1
00:00:01,520 --> 00:00:04,440
Hey, it's Eric Newcomer. 
Welcome to the Newcomer Podcast.

2
00:00:04,520 --> 00:00:06,640
A really exciting episode this 
week. 

3
00:00:07,160 --> 00:00:10,080
We're talking to a 
whistleblower, Francis Haugett, 

4
00:00:10,320 --> 00:00:12,720
the Facebook whistleblower is on
the show. 

5
00:00:12,880 --> 00:00:15,640
I feel like we've been building 
to this for a literal years. 

6
00:00:15,640 --> 00:00:18,520
I've been talking about 
Facebook's content moderation, 

7
00:00:18,520 --> 00:00:21,800
the implications of her release 
to the Wall Street Journal of 

8
00:00:21,800 --> 00:00:24,240
the Facebook Files, and her 
decision to go public. 

9
00:00:24,520 --> 00:00:26,720
And we really got into the 
substance of it. 

10
00:00:26,720 --> 00:00:29,560
I feel like so much of the 
discussion around social media. 

11
00:00:29,960 --> 00:00:34,160
Moderation is sort of built for 
a simplistic public consumption,

12
00:00:34,160 --> 00:00:37,640
and I really tried to get into 
the meat of these issues and 

13
00:00:37,640 --> 00:00:41,680
what potential policy solutions 
there are. 

14
00:00:42,040 --> 00:00:45,080
So give it a list. 
Francis Alga and Facebook 

15
00:00:45,080 --> 00:00:48,720
whistleblower, Welcome to the 
Newcomer podcast, author of The 

16
00:00:48,720 --> 00:00:50,560
Power of 1. 
Thanks for joining me. 

17
00:00:51,000 --> 00:00:52,760
Thank you for inviting. 
Happy to be here. 

18
00:00:53,080 --> 00:00:56,920
I have followed your story, the 
Wall Street Journal reporting 

19
00:00:56,920 --> 00:01:00,800
now the book super closely. 
So really excited to have the 

20
00:01:00,800 --> 00:01:04,040
conversation. 
You know, the Facebook Files 

21
00:01:04,080 --> 00:01:08,480
series and sort of the 
complaints about Facebook can be

22
00:01:08,480 --> 00:01:12,120
so sprawling, like there's so 
much to it. 

23
00:01:12,120 --> 00:01:14,400
So I just wanted to start off 
the conversation. 

24
00:01:14,400 --> 00:01:19,160
If you had to sum up the top 
Facebook sin or you're like main

25
00:01:19,560 --> 00:01:21,800
core objection, what is your 
top? 

26
00:01:22,180 --> 00:01:25,460
Facebook objection. 
So I would say the through line 

27
00:01:25,460 --> 00:01:28,140
that goes through all of the 
different kind of shocking, you 

28
00:01:28,140 --> 00:01:31,580
know, headline, exclamation mark
type things that came out of the

29
00:01:31,580 --> 00:01:34,020
Facebook files. 
And to remind people who maybe 

30
00:01:34,020 --> 00:01:36,500
don't remember exactly what 
happened about two years ago, 

31
00:01:36,780 --> 00:01:38,900
you know, it ranged everything 
from Facebook. 

32
00:01:38,900 --> 00:01:41,620
We knew they were human 
traffickers on the platform and 

33
00:01:41,620 --> 00:01:45,180
worried more about offending the
buyers than the people being 

34
00:01:45,180 --> 00:01:48,020
sold. 
Or Facebook had lots of research

35
00:01:48,020 --> 00:01:51,380
on their products, making kids 
unhappy, making them depressed. 

36
00:01:51,750 --> 00:01:54,310
Facilitating eating disorders, 
refused to hand it over to the 

37
00:01:54,310 --> 00:01:57,510
Senate when asked. 
You know what On and on and on. 

38
00:01:57,550 --> 00:02:00,270
But the thing that was the 
common through line, I would say

39
00:02:00,350 --> 00:02:04,110
through all those things is we 
live in a world where Facebook 

40
00:02:04,110 --> 00:02:08,270
has to report publicly it's 
profit and loss numbers, It's 

41
00:02:08,270 --> 00:02:12,270
expenses like how it got to that
profit and loss, but it doesn't 

42
00:02:12,270 --> 00:02:16,070
have to report a social balance 
sheet, you know, like what's the

43
00:02:16,070 --> 00:02:19,550
social costs of making those 
dollars and for most other 

44
00:02:19,550 --> 00:02:22,600
industries. 
Those public costs, those 

45
00:02:22,600 --> 00:02:25,680
externalities are very obvious. 
You know, it's pollution in a 

46
00:02:25,680 --> 00:02:28,880
river, it's pollution in the 
air, it's forced labor. 

47
00:02:28,880 --> 00:02:33,080
You know, we can see what's 
going on with digital products. 

48
00:02:33,160 --> 00:02:36,640
You can take from the social 
side of the balance sheet and 

49
00:02:36,640 --> 00:02:39,040
make your economic balance sheet
look a lot better. 

50
00:02:39,360 --> 00:02:42,360
And that's the core problem that
is repeated over and over again 

51
00:02:42,360 --> 00:02:44,520
across all of these, all these 
harms. 

52
00:02:44,960 --> 00:02:46,520
Yeah. 
And I mean, that really sort of 

53
00:02:46,520 --> 00:02:48,600
comes through in the book. 
I mean, you make the point that 

54
00:02:48,600 --> 00:02:51,240
like, you know, Apple, it would 
be sort of visible with a 

55
00:02:51,240 --> 00:02:54,920
hardware device, whereas with 
this social media company like 

56
00:02:54,920 --> 00:02:59,000
Facebook, it's less visible. 
Is it a matter of the fact that 

57
00:02:59,000 --> 00:03:04,040
Facebook is customizing feeds 
from person to person or is it 

58
00:03:04,040 --> 00:03:07,600
more than that, what makes 
Facebook's products so much, 

59
00:03:07,600 --> 00:03:11,000
like, less visible than 
something more tangible? 

60
00:03:11,500 --> 00:03:14,540
The point I make in the book is 
if you go search for Apple 

61
00:03:14,540 --> 00:03:17,260
whistleblower, you don't find a 
lot. 

62
00:03:17,340 --> 00:03:19,780
Like, there are occasionally 
Apple whistleblowers, but 

63
00:03:19,780 --> 00:03:21,380
they're not like Facebook 
whistleblowers. 

64
00:03:21,380 --> 00:03:24,140
Like I I always kind of bristle 
when people describe me as the 

65
00:03:24,180 --> 00:03:26,500
Facebook whistleblower because 
there's a new Facebook 

66
00:03:26,500 --> 00:03:30,500
whistleblower every two weeks, 
you know, often with really big 

67
00:03:30,500 --> 00:03:32,740
revelations. 
It's not like there's one person

68
00:03:32,740 --> 00:03:34,260
in the company that's feeding 
all that out there. 

69
00:03:34,260 --> 00:03:37,660
It's like a whole range of 
conscientious people who know 

70
00:03:37,660 --> 00:03:38,980
the public need to be brought 
in. 

71
00:03:39,410 --> 00:03:41,050
So you can ask the question, 
like, why is that? 

72
00:03:41,050 --> 00:03:44,850
Like, why is it Apple isn't 
producing the same volume or 

73
00:03:44,850 --> 00:03:48,370
intensity of whistleblowers, But
I think it's that when Apple 

74
00:03:48,450 --> 00:03:52,210
launches a new iPhone, you know,
within a couple of hours of that

75
00:03:52,210 --> 00:03:57,040
iPhone going on store shelves. 
Don't be YouTube videos live 

76
00:03:57,040 --> 00:04:00,400
like already live where people 
have taken apart those phones 

77
00:04:00,400 --> 00:04:03,320
and confirmed like yes that 
processor is in there. 

78
00:04:03,400 --> 00:04:04,920
Right. 
They'll be like, oh, they added 

79
00:04:04,920 --> 00:04:08,680
a hard drive and now the hard 
drive slightly slower because of

80
00:04:08,680 --> 00:04:10,120
it. 
Yeah, exactly. 

81
00:04:10,120 --> 00:04:13,920
It's very visible to people. 
But when it comes to what I call

82
00:04:13,920 --> 00:04:17,440
opaque systems. 
And more and more of our economy

83
00:04:17,440 --> 00:04:19,760
is going to be run by opaque 
systems. 

84
00:04:19,760 --> 00:04:23,040
So that's things like a large 
language model that lives on a 

85
00:04:23,040 --> 00:04:26,240
data center and all you see is 
the outputs, you don't see the 

86
00:04:26,240 --> 00:04:28,280
inputs, you don't see out there 
manipulated. 

87
00:04:28,720 --> 00:04:31,920
It could be a social media site.
So like you said before is the 

88
00:04:31,920 --> 00:04:34,960
problem because we all see 
something slightly different. 

89
00:04:35,400 --> 00:04:37,440
That's definitely a huge part of
it. 

90
00:04:37,680 --> 00:04:40,440
So if we were talking about 
Google, have you ever programmed

91
00:04:40,440 --> 00:04:42,000
before? 
No, I'm not a programmer. 

92
00:04:42,400 --> 00:04:44,640
I know you're not going to 
believe me when I say this, but 

93
00:04:44,720 --> 00:04:48,600
if you and I sat down for three 
weeks, I could teach you enough 

94
00:04:48,600 --> 00:04:52,640
programming that you could ask 
real, meaningful questions about

95
00:04:52,640 --> 00:04:54,920
how Google works. 
You know, what's getting 

96
00:04:54,920 --> 00:04:56,640
distributed? 
What's not getting distributed? 

97
00:04:56,880 --> 00:04:58,640
How prominent are different 
kinds of things? 

98
00:04:58,640 --> 00:05:02,240
What kinds of answers does 
Google give if we want to do the

99
00:05:02,240 --> 00:05:06,200
same kind of accountability? 
101 for Facebook? 

100
00:05:06,320 --> 00:05:09,080
Free Instagram. 
Don't even mention tick tock, 

101
00:05:09,080 --> 00:05:10,680
tick tock. 
It's way harder because it's 

102
00:05:10,680 --> 00:05:11,860
video. 
Right. 

103
00:05:11,940 --> 00:05:14,820
It's bandwidth intensive. 
You know how much hard it is to 

104
00:05:14,820 --> 00:05:17,020
a video podcast than an audio 
podcast? 

105
00:05:17,260 --> 00:05:21,980
We'd have to recruit 20,000 
volunteers and convince them to 

106
00:05:21,980 --> 00:05:25,940
install software on their phones
that would send back what they 

107
00:05:25,940 --> 00:05:30,100
saw on these platforms. 
It's a an entirely different 

108
00:05:30,540 --> 00:05:34,140
order of magnitude in terms of 
difficulty, and Facebook knows 

109
00:05:34,140 --> 00:05:36,700
that. 
With TikTok, sort of. 

110
00:05:36,740 --> 00:05:39,620
I really understand it right 
where it feels so tailored to 

111
00:05:39,620 --> 00:05:41,060
the human being. 
And it's. 

112
00:05:41,460 --> 00:05:44,220
Sort of machine learning driven 
where it feels like is there 

113
00:05:44,220 --> 00:05:48,300
even someone over at TikTok who 
really gets exactly why I'm 

114
00:05:48,300 --> 00:05:50,540
getting what I am? 
I mean but Facebook at least in 

115
00:05:50,540 --> 00:05:54,140
the beginning was supposed to be
more like network based and like

116
00:05:54,140 --> 00:05:55,620
friend based. 
Which? 

117
00:05:55,620 --> 00:05:58,860
Felt like, yeah, easier to 
trace. 

118
00:05:58,860 --> 00:06:01,900
Do you think it's totally moved 
away from that or why? 

119
00:06:02,180 --> 00:06:06,660
Why is Facebook so much harder 
to follow than at Google? 

120
00:06:07,340 --> 00:06:12,680
So by the time I got to Facebook
in 2019. 60% of everything 

121
00:06:12,720 --> 00:06:16,280
people in the United States 
viewed came from a Facebook 

122
00:06:16,280 --> 00:06:18,320
group. 
It wasn't content from your 

123
00:06:18,320 --> 00:06:21,600
family and friends anymore. 
It's one of these things where 

124
00:06:21,600 --> 00:06:24,640
one of the projects my nonprofit
is going to work on, let's go 

125
00:06:24,680 --> 00:06:28,040
beyond the screen because we 
help people see beyond their 

126
00:06:28,040 --> 00:06:30,800
screens, is doing a simulated 
social network. 

127
00:06:31,120 --> 00:06:34,320
Because, you know, it's like you
brought up before a lot of 

128
00:06:34,320 --> 00:06:37,400
people if you were to stop them 
on the street and say, what is 

129
00:06:37,400 --> 00:06:39,940
Facebook for? 
And they would say it's about 

130
00:06:39,940 --> 00:06:41,780
connecting with my family and 
friends. 

131
00:06:41,980 --> 00:06:46,700
But if you were to actually go 
and look under the hood that 

132
00:06:46,700 --> 00:06:49,740
products, you know, that product
where the only things you saw in

133
00:06:49,740 --> 00:06:52,660
your feed were things that your 
family and friends posted, that 

134
00:06:52,660 --> 00:06:55,860
world is long gone. 
And the reason why it changed 

135
00:06:55,860 --> 00:06:59,420
was Facebook is an advertising 
supported business. 

136
00:06:59,460 --> 00:07:03,180
You know, as you scroll, you see
an ad, they make money, you 

137
00:07:03,180 --> 00:07:04,700
click on that ad, they get 
money. 

138
00:07:05,490 --> 00:07:09,050
They have a motivation to get 
you to conceal more and more 

139
00:07:09,130 --> 00:07:11,770
higher value ads every single 
month. 

