1
00:00:06,200 --> 00:00:14,400
Welcome. 
Hey everybody. 

2
00:00:14,400 --> 00:00:16,500
Welcome to this week's episode 
of dead cat. 

3
00:00:16,500 --> 00:00:21,000
This is Cam joined as always by 
Eric and Katie joined this week 

4
00:00:21,000 --> 00:00:24,100
by our very special, guest Alex 
Stamos, the director of the 

5
00:00:24,100 --> 00:00:27,700
Stanford internet Observatory, 
though, you may all know him for

6
00:00:27,700 --> 00:00:29,900
his earlier work as the chief 
security officer. 

7
00:00:30,000 --> 00:00:32,800
Assert Facebook, where he 
oversaw Security on their 

8
00:00:32,800 --> 00:00:36,300
platforms at a time. 
That things got a little weird 

9
00:00:36,800 --> 00:00:39,500
over at Facebook spicen. 
Yes, by C. 

10
00:00:40,000 --> 00:00:42,400
So this week, we're going to go 
deep on Facebook and Russian 

11
00:00:42,400 --> 00:00:44,800
interference and more 
specifically if we can get 

12
00:00:44,800 --> 00:00:47,200
there. 
How the media reported on this 

13
00:00:47,700 --> 00:00:52,000
crazy brain breaking time that 
we probably will never recover 

14
00:00:52,000 --> 00:00:53,800
from before we get to that 
though. 

15
00:00:53,800 --> 00:00:55,500
Quick public congrats to Friend 
Of The Show. 

16
00:00:55,500 --> 00:00:58,700
Ken Griffin of Citadel Big 
Winners this week. 

17
00:00:58,700 --> 00:01:00,900
Dead cat sponsor chair. 
Auction to win. 

18
00:01:00,900 --> 00:01:04,700
Our in-house copy of the US. 
Constitution can beat out 

19
00:01:04,700 --> 00:01:08,800
constitutional Dao, better luck 
next time guys, and congrats to 

20
00:01:08,800 --> 00:01:10,700
Ken for winning. 
All right, Alex. 

21
00:01:10,700 --> 00:01:14,500
So this week we want to dive 
into a look back on Facebook and

22
00:01:14,500 --> 00:01:17,500
Russia and the media. 
The reason it's interesting now 

23
00:01:17,500 --> 00:01:21,600
is that there was this recent 
arrest of I Gordon chanko an 

24
00:01:21,600 --> 00:01:25,400
analyst who was indicted in an 
investigation led by the doj 

25
00:01:25,900 --> 00:01:29,400
that's looking into all the 
Trump Russia 2016. 

26
00:01:29,900 --> 00:01:32,300
Ian, what have you? 
There's actually a great piece 

27
00:01:32,300 --> 00:01:34,800
about it from Eric when people 
in Washington Post that we can 

28
00:01:34,800 --> 00:01:37,600
link to in the show description.
But to make this short, which is

29
00:01:37,607 --> 00:01:39,800
never easy to do with Russia 
gate stuff. 

30
00:01:39,800 --> 00:01:42,900
This indictment essentially led 
to the collapse of the steel 

31
00:01:42,900 --> 00:01:45,900
dossier and a number of 
attractions for media 

32
00:01:45,900 --> 00:01:49,300
organizations but was never in 
doubt throughout all of the 

33
00:01:49,300 --> 00:01:53,100
investigations was that Russia 
did interfere in the election 

34
00:01:53,600 --> 00:01:56,500
and continues to interfere in 
elections as do many, other 

35
00:01:56,500 --> 00:01:59,800
countries around the world and 
there's probably no one who 

36
00:02:00,000 --> 00:02:02,900
Dance, that better than you. 
Alex, while at Facebook, you 

37
00:02:02,900 --> 00:02:06,200
were one of the first people, to
grasp, really, how Russia use 

38
00:02:06,200 --> 00:02:09,400
the platform for their 
interference campaign. 

39
00:02:09,900 --> 00:02:14,100
And you work directly with the 
molar investigation to explain 

40
00:02:14,100 --> 00:02:16,900
that process. 
And then you got to see a 

41
00:02:16,900 --> 00:02:20,900
covered by the media. 
Basically, in real time poorly, 

42
00:02:20,900 --> 00:02:23,000
what's the score on? 
How is covered by the media in 

43
00:02:23,000 --> 00:02:24,800
real-time? 
I think that's what we're going 

44
00:02:24,800 --> 00:02:26,500
to discuss. 
Yeah, yeah. 

45
00:02:26,500 --> 00:02:31,600
Well, you know, Can you maybe 
walk us through? 

46
00:02:31,600 --> 00:02:34,900
What it was? 
You saw on Facebook's platform. 

47
00:02:34,900 --> 00:02:36,900
Yeah, we've got to get to check 
a launching Point here, the 

48
00:02:36,900 --> 00:02:38,700
weather. 
Yeah, what exactly did you see 

49
00:02:38,900 --> 00:02:40,500
while you're at Facebook? 
And how did you hear going to 

50
00:02:40,500 --> 00:02:43,000
cut right to the end? 
You guys gotta let me warm up 

51
00:02:43,000 --> 00:02:44,800
into the eggs. 
Like I don't want anybody to 

52
00:02:44,800 --> 00:02:47,000
actually know what happened. 
It's a special thing. 

53
00:02:47,000 --> 00:02:50,400
Nobody knows what happened. 
And it's all just hot. 

54
00:02:50,400 --> 00:02:54,000
Takes take first, smart 
blathers. 

55
00:02:54,100 --> 00:02:56,800
Yeah, maybe like, yeah, we will 
sit on the media center enough, 

56
00:02:56,800 --> 00:02:58,600
Eric, but we got to get a 
foundation here. 

57
00:02:58,700 --> 00:03:02,700
You can read all about it. 
Yeah, in an ugly truth by 

58
00:03:02,700 --> 00:03:06,200
Katie's colleagues which is soon
to be a miniseries which means 

59
00:03:06,200 --> 00:03:09,000
all of my friends are to eating 
are texting me with who should 

60
00:03:09,000 --> 00:03:12,100
play me in the, the adaptation 
of what have you heard so far? 

61
00:03:12,100 --> 00:03:13,600
Can we get some casting rumors 
going? 

62
00:03:13,600 --> 00:03:16,000
Well, apparently exact 
Galifianakis is the current 

63
00:03:16,000 --> 00:03:19,500
leader, making a dramatic turn 
as a Surefire way to 1 win, an 

64
00:03:19,500 --> 00:03:22,300
award, but also reinvent your 
career, so it's it went out 

65
00:03:22,300 --> 00:03:23,600
every. 
I think it would be a great move

66
00:03:23,600 --> 00:03:24,900
for him from a dramatic 
perspective. 

67
00:03:24,900 --> 00:03:27,100
I'm hoping for John Stamos, you 
know, even though he's like 10 

68
00:03:27,100 --> 00:03:29,700
years older, I feel like it's 
the same. 

69
00:03:29,900 --> 00:03:32,500
This on Stamos could be the name
of your podcast. 

70
00:03:32,500 --> 00:03:35,900
When you talk about the right, 
I'll have to do a real-time 

71
00:03:35,900 --> 00:03:38,000
podcast with the episode 
recapture. 

72
00:03:38,800 --> 00:03:41,900
I was criticized for making the 
making this to light. 

73
00:03:41,900 --> 00:03:43,200
Now you guys are. 
Okay. 

74
00:03:43,200 --> 00:03:44,900
Kitty, ask a real question. 
Okay. 

75
00:03:44,900 --> 00:03:47,400
Okay. 
Okay, so tell us what you saw in

76
00:03:47,400 --> 00:03:50,400
2016 so let's take. 
So look at the big picture. 

77
00:03:50,900 --> 00:03:54,500
There's like maybe four 
different themes of Russian 

78
00:03:54,500 --> 00:03:55,900
interference in the election, 
right? 

79
00:03:55,900 --> 00:03:58,300
And so the first is, when I'm 
not going to talk about, just to

80
00:03:58,300 --> 00:03:59,800
mention is like all of the 
reach. 

81
00:04:00,000 --> 00:04:02,000
It's in relationship to the 
Trump campaign, right? 

82
00:04:02,000 --> 00:04:05,600
So, you know, all of that stuff 
of which the Steel dossiers part

83
00:04:05,600 --> 00:04:06,900
of it. 
There's a lot of stuff in the 

84
00:04:06,900 --> 00:04:09,700
molar report and then the Senate
Intel reports. 

85
00:04:10,100 --> 00:04:12,700
And so, there's the physical 
stuff right from a propaganda 

86
00:04:12,700 --> 00:04:14,700
perspective. 
Another part we won't really 

87
00:04:14,700 --> 00:04:17,300
talk that much about is, what is
called, kind of white or gray 

88
00:04:17,300 --> 00:04:21,100
propaganda which is the 
propaganda that can be tied to a

89
00:04:21,100 --> 00:04:22,600
government. 
And so that's like Russia Today,

90
00:04:22,600 --> 00:04:24,900
Sputnik that's called White 
propaganda. 

91
00:04:25,200 --> 00:04:29,400
And then kind of the two darker 
more mysterious ways that they 

92
00:04:29,400 --> 00:04:30,400
influence. 
The election. 

93
00:04:31,000 --> 00:04:34,000
The first was the gru activity. 
So the gru is the main 

94
00:04:34,000 --> 00:04:37,200
intelligence, directorate of the
Russian military. 

95
00:04:37,500 --> 00:04:39,600
The gru has existed for a long 
period of time. 

96
00:04:39,600 --> 00:04:43,000
It's Military Intelligence. 
They have a hacking side and 

97
00:04:43,000 --> 00:04:45,400
they have a propaganda side and 
the hacking and propaganda sides

98
00:04:45,400 --> 00:04:47,500
work together. 
So this is what we see from kind

99
00:04:47,500 --> 00:04:50,500
of the really high-end State 
adversaries is these kinds of 

100
00:04:50,500 --> 00:04:54,700
hybrid operations, where you 
have both a offensive computer 

101
00:04:54,700 --> 00:04:57,700
network intrusion capability, 
combined, with the propaganda 

102
00:04:57,700 --> 00:05:00,300
capability. 
And so, with the gru Was they 

103
00:05:00,300 --> 00:05:02,300
broke into the dnc's email 
server. 

104
00:05:02,500 --> 00:05:04,800
They broke into John, podesta is
email, the email thousands of 

105
00:05:04,800 --> 00:05:06,800
other people. 
It turns out most of them didn't

106
00:05:06,800 --> 00:05:10,800
have much to do with this but 
the campaign that they uncovered

107
00:05:10,800 --> 00:05:14,400
Pizza gate, exactly right. 
We broke that thing out. 

108
00:05:14,500 --> 00:05:16,400
Eric's using this to spread 
disinformation. 

109
00:05:16,400 --> 00:05:18,400
Fortunately, if you have a whole
podcast but this information 

110
00:05:18,400 --> 00:05:20,000
Spotify will give you a hundred 
million dollars. 

111
00:05:20,000 --> 00:05:27,800
So there may be this is the way 
for you guys to we're working on

112
00:05:27,800 --> 00:05:31,000
it and so you have the gru 
activity T where they breaking 

113
00:05:31,000 --> 00:05:34,600
all that stuff, they steal all 
this information and then they 

114
00:05:35,100 --> 00:05:38,500
release it in a variety of 
different means, including via 

115
00:05:38,500 --> 00:05:41,100
Wikileaks, including via 
personas that they created 

116
00:05:41,100 --> 00:05:43,700
themselves. 
Their goal was to change the 

117
00:05:43,707 --> 00:05:46,300
conversation about Hillary 
Clinton, so that's the gru 

118
00:05:46,300 --> 00:05:49,100
activity and then there's the 
IRA activity. 

119
00:05:49,100 --> 00:05:51,700
So II RA in this context were 
talking about the Russian 

120
00:05:51,700 --> 00:05:55,600
internet research agency. 
That is a private group. 

121
00:05:55,700 --> 00:05:58,400
They belong to a guy named Ian 
vague deeper Goshen. 

122
00:05:58,700 --> 00:06:01,500
He is an was We considered an 
oligarch in Russia. 

123
00:06:01,800 --> 00:06:04,800
He has this troll Farm, which at
the time was in this building 

124
00:06:04,800 --> 00:06:06,300
st. 
Petersburg, they've actually 

125
00:06:06,300 --> 00:06:08,700
moved since then, but this 
famous building in st. 

