fans and it's in relation

40:03

to off-com. We're not starting an only

40:05

fans, but revenue diversification, Mike. We're going

40:07

to make this podcast pay somehow. You

40:10

promised. Offcom has handed a one million

40:12

pound fine to Only Fans for failing

40:14

to supply accurate information about how it

40:16

prevents underage users from accessing explicit content.

40:19

And this is a long-standing investigation which

40:21

has kind of finally come around this

40:23

week. It's a bit of a gaf

40:25

really. Only fans had told Offcom that

40:28

it's challenge age. So the age of

40:30

which it prompts users to tell them

40:32

how old it is, was 23, only

40:34

to find out from the tech provider

40:37

that provided that service, that it was

40:39

actually set to 20. So there was

40:41

a kind of gap of three years

40:43

between what it told off-com and what

40:46

was actually true. And that has led

40:48

it to be given and to accept

40:50

a one million pound fine. That in

40:52

itself is interesting. But I think what's...

40:54

really fascinating to me Mike is the

40:57

release of this story and the timing

40:59

of it okay so last week we

41:01

talked about the online safety act the

41:03

brand new but long in the making

41:06

legislation in the UK finally being rolled

41:08

out and intermediaries in the UK being

41:10

liable for the USA only a week

41:12

into that being true do we have

41:15

this announcement and The announcement is actually

41:17

not related to the online safety act

41:19

at all. It's in relation to a

41:21

regulation that predates the OSA. It was

41:24

in existence before the OSA came to

41:26

being. So basically Offcom could have brought

41:28

this and did obviously bring it against

41:30

only fans at any point. But it

41:33

waited until the OSA had been rolled

41:35

out. I think to potentially give the

41:37

impression of off-com kind of doing its

41:39

job and being the kind of enforcement

41:42

power that it wants to be seen

41:44

to be and I read around the

41:46

reports about this story the Guardian the

41:48

FT and others don't mention the OSA

41:51

or the kind of more niche regulation

41:53

that this enforcement is brought under so

41:55

it kind of gives the impression to

41:57

the unsuspecting eye. that actually this is

42:00

related to the OSA and the offcom

42:02

is suddenly kind of doing its job.

42:04

So the kind of cynical journalist in

42:06

me thinks that this has been coincided

42:09

very well with the OSA last week.

42:11

It's actually nothing to do with that.

42:13

But I have heard on the grapevine

42:15

that there is some enforcement being prepared

42:18

around the OSA and there's some naturally,

42:20

as we've talked about on the podcast,

42:22

there are some platforms that have been

42:24

closely looked at. What did you make

42:27

of the kind of timing of this?

42:29

Are you cynical as I am? Yeah,

42:31

I mean, I don't know. The timing

42:33

might be right just based on like

42:36

apparently only fans alerted off-com to the

42:38

error in January of 2024, which you

42:40

know, a little over a year ago,

42:42

there was the investigation in the back

42:45

and forth and then sort of figuring

42:47

out what it was going to be.

42:49

The timing seems about right. I mean,

42:51

having it come out now is not

42:54

like totally out of the ordinary. Under

42:56

that, look at you being all friendly

42:58

to off home. What have I made

43:00

you do? But yes, I mean it

43:03

is entirely possible that the exact timing

43:05

of the release may have been let's

43:07

say push back a few weeks or

43:09

a month or something Recognizing that the

43:12

OSA was about to go into effect

43:14

and that everybody would be looking to

43:16

offcom and to see how and when

43:18

they actually started enforcing things under the

43:21

OSA and so you know It wouldn't

43:23

surprise me if the timing was massaged

43:25

in some way to make this work

43:27

But, you know, we'll see when the

43:30

actual... enforcement's come out, but yeah, it'll

43:32

be interesting to see and to see

43:34

whether or not this has any impact.

43:36

And if people think like, oh, okay,

43:39

offcom is actually trying to enforce stuff,

43:41

it's possible. I do wonder, I mean,

43:43

you could argue to that offcom was.

43:45

Man, I'm going to defend off-com again.

43:48

They might reasonably have been concerned that

43:50

if they had announced this, you know,

43:52

three weeks ago, that it would confuse

43:54

people into because like it was pre-OSA.