140
00:07:11,770 --> 00:07:14,770
You know, the more time you 
spend on the platform, the more 

141
00:07:14,770 --> 00:07:17,090
content you view, the more ad 
dollars come in. 

142
00:07:18,170 --> 00:07:21,450
They had a huge motivation to 
figure out how can they make 

143
00:07:21,450 --> 00:07:24,490
sure your feed is always full of
really stimulating content. 

144
00:07:25,290 --> 00:07:27,810
What's interesting about TikTok 
is TikTok said. 

145
00:07:27,810 --> 00:07:31,570
Hey, we don't want to wait to 
get critical mass. 

146
00:07:32,070 --> 00:07:34,430
You know, in the case of 
Facebook or like creating a 

147
00:07:34,430 --> 00:07:38,750
Facebook competitor, when the 
expectation from the user is the

148
00:07:38,750 --> 00:07:42,630
things that I see come from my 
family and friends, you have the

149
00:07:42,630 --> 00:07:46,310
like a chicken and egg problem 
where you need to get all of the

150
00:07:46,310 --> 00:07:49,190
person's friends before the 
person wants to join, right? 

151
00:07:49,190 --> 00:07:52,590
Like how do you make that happen
with something like TikTok or 

152
00:07:52,830 --> 00:07:55,070
now Threads? 
Threads has a very similar 

153
00:07:55,070 --> 00:07:56,760
model. 
And that's Facebook's new 

154
00:07:56,760 --> 00:08:00,560
Twitter competitor. 
There's much less of a user 

155
00:08:00,560 --> 00:08:02,200
promise of what you're going to 
see. 

156
00:08:02,880 --> 00:08:06,040
So when you show up, they don't 
have to wait for critical mass. 

157
00:08:06,040 --> 00:08:09,520
They just need to entertain you.
But at the same time, now you're

158
00:08:09,520 --> 00:08:13,360
putting yourself in the hands of
an algorithm you know, You 

159
00:08:13,360 --> 00:08:16,320
better understand what the 
biases are in that algorithm, 

160
00:08:16,320 --> 00:08:18,920
what it shows you and doesn't 
show you, because now the 

161
00:08:18,920 --> 00:08:21,440
algorithm is in control, not 
your friends and family. 

162
00:08:21,990 --> 00:08:25,070
You brought up threads and I I 
too see the similarity with 

163
00:08:25,070 --> 00:08:27,830
TikTok where it's, you know, 
it's bite size. 

164
00:08:27,830 --> 00:08:31,710
You know, it's not as tied to 
even like a follower network as 

165
00:08:31,710 --> 00:08:34,070
Twitter because they needed it 
to work right away. 

166
00:08:34,350 --> 00:08:37,190
So it feels like, oh man, this 
is going to be really powered by

167
00:08:37,190 --> 00:08:41,190
machine learning pretty quickly.
Like as you know a reporter. 

168
00:08:41,190 --> 00:08:45,110
It's funny to see sort of the 
reporter class sort of embracing

169
00:08:45,110 --> 00:08:47,350
threads at the moment when I 
feel like. 

170
00:08:47,790 --> 00:08:50,630
You know two years ago or more 
than that, they would have been 

171
00:08:50,630 --> 00:08:55,670
so negative and apprehensive 
about trusting Facebook. 

172
00:08:55,670 --> 00:08:59,070
I'm just curious like watching 
the sort of pretty upbeat 

173
00:08:59,390 --> 00:09:02,390
response to threads, what do you
take from that? 

174
00:09:02,390 --> 00:09:07,430
And like, are you surprised that
there seems to be some Media 

175
00:09:07,430 --> 00:09:09,350
Trust of Facebook right now? 
I. 

176
00:09:09,950 --> 00:09:13,110
Think that's probably a sense 
where the trauma the Twitter 

177
00:09:13,110 --> 00:09:15,550
community has faced in the last 
year? 

178
00:09:16,400 --> 00:09:18,400
It's pretty intense. 
You know, it's one of the things

179
00:09:18,400 --> 00:09:21,840
where, you know, Twitter never 
really went down that often, now

180
00:09:21,840 --> 00:09:24,680
goes down regularly. 
It used to be, you know, as Jack

181
00:09:24,680 --> 00:09:27,880
Dorsey, the former CEO, has 
pointed out recently, you know, 

182
00:09:27,880 --> 00:09:30,440
it used to be basically when you
posted, things went live 

183
00:09:30,440 --> 00:09:32,840
instantaneously. 
And now they're, you know, he 

184
00:09:32,840 --> 00:09:35,520
flagged that they were now, you 
know, a minute or two delays 

185
00:09:35,520 --> 00:09:37,600
between when you posted and when
it went out to people. 

186
00:09:38,040 --> 00:09:42,440
I think Elon Musk really gave 
himself an incredibly hard hill 

187
00:09:42,440 --> 00:09:45,680
to climb when he bought the 
company using so much debt. 

188
00:09:46,410 --> 00:09:49,570
Because when that happened, it 
meant that he put a clock on 

189
00:09:49,570 --> 00:09:51,250
himself. 
You know, he has to make a 

190
00:09:51,290 --> 00:09:55,690
billion dollars of profit a year
and a product that was maybe 

191
00:09:55,690 --> 00:09:58,810
breaking even before he bought 
it, It was really hard. 

192
00:09:59,290 --> 00:10:01,010
And he's had lots of advertisers
flee. 

193
00:10:01,650 --> 00:10:05,210
So I think there's further the 
positive reporting on threads is

194
00:10:05,210 --> 00:10:10,850
that people really liked having 
a space to discuss ideas, to 

195
00:10:10,850 --> 00:10:14,810
discuss issues and the idea that
they could have a space again. 

196
00:10:15,700 --> 00:10:17,900
Feels really good. 
Just to say another way, it's 

197
00:10:17,900 --> 00:10:21,540
like people were so frustrated 
with Twitter that give them some

198
00:10:21,700 --> 00:10:23,940
openness to trust Facebook 
again. 

199
00:10:23,940 --> 00:10:26,460
Or Meadow, whatever we're 
calling it these days anyway. 

200
00:10:26,460 --> 00:10:28,580
Continue. 
It's interesting, like this is 

201
00:10:28,580 --> 00:10:31,540
still true, but people were 
reporting the, you know, the 

202
00:10:31,540 --> 00:10:34,820
first day to of threads. 
You weren't even given the 

203
00:10:34,860 --> 00:10:38,220
option to just receive content 
from people you chose. 

204
00:10:38,660 --> 00:10:42,100
Like it really is like talking 
that way where you have to put 

205
00:10:42,100 --> 00:10:43,700
yourselves in the hand of the 
algorithm. 

206
00:10:44,150 --> 00:10:45,870
And be like algorithm. 
May I please? 

207
00:10:45,870 --> 00:10:48,190
Yeah, there's no follower feed. 
Does that mean you haven't 

208
00:10:48,190 --> 00:10:49,670
downloaded it? 
Are you personally? 

209
00:10:49,670 --> 00:10:51,030
I have not. 
I have not. 

210
00:10:51,470 --> 00:10:52,870
Isn't that I'm unwilling to 
download it. 

211
00:10:52,910 --> 00:10:54,430
I just haven't got around to 
download it yet. 

212
00:10:54,630 --> 00:10:56,830
And I as someone who worked on 
Google Plus. 

213
00:10:57,110 --> 00:11:00,870
So for your listeners who may 
not remember, you know tech 12 

214
00:11:00,910 --> 00:11:04,030
years ago. 
Yeah, Google said, hey, there 

215
00:11:04,030 --> 00:11:06,430
are flaws with Facebook, like we
should make a version of 

216
00:11:06,430 --> 00:11:08,550
Facebook that doesn't have those
issues. 

217
00:11:09,150 --> 00:11:11,270
And in reality, you know, they 
were trying to go after the 

218
00:11:11,350 --> 00:11:14,270
personal social media market 
that's like you can actually 

219
00:11:14,270 --> 00:11:17,550
with people, you know, but they 
because they wanted to grow so 

220
00:11:17,550 --> 00:11:19,630
fast, they basically made 
something very similar to 

221
00:11:19,630 --> 00:11:22,990
Twitter. 
And you know, growing too fast 

222
00:11:22,990 --> 00:11:26,910
is actually a problem because it
means that when people go and 

223
00:11:26,910 --> 00:11:31,030
take that chance on you, they 
don't land in a functioning 

224
00:11:31,030 --> 00:11:32,670
community. 
They land in chaos. 

225
00:11:32,670 --> 00:11:34,750
And they don't really understand
what the point is. 

226
00:11:34,750 --> 00:11:37,030
They don't understand what the 
point of this place is, doesn't 

227
00:11:37,030 --> 00:11:38,830
feel like a place and they move 
on. 

228
00:11:39,260 --> 00:11:42,060
And at least some of the initial
data around engagement, It seems

229
00:11:42,060 --> 00:11:44,340
like that's a very similar 
problem to what Facebook is 

230
00:11:44,380 --> 00:11:46,540
facing right now. 
People were willing to download 

231
00:11:46,540 --> 00:11:50,700
it with threats, yeah. 
I mean, one thing Facebook did 

232
00:11:50,700 --> 00:11:53,140
differently is they, like, 
prepopulated it with all these 

233
00:11:53,140 --> 00:11:54,220
celebrities. 
Yeah. 

234
00:11:54,260 --> 00:11:57,060
So they had that. 
And they had obviously the 

235
00:11:57,060 --> 00:11:59,860
Instagram sort of connection to 
flow people in. 

236
00:12:00,100 --> 00:12:02,300
But I agree there's a lot of 
similarity, especially and, you 

237
00:12:02,300 --> 00:12:05,980
know, I mean, what's is it 
Google Plus, man, I can't even 

238
00:12:05,980 --> 00:12:07,020
remember the name. 
You just said it. 

239
00:12:07,020 --> 00:12:07,780
Google Plus. 
Yeah. 

240
00:12:09,100 --> 00:12:12,340
They like, I mean they got a ton
of users in the beginning I, you

241
00:12:12,340 --> 00:12:15,140
know, I was on it for a second 
and then it sort of faded. 

242
00:12:15,140 --> 00:12:19,300
So it's very possible. 
I mean Threads takes that same 

243
00:12:19,300 --> 00:12:21,260
arc, I don't think that's my 
personal view. 

244
00:12:21,260 --> 00:12:25,100
I mean is that a prediction or 
do you think this will be like a

245
00:12:25,100 --> 00:12:27,180
Google Plus situation? 
I think one of the things that's

246
00:12:27,180 --> 00:12:31,020
interesting about Twitter is the
way Einstein's like to frame 

247
00:12:31,020 --> 00:12:35,420
Twitter is. 
Twitter is fueled by content 

248
00:12:35,420 --> 00:12:38,300
from a relatively small fraction
of its users. 

249
00:12:38,990 --> 00:12:42,990
Who are most interested in the 
thoughts of that other small 

250
00:12:42,990 --> 00:12:46,710
fraction of users? 
So I talked to a law professor 

251
00:12:46,790 --> 00:12:50,990
who is regularly cited in tech 
discussions and she was like I 

252
00:12:50,990 --> 00:12:54,950
use Twitter to hear from 300 
other people and just happens to

253
00:12:54,950 --> 00:12:58,150
be that those people talking to 
each other and caring about what

254
00:12:58,150 --> 00:13:01,110
each other says everyone else 
gets to kind of be a fly on the 

255
00:13:01,110 --> 00:13:02,990
wall and follow along for the 
conversation. 

256
00:13:03,480 --> 00:13:06,720
There are people like Elon Musk 
who love like raising up their 

257
00:13:06,720 --> 00:13:09,600
followers and like having direct
conversations with a wide spot 

258
00:13:09,600 --> 00:13:12,720
that people. 
But I'd say the fuel, the real 

259
00:13:12,720 --> 00:13:17,520
core of Twitter is those, you 
know, communities of connection 

260
00:13:18,920 --> 00:13:21,360
and it'll be really interesting 
to see if threads can maintain 

261
00:13:21,360 --> 00:13:23,000
that. 
Like is it just going to end up 

262
00:13:23,000 --> 00:13:27,720
kind of a brand safe By brand 
safe, I mean for, you know even 

263
00:13:27,720 --> 00:13:31,790
sell Tide, Washington, GA. 
And Clorox, you know, bleach 

264
00:13:31,790 --> 00:13:33,470
one. 
Yeah, they're very programs, 

265
00:13:33,470 --> 00:13:34,550
definitely. 
Really. 

266
00:13:34,550 --> 00:13:37,510
Is it going to be just a space 
that is, like innocuous? 

267
00:13:37,670 --> 00:13:39,710
Or is it going to be a space 
that really does cultivate 

268
00:13:39,710 --> 00:13:43,710
community the way Twitter did? 
I mean it also, you know, it 

269
00:13:43,710 --> 00:13:48,270
feels like you could have one 
community on threads and one on 

270
00:13:48,270 --> 00:13:51,750
Twitter, especially with the 
part, you know, like I mean 

271
00:13:51,750 --> 00:13:54,510
you're seeing Elon right now, 
he's like making these payments 

272
00:13:54,510 --> 00:13:57,040
out to. 
Twitter personalities, it feels 

273
00:13:57,040 --> 00:13:59,920
like extremely conservative. 
I mean you could just see sort 

274
00:13:59,920 --> 00:14:04,320
of more left wing liberal, sort 
of, you know, people on threads 

275
00:14:04,320 --> 00:14:07,160
and sort of a right wing Twitter
world. 

276
00:14:07,280 --> 00:14:09,320
It's very possible. 
I can totally imagine that 

277
00:14:09,320 --> 00:14:11,880
happening. 
It makes me nervous because like

278
00:14:11,880 --> 00:14:15,800
when we do move into a space 
where we are entirely dependent 

279
00:14:15,800 --> 00:14:20,000
on an algorithm like, I have no 
idea if I post on threads like, 

280
00:14:20,160 --> 00:14:21,880
will my posts even get 
distributed? 