126
00:06:08,700 --> 00:06:13,300
Petersburg at which they would 
build propaganda and spread it 

127
00:06:13,300 --> 00:06:16,200
on social media, and they did 
that through the creation of 

128
00:06:16,200 --> 00:06:20,100
thousands of fake accounts of 
all of these personas that would

129
00:06:20,100 --> 00:06:23,100
pretend to be from the countries
where they spread it, and then 

130
00:06:23,100 --> 00:06:25,800
building up audiences and 
pushing it, part of that 

131
00:06:25,800 --> 00:06:28,200
includes advertising. 
And so, the internet research 

132
00:06:28,200 --> 00:06:32,000
agency were and about, Thousand 
dollars in ads on Facebook, the 

133
00:06:32,000 --> 00:06:34,000
ads were not really for the 
creation of the propaganda that 

134
00:06:34,000 --> 00:06:35,700
was for the creation of 
audiences. 

135
00:06:35,900 --> 00:06:40,000
So their goal was to use ads to 
Target groups that might be 

136
00:06:40,000 --> 00:06:43,600
interested in their content and 
then pull them into liking pages

137
00:06:43,900 --> 00:06:46,900
and then the most of the content
was pushed via Facebook pages. 

138
00:06:47,100 --> 00:06:49,600
So those are the identities for 
these groups. 

139
00:06:49,600 --> 00:06:51,700
They would create like a 
pro-immigration group and 

140
00:06:51,700 --> 00:06:54,100
anti-immigration group a pro 
black lives matter, group and 

141
00:06:54,108 --> 00:06:55,800
anti black lives matter group, 
stuff like that. 

142
00:06:56,200 --> 00:06:59,200
Their content in the end with 
something like 80,000 piece of 

143
00:06:59,200 --> 00:07:02,400
content. 
Seen in the end by over 130 

144
00:07:02,400 --> 00:07:05,100
million Americans. 
That was the number one topic 

145
00:07:05,100 --> 00:07:08,200
that they were talking about and
you're at Facebook when this is 

146
00:07:08,200 --> 00:07:13,000
happening or give us where you 
sort of fit into this time, 

147
00:07:13,200 --> 00:07:14,700
right? 
So I was at Facebook, when this 

148
00:07:14,700 --> 00:07:18,500
happened I joined Facebook in 
2015 and as Chief security 

149
00:07:18,500 --> 00:07:21,400
officer my primary job was 
keeping people from hacking 

150
00:07:21,400 --> 00:07:23,900
Facebook, right? 
So actually feel pretty good 

151
00:07:23,900 --> 00:07:26,200
about succeed at my piercing, 
right, right. 

152
00:07:26,300 --> 00:07:28,500
So we actually did a lot of 
really good work on security 

153
00:07:28,500 --> 00:07:31,000
stuff while I was there. 
And So that's like my primary 

154
00:07:31,000 --> 00:07:32,600
job. 
But then I also had a number of 

155
00:07:32,600 --> 00:07:35,400
groups that worked on the abuse 
of the platform to cause harm, 

156
00:07:35,600 --> 00:07:38,100
most specifically there. 
I had investigators that worked 

157
00:07:38,100 --> 00:07:40,400
on child. 
Safety, that worked on 

158
00:07:40,400 --> 00:07:42,000
counterterrorism. 
So we had the counterterrorism 

159
00:07:42,000 --> 00:07:44,400
investigators on my team and 
then we had to this threat 

160
00:07:44,400 --> 00:07:46,000
intelligence team. 
And so the threat intelligence 

161
00:07:46,000 --> 00:07:51,300
team is mostly X us and then a 
couple of other from Western, 

162
00:07:51,300 --> 00:07:54,500
Government intelligence analysts
work to NSA, what the CIA in 

163
00:07:54,500 --> 00:07:56,800
such whose job. 
It was to track different 

164
00:07:56,800 --> 00:07:59,000
government's activity, both 
attacking Facebook. 

165
00:07:59,100 --> 00:08:01,000
But then also, In Facebook to do
bad things. 

166
00:08:01,100 --> 00:08:05,900
And so kind of a normal work 
week for them would be, we found

167
00:08:05,900 --> 00:08:08,900
four accounts that we can tie to
the Islamic revolutionary guard 

168
00:08:08,900 --> 00:08:13,100
Corps and they are spearfishing,
State Department employees to 

169
00:08:13,108 --> 00:08:15,800
try to get access to their 
Facebook accounts, from which 

170
00:08:15,800 --> 00:08:19,600
they can then trick, Iranian 
dissidents into revealing 

171
00:08:19,600 --> 00:08:21,700
themselves and then arrest them.
So, that's like an actual thing 

172
00:08:21,700 --> 00:08:24,700
that we found was this big 
attack by the Iranians because 

173
00:08:24,700 --> 00:08:27,100
the state department where we 
had machine learning that 

174
00:08:27,100 --> 00:08:29,600
detected the kind of activity, 
we figured it all out. 

175
00:08:29,800 --> 00:08:32,299
We worked with the state 
department and the FBI to round 

176
00:08:32,299 --> 00:08:34,600
that all up in to stop it. 
And so, that was their normal 

177
00:08:34,600 --> 00:08:36,799
activity. 
But we got pulled, we got pulled

178
00:08:36,799 --> 00:08:37,600
into this from two different 
ways. 

179
00:08:37,600 --> 00:08:41,100
So first, we had a dedicated 
team working on a PT 28, which 

180
00:08:41,100 --> 00:08:43,600
is the, the hacking team. 
That's tied to the Gru. 

181
00:08:44,000 --> 00:08:48,800
And we saw in the spring of 2016
during the primary process 

182
00:08:49,100 --> 00:08:52,700
activity by Gru that look like 
they were interested in people 

183
00:08:52,700 --> 00:08:54,600
who work for the DN C and D 
Triple C. 

184
00:08:55,100 --> 00:08:59,000
If you're an intelligence agent,
you've been told I want to hack 

185
00:08:59,000 --> 00:09:00,900
somebody. 
Want to hack this organization 

186
00:09:00,900 --> 00:09:02,800
with the first things you do is 
you try to figure out all the 

187
00:09:02,800 --> 00:09:05,200
people that work there and then 
you learn everything you can 

188
00:09:05,200 --> 00:09:07,200
about them from their public 
social media profiles. 

189
00:09:07,400 --> 00:09:09,600
So, we had a bunch of accounts 
that we had previously 

190
00:09:09,600 --> 00:09:11,300
attributed to GRU because their 
activity. 

191
00:09:11,300 --> 00:09:13,700
So we saw them doing stuff in 
the Ukraine and we saw them 

192
00:09:13,700 --> 00:09:16,300
doing hacks against the world, 
anti-doping agency, your 

193
00:09:16,300 --> 00:09:20,600
Facebook, and Linkedin monetize 
that well or do you make exactly

194
00:09:21,900 --> 00:09:23,900
right? 
Yeah, we get ya LinkedIn. 

195
00:09:23,900 --> 00:09:26,300
I mean Facebook doesn't sell 
accounts, but LinkedIn could 

196
00:09:26,300 --> 00:09:28,700
probably sell some Premium 
Accounts to the gru. 

197
00:09:29,300 --> 00:09:32,100
I To most people that reach out 
to me on LinkedIn are affiliated

198
00:09:32,100 --> 00:09:33,600
with some sort of foreign State 
actor. 

199
00:09:33,600 --> 00:09:36,000
So, yeah, it's not, it's not a 
bad Theory. 

200
00:09:36,000 --> 00:09:38,500
And, and so we saw this 
activity, we saw them kind of 

201
00:09:38,500 --> 00:09:40,800
snooping around and so they 
weren't do any actual attacks on

202
00:09:40,800 --> 00:09:44,300
Facebook, but what they were 
doing reconnaissance, which is 

203
00:09:44,300 --> 00:09:47,600
the first step in this. 
The standard operating procedure

204
00:09:47,600 --> 00:09:49,900
from that point, was you go and 
you tell the FBI if they're 

205
00:09:49,900 --> 00:09:53,800
targeting Americans and the FBI 
handles it, we now know FBI 

206
00:09:53,800 --> 00:09:55,300
didn't really do much. 
Like they had all this 

207
00:09:55,300 --> 00:09:58,200
information about different 
kinds of targeting of DNC and 

208
00:09:58,200 --> 00:10:01,800
there's all this crazy stories. 
Like I mean, you know, the 

209
00:10:01,800 --> 00:10:04,400
Hoover building is whatever four
blocks from the DNC headquarters

210
00:10:04,400 --> 00:10:06,600
or something, and somehow the 
message didn't make it, but 

211
00:10:06,600 --> 00:10:08,800
okay. 
And then later in the year, 

212
00:10:08,800 --> 00:10:10,800
after the hacks happen, which 
don't happen on Facebook. 

213
00:10:10,800 --> 00:10:13,000
They happen on Gmail and then 
they happen directly against the

214
00:10:13,000 --> 00:10:15,700
dnc's infrastructure, which the 
DNC should never been running 

215
00:10:15,700 --> 00:10:17,800
their own email server. 
But that's a different issue. 

216
00:10:18,100 --> 00:10:21,200
They come back and they create 
these personas on Facebook 

217
00:10:21,200 --> 00:10:24,800
called DC leaks. 
So we'll Wikileaks is kind of a 

218
00:10:24,900 --> 00:10:27,600
useful idiot to them but they 
didn't have direct control over 

219
00:10:27,600 --> 00:10:29,400
Wikileaks. 
And so, they pretty clearly 

220
00:10:29,400 --> 00:10:32,100
wanted To have personas to leak 
information that were under 

221
00:10:32,100 --> 00:10:34,000
their direct control. 
And so they create these 

222
00:10:34,000 --> 00:10:37,000
personas, and then they start 
reaching out to journalists here

223
00:10:37,000 --> 00:10:38,900
are stolen documents. 
I have, these are the stories 

224
00:10:38,900 --> 00:10:41,500
you should write, right? 
And that is effective, it turns 

225
00:10:41,500 --> 00:10:42,900
out. 
Now, they also release the data 

226
00:10:42,900 --> 00:10:44,100
in a bunch of different ways, 
right? 

227
00:10:44,100 --> 00:10:47,200
And in the end, they get the 
stories written that they want. 

228
00:10:47,600 --> 00:10:51,100
Now, the internet research 
agency stuff, came after of, you

229
00:10:51,100 --> 00:10:53,700
know in the whole discussion of 
fake news and such that we had a

230
00:10:53,700 --> 00:10:56,600
big project to kind of answer 
the question of all of this fake

231
00:10:56,600 --> 00:10:58,800
news content. 
How much of it came from Russia 

232
00:10:58,800 --> 00:11:01,700
now in the long term. 
Just a very small fraction of 

233
00:11:01,700 --> 00:11:04,500
all that the vast majority of 
that content is domestically 

234
00:11:04,500 --> 00:11:07,700
created or family. 
We suck ourselves that we're 

235
00:11:07,700 --> 00:11:09,900
just right, right? 
And or it's created by like 

236
00:11:09,900 --> 00:11:13,600
financially motivated actors 
like the apocryphal but real 

237
00:11:13,600 --> 00:11:16,000
Macedonian teenagers, right? 
Like things like that. 

238
00:11:16,000 --> 00:11:18,100
But you know we did obviously 
found a big chunk of it. 

239
00:11:18,100 --> 00:11:19,800
We found the internet 
researching to see activity and 

240
00:11:19,800 --> 00:11:22,400
then announced that later in 
2017, but those are kind of two 

241
00:11:22,400 --> 00:11:24,000
different things. 
So, on the gru stuff, we were 

242
00:11:24,000 --> 00:11:25,600
ahead of the curve on the IRA 
stuff. 

243
00:11:25,600 --> 00:11:29,000
We were behind the curve. 
And you went to the molar team 

244
00:11:29,000 --> 00:11:32,500
to say, we Found what led to the
I re-indictment you reached out 

245
00:11:32,500 --> 00:11:35,700
to Jeanne re on the team. 
You know, what was her response?

246
00:11:35,700 --> 00:11:36,900
What was that first 
conversation? 

247
00:11:36,900 --> 00:11:39,700
Like yeah. 
So so there's an interesting 

248
00:11:39,700 --> 00:11:43,100
back and forth between the big 
tech companies and Department of

249
00:11:43,100 --> 00:11:46,700
Justice on all kinds of crimes. 
In that we have this this is one

250
00:11:46,700 --> 00:11:48,600
of the interesting things that 
we haven't really figured out. 