43:57

And so it's like, well, wait, I

43:59

thought this law isn't going into effect

44:01

for two more weeks. So, you know,

44:03

why are they enforcing it now? And

44:06

so there may be some reasons where

44:08

it actually did make sense just to

44:10

keep everybody else from being too confused

44:12

by it. the way that the coverage

44:15

has transpired and I know it's difficult

44:17

for journalists to necessarily know every nut

44:19

and bolt of every single piece of

44:21

legislation and what it refers to but

44:24

it was something that pricked my ears

44:26

at least. Let's round up Mike on

44:28

a slightly kind of more interesting quirky

44:30

story that you found a platform that

44:33

we may end up referring to in

44:35

a future episode of the podcast in

44:37

the opening section. We have to see

44:39

what the prompt is. Tell us says

44:42

EZ US. is this new platform. I

44:44

had heard about it, I think late

44:46

last year, there was some talk about

44:48

it, and it is a sort of

44:51

another Twitter-like platform. It was created by

44:53

Joe Trippy, who is a sort of

44:55

semi-famous political consultant figure, sort of became

44:57

famous in 2004 as the campaign manager

44:59

for Howard Dean and his sort of

45:02

upstart internet-fueled campaign. And then ever since

45:04

then has been sort of in and

45:06

around specifically democratic politics in the US.

45:08

And so late last year, there was

45:11

some talk about how he wanted to

45:13

set up his own social media platform

45:15

and he wanted to do something different

45:17

and it was going to be more

45:20

respectful and respectful. to some extent like

45:22

we've heard all that before you know

45:24

like lots of people said that especially

45:26

after either on must took over it

45:29

was this idea that like oh and

45:31

usually presented in a way where you're

45:33

just like, wow, this person is incredibly

45:35

naive about the realities of human beings,

45:38

and let alone getting a bunch of

45:40

them together. But there do seem to

45:42

be, as the app is now officially

45:44

launched, they do seem to have created

45:47

a few interesting elements to it that

45:49

I think will be worth seeing how

45:51

well they catch on. And so in

45:53

particular, rather than just relying on like

45:56

a... team of trust and safety officials

45:58

to determine who is violating the rules

46:00

and who isn't, there is an element

46:02

of sort of crowdsourcing stuff where they've

46:05

built in, they call it a reputation

46:07

engine that so users themselves get to

46:09

rate other people's posts. And so it's

46:11

sort of a mix of like community

46:14

notes and read it up votes and

46:16

down votes and even a little bit

46:18

of like Wikipedia elements to it where

46:20

it's like. sort of crowdsourcing reputation, and

46:23

the idea being that users who have

46:25

a high score, their content rises to

46:27

the top, users that have a low

46:29

score, their content will be not as

46:32

readily, it'll still be there, but it

46:34

will work into algorithms or be as

46:36

readily shared and viewable. And, you know,

46:38

it's an interesting idea. I have questions

46:41

about how well it'll work in practice.

46:43

You know, as soon as you get

46:45

into this kind of thing, you worry

46:47

about things about brigating and deliberative attacks

46:50

on certain kinds of speech and how

46:52

it can be abused and how, I

46:54

mean, there's always fears about things like

46:56

echo chambers and stuff which I think

46:59

might be a little bit overblown. But

47:01

it's an interesting and different approach. And

47:03

right now, the one thing that I

47:05

do believe very strongly, is that the

47:08

more experiments the better. And so I'm

47:10

happy to see an experiment. I'm happy

47:12

to see how it works. I might

47:14

be a little skeptical that this will

47:17

work out as well as they sort

47:19

of think it will. But again, what

47:21

we need right now is experiments, we

47:23

need differentiation, we need people to try

47:26

different things. And so I'm excited to

47:28

see them enter the space. Yeah. I

47:30

think it's an interesting feature and an

47:32

interesting idea that might kind of inversely

47:35

affect how platforms moderate and act as

47:37

a... way of slowing down harmful or

47:39

egregious content. It kind of made me

47:41

think a bit about the challenges of

47:44

breaking through on a platform like that

47:46

though. If somebody has a reputation and

47:48

they have built a reputation on a

47:50

platform, does that mean that a kind

47:53

of other voices can emerge and content

47:55

can emerge in a way that... You

47:57

definitely have a fear of like this

47:59

becomes a sort of winner-take-all situation and

48:02

the people with the most clout and

48:04

the most power sort of stay that

48:06

way. You have a little bit of

48:08

that no matter what on any social

48:11

media platform. Obviously people with bigger audiences

48:13

just have bigger audiences as this is

48:15

the natural way things are. But yeah,

48:17

there is a concern that this leads

48:20

into there's going to be a strata

48:22

of users who are like the royalty.