281
00:14:22,430 --> 00:14:25,190
You know, like my publisher 
couldn't even tag it to sell the

282
00:14:25,190 --> 00:14:28,230
book, right? 
So it's things like that. 

283
00:14:28,550 --> 00:14:33,070
Wait, you think Facebook is like
actively suppressing your book 

284
00:14:33,070 --> 00:14:36,470
on their platforms or? 
I know that for events I've 

285
00:14:36,470 --> 00:14:40,230
done, they couldn't use the word
Facebook to describe the event. 

286
00:14:40,310 --> 00:14:42,350
So I can't be the Facebook 
whistleblower. 

287
00:14:42,350 --> 00:14:44,790
I can be a whistleblower. 
Like, you know, they'll have to 

288
00:14:44,790 --> 00:14:47,070
sit there and try a bunch of 
different variation of ads to 

289
00:14:47,070 --> 00:14:50,710
run an ad I was told by my 
publisher was, you know. 

290
00:14:51,230 --> 00:14:54,030
They sell lots of books. 
They post about lots of books on

291
00:14:54,030 --> 00:14:57,070
Instagram. 
They went through and tried to 

292
00:14:57,070 --> 00:15:00,710
tag my book and they got back an
e-mail saying that it violated 

293
00:15:00,710 --> 00:15:02,990
the commerce policy for 
Instagram. 

294
00:15:03,830 --> 00:15:07,190
And they paint their like, 
concierge because they spent 

295
00:15:07,190 --> 00:15:10,350
enough money on ads that they 
get, you know, a human shot to. 

296
00:15:10,350 --> 00:15:13,150
And the person was like, I don't
know why this would violate the 

297
00:15:13,150 --> 00:15:15,030
commerce policy. 
I mean, there's not white 

298
00:15:15,030 --> 00:15:17,070
community. 
There's no violence, you know, 

299
00:15:17,670 --> 00:15:20,870
whatever. 
And yeah, so they they don't 

300
00:15:20,870 --> 00:15:23,590
know why. 
I mean, do you think Facebook 

301
00:15:23,590 --> 00:15:26,550
just has a blanket ban on people
using the word Facebook? 

302
00:15:26,750 --> 00:15:29,510
Ooh, I don't know. 
I've had other experiences, like

303
00:15:29,510 --> 00:15:34,830
the Nobel Peace Center had an 
event that I spoke at, and they 

304
00:15:34,830 --> 00:15:39,070
had never had an ad blocked for 
being a political ad until they 

305
00:15:39,070 --> 00:15:41,750
advertised my event. 
So think about that. 

306
00:15:41,750 --> 00:15:43,950
Like, do you think the events at
the Nobel Peace Center are 

307
00:15:43,950 --> 00:15:46,350
political? 
I think most of all are. 

308
00:15:46,350 --> 00:15:47,670
Right. 
So. 

309
00:15:48,440 --> 00:15:50,480
It is what it is. 
Have you interacted much with 

310
00:15:50,480 --> 00:15:53,000
Facebook otherwise or like? 
Have you had any direct 

311
00:15:53,000 --> 00:15:57,720
engagement since you left? 
I think if I was more of a 

312
00:15:57,720 --> 00:16:00,840
troll, you know I have a little 
bit of troll, not like a huge 

313
00:16:00,840 --> 00:16:03,760
amount of troll in my heart. 
I have like a enough troll for 

314
00:16:03,760 --> 00:16:06,960
spark, right. 
I would totally promote them 

315
00:16:06,960 --> 00:16:10,800
more because they so 
aggressively will not 

316
00:16:10,800 --> 00:16:13,400
acknowledge that I exist. 
So if like for example, if you 

317
00:16:13,440 --> 00:16:17,440
ever get to be in a Q&A session 
with a Facebook executive like 

318
00:16:17,440 --> 00:16:20,410
at a conference. 
Ask them about me because they 

319
00:16:20,410 --> 00:16:22,970
will not use my name. 
So that's the kind of thing 

320
00:16:22,970 --> 00:16:24,690
we're like. 
If I was a bigger troll I would 

321
00:16:24,690 --> 00:16:27,570
see more questions at their 
events, but you know, I have 

322
00:16:27,570 --> 00:16:31,330
other things to do so. 
Back to sort of the core 

323
00:16:31,490 --> 00:16:35,930
disclosures, you know, in the 
book and I mean clearly you talk

324
00:16:35,930 --> 00:16:40,170
about sort of them lying and I'm
just curious like what you think

325
00:16:40,170 --> 00:16:43,850
sort of the core or like great 
examples of like Facebook's 

326
00:16:44,490 --> 00:16:46,610
deceit? 
Have been because I mean there's

327
00:16:46,890 --> 00:16:49,890
obviously they have such an 
information advantage where it's

328
00:16:49,890 --> 00:16:52,770
like, yeah, they can just run 
circles around anybody trying to

329
00:16:52,770 --> 00:16:55,850
scrutinize them because they 
understand the platform so much 

330
00:16:55,850 --> 00:16:57,610
more. 
But the cases where they were, 

331
00:16:57,610 --> 00:17:00,530
you know, lying. 
Flagrant so I'll give you an 

332
00:17:00,530 --> 00:17:03,090
example. 
This is one of the core parts, 

333
00:17:03,090 --> 00:17:06,730
and I complain. 
Back in 2018, Facebook faced a 

334
00:17:06,770 --> 00:17:09,170
business problem. 
You know, they could see over 

335
00:17:09,170 --> 00:17:11,650
time people were making less and
less content. 

336
00:17:12,200 --> 00:17:15,240
This is a normal phenomenon on 
social networks like some people

337
00:17:15,240 --> 00:17:18,200
get really into it. 
Most people slowly start to self

338
00:17:18,200 --> 00:17:23,119
edit self censor and they tried 
a bunch of experiments on you 

339
00:17:23,119 --> 00:17:27,880
and I to see could they induce 
us to produce more content. 

340
00:17:29,320 --> 00:17:31,200
And one of the things I know, 
it's like guess what your 

341
00:17:31,200 --> 00:17:33,360
experiments are on every day 
every time you're on one of 

342
00:17:33,360 --> 00:17:36,400
these platforms. 
So it turns out if they 

343
00:17:36,520 --> 00:17:40,420
artificially. 
Get you twice as many likes and 

344
00:17:40,420 --> 00:17:43,300
the way they do this is they 
just keep showing your your post

345
00:17:43,300 --> 00:17:45,740
to more and more people until 
you get to that number of likes.

346
00:17:46,660 --> 00:17:50,700
If they can get you up to you 
know twice as many likes you 

347
00:17:50,700 --> 00:17:54,740
produce more content. 
You know it's a very clear like 

348
00:17:54,740 --> 00:17:58,100
dose response that if you feel 
more social rewards more little 

349
00:17:58,100 --> 00:18:03,740
hits of dopamine you make no. 
So they came out and said we 

350
00:18:03,740 --> 00:18:06,380
don't want people mindlessly 
scrolling. 

351
00:18:06,740 --> 00:18:09,060
You know, mindlessly scrolling, 
that's bad for people. 

352
00:18:09,060 --> 00:18:13,340
So we're no longer going to 
prioritize content on Facebook. 

353
00:18:13,660 --> 00:18:16,300
They start how much time you're 
going to spend on the platform. 

354
00:18:16,980 --> 00:18:19,300
That's like, you know, it could 
be a prophecy for like how much 

355
00:18:19,300 --> 00:18:24,340
you enjoy consuming things. 
Instead, we're going to reward 

356
00:18:24,540 --> 00:18:28,220
meaningful social interactions. 
So I think about a phrase like 

357
00:18:28,220 --> 00:18:29,940
that, it's like, what is a like 
to you? 

358
00:18:29,940 --> 00:18:31,900
What is a meaningful social 
interaction? 

359
00:18:32,490 --> 00:18:35,210
Something that, you know, like a
day later I'd reflect on and 

360
00:18:35,210 --> 00:18:37,010
say, oh, that was like a good 
thing. 

361
00:18:37,010 --> 00:18:38,650
And I'm like glad it happened 
you. 

362
00:18:39,130 --> 00:18:40,490
Know. 
Or maybe like your friends 

363
00:18:40,490 --> 00:18:43,130
shared something really 
revealing and you wrote a 

364
00:18:43,130 --> 00:18:44,970
comment saying, you know, that's
hard. 

365
00:18:44,970 --> 00:18:47,090
I'm really glad you shared that.
You like that sounds like a 

366
00:18:47,130 --> 00:18:51,930
meaningful social interaction. 
In reality it was just was there

367
00:18:51,970 --> 00:18:56,610
any social interaction, right? 
So you could put bullying or 

368
00:18:56,610 --> 00:18:59,010
hate speech in a comment and 
that would be considered 

369
00:18:59,250 --> 00:19:03,290
meaningful social interaction. 
And within six months of this, 

370
00:19:04,130 --> 00:19:07,850
researchers across Europe on the
left and the right, local 

371
00:19:07,850 --> 00:19:11,330
parties, local parties on the 
left and the right were telling 

372
00:19:11,330 --> 00:19:14,610
Facebook researchers, we know 
you changed the algorithm. 

373
00:19:15,170 --> 00:19:18,210
It used to be we could post like
a white paper on our 

374
00:19:18,210 --> 00:19:20,210
agricultural policy and we get 
it. 

375
00:19:20,210 --> 00:19:22,610
It's like, not the most 
thrilling thing in the world. 

376
00:19:22,610 --> 00:19:25,090
Didn't get a lot of comments. 
Well, we could see from the 

377
00:19:25,090 --> 00:19:27,090
stats that people spend time 
with it. 

378
00:19:27,090 --> 00:19:31,210
They read it. 
Now if we post that same kind 

379
00:19:31,210 --> 00:19:33,930
of, you know, bread and butter 
content, nothing. 

380
00:19:33,930 --> 00:19:36,570
It's crickets. 
Like, we're being forced to run 

381
00:19:36,570 --> 00:19:39,850
positions that are so extreme 
that we know our constituents 

382
00:19:39,850 --> 00:19:42,490
don't support them. 
But like, our job is to run 

383
00:19:42,490 --> 00:19:44,090
social media. 
Like we have to, like, put stuff

384
00:19:44,090 --> 00:19:46,690
out there and it has to work or 
we'll lose our jobs. 

385
00:19:47,370 --> 00:19:50,050
And what's kind of crazy about 
this is in some ways Facebook 

386
00:19:50,050 --> 00:19:52,730
acknowledged implicitly they had
a problem. 

387
00:19:53,550 --> 00:19:56,790
Because they put out, excuse me,
Mark Zuckerberg put out a white 

388
00:19:56,790 --> 00:19:59,110
paper, probably one of his 
employees. 

389
00:19:59,310 --> 00:20:02,270
It's like 5000 words long. 
I really doubt Mark wrote it. 

390
00:20:02,630 --> 00:20:06,390
It said, hey, this engagement 
based ranking thing we just 

391
00:20:06,390 --> 00:20:10,670
launched, there's a problem, 
which is people are drawn to 

392
00:20:10,670 --> 00:20:13,990
engage with extreme content. 
But don't worry, don't worry. 

393
00:20:14,270 --> 00:20:16,070
And you know, we ask people 
afterwards, did you like it? 

394
00:20:16,070 --> 00:20:19,270
They say, no, we're going to 
protect you from the lowest 

395
00:20:19,270 --> 00:20:20,790
extreme content by taking it 
down. 

396
00:20:20,790 --> 00:20:22,910
We're going to have these 
magical a I systems. 

397
00:20:23,570 --> 00:20:27,330
And even their solution to their
first lie was another lie, 

398
00:20:28,130 --> 00:20:31,130
because those systems that they 
claimed we take down all the bad

399
00:20:31,130 --> 00:20:34,970
stuff, they were only 
successfully removing 3 to 5% of

400
00:20:34,970 --> 00:20:39,330
things like hate speech were 
less than 1% of violence, right?

401
00:20:39,330 --> 00:20:41,610
It's kind of crazy when you 
think about it, but they told 

402
00:20:41,610 --> 00:20:45,650
everyone from Congress to, you 
know, the, you know, teachers 

403
00:20:45,650 --> 00:20:49,290
unions, like we're protecting 
people, But in fact, it was all 

404
00:20:49,410 --> 00:20:51,980
kind of smoke and mirrors. 
I think this is the policy 

405
00:20:51,980 --> 00:20:54,940
solution you want, or at least 
the one that would seem to sort 

406
00:20:54,940 --> 00:20:57,860
of come from what you're saying 
is like okay, some sort of 

407
00:20:57,860 --> 00:21:01,940
disclosure where, you know, 
outsiders can sort of track how 

408
00:21:01,940 --> 00:21:05,180
this is all happening. 
I mean, how possible is that? 

409
00:21:05,180 --> 00:21:08,100
I mean, every social media 
network is so different. 

410
00:21:08,100 --> 00:21:10,820
You know, I'm a capitalist. 
I want companies to be able to 

411
00:21:10,820 --> 00:21:14,020
change and react and like, you 
don't want a bunch of laws that 

412
00:21:14,020 --> 00:21:17,220
like, say, oh, you need to 
disclose things this way and so.