251
00:11:48,600 --> 00:11:51,600
Even since 2016 is, how do we 
want this kind of law 

252
00:11:51,600 --> 00:11:54,500
enforcement to happen on these 
companies because the truth is, 

253
00:11:54,500 --> 00:11:57,900
is that doj and FBI don't have 
the ability to find the things 

254
00:11:57,900 --> 00:11:59,600
that the companies can because 
they can't access. 

255
00:11:59,800 --> 00:12:02,800
Us petabytes, and petabytes of 
data without lawful process, 

256
00:12:03,000 --> 00:12:06,200
they have lots of power to get 
data on specific individuals, 

257
00:12:06,500 --> 00:12:08,700
either in a classified or 
unclassified setting, but they 

258
00:12:08,708 --> 00:12:11,300
have to know who to grab in the 
first place. 

259
00:12:11,400 --> 00:12:15,200
And so, this is actually a 
pretty common model of the 

260
00:12:15,200 --> 00:12:16,800
company's notice something bad 
happening. 

261
00:12:17,200 --> 00:12:20,700
You go and you send your lawyers
to go brief, somebody from the 

262
00:12:20,700 --> 00:12:23,900
right us attorney's office, you 
write sign out an Affidavit of 

263
00:12:24,300 --> 00:12:25,700
this. 
Facebook employee has witnessed 

264
00:12:25,700 --> 00:12:28,100
a crime being committed, so it's
structurally. 

265
00:12:28,100 --> 00:12:29,600
This is the kind of conversation
that happens all the time. 

266
00:12:29,700 --> 00:12:33,000
I'm from a political 
perspective, this was very 

267
00:12:33,000 --> 00:12:35,800
different right in that, you're 
not talking about reaching out 

268
00:12:35,800 --> 00:12:38,000
to like just the normal us, 
attorney's office in Kansas or 

269
00:12:38,000 --> 00:12:41,300
whatever because you found a 
child molester there, you're 

270
00:12:41,300 --> 00:12:43,200
talking to the special counsel's
office, right? 

271
00:12:43,200 --> 00:12:45,500
And it was interesting, I think 
in the end, I mean, they were 

272
00:12:45,500 --> 00:12:47,000
great and they were very 
thankful. 

273
00:12:47,500 --> 00:12:51,200
But yeah, we went and privately 
told them you personally or who 

274
00:12:51,200 --> 00:12:52,600
is no. 
No, the lawyers went. 

275
00:12:52,600 --> 00:12:55,100
So, I was part of preparing all 
this stuff, but like, in at that

276
00:12:55,100 --> 00:12:58,200
kind of level, you're not 
sending somebody like me in 

277
00:12:58,200 --> 00:13:02,300
there, all the big tech 
companies have Of X doj lawyers,

278
00:13:02,300 --> 00:13:04,800
who are in-house, who handle law
enforcement relationships. 

279
00:13:05,000 --> 00:13:08,200
So we had a relationship between
our threat Intel team and their 

280
00:13:08,200 --> 00:13:11,400
counterparts and doj. 
So, for these big groups like 

281
00:13:11,400 --> 00:13:15,200
apt 28, the FBI actually has 
coordinator sitting in the 

282
00:13:15,200 --> 00:13:17,400
Hoover building because, you 
know, the FBI's like a 

283
00:13:17,400 --> 00:13:19,300
franchise, right? 
Just like McDonald's. 

284
00:13:19,300 --> 00:13:21,700
There's some of them that have, 
like, great play places and some

285
00:13:21,700 --> 00:13:25,300
have the really clean bathrooms.
So, like some FBI offices are 

286
00:13:25,300 --> 00:13:26,600
really good at doing cyber 
stuff. 

287
00:13:26,600 --> 00:13:30,100
Some are not so great, so you'll
end up with 5 or 6 other Offices

288
00:13:30,100 --> 00:13:32,500
looking at a campaign 
separately. 

289
00:13:32,600 --> 00:13:34,900
And so the FBI is little have 
coordinators who sit in the 

290
00:13:34,900 --> 00:13:37,500
headquarters whose job it is to 
watch all of this data being 

291
00:13:37,500 --> 00:13:40,800
gathered and say, oh, these four
different hacks, against 

292
00:13:40,800 --> 00:13:43,900
different steel companies. 
In four different jurisdictions,

293
00:13:44,000 --> 00:13:46,600
are actually the same actor 
right there. 

294
00:13:46,608 --> 00:13:48,900
So, we had Direct relationships,
those those folks. 

295
00:13:49,100 --> 00:13:50,700
But when you get to this level 
in your talking, the special 

296
00:13:50,700 --> 00:13:51,900
counsel, it's like you're 
sending the lawyer. 

297
00:13:51,900 --> 00:13:54,800
So the lawyers go and brief her 
and them and then we send our 

298
00:13:54,800 --> 00:13:57,500
people to go talk. 
After we got lawful process to 

299
00:13:57,500 --> 00:14:00,500
go, fill them in and to give 
them all the data that we So 

300
00:14:00,500 --> 00:14:04,400
this investigation by the 
special counsel was probably the

301
00:14:04,400 --> 00:14:08,400
most watched and reported on 
Investigation, maybe in my 

302
00:14:08,400 --> 00:14:12,600
lifetime as your team is 
speaking to members of molars 

303
00:14:12,600 --> 00:14:14,600
team. 
How did you sort of view the way

304
00:14:14,600 --> 00:14:17,000
the media covered that? 
I mean, did you find that there 

305
00:14:17,000 --> 00:14:20,600
was fairly restrained and 
reasonable articles written 

306
00:14:20,600 --> 00:14:21,700
about it? 
I mean, what's the 

307
00:14:21,900 --> 00:14:23,100
sensationalism already 
surrendered? 

308
00:14:23,100 --> 00:14:25,100
Because we can talk about what 
happens after the fact because 

309
00:14:25,100 --> 00:14:27,700
that's when shit really gets 
crazy but even just in the 

310
00:14:27,700 --> 00:14:30,200
lead-up to the molar 
investigation, What was your 

311
00:14:30,200 --> 00:14:32,400
perspective from where you said 
right? 

312
00:14:32,500 --> 00:14:37,400
I mean, I think one, so we 
published a white paper, I think

313
00:14:37,400 --> 00:14:41,200
it was April or something. 
So in 2017, in Spring of 2017, 

314
00:14:41,200 --> 00:14:43,700
we publish a paper on what the 
GRE you did. 

315
00:14:44,000 --> 00:14:46,700
This is the paper that famously 
doesn't say it's Russia, but has

316
00:14:46,700 --> 00:14:50,300
a footnote that hints that it's 
Russia and there's almost no 

317
00:14:50,300 --> 00:14:52,900
coverage of it, right? 
Like we did media briefings, we 

318
00:14:52,900 --> 00:14:54,600
did this in that and there's 
only a couple of stories, right?

319
00:14:54,600 --> 00:14:57,100
A hint that it's Russia but not.
Not actually have it be Russia. 

320
00:14:57,100 --> 00:14:59,700
You just say a country and what 
All-Stars? 

321
00:14:59,900 --> 00:15:02,100
Backwards or something. 
So I can pull up the exact 

322
00:15:02,100 --> 00:15:03,300
footnote. 
But it says something along the 

323
00:15:03,300 --> 00:15:06,100
lines of our data is consistent 
with the attribution provided by

324
00:15:06,100 --> 00:15:08,900
the US intelligence agencies in 
their January memo. 

325
00:15:09,000 --> 00:15:12,400
So if you remember like, before 
Obama goes, there's a joint 

326
00:15:12,400 --> 00:15:15,000
communique from the intelligence
agencies, saying it was Russia. 

327
00:15:15,200 --> 00:15:16,400
Right? 
Like was clearly the 

328
00:15:16,400 --> 00:15:18,400
administration getting anywhere.
You're worried about getting 

329
00:15:18,400 --> 00:15:21,400
sued for libel or why the game. 
No, I mean, so that was a big 

330
00:15:21,400 --> 00:15:24,300
political fight internally which
I'm sure will be an entire 

331
00:15:24,300 --> 00:15:27,400
episode of this miniseries with 
Zach Galifianakis being yelled 

332
00:15:27,400 --> 00:15:29,700
at by by the queen. 
I was part of Facebook. 

333
00:15:29,800 --> 00:15:31,600
Except her to not get as much 
news. 

334
00:15:31,600 --> 00:15:34,800
So, Facebook, explain any other 
not the lack of news coverage? 

335
00:15:34,900 --> 00:15:37,500
I think it's, I mean, there was 
a bunch of different things, but

336
00:15:37,500 --> 00:15:40,000
there's comfy people who did not
want Facebook volunteering that 

337
00:15:40,000 --> 00:15:41,700
bad things. 
Had happened that Facebook had 

338
00:15:41,700 --> 00:15:44,600
any part of this, which was 
clearly not going to work, 

339
00:15:44,600 --> 00:15:45,000
right? 
Yeah. 

340
00:15:45,000 --> 00:15:46,900
Thatís a, and it worked 
perfectly and no one's blame 

341
00:15:46,900 --> 00:15:48,400
Facebook since, right? 
Exactly. 

342
00:15:48,400 --> 00:15:50,000
And so you have to calm people 
were pushing that, but then you 

343
00:15:50,000 --> 00:15:52,500
also have like the d.c. 
government Affairs. 

344
00:15:52,500 --> 00:15:55,100
People who don't want to take a 
side, right? 

345
00:15:55,100 --> 00:15:57,100
And so, I think that was one of 
the big things is like, if you 

346
00:15:57,100 --> 00:15:59,700
say Russia, you know, you got 
those new Trump Administration, 

347
00:15:59,900 --> 00:16:02,400
Didn't the right thing to do is 
to turn this over to the special

348
00:16:02,400 --> 00:16:05,100
counsel and eternal all the data
over to the people who are 

349
00:16:05,100 --> 00:16:07,000
investigating. 
And our job is not to get 

350
00:16:07,000 --> 00:16:10,400
involved in the public fight, 
which just happened to coincide 

351
00:16:10,400 --> 00:16:12,700
with the, like, coms goals of 
the company. 

352
00:16:12,700 --> 00:16:15,100
And not yet like involved in 
anything, right? 

353
00:16:15,300 --> 00:16:18,200
So, yes, there's a big fight. 
Like it was a lot, just to get 

354
00:16:18,200 --> 00:16:19,900
it out. 
Like the fact that we admitted 

355
00:16:19,900 --> 00:16:22,400
that anything happened, but 
everything in there is accurate,

356
00:16:22,400 --> 00:16:25,200
just like the whole Rush 
attribution section was cut out 

357
00:16:25,200 --> 00:16:29,100
as in one of 85 different 
versions that went out, but it 

358
00:16:29,100 --> 00:16:30,400
got very little Coverage. 
Right? 

359
00:16:30,400 --> 00:16:32,300
And there's anything we've 
learned from the whole mother 

360
00:16:32,300 --> 00:16:34,900
thing. 
It is that the rollout of stuff 

361
00:16:34,900 --> 00:16:37,000
matters. 
Almost more than sort of all the

362
00:16:37,000 --> 00:16:40,400
technical footnoted. 
It's not like, oh, the public 

363
00:16:40,400 --> 00:16:44,200
eventually consumes it the 
smartest way as long as it's in 

364
00:16:44,200 --> 00:16:46,300
there. 
It's that the the fucking 

365
00:16:46,300 --> 00:16:49,800
headline is the most important 
thing and you can say, oh is in 

366
00:16:49,800 --> 00:16:51,300
there, I don't know. 
That's that's no. 

367
00:16:51,300 --> 00:16:53,500
I think you're right. 
I think you're right and 

368
00:16:53,500 --> 00:16:55,700
certainly in so it didn't cover 
the much. 

369
00:16:55,800 --> 00:16:58,000
I mean, looking back it's not 
that great. 

370
00:16:58,100 --> 00:17:00,800
I think like the treatment of 
Facebook has to Seen in the 

371
00:17:00,808 --> 00:17:04,000
context of the overall coverage 
of trump in Russia, which I 

372
00:17:04,000 --> 00:17:07,300
think, you know, and Smith came 
up with this term resistance 

373
00:17:07,300 --> 00:17:09,900
journalism and I think there's a
lot of that going on right in 

374
00:17:09,900 --> 00:17:11,800
that. 
There's a bit of a moral panic 

375
00:17:12,000 --> 00:17:15,400
and as long as you say stuff 
that is about the targets, 

376
00:17:15,599 --> 00:17:18,000
there's not a lot of pressure 
from other journalists to get it

377
00:17:18,000 --> 00:17:19,800
totally right. 
And so there's a lot of kind of 

378
00:17:19,800 --> 00:17:23,200
logical leaping and I think that
happened both before and after 

379
00:17:23,200 --> 00:17:26,000
our announcement September of 
2017, obviously after we 

380
00:17:26,000 --> 00:17:28,099
announce people kind of lose 
their minds what we'll get to 

381
00:17:28,099 --> 00:17:29,500
that in a bit because II 
actually. 