48:24

and have all the power. And we

48:26

have seen how that has failed on

48:29

other platforms. And so most notably, Dig,

48:31

which is now coming back, apparently, you

48:33

know, which was like the early version

48:35

of Reddit, where people would vote up

48:38

and vote down stories. They had some

48:40

sort of ranking system where particular users,

48:42

their votes, if they were considered. good

48:44

signalers, their votes counted more. And that

48:47

got to a point where it was

48:49

actually kind of crazy where like the

48:51

leading users on dig. We had this

48:53

happen to us. In fact, someone came

48:56

to us and said, hey, I have

48:58

like strong power signal on dig. Do

49:00

you want me to promote tectored articles?

49:02

And I know that in some cases,

49:04

there were people who were like. sort

49:07

of selling their ability to do that.

49:09

That wasn't the person who approached us,

49:11

was just like, I like texture articles,

49:13

I would wrote them. And I actually

49:16

told him no, because I felt that

49:18

that was, I felt like cheating to

49:20

rely on someone like that. But like,

49:22

once you have that kind of power,

49:25

then there's corruption potential there. Yeah, I

49:27

remember talking to a guy called Rob

49:29

Allam, who's username on Red It is

49:31

Gallo Boob, and he had a ton

49:34

of karma. He's like one of the

49:36

most kind of the most kind of

49:38

like... largest ready-uses for common. He also

49:40

used to get loads of approaches from

49:43

companies and brands, in which he was

49:45

kind of invited to basically shell on

49:47

behalf of the company and he was

49:49

pretty principled about it. But there are

49:52

all these kind of unintended consequences of

49:54

focus. on reputation and it relies on

49:56

having mitigating systems in place to, I

49:58

guess, avoid that. So really interesting experiment

50:01

will be interesting to see if that

50:03

can scale and if people enjoy or

50:05

see if those ideas proliferate onto other

50:07

platforms as well. Yeah, and I'll note

50:10

too that I do appreciate the fact

50:12

that they're trying to build this in

50:14

a decentralized way, using the decentralized social

50:16

networking protocol, which is the Project Liberty

50:19

Project, and there's a few different social

50:21

media apps that are using that. as

50:23

someone who believes in protocols over platforms

50:25

and decentralizedization. I'm excited that they're doing

50:28

that rather than trying to build up

50:30

a brand new thing from scratch. Yeah,

50:32

indeed. Great. Thanks, Mike. That brings us

50:34

to the end of today's episode. Thanks

50:37

to our listeners for tuning in. If

50:39

you enjoyed today's episode or if you

50:41

didn't hate it, you know what to

50:43

do. Let's raise the bar, okay? Last

50:46

week, it was, it was, didn't hate

50:48

it. This week, this week. Let's get

50:50

some, we really, really like the podcast.

50:52

Okay, if you really, really like the

50:55

podcast, I think these three reviews have

50:57

gone to your head. But yeah, okay,

50:59

if you really, really like the podcast,

51:01

leave us a review in which you

51:04

tell us that you really, really like

51:06

it and we will really, really like

51:08

you. And if you like the podcast

51:10

enough to sponsor an episode, we are

51:13

in the market for sponsors, you get

51:15

a mention at the end of the

51:17

podcast and an excellent. 10-minute interview with

51:19

one of us and our listeners are

51:22

growing all the time and we're getting

51:24

lots of really great feedback so get

51:26

in touch podcast at control alt-speech.com. Thanks

51:28

for your time as ever Mike. It's

51:31

great to chat to you. Thanks for

51:33

all the listeners tuning in and we'll

51:35

speak you next week. Thanks for listening

51:37

to Control Alt Speech. Subscribe now to

51:40

get our weekly episodes as soon as

51:42

they're released. If your company or organization

51:44

is interested in sponsoring the podcast, contact

51:46

us by visiting controlalt speech.com, that's CTRL

51:49

alt speech.com. This podcast is produced with

51:51

financial support from the future of online

51:53

trust and safety. Fund, a a

51:55

fiscally multi-doner fund that

51:58

global impact, that

52:00

supports charitable activities to

52:02

build a more

52:04

robust, capable, and inclusive

52:07

trust and safety

52:09

ecosystem. ecosystem.