413
00:21:17,690 --> 00:21:20,970
Now you're forced with a certain
type of the same network you've 

414
00:21:20,970 --> 00:21:24,490
had, Instead of evolving, how 
would you see a disclosure 

415
00:21:24,490 --> 00:21:28,090
regime working that still allows
companies like Facebook to be 

416
00:21:28,370 --> 00:21:31,690
flexible and to change? 
Let's actually unpack for a 

417
00:21:31,690 --> 00:21:34,730
second something you just said, 
which was, I think a lot of 

418
00:21:34,730 --> 00:21:38,450
people don't sit and think about
what's the menu of options when 

419
00:21:38,450 --> 00:21:41,530
it comes to intervening in a 
problem as complicated as this, 

420
00:21:41,530 --> 00:21:43,770
right? 
I'm really glad that you brought

421
00:21:43,770 --> 00:21:45,850
up the idea that these companies
grow a change. 

422
00:21:46,220 --> 00:21:48,420
Where that, well, you know the 
next one to come along might not

423
00:21:48,420 --> 00:21:50,220
fit the exact same walls of this
one. 

424
00:21:50,980 --> 00:21:54,580
One of the ways the European 
Union handles that flexibility 

425
00:21:54,700 --> 00:21:58,180
and and to be really clear like 
this kind of way of doing 

426
00:21:58,180 --> 00:22:02,660
regulation of saying disclosure 
and transparency is in instead 

427
00:22:02,660 --> 00:22:05,820
of something like what's 
happening in say Utah, where 

428
00:22:05,820 --> 00:22:08,100
Utah is coming in and saying 
this is how you will write your 

429
00:22:08,100 --> 00:22:09,540
company. 
If people are under 18, they 

430
00:22:09,540 --> 00:22:12,260
have to have parent supervision 
no privacy for kids. 

431
00:22:12,260 --> 00:22:13,500
Their parents can see 
everything. 

432
00:22:13,980 --> 00:22:16,100
We're like Montana coming out 
and just flat out banning 

433
00:22:16,100 --> 00:22:18,580
Tiktok. 
Those are kind of we'll call 

434
00:22:18,580 --> 00:22:21,700
them building fences, type rules
where we're like, oh, this is 

435
00:22:21,700 --> 00:22:24,220
the fence you can't cross. 
And the thing about technology 

436
00:22:24,220 --> 00:22:26,820
is they it moves and changes and
they're very good at running 

437
00:22:26,820 --> 00:22:30,260
around fences. 
So the alternative is something 

438
00:22:30,260 --> 00:22:33,220
like what the European Union 
passed last year, which is 

439
00:22:33,220 --> 00:22:36,940
called the Digital Services Act.
And the Digital Services Act 

440
00:22:36,940 --> 00:22:39,660
says, hey, if the core problem, 
you know we started at the 

441
00:22:39,660 --> 00:22:42,620
beginning of this conversation, 
if the core problem. 

442
00:22:42,990 --> 00:22:46,670
Is a power imbalance, right, 
Like the fact that you can know 

443
00:22:46,670 --> 00:22:48,110
what's going on and I can't 
know. 

444
00:22:49,070 --> 00:22:52,630
Let's address that core problem 
because a lot of other things 

445
00:22:52,630 --> 00:22:56,510
will flow downstream from it. 
So they say, hey, if you know 

446
00:22:56,510 --> 00:23:00,030
there's a risk of your product, 
you need to tell us about it. 

447
00:23:00,830 --> 00:23:04,110
You know, if you discover one, 
if you can, you know, imagine 

448
00:23:04,110 --> 00:23:05,110
one, you need to tell us about 
it. 

449
00:23:05,230 --> 00:23:07,510
You need to tell us your plan 
for mitigating it, because it's 

450
00:23:07,510 --> 00:23:09,150
going to be different for every 
platform. 

451
00:23:09,630 --> 00:23:13,150
We want to unleash innovation. 
And you need to give us enough 

452
00:23:13,150 --> 00:23:16,430
data that we could see that 
there was progress going on to 

453
00:23:16,430 --> 00:23:20,150
meet that goal. 
And if we ask you a question, we

454
00:23:20,150 --> 00:23:23,310
deserve to get an answer, which 
sounds really basic but is not 

455
00:23:23,310 --> 00:23:25,030
true today. 
You know, I've been asked 

456
00:23:25,310 --> 00:23:28,390
questions by governments around 
the world that are very basic, 

457
00:23:28,390 --> 00:23:31,670
like how many moderators speak 
German. 

458
00:23:32,150 --> 00:23:34,870
I've gotten that question but 
for different languages all 

459
00:23:34,870 --> 00:23:36,750
around the world. 
And it's because Facebook 

460
00:23:36,750 --> 00:23:38,990
doesn't have to answer. 
They don't even have to answer 

461
00:23:38,990 --> 00:23:41,270
things like how many users are 
there in your country? 

462
00:23:41,820 --> 00:23:44,860
Which is extremely frustrating 
given that Facebook stance has 

463
00:23:44,860 --> 00:23:47,540
always been where their move has
been like, regulate us. 

464
00:23:47,540 --> 00:23:50,580
Like, you know, not all these 
decisions should be made by a 

465
00:23:50,580 --> 00:23:53,180
private company, like. 
And then it's like, oh, well, 

466
00:23:53,180 --> 00:23:56,780
then why don't you give at least
the governments around the world

467
00:23:56,780 --> 00:23:59,820
or the information they would 
need to write good regulations? 

468
00:23:59,820 --> 00:24:02,180
So, I mean, obviously you're 
talking to a journalist, so 

469
00:24:02,260 --> 00:24:04,460
disclosure is always gonna be 
something I'm gonna cheer for, 

470
00:24:04,460 --> 00:24:06,740
but certainly a frustrating 
situation. 

471
00:24:06,860 --> 00:24:09,420
To provide a road map, though, 
so my nonprofit, it's called 

472
00:24:09,420 --> 00:24:12,780
Beyond the Screen. 
One of our core projects is 

473
00:24:12,780 --> 00:24:16,740
called Standard of Care where we
are working to build out a wiki 

474
00:24:16,740 --> 00:24:20,300
that people can contribute to 
around identifying the problems 

475
00:24:20,300 --> 00:24:23,540
of social media. 
What are the surface areas or 

476
00:24:23,540 --> 00:24:27,020
what we call the levers for 
preventing or mitigating harm 

477
00:24:27,020 --> 00:24:30,460
and then what are the strategies
for pulling those levers. 

478
00:24:30,700 --> 00:24:34,100
So just to give context on that,
you know a lot of the problems 

479
00:24:34,100 --> 00:24:37,360
around kids. 
A lever that is common across 

480
00:24:37,360 --> 00:24:41,040
them is let's keep under 13 year
olds off these systems. 

481
00:24:41,720 --> 00:24:45,840
But when the kids advocates talk
about technology, they often 

482
00:24:46,280 --> 00:24:49,440
don't know what's possible and 
so they'll settle on solutions 

483
00:24:49,440 --> 00:24:51,200
that might seem obvious but have
problems. 

484
00:24:51,280 --> 00:24:54,200
So for example, in the case of 
the lever of keeping under 13 

485
00:24:54,200 --> 00:24:57,080
year olds off the platform, 
they'll say let's check ID's, 

486
00:24:57,480 --> 00:24:59,360
which I don't think you want, I 
don't think I want. 

487
00:24:59,360 --> 00:25:02,400
It also just doesn't work. 
But if you've gone to a 

488
00:25:02,400 --> 00:25:05,280
technologist and said, hey, I 
have this lever. 

489
00:25:06,120 --> 00:25:08,440
Let's keep under 13 year olds 
off the platform. 

490
00:25:09,200 --> 00:25:11,760
That technologist could come 
back and say here's 10 or 15 

491
00:25:11,760 --> 00:25:14,120
different ways to find under 13 
year olds. 

492
00:25:14,480 --> 00:25:16,400
You know, some of it's really 
basic, like kids will say I'm a 

493
00:25:16,400 --> 00:25:18,840
fourth grader or things like I 
learned this one from my 

494
00:25:18,840 --> 00:25:21,960
principal. 
Kids report each other, like to 

495
00:25:21,960 --> 00:25:24,240
punish each other. 
So like you're mean to me on the

496
00:25:24,240 --> 00:25:27,680
playground and I'll report your 
Instagram account and as soon as

497
00:25:27,680 --> 00:25:31,400
you find 10,000 kids, you can 
find all the other ones. 

498
00:25:32,040 --> 00:25:35,200
And so the way I think this 
could tie into transparency. 

499
00:25:35,760 --> 00:25:39,640
Is once we have a menu saying 
these are the harms, these are 

500
00:25:39,640 --> 00:25:43,040
the surface areas, the levers 
for addressing each of those 

501
00:25:43,040 --> 00:25:45,160
harms. 
You can come in and say okay. 

502
00:25:45,160 --> 00:25:48,080
Then we at a minimum like I 
think there should be raw data 

503
00:25:48,080 --> 00:25:50,440
assets for researchers. 
But if you don't want to go that

504
00:25:50,440 --> 00:25:54,600
far, at a minimum you can say we
need data on how bad each of 

505
00:25:54,600 --> 00:25:59,840
those harms are and how hard 
you're pulling the levers to try

506
00:25:59,840 --> 00:26:02,000
to reduce those problems. 
You can figure out lots of 

507
00:26:02,000 --> 00:26:03,960
different strategies to pull 
those levers. 

508
00:26:04,390 --> 00:26:07,830
But you need to show U.S. data 
on things like under 13 year 

509
00:26:07,830 --> 00:26:10,950
olds, how late are they on the 
platforms at night, that kind of

510
00:26:10,950 --> 00:26:14,710
thing. 
I mean, a core response that 

511
00:26:14,710 --> 00:26:17,870
Facebook would give in this 
situation would just be, they 

512
00:26:17,870 --> 00:26:19,670
might not say this outright 
because it's probably not 

513
00:26:19,670 --> 00:26:22,710
political, but like some of 
these problems that you've 

514
00:26:22,710 --> 00:26:24,470
identified are just human 
problems, right? 

515
00:26:24,470 --> 00:26:27,750
If you talk about sort of the 
Instagram critique with it 

516
00:26:27,750 --> 00:26:32,090
potentially making sort of. 
Young teenage women, sort of, 

517
00:26:32,090 --> 00:26:36,050
some segment of them unhappy. 
I mean, you could say, like, was

518
00:26:36,050 --> 00:26:39,130
that so different from Vogue? 
Is this really an algorithmic 

519
00:26:39,130 --> 00:26:42,770
problem, or is this just like 
how humans are? 

520
00:26:42,770 --> 00:26:46,930
I mean, and I would probably 
attach more on the a lot of the 

521
00:26:46,930 --> 00:26:50,690
sort of Democratic liberal sort 
of anger at Facebook was about 

522
00:26:51,290 --> 00:26:54,010
just what Trump supporters are 
like and like their views and 

523
00:26:54,010 --> 00:26:56,090
the fact that there's like an 
audience for them. 

524
00:26:56,510 --> 00:26:59,350
And not always the fact that 
Facebook would give them 

525
00:26:59,350 --> 00:27:02,110
distribution. 
So I guess across a lot of these

526
00:27:02,110 --> 00:27:04,550
categories and we can get into 
the particulars of totally I 

527
00:27:04,550 --> 00:27:06,310
mentioned. 
But yeah, what would you say 

528
00:27:06,310 --> 00:27:08,270
about the just like some of 
these things people are mad 

529
00:27:08,270 --> 00:27:11,270
about are just things that human
beings do that happen to happen 

530
00:27:11,270 --> 00:27:13,870
on Facebook. 
But it's not necessarily their 

531
00:27:13,870 --> 00:27:17,150
levers that are moving people to
do those things. 

532
00:27:17,950 --> 00:27:20,310
So I think one way to think 
about this is technology can 

533
00:27:20,390 --> 00:27:22,610
either. 
Amplify and bring out the 

534
00:27:22,610 --> 00:27:25,170
worstness. 
Or it can act as a bridge that 

535
00:27:25,170 --> 00:27:28,010
helps us seek our best and 
happiest selves. 

536
00:27:28,530 --> 00:27:31,570
So I totally agree with you that
you know there are always been 

537
00:27:31,570 --> 00:27:35,370
teen girls that were unhappy 
about their bodies or how nice 

538
00:27:35,370 --> 00:27:39,130
their clothes were, but there 
are a limited number of pages of

539
00:27:39,130 --> 00:27:41,810
Vogue every month. 
You know, the second time you 

540
00:27:41,810 --> 00:27:44,610
read Vogue, you're going to have
a different impact on you than 

541
00:27:44,610 --> 00:27:46,690
the third time you read Vogue, 
or you're going to get bored of 

542
00:27:46,690 --> 00:27:48,420
it. 
And in the case of something 

543
00:27:48,420 --> 00:27:52,020
like Instagram, you know, 
Instagram progressively pushes 

544
00:27:52,020 --> 00:27:54,220
you towards more and more 
extreme content. 

545
00:27:54,540 --> 00:27:57,620
So I'll give an example. 
I had a journalist reach out to 

546
00:27:57,620 --> 00:28:00,740
me for an interview and he 
explained to me that he had just

547
00:28:00,740 --> 00:28:03,860
had a new baby. 
So this is a healthy, happy baby

548
00:28:03,860 --> 00:28:06,780
boy. 
He's a modern father. 

549
00:28:06,780 --> 00:28:08,580
He made an Instagram account for
the baby. 

550
00:28:09,220 --> 00:28:12,140
That baby had maybe five or six 
baby friends. 

551
00:28:12,660 --> 00:28:15,620
Everyone here is healthy, happy,
cute baby. 

552
00:28:16,660 --> 00:28:19,220
He's only ever clicked on or 
commented on happy, cute, 

553
00:28:19,220 --> 00:28:24,100
healthy babies and yet about 10%
of his feed with children who 

554
00:28:24,100 --> 00:28:27,900
were visibly suffering. 
So kids who had gone horrible 

555
00:28:27,900 --> 00:28:31,340
accidents and were just figured 
disabilities and deformities 

556
00:28:31,340 --> 00:28:34,780
that looked really painful. 
Kids died of cancer in hospital 

557
00:28:34,780 --> 00:28:36,100
beds with tubes coming out of 
them. 