382
00:17:29,500 --> 00:17:30,400
Yeah. 
We'll get to that in a bit 

383
00:17:30,400 --> 00:17:33,700
because I pulled some, some 
clips from that era which was a 

384
00:17:33,800 --> 00:17:36,000
great thing. 
I spent the, I spent like two 

385
00:17:36,000 --> 00:17:37,900
hours last night, rolling 
through articles. 

386
00:17:38,400 --> 00:17:40,200
We'll get to that in a because I
definitely want your things. 

387
00:17:40,500 --> 00:17:44,000
So, so anyway, so you guys work 
with the council and it is being

388
00:17:44,000 --> 00:17:47,800
reported on simultaneously, what
will we briefed molar way before

389
00:17:47,800 --> 00:17:50,300
the Public Announcement, right? 
And in fact, when things get 

390
00:17:50,300 --> 00:17:52,000
complicated, the Public 
Announcement is that we had a 

391
00:17:52,000 --> 00:17:53,900
gag order from the molar team, 
right? 

392
00:17:53,900 --> 00:17:57,200
So like we go and tell them and 
then they hit us with a gag to 

393
00:17:57,200 --> 00:17:59,700
not talk about any of this. 
And so everything that Facebook 

394
00:17:59,800 --> 00:18:01,600
Okay, announced had to be 
negotiated with the special 

395
00:18:01,600 --> 00:18:04,300
counsel's office, which turns 
into this whole kind of 

396
00:18:04,300 --> 00:18:07,100
nightmare when we do the 
announcement, it doesn't come 

397
00:18:07,100 --> 00:18:10,100
with any of the data and that 
was specifically due to this 

398
00:18:10,300 --> 00:18:13,800
order from Miller's office. 
So you know and I'm not a lawyer

399
00:18:13,800 --> 00:18:16,500
but like these lawyers were 
having some pretty honest and 

400
00:18:17,500 --> 00:18:21,300
big arguments both with molars 
team and then the lawyers in 

401
00:18:21,300 --> 00:18:24,100
Congress and side like you said,
Eric the rollout is what 

402
00:18:24,100 --> 00:18:26,000
matters. 
And so you know Facebook ends up

403
00:18:26,000 --> 00:18:29,000
putting out this mealy-mouthed 
blog post with my name on it. 

404
00:18:29,000 --> 00:18:31,900
That had pretty Much nothing to 
do with the first draft that I 

405
00:18:31,908 --> 00:18:34,600
wrote like every single word was
changed by a lawyer. 

406
00:18:34,600 --> 00:18:36,500
A comes first. 
Before when I say every reporter

407
00:18:36,500 --> 00:18:39,500
would tell you you can't let 
people put stuff on your byline.

408
00:18:39,500 --> 00:18:42,300
The you know yes, I learned that
too late. 

409
00:18:42,300 --> 00:18:44,600
Yeah, I should not have let them
get my name on it. 

410
00:18:44,600 --> 00:18:46,000
Really you regret the whole 
thing? 

411
00:18:46,000 --> 00:18:48,500
Like I mean isn't it better 
something than nothing? 

412
00:18:49,100 --> 00:18:51,400
I don't regret the white paper I
think getting the white paper 

413
00:18:51,400 --> 00:18:54,100
out. 
So on the white paper side I 

414
00:18:54,100 --> 00:18:56,300
wish it said Russia. 
This is the other part. 

415
00:18:56,400 --> 00:18:58,800
So when you talk about the 
treatment of the media, you 

416
00:18:58,800 --> 00:18:59,700
know, again there's two totally 
different. 

417
00:18:59,800 --> 00:19:02,100
Campaigns the IRA campaign. 
Absolutely. 

418
00:19:02,100 --> 00:19:04,600
The responsibility of Facebook 
and Twitter, right? 

419
00:19:04,600 --> 00:19:06,800
It is our responsibility to 
catch that and we did not my 

420
00:19:06,800 --> 00:19:08,700
team did not catch that. 
I did not catch that you didn't 

421
00:19:08,700 --> 00:19:12,400
stop it. 
The gru campaign was targeted at

422
00:19:12,400 --> 00:19:15,700
the media and that absolutely 
worked, right? 

423
00:19:15,800 --> 00:19:17,300
I wrote. 
There's a there was a story 

424
00:19:17,300 --> 00:19:19,700
about John podesta as pasta 
sauce recipe, which one could 

425
00:19:19,700 --> 00:19:22,100
argue probably wasn't really 
relevant to the election issues 

426
00:19:22,100 --> 00:19:24,700
at hand. 
And yet, there was political ran

427
00:19:24,700 --> 00:19:26,900
a live blog at start. 
We all know we all know that he 

428
00:19:26,900 --> 00:19:28,300
had risotto recipes, or 
whatever. 

429
00:19:28,300 --> 00:19:31,300
Yeah, I mean, yeah, like Luke. 
I ran a live blog of, like, what

430
00:19:31,300 --> 00:19:33,400
embarrassing stuff can we find 
in these emails? 

431
00:19:33,400 --> 00:19:35,400
And maybe it's a Russian plot, 
but doesn't really matter 

432
00:19:35,400 --> 00:19:37,900
because it's newsworthy. 
This is where I spent most of my

433
00:19:37,900 --> 00:19:40,700
evening yesterday reading 
through old articles because as 

434
00:19:40,700 --> 00:19:44,300
Facebook, gets closer to, I 
guess it was just delivering 

435
00:19:44,300 --> 00:19:47,500
testimony front of Congress 
about what you found in your 

436
00:19:47,700 --> 00:19:50,600
research on the IRA. 
There's kind of leaks that come 

437
00:19:50,600 --> 00:19:53,600
out did the day before hand and 
so just pulling through a 

438
00:19:53,608 --> 00:19:56,300
sampling of Articles, you know, 
this one from the New York Times

439
00:19:56,300 --> 00:20:00,700
actually October 2nd 2017 
headline Facebook's Russia link 

440
00:20:00,700 --> 00:20:04,500
ads came in many disguises and, 
you know, has a lead in there. 

441
00:20:04,800 --> 00:20:07,300
The Russians, who poses 
Americans on Facebook last year 

442
00:20:07,300 --> 00:20:09,200
tried on quite an array of 
disguises. 

443
00:20:09,200 --> 00:20:13,900
There's an article in CNN from 
Dylan Byers, September 28th, 

444
00:20:13,900 --> 00:20:18,700
2017 that refers to the IRA as a
shadowy agency, which comes up 

445
00:20:18,700 --> 00:20:21,600
very frequently, describing them
as a shadowy agency that's 

446
00:20:21,600 --> 00:20:25,600
extremely sophisticated in their
ways of disseminating. 

447
00:20:25,700 --> 00:20:28,500
They're not necessary to visit 
of information things of that 

448
00:20:28,500 --> 00:20:30,800
nature. 
There's also So, a column from 

449
00:20:30,900 --> 00:20:32,600
Margaret Sullivan, I want to get
to later. 

450
00:20:32,700 --> 00:20:34,900
But again, from your 
perspective, as these articles 

451
00:20:34,900 --> 00:20:38,200
are coming out in advance of 
Facebook employees testimony in 

452
00:20:38,208 --> 00:20:40,700
front of Congress. 
What's your sense as this is 

453
00:20:40,700 --> 00:20:42,700
being rolled out? 
Because it definitely fed a 

454
00:20:42,700 --> 00:20:45,100
frenzy happening in the public 
that. 

455
00:20:45,400 --> 00:20:48,500
Russia fucked everything up. 
They came in there. 

456
00:20:48,500 --> 00:20:52,000
They broke American democracy 
and they did it through to visit

457
00:20:52,000 --> 00:20:54,900
Facebook posts. 
Yeah, I mean, I think so. 

458
00:20:54,900 --> 00:20:57,400
Facebook obviously takes a huge 
should take a huge amount of 

459
00:20:57,408 --> 00:20:59,600
blame on the way that all this 
information. 

460
00:20:59,800 --> 00:21:03,200
Was released and framed and the 
desire to keep everything 

461
00:21:03,200 --> 00:21:04,900
secret. 
And then to only allow things to

462
00:21:04,900 --> 00:21:06,700
come out drips and drabs really 
hurt the company. 

463
00:21:07,200 --> 00:21:10,200
That being said, the overall 
reaction seemed crazy because it

464
00:21:10,200 --> 00:21:15,200
was completely out of touch with
the actual quantitative size of 

465
00:21:15,200 --> 00:21:19,100
this content because something 
like 80,000 post sounds big 

466
00:21:19,300 --> 00:21:22,700
Until you realize the 
denominator on that is in the 

467
00:21:22,700 --> 00:21:25,500
hundreds of millions, right? 
And so the vast majority, this 

468
00:21:25,500 --> 00:21:28,600
content had almost no reach like
on the stuff where the Russians 

469
00:21:28,600 --> 00:21:31,000
posted the content themselves. 
All right, so II ra, they're 

470
00:21:31,000 --> 00:21:32,500
making propaganda, they're 
posting it. 

471
00:21:32,600 --> 00:21:36,400
So this is like a picture of 
people at a fence that is an 

472
00:21:36,400 --> 00:21:39,300
anti-immigration add or it's a 
pro. 

473
00:21:39,300 --> 00:21:41,100
Something that looks like it's 
Pro black lives matter. 

474
00:21:41,100 --> 00:21:45,800
But made by a fake group, right?
For that stuff, its overall 

475
00:21:45,800 --> 00:21:49,900
reach compared to other, both 
organic, and paid content was 

476
00:21:49,900 --> 00:21:52,100
nowhere near, right? 
There's hundreds of millions of 

477
00:21:52,100 --> 00:21:54,100
dollars were spent on the 
election on Facebook and they 

478
00:21:54,100 --> 00:21:56,800
spent 150,000 dollars, right? 
Which it's like over the course 

479
00:21:56,800 --> 00:21:59,600
of two years, which do the math.
That's what less than 5. 

480
00:21:59,800 --> 00:22:01,900
I was in a month. 
I mean like there are they are 

481
00:22:01,900 --> 00:22:05,900
like, you know, d2c makeup 
brands for dogs, that probably 

482
00:22:05,900 --> 00:22:09,600
spend more on Facebook and 
Instagram than the IRA did. 

483
00:22:09,600 --> 00:22:12,800
So the idea that this was a huge
spend was was, was always a 

484
00:22:12,808 --> 00:22:16,700
fallacy and I don't know how 
many people that, you know, got 

485
00:22:16,800 --> 00:22:19,200
whipped up into a frenzy, about 
it, really understood that. 

486
00:22:19,200 --> 00:22:22,600
Well, because it also got 
combined completely without any 

487
00:22:22,600 --> 00:22:25,600
evidence with the Cambridge 
analytic issues, which is this 

488
00:22:25,600 --> 00:22:28,900
kind of imaginary world in which
Cambridge analytical is this 

489
00:22:28,900 --> 00:22:32,400
magical Group of bond villains 
that have mind-control powers 

490
00:22:32,400 --> 00:22:35,400
based upon, right? 
Yo, people's likes that they 

491
00:22:35,408 --> 00:22:39,300
took yo, illegally via an API to
be totally clear Cambridge 

492
00:22:39,300 --> 00:22:41,300
analytical themselves as a scam,
right? 

493
00:22:41,300 --> 00:22:43,800
So they said they had these 
magical psychographic models of 

494
00:22:43,800 --> 00:22:46,600
people and there are 
quantitative metrics as to 

495
00:22:46,600 --> 00:22:50,200
whether online ads work. 
There's a whole set of people 

496
00:22:50,200 --> 00:22:52,100
whose entire job. 
It is to think about these 

497
00:22:52,100 --> 00:22:54,900
metrics and I know of those met 
none of those metrics is came a 

498
00:22:54,908 --> 00:22:58,200
general like a better than just 
any other kinds of AD targeting 

499
00:22:58,200 --> 00:22:59,600
that people do on Facebook, 
right? 

500
00:22:59,800 --> 00:23:02,600
And they had nothing to do with 
Ira to be clear, Kim Joon wook. 

501
00:23:02,600 --> 00:23:04,500
It had absolutely nothing to do 
with internet, research to see 

502
00:23:04,700 --> 00:23:07,800
the research internet research 
agency did zero data upload 

503
00:23:07,900 --> 00:23:11,000
custom audiences. 
So they didn't even do what we 

504
00:23:11,000 --> 00:23:14,300
consider micro-targeting. 
They targeted like you know, 

505
00:23:14,500 --> 00:23:19,200
young men in this ZIP code in 
Baltimore for pro black lives 

506
00:23:19,200 --> 00:23:22,600
matter messages, but none of 
that mattered to the narrative 

507
00:23:22,800 --> 00:23:25,200
which conflating these things at
these Russians have kind of 

508
00:23:25,200 --> 00:23:28,100
magical mind control powers that
they were able to use through a 

509
00:23:28,108 --> 00:23:29,500
hundred fifty thousand dollars 
of ads. 

510
00:23:29,800 --> 00:23:31,900
When Hillary Clinton campaign, 
spent like a hundred million 

511
00:23:31,900 --> 00:23:34,500
dollars, right answers? 
Like the idea that they were a 

512
00:23:34,500 --> 00:23:36,700
thousand times. 
But what about this idea that 

513
00:23:36,700 --> 00:23:39,900
you guys were like hand holding 
the Trump campaign in a way, you

514
00:23:39,900 --> 00:23:43,200
weren't the cloaking campaign. 
So there's I think totally 

515
00:23:43,200 --> 00:23:47,900
legitimate arguments on whether 
Facebook ads had an effect on 

516
00:23:47,908 --> 00:23:49,700
the campaign. 
Almost certainly Facebook ads 

517
00:23:49,800 --> 00:23:51,900
had an effect on the campaign 
but they were to the 

518
00:23:51,900 --> 00:23:55,300
quote-unquote legitimate ads 
paid for by the campaigns and 

519
00:23:55,300 --> 00:23:58,900
the official committees and such
using the money that our country

520
00:23:58,900 --> 00:24:00,700
allows people. 
To give, right? 

521
00:24:00,700 --> 00:24:04,000
Yo, to two candidates in the 
parties, for which there are no 

522
00:24:04,000 --> 00:24:05,900
legal controls, I had nothing to
do this. 

523
00:24:05,900 --> 00:24:08,900
So all of this I know from the 
investigation afterwards, right?

524
00:24:08,900 --> 00:24:11,600
But, my understanding is that 
both campaigns, just like any of

525
00:24:11,600 --> 00:24:14,500
the, if you're a Procter & 
Gamble, you get a personal, you 

526
00:24:14,500 --> 00:24:16,800
get a Facebook employee whose 
entire job it is to help your 

527
00:24:16,800 --> 00:24:19,200
campaigns run. 
And so these campaigns are 

528
00:24:19,400 --> 00:24:21,100
spending tens hundreds millions 
of dollars. 

529
00:24:21,400 --> 00:24:24,800
So they got offers of these 
people will help you do it and 

530
00:24:24,800 --> 00:24:27,600
the Trump campaign not knowing 
what they're doing said, yes. 

531
00:24:27,600 --> 00:24:30,800
And the Clean Campaign was like 
now we At this and said no. 

532
00:24:31,100 --> 00:24:33,700
And so it is true and it is 
quite possible that trumps 

533
00:24:33,700 --> 00:24:36,300
campaign was way better than 
Clinton's campaign because of 

534
00:24:36,300 --> 00:24:39,200
that, we partially don't have 
this data because the data is 

535
00:24:39,200 --> 00:24:42,200
kind of Legally locked up and 
it's never come out and there's 

536
00:24:42,200 --> 00:24:44,900
never been any changes to the 
law to allow it to come out. 

537
00:24:44,900 --> 00:24:46,400
I'm actually think this is where
Congress has failed. 

538
00:24:46,400 --> 00:24:49,800
The most is that there have been
zero changes to online 

539
00:24:49,800 --> 00:24:52,700
advertising law, since 2016. 
I would like to ban 

540
00:24:52,700 --> 00:24:55,200
micro-targeting for political 
ads, because I think it's 

541
00:24:55,200 --> 00:24:59,800
innately corrupting because it 
both drives campaigns just All 

542
00:24:59,800 --> 00:25:03,000
this information but it also 
drives them to Target message 

543
00:25:03,000 --> 00:25:04,600
different messages to different 
audiences. 

544
00:25:05,000 --> 00:25:06,500
Both political parties. 
Do this in both political 

545
00:25:06,500 --> 00:25:09,000
parties, think they're better. 
So they will make all these 

546
00:25:09,000 --> 00:25:11,600
noises publicly about how 
horrible ads are incorrigible 

547
00:25:11,600 --> 00:25:13,300
lake or river. 
It's all Bs right? 

548
00:25:13,300 --> 00:25:16,400
In private, all of these 
senators and congressmen are 

549
00:25:16,400 --> 00:25:19,200
being told by their Consultants 
were getting paid to run 

550
00:25:19,200 --> 00:25:20,800
political ads. 
They're being told by their 

551
00:25:20,800 --> 00:25:24,300
Consultants were better, do not 
unilaterally disarm, and so they

552
00:25:24,300 --> 00:25:25,600
won't pass any laws. 
Europe. 

553
00:25:25,600 --> 00:25:27,600
Europe is starting to make some 
moves here, but in the United 

554
00:25:27,600 --> 00:25:28,900
States, there's been nothing 
Kim. 

555
00:25:28,900 --> 00:25:32,000
Can we take a quick Step back 
for a second to to the IRA stuff

556
00:25:32,000 --> 00:25:34,800
because I mean you sort of 
brought it up and I have this 

557
00:25:34,800 --> 00:25:36,300
column here. 
This is again for that. 

558
00:25:36,300 --> 00:25:39,200
Same time period in 2017, this 
is from Margaret Sullivan, who 

559
00:25:39,200 --> 00:25:41,800
is a great media columnist. 
This is not a slam on her at 

560
00:25:41,800 --> 00:25:44,700
all, but I think it looks very 
interesting in retrospect and 

561
00:25:44,700 --> 00:25:46,800
really is a good summary of the 
era. 

562
00:25:47,100 --> 00:25:49,500
So the headline is collusion, 
who needs it? 

563
00:25:49,500 --> 00:25:51,600
When Facebook was allowing 
Russians to help Trump. 

564
00:25:52,000 --> 00:25:54,400
This is a key paragraph from 
this column much of that 

565
00:25:54,400 --> 00:25:56,500
content. 
This is the IRAs content was 

566
00:25:56,500 --> 00:25:59,100
expressly designed to widen the 
cultural divides, the United 

567
00:25:59,100 --> 00:26:02,200
States to Of wedges among its 
citizens and in doing so, to 

568
00:26:02,200 --> 00:26:04,400
help elect the Russian, 
government's preferred candidate

569
00:26:04,400 --> 00:26:06,800
Donald Trump. 
It worked. 

570
00:26:07,000 --> 00:26:09,300
And while those more recent 
numbers are astonishing, I'm 

571
00:26:09,300 --> 00:26:11,600
assuming this is the numbers. 
Unlike reach the reality is 

572
00:26:11,600 --> 00:26:15,200
probably far worse research from
Columbia University's Tau Center

573
00:26:15,200 --> 00:26:17,700
suggests that Russian, linked 
information or disinformation 

574
00:26:18,000 --> 00:26:20,400
with shared hundreds of millions
of times on Facebook. 

575
00:26:20,400 --> 00:26:22,800
The numbers Boggle, the mind of 
what here. 

576
00:26:22,800 --> 00:26:24,600
No, let me, let me read the key 
paragraph here because this is 

577
00:26:24,600 --> 00:26:26,100
where I think it really gets to 
the point. 

578
00:26:26,200 --> 00:26:29,100
We need to admit the obvious. 
If there had been no Facebook 

579
00:26:29,100 --> 00:26:31,900
spreading, Russia propaganda. 
There might as well be no 

580
00:26:31,900 --> 00:26:35,200
President Trump. 
Now looking at this. 

581
00:26:35,200 --> 00:26:38,400
Now from 2021, I don't want to 
seem too much smarter now 

582
00:26:38,400 --> 00:26:40,300
because it's easy to be critical
in hindsight. 

583
00:26:40,300 --> 00:26:44,600
But is that an insane conclusion
to reach based on everything 

584
00:26:44,600 --> 00:26:47,500
that you saw from the IRA? 
And is the fact that a major 

585
00:26:47,800 --> 00:26:51,000
newspaper is publishing this, in
some way, a failure on the part 

586
00:26:51,000 --> 00:26:54,400
of the media, or the interaction
between the media and these tech

587
00:26:54,400 --> 00:26:58,100
companies to accurately, assess 
the role that the I replayed in 

588
00:26:58,100 --> 00:27:01,200
the 2016 election. 
Yes, that's wrong from my 

589
00:27:01,200 --> 00:27:04,100
perspective, the two kinds of 
online propaganda that were most

590
00:27:04,100 --> 00:27:07,300
effective was one legitimate 
Facebook ads. 

591
00:27:07,300 --> 00:27:10,200
The ads that were run by the 
candidates in, which Trump did a

592
00:27:10,208 --> 00:27:11,600
better job than the Clinton 
headed. 

593
00:27:11,600 --> 00:27:14,200
Straight up above board, using 
custom audiences and all the 

594
00:27:14,400 --> 00:27:17,200
Procter & Gamble shit. 
Which again, I would like laws 

595
00:27:17,200 --> 00:27:19,800
to change that but we need laws 
to change it, right, right. 

596
00:27:19,800 --> 00:27:22,700
And then the second was the gru 
activity because the gru 

597
00:27:22,700 --> 00:27:26,100
activity completely changed the 
tenor of the coverage of Hillary

598
00:27:26,100 --> 00:27:29,100
Clinton, that was incredibly 
effective because what they used

599
00:27:29,100 --> 00:27:32,400
was. 
Olin emails that allowed them to

600
00:27:32,400 --> 00:27:35,000
create stories that were based 
upon a kernel of Truth. 

601
00:27:35,000 --> 00:27:38,700
Where the story itself was not 
accurate and be frank. 

602
00:27:38,800 --> 00:27:43,000
The US media has never ever ever
looked inside of itself of the 

603
00:27:43,000 --> 00:27:45,700
fact that every major news are, 
you know, the New York Times The

604
00:27:45,708 --> 00:27:48,800
Washington Post, CNN MSNBC, not 
just Fox, which like, we'll just

605
00:27:48,800 --> 00:27:51,800
have to write off Fox of like 
what we can do about Fox. 

606
00:27:52,000 --> 00:27:54,300
But for all the, what I 
considered the legitimate media,

607
00:27:54,500 --> 00:27:57,200
every single one of them was 
played by the gru they wrote 

608
00:27:57,200 --> 00:27:59,900
exactly the stories of Gru 
wanted and they Done. 

609
00:28:00,200 --> 00:28:02,900
No soul-searching on that at 
all. 

610
00:28:03,200 --> 00:28:06,200
And so at, yes, we screwed up 
like I will absolutely admit 

611
00:28:06,200 --> 00:28:09,000
that we screwed up but the idea 
that a hundred thousand dollars 

612
00:28:09,000 --> 00:28:13,900
in ads is as important as the 
entire legitimate, American 

613
00:28:13,900 --> 00:28:17,200
Media changing the way it 
covered Hillary Clinton is just 

614
00:28:17,300 --> 00:28:21,000
ridiculous. 
I want to just to say Tom 

615
00:28:21,000 --> 00:28:23,500
response to your comment and 
then it was must Alex in a 

616
00:28:23,508 --> 00:28:25,300
minute. 
That Margaret sullavan column is

617
00:28:25,308 --> 00:28:29,100
so interesting because it kind 
of erases history, you know, 

618
00:28:29,100 --> 00:28:34,100
Russia 'He had used propaganda 
to widen the divisions between 

619
00:28:34,100 --> 00:28:37,600
Americans before there were very
active in the 60s and 70s the 

620
00:28:37,600 --> 00:28:39,400
United States when the country 
was going through an 

621
00:28:39,400 --> 00:28:42,100
extraordinary Civil Rights 
Movement. 

622
00:28:42,200 --> 00:28:45,100
This is not the first time that 
Russia and other countries but 

623
00:28:45,100 --> 00:28:48,200
especially Russia have used 
internal divisions in a country 

624
00:28:48,200 --> 00:28:50,500
to try to destabilise the 
country itself. 