558
00:28:36,580 --> 00:28:40,380
And he was like, Francis, how 
did we get from healthy, happy 

559
00:28:40,380 --> 00:28:43,100
babies to this? 
Like, what happened? 

560
00:28:43,700 --> 00:28:45,940
I've only ever clicked on the 
Happy Fun tent. 

561
00:28:46,310 --> 00:28:50,310
And I was like, well, the A I 
knows very clearly you like 

562
00:28:50,310 --> 00:28:51,990
babies. 
You know you've made this whole 

563
00:28:51,990 --> 00:28:55,190
little baby centric world. 
So it's showing you content and 

564
00:28:55,190 --> 00:28:58,630
knows that people who like 
babies have trouble not engaging

565
00:28:58,630 --> 00:29:00,750
with. 
And I want to be honest, even if

566
00:29:00,750 --> 00:29:04,630
you're not clicking like on that
content or sad or whatever you 

567
00:29:04,630 --> 00:29:09,430
probably are lingering and a lot
of these A I's have dwell you're

568
00:29:09,430 --> 00:29:12,670
just you're the fact that you 
paused is a single signal that 

569
00:29:12,670 --> 00:29:15,880
you like that content. 
And so, you know, he's old 

570
00:29:15,880 --> 00:29:19,680
enough and, you know, 
cognitively mature enough to see

571
00:29:19,680 --> 00:29:23,600
something weird is happening 
with a 13 year old girl. 

572
00:29:23,640 --> 00:29:25,800
You know, she might start out by
looking for something like 

573
00:29:25,800 --> 00:29:29,360
healthy recipes and just by 
clicking all the content get 

574
00:29:29,520 --> 00:29:33,080
pushed over time towards more 
and more extreme materials. 

575
00:29:33,640 --> 00:29:36,920
And we see these reports from 
things like child psychologists 

576
00:29:36,920 --> 00:29:38,960
who say I have these kids in my 
practice. 

577
00:29:39,480 --> 00:29:41,320
They come into their 
appointments and say I'm trying 

578
00:29:41,400 --> 00:29:44,160
to make better choices. 
Like I'm trying to follow the 

579
00:29:44,160 --> 00:29:47,640
program, but it follows me 
around Instagram. 

580
00:29:48,680 --> 00:29:51,880
And right now we live in a world
where we don't have consumer 

581
00:29:51,880 --> 00:29:55,000
rights to really basic things 
like should you be allowed to 

582
00:29:55,000 --> 00:29:58,360
reset an algorithm without 
losing all your past. 

583
00:29:59,080 --> 00:30:01,320
So like, those kids are being 
forced to choose from their 

584
00:30:01,320 --> 00:30:03,040
past. 
You know, all their memories, 

585
00:30:03,040 --> 00:30:06,830
their friends and their futures.
Because an algorithm wants to 

586
00:30:06,830 --> 00:30:10,110
keep them from moving away from 
something that was hurting them.

587
00:30:10,270 --> 00:30:12,110
I'm sure you saw. 
You know if you want to delete 

588
00:30:12,110 --> 00:30:13,830
threads, they say. 
You delete. 

589
00:30:13,910 --> 00:30:17,510
Instagram too, which is just 
like insanity. 

590
00:30:17,510 --> 00:30:19,910
I mean, it's like, oh, we know 
you love this other thing. 

591
00:30:19,910 --> 00:30:24,150
I'm much more supportive of the 
we should regulate tech through 

592
00:30:24,150 --> 00:30:26,750
laws that, like correct it 
rather than the sort of huge 

593
00:30:26,750 --> 00:30:29,390
antitrust push. 
I mean to me like things as 

594
00:30:29,390 --> 00:30:32,750
simple as just like then push 
notification hacking. 

595
00:30:32,750 --> 00:30:34,590
Like we need to escape a world 
where like. 

596
00:30:35,170 --> 00:30:38,490
Tech companies are allowed to 
use sort of little badges that 

597
00:30:38,490 --> 00:30:42,490
psychologically drive us crazy 
that don't show, like actual 

598
00:30:42,490 --> 00:30:44,210
alerts. 
But the problem is you write 

599
00:30:44,210 --> 00:30:47,250
that rule, you know, like I'm 
sitting here like I'm frustrated

600
00:30:47,250 --> 00:30:49,610
with like Facebook. 
And I'm like, why is Facebook 

601
00:30:49,610 --> 00:30:51,690
showing me this group that I 
never use? 

602
00:30:51,690 --> 00:30:53,570
But it knows I love to clear 
algorithms. 

603
00:30:53,570 --> 00:30:57,530
We should ban it. 
But then, you know, for the next

604
00:30:57,530 --> 00:31:01,290
startup that's trying to build 
and doesn't know anything about,

605
00:31:01,290 --> 00:31:03,770
you know, it's just like, it 
creates a huge regulatory 

606
00:31:03,770 --> 00:31:05,930
burden. 
They could end up helping 

607
00:31:06,210 --> 00:31:09,370
Facebook. 
So this is part of why we want 

608
00:31:09,370 --> 00:31:11,850
to do our standard of care 
project, right. 

609
00:31:11,850 --> 00:31:15,890
Like right now Facebook has a 
really huge advantage in that 

610
00:31:15,890 --> 00:31:19,490
they have done research in this 
space for decades and decades. 

611
00:31:19,570 --> 00:31:23,450
15 years, 17 years somewhere in 
there I guess they are coming up

612
00:31:23,450 --> 00:31:27,330
on 20 years cuz it's 2023 and 
they were founded in 2004. 

613
00:31:28,290 --> 00:31:30,410
But you want to talk about real 
deep cuts. 

614
00:31:30,570 --> 00:31:33,690
Do you remember Friendster or 
you not quite old enough? 

615
00:31:33,690 --> 00:31:36,850
Yeah, that's. 
So Facebook was my first real 

616
00:31:37,050 --> 00:31:39,610
social media I graduated. 
You weren't a life spacer. 

617
00:31:39,730 --> 00:31:42,970
In 2009 so, but I'm aware of 
sort of the history. 

618
00:31:42,970 --> 00:31:47,290
Of yeah, the first web crawler. 
So that's a thing for 

619
00:31:47,290 --> 00:31:49,970
downloading data off the 
Internet so you can analyze it 

620
00:31:50,010 --> 00:31:53,130
that I ever wrote. 
I wrote Friendster because I was

621
00:31:53,130 --> 00:31:55,890
using really early graph mapping
algorithms because I guess I've 

622
00:31:55,930 --> 00:31:57,650
always loved graph based 
problems. 

623
00:31:58,280 --> 00:32:01,040
But yeah, part of why we're 
doing standard of care is we 

624
00:32:01,040 --> 00:32:04,680
know we want to make sure that 
the next generation of social 

625
00:32:04,880 --> 00:32:07,680
platforms and you know, social 
platforms take lots and lots of 

626
00:32:07,680 --> 00:32:09,960
different forms, things like 
roadblocks. 

627
00:32:10,000 --> 00:32:13,280
Roadblocks is a social network. 
You know, we're going to always 

628
00:32:13,280 --> 00:32:16,440
have new social platforms. 
How can we make sure that people

629
00:32:16,640 --> 00:32:20,320
have the best head start, you 
know, the best platform to build

630
00:32:20,320 --> 00:32:23,440
off of to be like, oh 
interesting, we can get a very 

631
00:32:23,440 --> 00:32:25,800
robust education on what we 
should be worrying about. 

632
00:32:26,060 --> 00:32:28,820
Very quickly and an 
understanding of like what 

633
00:32:28,940 --> 00:32:32,300
options exist for being able to 
afford all those problems. 

634
00:32:33,180 --> 00:32:36,500
Is privacy a big part of your 
advocacy? 

635
00:32:36,540 --> 00:32:40,300
Or like how optimistic are to 
me, I'm sort of like, oh man, 

636
00:32:40,300 --> 00:32:43,140
privacy sort of lost in this 
world once you're on them. 

637
00:32:43,140 --> 00:32:48,020
Like, I'm just not like a 
privacy that's I'm much more 

638
00:32:48,020 --> 00:32:51,020
aligned on this sort of like, oh
man, them hacking our brains and

639
00:32:51,020 --> 00:32:53,580
like, using actual, like, 
psychological tricks to get us 

640
00:32:53,580 --> 00:32:56,460
to engage with them. 
That's really troubling. 

641
00:32:56,460 --> 00:32:59,340
But like, yeah, I don't know how
much time are you spending on 

642
00:32:59,340 --> 00:33:03,100
sort of privacy advocacy and 
what's your sort of view on on 

643
00:33:03,100 --> 00:33:05,300
that part of? 
It so privacy is not what I 

644
00:33:05,300 --> 00:33:08,660
would say, one of my tentpole 
issues, I'm very open to it as a

645
00:33:08,660 --> 00:33:12,780
technique in terms of, you know,
people emotionally connect with 

646
00:33:12,900 --> 00:33:17,780
privacy as an issue and it is a 
way of decreasing the ability of

647
00:33:17,780 --> 00:33:19,860
algorithms to be able to act 
towards you. 

648
00:33:19,860 --> 00:33:22,220
Like if they're not allowed to 
record information about you, 

649
00:33:22,540 --> 00:33:25,900
it's harder to manipulate you. 
At the same time, like, you 

650
00:33:25,900 --> 00:33:28,300
know, we're developing AIS that 
are getting better and better at

651
00:33:28,300 --> 00:33:31,580
doing implicit inferences. 
And so it's this question of, 

652
00:33:31,580 --> 00:33:34,700
you know, how little data do you
need in order to actually still 

653
00:33:34,700 --> 00:33:37,380
see a lot of these phenomena. 
And also, I think it's only 

654
00:33:37,380 --> 00:33:40,580
things where it's like, I don't 
think you can write privacy laws

655
00:33:40,580 --> 00:33:44,460
that are going to strip enough 
data to actually neutralize the 

656
00:33:44,460 --> 00:33:45,540
problems that we're talking 
about. 

657
00:33:45,540 --> 00:33:46,780
Here. 
I'll give you a really basic 

658
00:33:46,780 --> 00:33:50,600
example. 
One of the big ways you reduce 

659
00:33:50,600 --> 00:33:53,880
misinformation on something like
WhatsApp, which is end to end 

660
00:33:53,880 --> 00:33:58,240
encrypted, it's private chat is 
he's saying, hey you know, as 

661
00:33:58,240 --> 00:34:01,240
something gets reshared, so 
there's a chain of reshares. 

662
00:34:01,240 --> 00:34:04,440
So I got it, I reshared it. 
My friend reasshared it on and 

663
00:34:04,440 --> 00:34:07,360
on. 
One is a chain of reshares. 

664
00:34:07,360 --> 00:34:11,400
Once it gets 5 hops away from 
the person who created it, say 

665
00:34:11,400 --> 00:34:15,159
to them, hey, you can totally 
spread this further, but you 

666
00:34:15,159 --> 00:34:17,860
have to copy and paste it. 
You have to make an affirmative 

667
00:34:17,860 --> 00:34:22,860
action to continue forward. 
That kind of change doesn't, you

668
00:34:22,860 --> 00:34:26,139
know, you know, allowing that, 
not allowing. 

669
00:34:26,139 --> 00:34:28,940
That's not really a privacy 
topic, but it's one of the most 

670
00:34:28,940 --> 00:34:32,780
impactful things for safety. 
So I'm very pro privacy 

671
00:34:32,780 --> 00:34:34,659
legislation. 
If people want to push it, I 

672
00:34:34,659 --> 00:34:37,100
think it can have a really 
positive impact on a lot of the 

673
00:34:37,100 --> 00:34:39,739
problems I talk about. 
But it's not being trinsically 

674
00:34:39,739 --> 00:34:41,500
solve all the things that we're 
talking about here. 

675
00:34:42,340 --> 00:34:45,260
No, that makes a lot of sense. 
I sort of alluded to the 

676
00:34:45,260 --> 00:34:49,620
political piece of this sort of 
digging into that section. 

677
00:34:49,620 --> 00:34:55,340
I mean the political asymmetry 
in the United States and like 

678
00:34:55,340 --> 00:34:58,540
how that affects conversations 
about Facebook. 

679
00:34:58,700 --> 00:35:02,900
Like to me it's like pretty 
obvious that like Trump 

680
00:35:02,900 --> 00:35:05,740
supporters, Republicans like are
behaving in sort of different 

681
00:35:05,740 --> 00:35:09,380
ways than Democrats and 
spreading generally more false 

682
00:35:09,380 --> 00:35:12,220
information. 
And like, it's very awkward. 

683
00:35:12,710 --> 00:35:16,710
For Facebook, certainly. 
And for politicians or, well, I 

684
00:35:16,710 --> 00:35:19,070
don't know, journalists can 
sometimes sort of have trouble 

685
00:35:19,070 --> 00:35:21,870
sort of highlighting that 
asymmetry. 

686
00:35:22,910 --> 00:35:23,990
I don't know. 
What's your view on it? 

687
00:35:23,990 --> 00:35:26,990
Do you agree with me that 
there's like a sort of partisan 

688
00:35:26,990 --> 00:35:30,430
asymmetry there and that that is
sort of creating some of the 

689
00:35:30,430 --> 00:35:32,590
problems about how Facebook is 
able. 

690
00:35:32,910 --> 00:35:35,510
Facebook doesn't want to target 
Republicans because then they're

691
00:35:35,510 --> 00:35:37,830
going to get a lot of heat from 
Republicans. 

692
00:35:37,830 --> 00:35:41,470
And so then they're not willing 
to do sort of things that would 

693
00:35:41,470 --> 00:35:44,400
have. 
Asymmetric outcome, I guess. 