625
00:28:50,700 --> 00:28:53,400
So I think that, what is missed 
in that column is that this is 

626
00:28:53,400 --> 00:28:57,200
not a new strategy, it's simply 
a new tool and I think we could 

627
00:28:57,200 --> 00:28:59,400
debate whether or not the tool 
is more. 

628
00:28:59,600 --> 00:29:03,100
Active than any other tool 
they've had before and that is 

629
00:29:03,100 --> 00:29:05,700
an interesting debate but 
certainly this is not new. 

630
00:29:05,700 --> 00:29:08,800
I mean, if you're a student of 
history and you've read anything

631
00:29:08,800 --> 00:29:13,600
about Russia's efforts to cause 
the collapse of the United 

632
00:29:13,600 --> 00:29:16,700
States in the 60s, you know, 
this isn't new. 

633
00:29:16,800 --> 00:29:18,400
Yeah. 
Tom Tom ridge book on this is 

634
00:29:18,400 --> 00:29:21,600
fantastic right. 
Active measures like absolutely 

635
00:29:22,200 --> 00:29:24,200
CIA invented AIDs that whole 
thing. 

636
00:29:24,500 --> 00:29:26,700
Yes, I think this is a good 
because we're now promoted 

637
00:29:26,700 --> 00:29:29,400
Cecilia and share has book and 
soon-to-be TV show. 

638
00:29:29,500 --> 00:29:30,900
Ow. 
And now, this really great book 

639
00:29:30,900 --> 00:29:33,600
after measures were really more 
of kind of like a marketing, a 

640
00:29:34,200 --> 00:29:37,300
marketing agency owner. 
Do you guys have a code that 

641
00:29:37,300 --> 00:29:40,000
unites discount on these books? 
You need to you need to line 

642
00:29:40,000 --> 00:29:42,000
that up before this goes. 
Yeah, but I think that on the 

643
00:29:42,000 --> 00:29:46,200
question of whether or not 
newsrooms have really reckoned 

644
00:29:46,200 --> 00:29:50,900
with this idea of being played 
by the gru, there were moments 

645
00:29:50,900 --> 00:29:55,100
where at least Dean baquet at 
the times, he came out different

646
00:29:55,100 --> 00:29:56,800
panels. 
He was asked about this and he 

647
00:29:56,800 --> 00:29:59,100
did say that it had led to some 
soul-searching. 

648
00:29:59,200 --> 00:30:03,100
I You know, the internally 
certainly this conversation 

649
00:30:03,600 --> 00:30:07,800
husband had an is being had and 
those are really important and 

650
00:30:07,800 --> 00:30:10,800
ongoing conversations, 
especially we saw going into the

651
00:30:10,800 --> 00:30:13,600
2020 election, will see you 
again going to the 2022. 

652
00:30:15,100 --> 00:30:18,100
Midterms like this, this is an 
ongoing conversation. 

653
00:30:18,100 --> 00:30:21,100
Even if it's not necessarily 
happening as publicly as it is 

654
00:30:21,100 --> 00:30:24,000
for Facebook and whether or not 
that's fair, I think is very is 

655
00:30:24,000 --> 00:30:28,000
a very good question, really? 
I mean, there are there are 

656
00:30:28,000 --> 00:30:30,800
great journalists at All of 
these organizations who I've had

657
00:30:30,800 --> 00:30:33,200
private conversations with where
they absolutely recognize all 

658
00:30:33,200 --> 00:30:35,200
this. 
And I know that they are part of

659
00:30:35,300 --> 00:30:37,600
the internal conversations of. 
Like, let's not get played again

660
00:30:37,800 --> 00:30:40,700
on D McKay, I remember very 
distinctly a week because I 

661
00:30:40,700 --> 00:30:43,300
wrote something about this, 
where Mark Zuckerberg gave his, 

662
00:30:43,300 --> 00:30:45,300
like, the speech were like, he 
changes history. 

663
00:30:45,300 --> 00:30:47,400
That Facebook was invented 
because of the Iraq War, which 

664
00:30:47,400 --> 00:30:49,800
is, yeah, yeah. 
I was not in that Harbor Group 

665
00:30:49,900 --> 00:30:52,100
which is why I don't own an 
island, but like, that does not 

666
00:30:52,100 --> 00:30:55,800
seem accurate to me, but the 
same week Dean Beck a was on 

667
00:30:56,000 --> 00:30:59,300
Michael Voris podcast, which I'm
a huge fan of the times podcast.

668
00:30:59,700 --> 00:31:04,100
And borrow to his credit is like
interviewing his boss's boss's 

669
00:31:04,100 --> 00:31:06,900
boss, and ask them. 
These really tough questions. 

670
00:31:06,900 --> 00:31:10,000
And Bouquet says, if it's 
newsworthy, we will publish it, 

671
00:31:10,000 --> 00:31:12,300
right? 
And so I saw both Zach and Decay

672
00:31:12,300 --> 00:31:16,600
or kind of actually to me 
similar figures in that from 

673
00:31:16,600 --> 00:31:19,800
bekay everything is about 
newsworthiness right? 

674
00:31:19,900 --> 00:31:22,100
And even if you're getting 
played, even if the leaked 

675
00:31:22,100 --> 00:31:25,500
documents were leaked from 
hacking from a Russian hacking 

676
00:31:25,500 --> 00:31:28,600
group, they're going to run. 
It whereas duck was all about 

677
00:31:28,600 --> 00:31:30,500
kind of free speech. 
In the freedom of individuals 

678
00:31:30,500 --> 00:31:32,600
and I saw them as like in this 
parallel that both of them are 

679
00:31:32,608 --> 00:31:35,500
getting played, both of them 
have like this deep belief in 

680
00:31:35,500 --> 00:31:37,800
what they're doing. 
And can't understand how that 

681
00:31:37,800 --> 00:31:39,800
deep belief is being used 
against them, so they Facebook, 

682
00:31:39,900 --> 00:31:43,600
like we at least published I? 
Yes, I failed to get the company

683
00:31:43,600 --> 00:31:47,300
to put Russia into that paper. 
The New York Times wrote nothing

684
00:31:47,800 --> 00:31:50,100
right. 
The York Times has never ever 

685
00:31:50,200 --> 00:31:51,800
written anything. 
About we got played by the 

686
00:31:51,800 --> 00:31:54,800
Jerry, you were sorry. 
And so yes, did I fail? 

687
00:31:54,800 --> 00:31:56,500
Yes, but I, at least will say 
it. 

688
00:31:56,500 --> 00:31:59,200
We need to realize, we need to 
rely on media columnist for 

689
00:31:59,200 --> 00:32:00,800
that. 
We need to rely on Margaret 

690
00:32:00,800 --> 00:32:03,300
Sullivan for those sorts of 
Articles anymore. 

691
00:32:03,800 --> 00:32:06,900
But Margaret Sullivan is as 
subject to the, you know, the 

692
00:32:06,900 --> 00:32:09,200
whims of the moment. 
And that means like in general, 

693
00:32:09,200 --> 00:32:09,600
right? 
Sure. 

694
00:32:09,600 --> 00:32:12,200
It's really hard to get the 
media to write a story saying, 

695
00:32:12,200 --> 00:32:14,100
we messed up. 
And I think the red and many 

696
00:32:14,100 --> 00:32:18,200
moments where various people 
have said, don't you think that 

697
00:32:18,200 --> 00:32:23,200
BuzzFeed should apologize for 
publishing raw unvetted 

698
00:32:23,300 --> 00:32:25,500
intelligence that completely 
changed the course of how we 

699
00:32:25,500 --> 00:32:27,500
think about Donald Trump, that's
falling apart. 

700
00:32:27,600 --> 00:32:28,900
These questions have come up 
again. 

701
00:32:28,900 --> 00:32:31,100
And again, Multiple media 
organizations and I'm not 

702
00:32:31,100 --> 00:32:33,000
excusing them. 
I'm just going to agree with you

703
00:32:33,000 --> 00:32:35,200
that we haven't seen the 
equivalent of a white paper 

704
00:32:35,200 --> 00:32:37,000
where a media organization comes
out. 

705
00:32:37,000 --> 00:32:40,100
And says, this is, this is how 
we messed up and this is what we

706
00:32:40,100 --> 00:32:42,400
want to do better. 
So I think, because of the way 

707
00:32:42,400 --> 00:32:45,600
the media works, we are all 
really dependent on columnist. 

708
00:32:45,600 --> 00:32:47,900
Whether it's a Margaret sullavan
or whether it's a Columbia, 

709
00:32:47,900 --> 00:32:52,100
journalism review, some sort of 
outside forcing function to say 

710
00:32:52,100 --> 00:32:55,200
this is important to recognize 
for sure. 

711
00:32:55,200 --> 00:32:57,100
For sure. 
I mean the handling of like the 

712
00:32:57,100 --> 00:33:00,300
hunter by knee males Joe His 
daughters diary. 

713
00:33:00,300 --> 00:33:05,300
I mean those would seem to be 
early signs that the media is 

714
00:33:05,300 --> 00:33:09,100
having a much more command and 
control approach which is also 

715
00:33:09,100 --> 00:33:10,900
criticized. 
Well I mean it is sort of a 

716
00:33:10,900 --> 00:33:14,600
no-win case where if there's 
information out there that seems

717
00:33:14,600 --> 00:33:18,400
it's absolutely no win and it's 
no win for both the tech 

718
00:33:18,400 --> 00:33:20,000
companies in the media. 
I mean, there's no direct hit 

719
00:33:20,008 --> 00:33:23,500
both sides screwed up right and 
sort of the government. 

720
00:33:23,500 --> 00:33:26,200
So, like you're asking about 
what is like, so what did it 

721
00:33:26,200 --> 00:33:27,700
feel? 
Like for us inside Facebook, 

722
00:33:27,700 --> 00:33:29,400
what it felt like was there. 
Three? 

723
00:33:29,500 --> 00:33:31,200
Three people, there's Mark 
Zuckerberg in the hoodie, 

724
00:33:31,600 --> 00:33:36,100
there's, you know, the FBI G-Man
in his suit and there's, you 

725
00:33:36,100 --> 00:33:38,500
know, a New York Times Reporter 
with like the Press hat on like 

726
00:33:38,500 --> 00:33:41,300
the little, you know, the Fedora
or whatever and both the 

727
00:33:41,300 --> 00:33:44,100
government guy in the Press car.
Like, oh man, that guy in the 

728
00:33:44,108 --> 00:33:47,600
hoodie screwed up, right? 
We did nothing wrong, but they 

729
00:33:47,600 --> 00:33:49,400
really messed up. 
All these things were in bed, 

730
00:33:49,400 --> 00:33:52,900
Facebook are accurate and true. 
And it is the mistakes of the 

731
00:33:52,908 --> 00:33:55,500
company that need to be fixed, 
but they are in the context of 

732
00:33:55,500 --> 00:34:00,700
an overall, failure of American 
society to deal with this new Of

733
00:34:00,800 --> 00:34:02,600
attack by one of our 
adversaries. 

734
00:34:02,900 --> 00:34:05,500
And unless you look at the big 
picture, you can't solve it. 

735
00:34:05,500 --> 00:34:07,600
And that's like one of my 
lessons come out of the 2020. 

736
00:34:07,600 --> 00:34:10,400
Election was the context of 
propaganda. 

737
00:34:10,400 --> 00:34:12,400
Mind is completely changed in 
the vast majority of it. 

738
00:34:12,400 --> 00:34:14,699
Now comes from verified American
voices. 

739
00:34:14,699 --> 00:34:16,400
Tuna, we know exactly what they 
are. 

740
00:34:16,400 --> 00:34:18,699
They're not being Amplified by 
Russians are not fake accounts 

741
00:34:19,000 --> 00:34:22,400
and they have the ability to 
control the media context as 

742
00:34:22,400 --> 00:34:24,800
well as to get their messages 
out on social media. 

743
00:34:24,800 --> 00:34:27,000
And until we kind of deal with 
the fact that this is a 

744
00:34:27,000 --> 00:34:31,600
multimedia issue, That that it 
flows between kind of the 

745
00:34:31,600 --> 00:34:34,000
traditional media and the online
platforms. 

746
00:34:34,000 --> 00:34:35,500
We're not really gonna be able 
to do anything. 

747
00:34:35,699 --> 00:34:37,199
I don't think any of us would 
make this case. 

748
00:34:37,199 --> 00:34:40,500
But some you could say that it 
was in the media's interest to 

749
00:34:40,500 --> 00:34:44,100
put the blame on Facebook to 
shift the blame from the media. 