694
00:35:44,640 --> 00:35:46,200
Do you agree with that is sort 
of the question. 

695
00:35:46,240 --> 00:35:49,720
Facebook has spent a lot of 
money trying to frame the issue 

696
00:35:49,760 --> 00:35:53,920
of what can be done about any of
these online safety bras in 

697
00:35:53,920 --> 00:35:56,360
terms of freedom of speech 
versus safety. 

698
00:35:56,920 --> 00:35:59,400
And they get up there, and this 
is a real quote. 

699
00:35:59,960 --> 00:36:03,800
Mark Zuckerberg went on a 
podcast and he was like, you 

700
00:36:03,800 --> 00:36:06,920
know, I've really grown a lot in
the last year. 

701
00:36:06,920 --> 00:36:09,760
This was like, I don't know, a 
year after I came out nine 

702
00:36:09,760 --> 00:36:11,840
months. 
I know I've really grown a lot 

703
00:36:11,840 --> 00:36:15,680
in the last year because I've 
realized sometimes if you stand 

704
00:36:15,680 --> 00:36:19,000
up for what you believe in, 
you're going to be unpopular. 

705
00:36:19,560 --> 00:36:22,000
And I'm a defender of freedom of
speech. 

706
00:36:22,640 --> 00:36:25,400
And what I found so aggravating 
about this is like, you know, 

707
00:36:25,400 --> 00:36:27,560
that thing we just talked about 
with WhatsApp, You know, when 

708
00:36:27,560 --> 00:36:31,080
you cut the research chain at 5,
You know, if you do that same 

709
00:36:31,080 --> 00:36:33,560
thing on Facebook when you cut 
the research chain at 2:00. 

710
00:36:34,070 --> 00:36:37,350
And has the same impact on 
misinformation as the entire 

711
00:36:37,350 --> 00:36:39,150
third party fact checking 
organization. 

712
00:36:39,550 --> 00:36:42,470
And you're not picking and 
choosing what are the good ideas

713
00:36:42,470 --> 00:36:45,710
and the bad ideas. 
And I think right now, you know,

714
00:36:45,910 --> 00:36:50,030
a pattern we see across the 
world is if you are a political 

715
00:36:50,030 --> 00:36:54,390
party that is not in power, you 
have more of a motivation to 

716
00:36:54,390 --> 00:36:58,510
figure out new technologies, new
ways of reaching people because 

717
00:36:58,550 --> 00:36:59,990
you're the party that's out of 
power. 

718
00:37:00,110 --> 00:37:02,750
You know you don't have the 
advantages of being an 

719
00:37:02,750 --> 00:37:04,750
incumbent. 
And so if we were to roll back 

720
00:37:04,750 --> 00:37:07,790
in time to say, the Obama 
presidency, like the first one, 

721
00:37:08,390 --> 00:37:11,310
you know, when Obama got 
elected, he did a lot of 

722
00:37:11,310 --> 00:37:14,550
techniques with technology that 
no one else was doing. 

723
00:37:15,230 --> 00:37:18,270
You know, they were monitoring, 
you know, what's not as a, B, 

724
00:37:18,270 --> 00:37:20,990
testing emails. 
So they would say, you know, if 

725
00:37:21,190 --> 00:37:23,270
Obama holds a puppy, how much 
money do we make? 

726
00:37:23,270 --> 00:37:25,510
If Obama is their first wife, 
how much money do we make? 

727
00:37:25,870 --> 00:37:28,510
You know, do people sign up for 
the XYZ thing? 

728
00:37:29,110 --> 00:37:32,150
He had a a motivation to do that
because he was the scrappy 

729
00:37:32,150 --> 00:37:34,030
incumbent. 
And so I think when we think 

730
00:37:34,030 --> 00:37:37,790
about these things as like there
is a partisan lean that you know

731
00:37:37,790 --> 00:37:41,510
one side is maybe playing the 
game a little harder than the 

732
00:37:41,510 --> 00:37:43,430
other side. 
I think when I think of it is 

733
00:37:43,430 --> 00:37:45,550
that it's just a question of 
like who it is or isn't in 

734
00:37:45,550 --> 00:37:48,110
power. 
And So what I always try to say 

735
00:37:48,110 --> 00:37:51,230
when I speak with right leaning 
voters or right leaning 

736
00:37:51,590 --> 00:37:54,510
podcasters, if you have any 
suggestions for ones I should go

737
00:37:54,510 --> 00:37:57,350
on, I'm always trying to reach 
out to a more and more diverse 

738
00:37:57,350 --> 00:37:59,820
audience. 
If your listeners can think of 

739
00:37:59,820 --> 00:38:01,820
great ones for me to go on, 
e-mail me at 

740
00:38:01,820 --> 00:38:06,540
francis@francishogan.com. 
What I would say is, if you care

741
00:38:06,540 --> 00:38:10,260
about freedom of speech, you 
should be demanding transparency

742
00:38:10,260 --> 00:38:14,380
about these censorship systems. 
You know, when I talk to women's

743
00:38:14,380 --> 00:38:18,460
rights advocates around the 
world, all of the have been 

744
00:38:18,460 --> 00:38:22,420
kicked off Facebook because 
these A I systems are so crude 

745
00:38:22,940 --> 00:38:26,700
that if you talk about violence 
against women, the A I thinks 

746
00:38:26,700 --> 00:38:28,100
you're committing violence 
against women. 

747
00:38:29,200 --> 00:38:31,400
Like, if you really care about 
for no space, you should be 

748
00:38:31,400 --> 00:38:33,400
marching on the streets for 
transparency. 

749
00:38:34,160 --> 00:38:36,440
And I think that's a space that 
we should all be willing to work

750
00:38:36,480 --> 00:38:39,080
at. 
I totally get sort of the 

751
00:38:39,080 --> 00:38:42,280
messaging there and that you 
want everyone to be sort of on 

752
00:38:42,280 --> 00:38:44,360
board with these issues. 
I mean, you look at even the 

753
00:38:44,360 --> 00:38:48,560
threads roll out when they 
launched it, they had some, I 

754
00:38:48,560 --> 00:38:52,520
think, tools in there early on 
that said, oh, you know, this 

755
00:38:52,520 --> 00:38:55,560
person, I forget what the exact 
language was, it was like. 

756
00:38:56,080 --> 00:38:59,680
Oh, you know, they, I forget if 
it was the past false things or 

757
00:38:59,680 --> 00:39:00,920
I'm not going to get it totally 
right. 

758
00:39:00,920 --> 00:39:04,920
But there were warnings before 
people reshared them, which then

759
00:39:05,040 --> 00:39:08,640
like can have sort of a 
perceived sort of partisan 

760
00:39:08,640 --> 00:39:10,040
focus. 
And obviously, I mean, Elon 

761
00:39:10,040 --> 00:39:15,200
Musk, whole campaign on Twitter 
was the idea that Twitter was 

762
00:39:15,760 --> 00:39:19,400
sort of shadow banning, you 
know, people and the replies and

763
00:39:19,400 --> 00:39:22,040
all that. 
Yeah, yeah, I don't know. 

764
00:39:22,120 --> 00:39:23,920
I guess the question there is 
like. 

765
00:39:24,990 --> 00:39:27,830
Clearly what we're seeing is 
there is some sort of right wing

766
00:39:27,830 --> 00:39:31,670
backlash to cases where social 
media companies try to do what 

767
00:39:31,670 --> 00:39:35,670
you're saying, which is sort of 
either flag or not sort of a 

768
00:39:35,670 --> 00:39:37,870
size things that violate their 
policies. 

769
00:39:37,950 --> 00:39:40,630
I think this again, it's funny 
since I actually worry about 

770
00:39:40,750 --> 00:39:44,670
this a lot for AI safety. 
If you already feel distressed, 

771
00:39:44,950 --> 00:39:47,510
you know you feel like you have 
been left behind or you're a 

772
00:39:47,510 --> 00:39:50,440
marginalist. 
And, you know, I grew up in 

773
00:39:50,440 --> 00:39:52,440
Iowa. 
Like, I have a lot of empathy 

774
00:39:52,440 --> 00:39:56,480
for Republican voters today. 
You know, Iowa has been left 

775
00:39:56,480 --> 00:39:58,800
behind economically in a pretty 
dramatic way. 

776
00:39:59,320 --> 00:40:03,360
That if you already are feeling 
a little that anxiety about the 

777
00:40:03,360 --> 00:40:06,000
idea that people with power 
don't really care about you, 

778
00:40:06,480 --> 00:40:11,200
it's very easy to read into when
a moderation system takes an 

779
00:40:11,320 --> 00:40:13,280
impact that you're being singled
out. 

780
00:40:13,880 --> 00:40:16,800
And I think this is also true 
for, say, African American or 

781
00:40:16,800 --> 00:40:19,720
less affluent people who 
participate, only that you have 

782
00:40:20,120 --> 00:40:23,080
very similar things where like 
African Americans will get 

783
00:40:23,080 --> 00:40:26,160
sanctioned for hate speech 
because these systems are not 

784
00:40:26,160 --> 00:40:29,000
very well done. 
And so I think there's a fear, 

785
00:40:29,120 --> 00:40:32,600
and this is a fear that could be
addressed by being more 

786
00:40:32,600 --> 00:40:34,760
transparent of saying, hey, 
we're going to actually let you 

787
00:40:34,760 --> 00:40:38,480
see what we're doing because you
know the case of a I safety, 

788
00:40:38,880 --> 00:40:41,520
you're already seeing people 
come out with calls of saying I 

789
00:40:41,520 --> 00:40:45,010
don't want. 
Systems that have been aligned 

790
00:40:45,010 --> 00:40:48,370
with the public good. 
I want true or real AI. 

791
00:40:48,930 --> 00:40:52,090
And I think that in both cases, 
either what conservatives are 

792
00:40:52,090 --> 00:40:56,570
saying about content, moderation
on threats, or AI safety, you 

793
00:40:56,570 --> 00:40:59,730
know, when people feel like it's
out of their control, when they 

794
00:40:59,730 --> 00:41:02,650
feel like something's being done
behind the scenes, they object 

795
00:41:02,650 --> 00:41:06,530
to it. 
So giving the user control 

796
00:41:06,570 --> 00:41:09,650
basically. 
Or having enough transparency 

797
00:41:09,650 --> 00:41:11,650
that you can build trust on 
these things. 

798
00:41:12,040 --> 00:41:13,560
Right. 
Like it should be possible for 

799
00:41:13,560 --> 00:41:16,200
researchers to come out and say,
like, hey, actually, or if you 

800
00:41:16,200 --> 00:41:18,400
yourself. 
So right now, if your content 

801
00:41:18,400 --> 00:41:22,600
gets taken down on Facebook, the
only people you can appeal to or

802
00:41:22,600 --> 00:41:25,120
you can send your thing off to 
is the oversight board. 

803
00:41:25,880 --> 00:41:28,240
Imagine if you would say, I'd 
like to be put in a public 

804
00:41:28,240 --> 00:41:31,080
research data set. 
Like, I would like reporters to 

805
00:41:31,080 --> 00:41:34,000
be able to look at this and say,
oh, interesting, you know, this 

806
00:41:34,000 --> 00:41:36,640
whole political candidate is 
getting taken down at a much 

807
00:41:36,640 --> 00:41:39,450
higher rate than this other one.
Like that would be a very 

808
00:41:39,450 --> 00:41:41,370
different world. 
Facebook would have to work 

809
00:41:41,370 --> 00:41:44,730
harder to make sure that its 
systems were objective and 

810
00:41:44,730 --> 00:41:48,370
effective. 
In the book Power One which 

811
00:41:48,370 --> 00:41:51,050
everyone should go read and you 
talk about like the rise of 

812
00:41:51,050 --> 00:41:55,370
COVID and like I mean COVID sort
of is a great way to get into 

813
00:41:55,370 --> 00:42:00,010
the misinformation question and 
I think like how misinformation 

814
00:42:00,090 --> 00:42:03,610
like the view on sort of that 
word and sort of the ideas 

815
00:42:03,610 --> 00:42:06,690
there. 
What is your view on like how 

816
00:42:06,770 --> 00:42:10,640
Facebook. 
Handled COVID, like, are you 

817
00:42:10,640 --> 00:42:16,160
supportive of, you know, tamping
down on people who were like, 

818
00:42:16,440 --> 00:42:19,080
you know, skeptical about the 
origins of the vaccine or 

819
00:42:19,080 --> 00:42:21,640
skeptical about the vaccine? 
Because that fits right into 

820
00:42:21,640 --> 00:42:24,800
this sort of partisan thing 
where in there end up being sort

821
00:42:24,800 --> 00:42:28,400
of sides that align with parties
that fit onto these. 

822
00:42:28,400 --> 00:42:32,000
So yeah, how do you score 
Facebook's handling of COVID 

823
00:42:32,000 --> 00:42:35,640
misinformation? 
Oh, great question, God. 

824
00:42:36,360 --> 00:42:38,880
It's interesting. 
My criticisms on Facebook are 

825
00:42:38,880 --> 00:42:42,360
much more like the part of the 
book that I I think you're 

826
00:42:42,360 --> 00:42:46,440
alluding to is, you know, when 
they went and divided up the 

827
00:42:46,440 --> 00:42:49,520
United States into 600 
communities. 

828
00:42:50,000 --> 00:42:53,280
So think of a community as you 
enjoy the same kind of groups, 

829
00:42:53,280 --> 00:42:56,280
You follow the same kind of 
pages, you post on the same 

830
00:42:56,280 --> 00:42:58,600
kinds of topics, you click on 
the same kinds of topics. 

831
00:42:58,880 --> 00:43:01,240
You know, imagine you put people
into communities that were 

832
00:43:01,240 --> 00:43:03,640
between 500,000 people and 3 
million people. 