750
00:34:44,100 --> 00:34:48,699
If the bigger hack was in 
retrospect about using the media

751
00:34:48,800 --> 00:34:52,400
to spread disinformation 
pointing the finger at Facebook,

752
00:34:52,600 --> 00:34:56,500
was in the media, is like 
Financial corporate interest. 

753
00:34:56,699 --> 00:34:59,300
I mean, obviously I disagree 
with that. 

754
00:34:59,900 --> 00:35:02,700
Really, really the only reason I
disagree with that is because 

755
00:35:02,700 --> 00:35:06,100
you're talking about a period 
when the media was so beloved 

756
00:35:06,100 --> 00:35:08,500
that nobody was shifting blame. 
Well, bye. 

757
00:35:08,500 --> 00:35:11,600
If you want to Kratz. 
So loved by the left, if you 

758
00:35:11,600 --> 00:35:16,800
were walking down the street as 
a Washington reporter, you went 

759
00:35:16,800 --> 00:35:20,100
to an airport and somebody 
recognized in Washington, not 

760
00:35:20,100 --> 00:35:22,500
just in Washington like all over
the country. 

761
00:35:22,500 --> 00:35:24,600
People were thinking you were 
getting letters. 

762
00:35:24,600 --> 00:35:26,200
This. 
I'm talking about the period 

763
00:35:26,200 --> 00:35:28,500
right after Trump was elected. 
Oh, I know what you're talking 

764
00:35:28,500 --> 00:35:29,300
about. 
Look my mom. 

765
00:35:29,400 --> 00:35:31,600
I signed up for a subscription 
to The Washington Post just to 

766
00:35:31,600 --> 00:35:32,900
support them. 
She's like right now. 

767
00:35:32,900 --> 00:35:35,500
I'm no, I'm not. 
I'm talking about the Trump, the

768
00:35:35,500 --> 00:35:38,600
Trump term of his presidency. 
I mean, people are wearing 

769
00:35:38,600 --> 00:35:40,900
democracy dies in darkness 
shirts, and DC. 

770
00:35:40,900 --> 00:35:45,000
Yeah, there is no blame to shift
who was blaming the media for 

771
00:35:45,000 --> 00:35:48,500
Trump at that point. 
Now, people are now might be the

772
00:35:48,500 --> 00:35:52,600
time when that thesis would make
sense, but literally in the days

773
00:35:52,600 --> 00:35:57,000
months after the election, 
basically, from the time, Donald

774
00:35:57,000 --> 00:36:00,700
Trump was elected through the 
issuance of Of the molar report.

775
00:36:00,800 --> 00:36:03,200
They'll nobody was blaming the 
media for anything, just like 

776
00:36:03,200 --> 00:36:06,400
nobody is blaming the FBI 
another group with its hated but

777
00:36:06,400 --> 00:36:08,800
suddenly was not hated for like 
a couple of years. 

778
00:36:09,200 --> 00:36:11,900
I just want to ask what this 
social psychological story for 

779
00:36:11,900 --> 00:36:15,000
the media is if it's not that 
sort of narrative, like this is 

780
00:36:15,000 --> 00:36:17,800
a good way to deflect it. 
Like what is your mental model? 

781
00:36:18,200 --> 00:36:21,300
Is it just like the media is not
top down enough and there isn't 

782
00:36:21,300 --> 00:36:24,400
enough like command-and-control,
it's like sort of that it's an 

783
00:36:24,400 --> 00:36:27,900
organ or just like do you have a
thesis on like where the media 

784
00:36:28,300 --> 00:36:29,800
fails on this? 
Stuff. 

785
00:36:30,000 --> 00:36:32,100
Well, I think Ben Smith was 
right about resistance 

786
00:36:32,100 --> 00:36:34,900
journalism, right? 
Which is like we end up in a 

787
00:36:34,900 --> 00:36:38,400
situation where as long as you 
publish something that fits 

788
00:36:38,400 --> 00:36:40,800
preconceived notions about, who 
the bad guy is like, if you can 

789
00:36:40,800 --> 00:36:44,900
tell something in the in the Arc
of this is this is the bad guy 

790
00:36:44,900 --> 00:36:47,300
and they did bad things then 
getting the details, right? 

791
00:36:47,300 --> 00:36:49,600
Isn't that important, right? 
Like, there's not push back and,

792
00:36:50,000 --> 00:36:51,700
and we see that with Facebook 
all the time, we're seeing that 

793
00:36:51,700 --> 00:36:54,800
now with the Hagen documents, 
we're about to the media reports

794
00:36:54,800 --> 00:36:57,400
about those documents, do not 
match up to the documents 

795
00:36:57,400 --> 00:36:58,600
themselves. 
Once they are released which is 

796
00:36:58,600 --> 00:37:01,000
one of the reasons. 
I am quite upset that this is 

797
00:37:01,000 --> 00:37:02,200
the way. 
These documents were released is

798
00:37:02,200 --> 00:37:07,300
like in a way to create kind of 
a Feeding Frenzy only for the 

799
00:37:07,300 --> 00:37:09,200
headlines. 
There's a bunch of stuff in 

800
00:37:09,200 --> 00:37:10,100
there. 
That's going to be super 

801
00:37:10,100 --> 00:37:13,100
important for academic, study of
social media. 

802
00:37:13,200 --> 00:37:14,500
In the long run. 
This is going to be great 

803
00:37:14,500 --> 00:37:17,400
because this is going to kick 
off, kind of a much more 

804
00:37:17,400 --> 00:37:19,100
rigorous. 
Empirical study of the things 

805
00:37:19,100 --> 00:37:21,500
that happened on line that is 
going to see a lot of great 

806
00:37:21,500 --> 00:37:23,500
work. 
But in the meantime, we have to 

807
00:37:23,500 --> 00:37:26,500
deal with the headline grabbing 
of like the Wall Street Journal,

808
00:37:26,500 --> 00:37:29,100
writing the story about the 
Instagram slides that did not 

809
00:37:29,100 --> 00:37:30,400
match. 
If the slides once the slides 

810
00:37:30,400 --> 00:37:33,500
came out, the Washington Post 
had a headline about Facebook. 

811
00:37:33,600 --> 00:37:37,500
Uprating, the, you know, the 
angry Emoji, which was 

812
00:37:37,500 --> 00:37:39,700
intentionally misleading in that
in the story. 

813
00:37:39,700 --> 00:37:43,000
If you read it, all of the 
Emojis were upright, all the 

814
00:37:43,000 --> 00:37:45,400
reactions other than, like, were
up ranked and then, and then 

815
00:37:45,400 --> 00:37:47,600
there is an internal study and 
that turned out to be causing 

816
00:37:47,600 --> 00:37:49,300
harm. 
So they fixed it, which is 

817
00:37:49,300 --> 00:37:51,900
exactly what you want. 
People who wrote the read The 

818
00:37:51,900 --> 00:37:54,600
Washington Post story where, 
less educated about how faithful

819
00:37:54,600 --> 00:37:57,100
camels these issues, than before
they read the story. 

820
00:37:57,300 --> 00:37:59,900
And so that is resistance 
journalism to me, because Nobody

821
00:37:59,900 --> 00:38:01,400
pushed back on them, the Wall 
Street Journal. 

822
00:38:01,400 --> 00:38:03,900
The Instagram one is been, I 
mean we've talked about a lot. 

823
00:38:03,900 --> 00:38:06,300
What's your case? 
You just quickly articulate. 

824
00:38:06,500 --> 00:38:08,500
Now your objection though. 
So there's a couple things. 

825
00:38:08,500 --> 00:38:12,500
One, this is like very initial 
work by an Instagram team to try

826
00:38:12,500 --> 00:38:15,200
to ask people how they feel in 
different circumstances to then 

827
00:38:15,200 --> 00:38:17,400
lead to more research into your 
talking about a very small 

828
00:38:17,400 --> 00:38:20,100
sample size. 
And this kind of just asking 

829
00:38:20,100 --> 00:38:22,900
people how they feel is not that
great, and it's known. 

830
00:38:23,000 --> 00:38:25,800
And then the second is, there's 
Parts where it's much more 

831
00:38:25,800 --> 00:38:28,400
positive, and it's great makes 
people really happy, right? 

832
00:38:28,500 --> 00:38:29,900
When you read this, you know, 
Okay. 

833
00:38:30,100 --> 00:38:32,600
It turns out that, yes, there 
are problems with teenagers and 

834
00:38:32,600 --> 00:38:34,500
social media and that is 
something that needs to be 

835
00:38:34,500 --> 00:38:36,800
studied. 
And this is the beginning of 

836
00:38:36,800 --> 00:38:38,600
that. 
The best part of this was like, 

837
00:38:38,700 --> 00:38:41,300
there was an article in Wired 
Magazine, which is owned by 

838
00:38:41,300 --> 00:38:44,500
condé Nast, who are the people 
who publish Vogue, which is like

839
00:38:44,500 --> 00:38:48,100
to get lectured by condé Nast 
on, you know, beauty standards 

840
00:38:48,100 --> 00:38:52,000
for teenage girls is is amazing 
and just kind of encapsulate. 

841
00:38:52,000 --> 00:38:54,100
The overall problem here, which 
is, there is a problem, both in 

842
00:38:54,107 --> 00:38:57,000
social media, and media, in a 
bunch of these areas, but only 

843
00:38:57,000 --> 00:38:58,900
one of those sides ever gets 
looked at in the media, never 

844
00:38:58,900 --> 00:39:00,600
looks them. 
Condé Nast will never ask 

845
00:39:00,600 --> 00:39:04,700
itself, have generations of our 
publishing, but they're really 

846
00:39:04,700 --> 00:39:07,300
obviously doesn't work that way.
A reporter at wired isn't even 

847
00:39:07,300 --> 00:39:09,500
connected, to what the 
editor-in-chief of wired things.

848
00:39:09,500 --> 00:39:12,700
Let alone with somebody it voted
but nobody writes it. 

849
00:39:12,900 --> 00:39:14,700
Nobody will write any of these 
things in the bigger context. 

850
00:39:14,700 --> 00:39:16,900
No, but I understand the 
argument in part, because the 

851
00:39:16,900 --> 00:39:19,400
media companies themselves 
aren't doing those studies to. 

852
00:39:19,400 --> 00:39:21,400
So, we're in an unusual 
situation, we're probably for 

853
00:39:21,400 --> 00:39:24,400
the first time. 
We have studies about how media 

854
00:39:24,400 --> 00:39:26,900
products. 
If we want to call, you know, 

855
00:39:26,900 --> 00:39:30,100
Instagram and meteor product. 
We want to call copy of Vogue 

856
00:39:30,100 --> 00:39:32,700
and media product, how they're 
actually making people feel, you

857
00:39:32,700 --> 00:39:36,300
know, Vogue does reader surveys 
all Publications do reader 

858
00:39:36,300 --> 00:39:38,800
research but it's really mostly 
about subscription. 

859
00:39:39,100 --> 00:39:42,000
How did this convert a 
subscription or did this not 

860
00:39:42,000 --> 00:39:44,600
convert a subscription? 
What are our most loyal readers 

861
00:39:44,600 --> 00:39:47,000
do? 
And how do we get more of them? 

862
00:39:47,100 --> 00:39:50,700
I think to your point about why 
the Facebook research will 

863
00:39:50,700 --> 00:39:53,600
continue to be important beyond 
the stories is that we'd have 

864
00:39:53,600 --> 00:39:57,800
never seen really this much 
research done on media products 

865
00:39:57,800 --> 00:39:59,200
before and I think would be 
great. 

866
00:39:59,400 --> 00:40:02,300
More was done. 
I think it'd be great if the 

867
00:40:02,300 --> 00:40:04,200
media itself, didn't more of 
that kind of research. 

868
00:40:04,200 --> 00:40:07,400
But right now, it does 0 of it. 
So, right. 

869
00:40:07,400 --> 00:40:09,200
Well, which is why the big irony
series, it's coming out the 

870
00:40:09,207 --> 00:40:11,600
Wall, Street Journal, a Murdoch 
property, and the idea of Rupert

871
00:40:11,600 --> 00:40:14,100
Murdoch, having like a Civic 
Integrity team looking at his 

872
00:40:14,100 --> 00:40:16,300
impact on Democracy is kind of 
hilarious. 

873
00:40:16,500 --> 00:40:19,400
I sort of realized that, you 
know, when Ben wrote that column

874
00:40:19,400 --> 00:40:20,800
and by the way, it must be said 
that. 

875
00:40:20,800 --> 00:40:23,100
If you want to talk about the 
various purveyors of resistance 

876
00:40:23,100 --> 00:40:26,700
journalism, the release of the 
steel, dossier could easily fall

877
00:40:26,700 --> 00:40:28,900
within that category, that was 
Ben's decision. 