833
00:43:04,700 --> 00:43:09,020
If you went and said, you know 
how many of these communities 

834
00:43:09,260 --> 00:43:14,740
make up 80% of all the received 
COVID misinformation, it turned 

835
00:43:14,740 --> 00:43:18,940
out that 4% of the US population
fell into communities that got 

836
00:43:19,180 --> 00:43:22,700
80% of the COVID misinformation,
right? 

837
00:43:22,700 --> 00:43:27,180
So you and me, average person 
gets maybe one piece of COVID 

838
00:43:27,180 --> 00:43:30,580
misinformation every couple 
days, once a week. 

839
00:43:31,280 --> 00:43:34,720
For a small fraction of people, 
they were getting whole streams 

840
00:43:34,720 --> 00:43:35,960
of it. 
That's what happened with the 

841
00:43:35,960 --> 00:43:38,920
January 6th people, but for 
different groups, different 

842
00:43:38,920 --> 00:43:41,760
ideas. 
And and part of that was because

843
00:43:41,920 --> 00:43:45,440
the way Facebook was designed 
was if you have a post that is 

844
00:43:45,440 --> 00:43:49,320
really controversial, you know 
has a big fight in the comments,

845
00:43:49,760 --> 00:43:54,120
Every new comment makes that 
post new and can show up at the 

846
00:43:54,120 --> 00:43:57,590
top of your home feed again. 
And so there were a few 

847
00:43:57,590 --> 00:44:00,910
communities where, you know, 
COVID had a really intense 

848
00:44:00,910 --> 00:44:04,990
emotional balance. 
Those communities actively 

849
00:44:04,990 --> 00:44:08,510
censored out voices that said 
anything different and they 

850
00:44:08,510 --> 00:44:10,150
became these kind of echo 
chambers. 

851
00:44:10,590 --> 00:44:12,430
And so I think there's 
interesting conversations to be 

852
00:44:12,430 --> 00:44:16,030
had around, like how do 
phenomena like that occur and 

853
00:44:16,030 --> 00:44:18,030
like what other contexts are 
they happening in? 

854
00:44:18,630 --> 00:44:21,150
And so like, let's imagine 
you're seeing something like 

855
00:44:21,150 --> 00:44:24,070
that where you know, the 
algorithm and the product design

856
00:44:24,070 --> 00:44:27,620
are pushing people towards. 
These kind of parallel 

857
00:44:27,620 --> 00:44:29,540
realities. 
And like, that's why you had 

858
00:44:29,540 --> 00:44:32,740
people showing up and like, you 
know, threatening teachers with 

859
00:44:32,740 --> 00:44:35,500
guns or like people showing up 
at school board meetings 

860
00:44:35,500 --> 00:44:37,980
screaming because literally when
they looked at their Facebook 

861
00:44:37,980 --> 00:44:41,620
feeds, all they saw was stuff 
about how teachers were trying 

862
00:44:41,620 --> 00:44:44,540
to kill their kids, Right? 
So that's kind of the 

863
00:44:44,540 --> 00:44:47,980
environment we're entering into 
when we have to say, like, what 

864
00:44:47,980 --> 00:44:50,660
is Facebook's role or what 
should Facebook's policy be? 

865
00:44:51,420 --> 00:44:53,380
And I think it's really 
complicated. 

866
00:44:53,620 --> 00:44:55,500
Like, I think they did a bad 
job. 

867
00:44:56,030 --> 00:44:59,910
In that they had, you know, 
these blacklists, they had the 

868
00:44:59,910 --> 00:45:02,030
concepts that you couldn't talk 
about, but they never told 

869
00:45:02,030 --> 00:45:06,070
anyone, you know, They never let
us see how well these systems 

870
00:45:06,070 --> 00:45:08,590
performed. 
And it meant that you had people

871
00:45:08,590 --> 00:45:12,510
feel like they weren't, like it 
was a hidden truth and hidden 

872
00:45:12,510 --> 00:45:16,630
truths are very alluring. 
And so, you know, effective 

873
00:45:16,710 --> 00:45:20,190
social software should be 
designed such that if I write a 

874
00:45:20,190 --> 00:45:24,070
thoughtful reply, if I go do 
research, if I come back to your

875
00:45:24,070 --> 00:45:26,920
allegation with. 
Something meaningful for a 

876
00:45:26,920 --> 00:45:29,240
conversation. 
I should get similar amounts of 

877
00:45:29,240 --> 00:45:32,120
distribution to your 
inflammatory statement, and that

878
00:45:32,120 --> 00:45:34,320
just doesn't happen today. 
Like, the systems aren't 

879
00:45:34,320 --> 00:45:36,600
designed to reward extreme 
ideas. 

880
00:45:36,600 --> 00:45:39,960
They're not designed to reward 
thoughtful, moderate ideas. 

881
00:45:39,960 --> 00:45:41,920
So. 
Your solution there would be 

882
00:45:41,920 --> 00:45:45,070
just to. 
To give more distribution to 

883
00:45:45,070 --> 00:45:46,950
sort of counter messaging 
basically. 

884
00:45:48,350 --> 00:45:50,910
So like, they've done research 
inside of Facebook, and one 

885
00:45:50,910 --> 00:45:53,350
thing that's kind of exciting 
about the next few months is 

886
00:45:53,350 --> 00:45:57,510
like Harvard has an archive of 
most of the Facebook files and 

887
00:45:57,510 --> 00:46:00,630
they're starting to make access 
for those documents to academic 

888
00:46:00,630 --> 00:46:04,430
researchers. 
One of the papers in there talks

889
00:46:04,430 --> 00:46:10,510
about how, if you are good at 
writing such that people diverse

890
00:46:10,510 --> 00:46:13,670
from you different from you. 
Can engage with that content. 

891
00:46:13,670 --> 00:46:18,350
Are those ideas You're doing a 
very complicated thing, right? 

892
00:46:18,350 --> 00:46:21,830
You know, if I can write a 
political post that Republicans 

893
00:46:21,830 --> 00:46:23,990
and Democrats can like 
instructively have a 

894
00:46:23,990 --> 00:46:26,670
conversation in the comments, 
you know, people are thumbs up 

895
00:46:26,670 --> 00:46:28,430
and each other. 
It's like a positive 

896
00:46:28,430 --> 00:46:32,070
conversational template. 
I should be rewarded for that 

897
00:46:32,230 --> 00:46:35,310
because it's not obvious. 
Like I might not get the most 

898
00:46:35,870 --> 00:46:37,830
comments, I might not get the 
most likes. 

899
00:46:38,410 --> 00:46:41,250
But if I can get a diverse 
community of people, a diverse 

900
00:46:41,250 --> 00:46:43,770
audience to engage with it, that
should be rewarded. 

901
00:46:44,410 --> 00:46:47,290
If you come in there and say, 
hey, we're going to boost 

902
00:46:47,290 --> 00:46:50,610
content from people who can 
successfully reach across the 

903
00:46:50,610 --> 00:46:54,770
aisle, you end up getting less 
hate speech, less violence for 

904
00:46:54,770 --> 00:46:58,090
free, less misinformation. 
So it's a lot of these things 

905
00:46:58,090 --> 00:47:02,530
where if you start making 
transparent, what is distributed

906
00:47:02,810 --> 00:47:05,650
versus what is created on these 
platforms. 

907
00:47:06,610 --> 00:47:09,810
And you start saying, hey, like 
right now what you're giving 

908
00:47:09,810 --> 00:47:11,690
distribution to is very 
different than what's being 

909
00:47:11,690 --> 00:47:13,370
created. 
Can we get those a little bit 

910
00:47:13,370 --> 00:47:16,290
more in line? 
I have a feeling what would end 

911
00:47:16,290 --> 00:47:18,650
up happening as you'd start 
finding more techniques like 

912
00:47:18,650 --> 00:47:22,410
that where you could come in and
say, hey, people who are fear 

913
00:47:22,410 --> 00:47:26,130
mongering, you know, they 
shouldn't be the only ones who 

914
00:47:26,130 --> 00:47:28,970
get to stand on the stage. 
And that's kind of what's 

915
00:47:28,970 --> 00:47:32,330
happening today. 
So it's still a world where like

916
00:47:32,330 --> 00:47:36,500
there's like sort of? 
A hand on the till in terms of 

917
00:47:36,860 --> 00:47:40,580
what's getting reach and like 
it's just sort of in some ways 

918
00:47:40,580 --> 00:47:43,100
just doing things that are less 
oriented around sort of their 

919
00:47:43,100 --> 00:47:46,580
business interests necessarily, 
which are Max engagement, right.

920
00:47:46,580 --> 00:47:51,100
I mean it's certainly not saying
like don't sort of weigh certain

921
00:47:51,100 --> 00:47:54,100
things differently than others, 
right, Because if you just tried

922
00:47:54,100 --> 00:47:57,420
to create this sort of like 
neutral algorithm, you would 

923
00:47:57,420 --> 00:48:00,180
just then be deferring to sort 
of. 

924
00:48:00,690 --> 00:48:02,730
Negative aspects of humanity 
potentially. 

925
00:48:02,730 --> 00:48:04,130
Am I understanding that 
correctly? 

926
00:48:04,210 --> 00:48:07,650
Yeah, algorithms can only 
capture the level of complexity 

927
00:48:07,650 --> 00:48:11,410
that you put in the algorithm. 
So if you come in there and say,

928
00:48:11,410 --> 00:48:16,170
hey, people developing 
compulsive behaviors, right? 

929
00:48:16,170 --> 00:48:20,010
So in the case of kids, you 
know, something like 8% of kids 

930
00:48:20,010 --> 00:48:23,850
say that they take control of 
usage and it's hurting either 

931
00:48:23,850 --> 00:48:26,410
their employment, school or 
their health, right? 

932
00:48:26,410 --> 00:48:30,210
Think about that 8%. 
How selfaware are 14 year olds 

933
00:48:30,290 --> 00:48:32,450
like, not super selfaware. 
How honest are they with 

934
00:48:32,450 --> 00:48:34,250
themselves? 
Not super honest. 

935
00:48:34,490 --> 00:48:36,490
It's probably a lot more than 
8%, right? 

936
00:48:36,810 --> 00:48:39,850
They're suffering like that. 
You know, if you don't have a 

937
00:48:39,850 --> 00:48:44,290
system that says, hey, sleep 
deprived kids are a longterm 

938
00:48:44,330 --> 00:48:46,770
harm to society, you know, they 
do worse in school. 

939
00:48:47,050 --> 00:48:49,970
They are at higher risk for 
developing mental health issues 

940
00:48:49,970 --> 00:48:51,810
that will last so long 
throughout their life or put 

941
00:48:51,810 --> 00:48:53,330
them at higher risk of 
recurrence. 

942
00:48:53,690 --> 00:48:55,770
They're more likely to use 
drugs, both uppers because 

943
00:48:55,770 --> 00:48:57,730
they're tired, Downers because 
they're depressed. 

944
00:48:58,200 --> 00:49:00,640
You know, if you have an 
algorithm that just is like, how

945
00:49:00,640 --> 00:49:02,880
many clicks can we get? 
Like how much ad revenue can we 

946
00:49:02,880 --> 00:49:05,440
get? 
You don't capture those kinds of

947
00:49:05,440 --> 00:49:09,000
social costs. 
And so I think there's a huge 

948
00:49:09,000 --> 00:49:12,920
opportunity where if you just 
come in and say like, hey, with 

949
00:49:12,920 --> 00:49:16,680
cars, if you have a car 
accident, everyone gets to see 

950
00:49:16,680 --> 00:49:19,120
the car accident. 
You know, everyone gets to see 

951
00:49:19,120 --> 00:49:21,320
the body on the ground and see 
that your, you know, your seat 

952
00:49:21,320 --> 00:49:25,360
belt didn't work. 
We don't have a similar feedback

953
00:49:25,360 --> 00:49:28,480
cycle. 
On these social platforms, and 

954
00:49:28,480 --> 00:49:31,920
they can keep kind of having 
their problems and they don't 

955
00:49:31,920 --> 00:49:33,480
have anything that brings them 
back to center. 

956
00:49:33,760 --> 00:49:35,840
For the last part of the 
conversation, just talk about, 

957
00:49:36,120 --> 00:49:41,360
you know, the decision to 
release information and to go to

958
00:49:41,360 --> 00:49:44,680
the Wall Street Journal. 
And, yeah, the reporter, it's 

959
00:49:44,680 --> 00:49:46,880
Jeff Horowitz, right? 
He reached out to you on 

960
00:49:46,880 --> 00:49:48,600
LinkedIn. 
Or tell the story a little bit 

961
00:49:48,600 --> 00:49:51,680
about how this happened. 
Like, did you see yourself? 

962
00:49:51,800 --> 00:49:55,680
Sure, like as a whistleblower 
before you heard from him? 

963
00:49:56,410 --> 00:49:59,410
I had been contemplating for 
quite a while. 

964
00:50:00,170 --> 00:50:02,730
Like, would I have to go forward
at some point? 

965
00:50:03,090 --> 00:50:05,570
Like, I had a chance to talk to 
my parents about it a large 

966
00:50:05,570 --> 00:50:08,330
number of times, just because, 
like, what I was seeing on a 

967
00:50:08,330 --> 00:50:11,890
daytoday basis while I lived 
with them during COVID was just 

968
00:50:12,090 --> 00:50:14,890
so different than what the 
Facebook public narrative was on

969
00:50:14,890 --> 00:50:17,370
these issues. 
But the moment where I was like,

970
00:50:17,370 --> 00:50:21,690
OKI have no other options was 
right after the 2020 election. 

971
00:50:21,690 --> 00:50:24,370
So this is in December. 
It's like less than 30 days 

972
00:50:24,370 --> 00:50:27,440
after the election. 
They pulled us all together on 

973
00:50:27,440 --> 00:50:31,080
Zoom, and they said, hey, you 
know how for the last four 

974
00:50:31,080 --> 00:50:35,760
years, the only part of Facebook
that was growing was the civic 

975
00:50:35,760 --> 00:50:39,080
integrity team. 
So it's the team that was for 

976
00:50:39,080 --> 00:50:42,520
facebook.com, at least. 
So civic Integrity's job was to 

977
00:50:42,520 --> 00:50:45,240
make sure that Facebook was a 
positive force in the world, 

978
00:50:45,240 --> 00:50:46,920
like a positive social force in 
the world. 

979
00:50:47,480 --> 00:50:49,160
You know, it wasn't going to 
disrupt elections. 

980
00:50:49,200 --> 00:50:52,000
It wasn't going to cause any 
more genocides, because by that 

981
00:50:52,000 --> 00:50:54,920
point there'd been 2. 
You know, it was going to be a 

982
00:50:54,920 --> 00:50:58,480
positive force. 
And they said, hey, you are so 

983
00:50:58,480 --> 00:51:01,640
important, we're going to 
dissolve your team and integrate

984
00:51:01,640 --> 00:51:06,880
it into the rest of Facebook. 
And when they did that, that was

985
00:51:06,880 --> 00:51:08,880
kind of the moment where I 
realized Facebook had given up, 

986
00:51:09,280 --> 00:51:12,360
that the only way Facebook was 
going to save itself was if the 

987
00:51:12,360 --> 00:51:15,160
public got involved, that the 
public had to come and save 

988
00:51:15,160 --> 00:51:19,080
Facebook. 
And so by chance, so that day I 

989
00:51:19,080 --> 00:51:22,880
went and opened up LinkedIn 
because, you know, I was kind 

990
00:51:22,880 --> 00:51:24,710
of. 
When you have instability at 

991
00:51:24,710 --> 00:51:27,030
work, that's what you do. 
You know, you open up LinkedIn. 

992
00:51:27,950 --> 00:51:33,510
And I saw that I had a message 
from this guy, Jeff Horwitz, and

993
00:51:33,670 --> 00:51:35,790
he did a lot of reporting for 
the Wall Street Journal about 

994
00:51:35,790 --> 00:51:38,590
the violence that Facebook had 
facilitated in India, 

995
00:51:38,830 --> 00:51:40,830
particularly like Muslim Hindu 
violence. 

996
00:51:41,430 --> 00:51:46,670
And, you know, he said, do you 
want to talk? 

997
00:51:47,150 --> 00:51:51,270
And I was like, oh, interesting,
like of all the places that I 

998
00:51:51,270 --> 00:51:53,840
would want to work with. 
I wanted to work with the Wall 

999
00:51:53,840 --> 00:51:57,240
Street Journal because I really 
view all of these topics as 

1000
00:51:57,240 --> 00:51:59,520
bipartisan. 
You know, they're not left, 

1001
00:51:59,520 --> 00:52:02,240
they're not right. 
They're like basic rules of the 

1002
00:52:02,240 --> 00:52:04,680
road. 
And I knew that if the reporting

1003
00:52:04,680 --> 00:52:08,160
had come from the New York Times
that there would be a large 

1004
00:52:08,160 --> 00:52:10,960
swath of right leaning voters 
that would never be able to 

1005
00:52:10,960 --> 00:52:12,840
trust it. 
But if it came from the Wall 

1006
00:52:12,840 --> 00:52:16,000
Street Journal, it was likely 
that they would be able to at 

1007
00:52:16,000 --> 00:52:19,280
least consider it. 
And I think that's part of why 

1008
00:52:19,280 --> 00:52:22,320
the Senate hearing was so 
bipartisan because. 

1009
00:52:22,750 --> 00:52:25,630
It was something that came out 
of, you know, it was trusted 

1010
00:52:25,910 --> 00:52:27,710
like that. 
A lot of the things that were 

1011
00:52:27,710 --> 00:52:29,710
said in those articles sounded 
crazy. 

1012
00:52:29,910 --> 00:52:32,590
They sounded super. 
There's like, no way, there's no

1013
00:52:32,590 --> 00:52:34,990
way the company could be this 
bad, like the human trafficking 

1014
00:52:34,990 --> 00:52:38,550
thing of the Facebook employees 
being worried about offending 

1015
00:52:38,550 --> 00:52:40,590
buyers over the people being 
trafficked. 

1016
00:52:41,190 --> 00:52:45,790
But because it came from a very,
you know, center of the road, 

1017
00:52:45,910 --> 00:52:48,590
cautious publication, people 
were like, oh, maybe this is 

1018
00:52:48,590 --> 00:52:52,580
actually true. 
And yeah, I mean, I think this 

1019
00:52:52,580 --> 00:52:56,340
journal is so credible and sort 
of careful and how they present 

1020
00:52:56,340 --> 00:53:00,900
things. 
Initially, you were not going to

1021
00:53:00,900 --> 00:53:02,820
come out, right? 
Like what motivated your 

1022
00:53:02,820 --> 00:53:06,140
decision to go public? 
So wondering why the disclosures

1023
00:53:06,140 --> 00:53:07,940
are so large. 
So I want them to be able to 

1024
00:53:07,940 --> 00:53:10,340
stand on their own like I always
expected to do. 

1025
00:53:10,340 --> 00:53:13,220
Like closed door briefings but 
like governments to be able to 

1026
00:53:13,220 --> 00:53:16,100
explain them. 
But I never intended to be like 

1027
00:53:16,100 --> 00:53:20,450
part of the story. 
And right before I came out, so 

1028
00:53:20,450 --> 00:53:24,610
maybe a couple months before, my
lawyers started really putting 

1029
00:53:24,610 --> 00:53:27,330
pressure on me where they were 
saying, you know, Facebook knows

1030
00:53:27,330 --> 00:53:30,290
it's and they can look at all 
the different documents, right? 

1031
00:53:30,290 --> 00:53:32,370
They know it's you. 
You know, the report is going to

1032
00:53:32,370 --> 00:53:33,690
get started. 
They're going to figure out real

1033
00:53:33,690 --> 00:53:35,330
fast or as soon as they start 
going to ask for comment, 

1034
00:53:35,330 --> 00:53:36,570
they're going to figure out real
fast. 

1035
00:53:37,290 --> 00:53:42,650
And the Wall Street Journal, 
they said here's the deal. 

1036
00:53:43,060 --> 00:53:47,500
Like either you can wait for the
rest of your life for Facebook 

1037
00:53:47,500 --> 00:53:51,180
to present you to the public. 
You know, every day you're like 

1038
00:53:51,220 --> 00:53:53,500
open, you know Google News. 
You're going to open your 

1039
00:53:53,500 --> 00:53:56,500
newspaper and be like, is today 
the day that Facebook is going 

1040
00:53:56,500 --> 00:53:58,340
to introduce me in the worst 
possible way? 

1041
00:53:58,900 --> 00:54:00,900
Or you can take responsibility 
for what you did. 

1042
00:54:01,620 --> 00:54:03,420
You know, I know that you don't 
want to be out. 

1043
00:54:03,420 --> 00:54:04,420
You don't want people looking at
you. 

1044
00:54:04,420 --> 00:54:07,740
You don't want to be out there. 
Like, you know, go and take 

1045
00:54:07,740 --> 00:54:10,820
responsibility because if you 
do, close door briefings. 

1046
00:54:11,540 --> 00:54:14,940
Every single briefing you do 
will expand the circle of trust,

1047
00:54:15,940 --> 00:54:19,180
and you're going to be like the 
juiciest story for some reporter

1048
00:54:19,180 --> 00:54:20,580
to break. 
Like if they can. 

1049
00:54:20,700 --> 00:54:22,780
Find post or somebody would have
just. 

1050
00:54:23,720 --> 00:54:26,640
And one of my friends joked, 
like, after I came out, they 

1051
00:54:26,640 --> 00:54:29,680
were like, I don't understand 
how this story stayed quiet for 

1052
00:54:29,680 --> 00:54:31,200
so long. 
I thought, everyone in San 

1053
00:54:31,200 --> 00:54:32,960
Francisco, I knew Francis was 
doing this. 

1054
00:54:33,360 --> 00:54:35,720
And I was like, everyone in our 
circle of friends knew I was 

1055
00:54:35,720 --> 00:54:37,640
doing this. 
Not everyone in San Francisco. 

1056
00:54:37,840 --> 00:54:40,360
But still, it's one of those 
things, like, someone brags to 

1057
00:54:40,360 --> 00:54:42,680
their friend about how I know 
where the Facebook whistleblower

1058
00:54:42,680 --> 00:54:48,480
is and, you know, I just goes. 
And so I decided to step out 

1059
00:54:48,480 --> 00:54:50,980
into the light. 
And it's actually been a really 

1060
00:54:50,980 --> 00:54:53,060
transformative experience. 
You know, it's one of these 

1061
00:54:53,060 --> 00:54:57,260
things where I spent a lot of my
life, like, really trying to 

1062
00:54:57,260 --> 00:54:59,940
avoid being seen. 
You know, like, I've gotten 

1063
00:54:59,940 --> 00:55:02,140
married twice. 
I eloped both times. 

1064
00:55:02,500 --> 00:55:05,220
The idea of standing in front of
a crowd, having them just, like,

1065
00:55:05,220 --> 00:55:07,300
stare at me for my wedding 
sounds very stressful. 

1066
00:55:07,540 --> 00:55:09,580
You know, I've had two birthday 
parties in the last, like, 20 

1067
00:55:09,580 --> 00:55:11,500
years. 
Like, it's just not my jail. 

1068
00:55:11,820 --> 00:55:14,620
But it's been really interesting
having to, like, stand up and 

1069
00:55:14,620 --> 00:55:16,860
try to educate people on these 
issues. 

1070
00:55:17,290 --> 00:55:21,050
Because, you know, I really 
believe in democracy and the way

1071
00:55:21,050 --> 00:55:23,170
the world changes is we change 
it. 

1072
00:55:24,010 --> 00:55:27,010
You know, we get out there and 
we say, hey, you know, I'm going

1073
00:55:27,010 --> 00:55:29,130
to keep repeating this until I 
see something different in the 

1074
00:55:29,130 --> 00:55:31,930
world. 
And having to show up in my own 

1075
00:55:31,930 --> 00:55:35,290
life has been a huge blessing. 
And I, something I never 

1076
00:55:35,290 --> 00:55:37,970
expected to be one of the things
that I would walk away, you 

1077
00:55:37,970 --> 00:55:40,050
know, two years later and say 
I'm so grateful for this. 

1078
00:55:41,200 --> 00:55:43,040
Awesome. 
That is a great ending to the 

1079
00:55:43,040 --> 00:55:45,400
podcast. 
Thank you so much for coming on 

1080
00:55:45,400 --> 00:55:47,200
and I really appreciate talking 
to. 

1081
00:55:47,960 --> 00:55:48,680
You. 
My pleasure. 

1082
00:55:49,040 --> 00:55:52,520
The honest thing I leave people 
with is if you think all these 

1083
00:55:52,520 --> 00:55:55,400
things are too complicated, you 
know, like these are, you know I

1084
00:55:55,400 --> 00:55:56,920
couldn't understand a technical 
book. 

1085
00:55:57,600 --> 00:56:01,760
The whole point of power of 1 is
that democracy requires an 

1086
00:56:01,760 --> 00:56:04,880
informed population. 
So it is written so that if you 

1087
00:56:04,880 --> 00:56:08,040
could follow along today, I 
guarantee you'll be entertained 

1088
00:56:08,120 --> 00:56:11,120
and follow along the power of 1.
And there's lots of fun stories 

1089
00:56:11,240 --> 00:56:13,480
and, you know, crazy hijinks 
along the way. 

1090
00:56:13,480 --> 00:56:15,120
Because I've always got into 
trouble. 

1091
00:56:15,480 --> 00:56:16,560
You know what? 
I didn't mean to. 

1092
00:56:16,840 --> 00:56:18,640
And so I hope you come along on 
this journey too. 

1093
00:56:19,240 --> 00:56:21,560
Thank you so much. 
That's our episode. 

1094
00:56:21,560 --> 00:56:23,800
Thanks so much to Tommy Herron, 
our audio editor. 

1095
00:56:23,920 --> 00:56:26,400
I want to shout out. 
Annie Wen, our intern for the 

1096
00:56:26,400 --> 00:56:30,040
summer, has been helping me prep
for the podcast and working on 

1097
00:56:30,400 --> 00:56:33,000
punching up the show. 
Shout out to Riley Kinsella, my 

1098
00:56:33,000 --> 00:56:36,840
chief of staff, and young 
Chomsky for the theme music. 

1099
00:56:37,240 --> 00:56:39,080
This has been the newcomer 
podcast. 

1100
00:56:39,080 --> 00:56:43,440
Please like comments, subscribe 
on YouTube, give us a review on 

1101
00:56:43,440 --> 00:56:46,520
Apple Podcasts, and most 
importantly, most important, 

1102
00:56:46,680 --> 00:56:50,040
subscribe to the Sub Stack 
newcomer.co Become a paid 

1103
00:56:50,040 --> 00:56:53,280
subscriber today. 
It makes this all possible. 

1104
00:56:53,560 --> 00:56:55,800
Thank you so much. 
See you next week. 

1105
00:56:56,120 --> 00:56:57,320
Goodbye. 
Goodbye. 

1106
00:56:57,480 --> 00:56:58,680
Goodbye. 
Goodbye. 

1107
00:56:58,800 --> 00:56:59,680
Goodbye. 
Goodbye. 

1108
00:56:59,800 --> 00:57:00,680
Goodbye, Goodbye, Goodbye.