878
00:40:29,300 --> 00:40:31,900
It's amazing. 
Yeah it was and look he can 

879
00:40:31,900 --> 00:40:34,300
defend himself and I'm sure he 
will on any has several times. 

880
00:40:34,300 --> 00:40:36,700
So he has, yeah, there's no 
reason to call him out. 

881
00:40:36,700 --> 00:40:40,400
He's a friend of the show but it
absolutely was friend of my, 

882
00:40:40,700 --> 00:40:44,000
friend of up, he was a great 
cast and we like, then it was a 

883
00:40:44,000 --> 00:40:49,600
great guy love, but I started 
fighting people Eric, you are 

884
00:40:49,600 --> 00:40:52,500
more like been than any of the 
other people on this show. 

885
00:40:52,500 --> 00:40:56,000
And this is why you can't get 
all of those Farm on side 

886
00:40:56,000 --> 00:40:57,400
Channel. 
My Discord thread. 

887
00:40:57,400 --> 00:41:01,400
I write, I wrote it like, multi,
A screed defending Ben Smith 

888
00:41:01,400 --> 00:41:03,100
owning shares. 
I've written on Twitter. 

889
00:41:03,100 --> 00:41:05,600
I'm I'm the most Ben Smith Toady
around. 

890
00:41:05,600 --> 00:41:08,000
I agree with Ben Smith on the 
steel dossier. 

891
00:41:08,000 --> 00:41:10,700
The only reason you're the only 
reason you're ambivalent is 

892
00:41:10,700 --> 00:41:12,500
because you guys are the same 
person. 

893
00:41:12,500 --> 00:41:13,400
I'm just saying. 
Yeah. 

894
00:41:13,400 --> 00:41:15,700
They if you playing the banker 
is more fun. 

895
00:41:15,700 --> 00:41:18,000
If he's a friend of me then he's
just a blind friend. 

896
00:41:18,000 --> 00:41:19,600
Yeah. 
Weirdly Eric was not there that 

897
00:41:19,600 --> 00:41:21,700
week, we interviewed been. 
So you make your own decisions 

898
00:41:21,700 --> 00:41:23,100
on what that? 
I know I missed out. 

899
00:41:23,100 --> 00:41:25,400
I've talked to many, many ways. 
Oh, I just mean you're the same 

900
00:41:25,400 --> 00:41:29,200
person, but that's not going to 
make a substitution. 

901
00:41:29,300 --> 00:41:32,600
Stack joke aren't you supposed 
to make some tax jokes like 

902
00:41:32,600 --> 00:41:35,300
every 20 minutes or something? 
And that required traditional 

903
00:41:35,300 --> 00:41:38,400
journalists. 
But no in my mind everything has

904
00:41:38,400 --> 00:41:40,400
become resistant journalism, not
everything. 

905
00:41:40,400 --> 00:41:44,000
A lot of things have become 
resistant journalism now because

906
00:41:44,300 --> 00:41:48,400
I think people recognize that as
subscriptions have driven, you 

907
00:41:48,400 --> 00:41:51,100
know, to be the predominant, 
business model of many things of

908
00:41:51,100 --> 00:41:53,600
many of these news 
organizations, there is real 

909
00:41:53,600 --> 00:41:56,500
need to serve that audience. 
Do you think that that has 

910
00:41:56,500 --> 00:41:59,900
gotten to be more and more the 
case, you know, post, Teen 

911
00:42:00,200 --> 00:42:03,500
post-2020. 
And if you agree with that Alex,

912
00:42:03,800 --> 00:42:06,600
what role do you think? 
The tech companies and Facebook 

913
00:42:06,600 --> 00:42:10,500
specifically play in in that 
being the way things are? 

914
00:42:10,900 --> 00:42:13,400
I mean, yeah. 
So the Dynamics of like the 

915
00:42:13,400 --> 00:42:16,100
change way from the advertising 
model towards subscriptions that

916
00:42:16,100 --> 00:42:19,200
is obviously being driven by the
tech companies taking up so much

917
00:42:19,200 --> 00:42:21,900
revenue and I think, and I've 
said this probably for I think a

918
00:42:21,900 --> 00:42:24,400
huge failure by the tech 
companies was never to figure 

919
00:42:24,400 --> 00:42:27,500
out a way to revenue share with 
legitimate media, whatever you 

920
00:42:27,500 --> 00:42:29,700
call legitimate media, right? 
Like a bunch of The stuff has 

921
00:42:29,700 --> 00:42:31,200
happened and it happened kind of
too late. 

922
00:42:31,400 --> 00:42:34,600
After media was hollowed out, 
especially local. 

923
00:42:34,700 --> 00:42:37,400
I saw some stats of like, 10% of
people who work in journalism 

924
00:42:37,400 --> 00:42:39,700
work for the New York Times. 
Now, it's like four percent of 

925
00:42:39,700 --> 00:42:43,100
the actual reporters and 10% of 
people overall. 

926
00:42:43,500 --> 00:42:45,300
Officially Wonder, like how much
support staff in your time, 

927
00:42:45,300 --> 00:42:49,300
says, I'm skeptical that I would
need to see ya insulted Dean 

928
00:42:49,300 --> 00:42:50,900
McKay, I talk about that your 
time. 

929
00:42:52,100 --> 00:42:55,700
This is like the game Katie 
fire, well, that would get rid 

930
00:42:55,700 --> 00:42:59,600
of one employee at the, I mean, 
that, like, consolidation, Is 

931
00:42:59,600 --> 00:43:03,400
clearly both from advertising 
being taken up by the companies 

932
00:43:03,400 --> 00:43:06,700
as well as the decisions of some
journalistic Outlets. 

933
00:43:06,700 --> 00:43:09,800
Plus just the success of the 
times of building up an audience

934
00:43:09,800 --> 00:43:12,200
that is willing to pay all this 
money but is a smaller smaller 

935
00:43:12,200 --> 00:43:14,000
audience, right? 
So that's what sub s 

936
00:43:14,000 --> 00:43:15,400
demonstrate. 
I mean, the New York Times thing

937
00:43:15,400 --> 00:43:17,900
is the same as the subject 
demonstration, which is if you 

938
00:43:17,900 --> 00:43:20,800
have a small number of people 
who love you, then, that is way 

939
00:43:20,800 --> 00:43:24,200
more profitable than making it, 
you know, trying to make it in a

940
00:43:24,400 --> 00:43:27,700
per click on an ad basis, right?
But our subscription base is 

941
00:43:27,700 --> 00:43:29,200
growing. 
It's not a small number of 

942
00:43:29,300 --> 00:43:32,200
People who are subscribing so 
it's hardcore. 

943
00:43:32,200 --> 00:43:34,600
Dedicated you mean compared to 
like the entire universe of 

944
00:43:34,600 --> 00:43:38,700
people who use Facebook for TV? 
What I'm saying is like you have

945
00:43:38,700 --> 00:43:42,100
a there's a small number of 
people who really all of these 

946
00:43:42,100 --> 00:43:44,300
media outlets and moved to 
subscription are absolutely 

947
00:43:44,300 --> 00:43:46,300
motivated to keep their small 
number of subscribers have. 

948
00:43:46,300 --> 00:43:47,900
Okay. 
I think we have more than 8 

949
00:43:47,900 --> 00:43:49,400
million subscribers. 
I don't think it's like the 

950
00:43:49,400 --> 00:43:52,300
tiniest number and I was just 
trying to be clear. 

951
00:43:52,300 --> 00:43:55,700
I get the times I'm you know but
like I just have to like we're 

952
00:43:55,700 --> 00:43:58,200
all liberal Elites Coastal 
Elites, right? 

953
00:43:58,400 --> 00:44:00,700
And don't you trust At the most,
this is what I want to ask. 

954
00:44:00,800 --> 00:44:03,600
What do you trust like, what if 
not the times? 

955
00:44:03,600 --> 00:44:06,400
What, you know, like oh I do 
trust the times most of the time

956
00:44:06,400 --> 00:44:07,700
except like, I'll read the 
story. 

957
00:44:07,700 --> 00:44:10,300
I mean, again, there's a 
difference between like things 

958
00:44:10,300 --> 00:44:13,100
being factually accurate or not 
and things being applied in the 

959
00:44:13,100 --> 00:44:14,900
right context or the amount of 
coverage they get. 

960
00:44:14,900 --> 00:44:16,500
Right? 
Yeah, I understand. 

961
00:44:16,500 --> 00:44:19,100
I mean, like, this idea of 
resistance journalism, I'm 

962
00:44:19,200 --> 00:44:22,800
opposed to it just in practice, 
just, that's just me. 

963
00:44:23,100 --> 00:44:25,000
I mean, if you look at most of 
the things that came out of the 

964
00:44:25,000 --> 00:44:27,900
Frances Hogan stuff, you can 
agree or disagree with the 

965
00:44:27,900 --> 00:44:31,300
conclusions and, and either 
they, you know, illustrate a 

966
00:44:31,308 --> 00:44:34,300
completely fucked up company. 
Or the question is, like, what 

967
00:44:34,300 --> 00:44:36,600
should they just have released 
that information publicly? 

968
00:44:36,900 --> 00:44:39,200
And we could have seen, oh, 
there were research internally 

969
00:44:39,200 --> 00:44:42,400
at Facebook, showing that there 
a percentage of young girls that

970
00:44:42,400 --> 00:44:45,300
use Instagram that are unhappy. 
Obviously, you don't want to put

971
00:44:45,300 --> 00:44:48,800
out the internal deliberations 
at the New York Times, but some 

972
00:44:48,800 --> 00:44:52,700
sort of public discussion of a 
decision as to why they chose 

973
00:44:52,700 --> 00:44:56,000
not to run certain things would 
sort of benefit people to 

974
00:44:56,000 --> 00:45:00,600
understand what the process is 
that, you know, Separates a 

975
00:45:00,600 --> 00:45:03,200
mainstream news organization 
from one that has no 

976
00:45:03,200 --> 00:45:05,500
credibility. 
I mean, the Facebook side. 

977
00:45:05,500 --> 00:45:07,200
I think absolutely, they should 
publish it. 

978
00:45:07,200 --> 00:45:09,500
We go two directions from this 
one either. 

979
00:45:09,500 --> 00:45:12,100
Tech companies start to publish 
these things because they 

980
00:45:12,100 --> 00:45:14,400
realize that doing it 
themselves. 

981
00:45:14,400 --> 00:45:16,800
Makes it not a scandal like a 
bunch of these reports are just 

982
00:45:16,800 --> 00:45:18,800
reports in like there are good 
things and bad things like the 

983
00:45:18,808 --> 00:45:21,000
Instagram one. 
Honestly is both sides and so 

984
00:45:21,000 --> 00:45:23,200
Facebook, published that it 
would not be seen as like this 

985
00:45:23,200 --> 00:45:25,200
huge Scandal for what you need. 
All these investigations. 

986
00:45:25,500 --> 00:45:27,800
We actually launched a journal 
at Stanford which was totally 

987
00:45:27,800 --> 00:45:29,800
coincidental at the hog and 
stuff but I It's pretty the 

988
00:45:29,808 --> 00:45:31,800
Journal of online trust and 
safety specifically. 

989
00:45:31,800 --> 00:45:34,500
Because we want to get platforms
to published peer-reviewed 

990
00:45:34,500 --> 00:45:37,400
research in places that they can
interact with academics and 

991
00:45:37,400 --> 00:45:39,600
others. 
So I think that's a good thing. 

992
00:45:39,700 --> 00:45:41,700
There's also the possibility 
from here is that companies just

993
00:45:41,700 --> 00:45:43,700
don't do the research because 
the truth is that Facebook has 

994
00:45:43,700 --> 00:45:45,400
more people working on this than
the rest of the industry 

995
00:45:45,400 --> 00:45:47,700
combined and is now being 
punished for that fact. 

996
00:45:47,700 --> 00:45:50,200
And so there's a possibility 
that if you're at Tick-Tock, 

997
00:45:50,300 --> 00:45:53,900
you're like, oh no, never look, 
never create a document that 

998
00:45:53,900 --> 00:45:56,900
says anything is bad because 
eventually if it leaks it 

999
00:45:56,900 --> 00:45:59,100
becomes a scandal and it's 
better not to look to go. 

1000
00:45:59,400 --> 00:46:01,300
Also to like kind of the 2016 
story. 

1001
00:46:01,500 --> 00:46:04,200
Facebook is not the largest 
Advertiser on the internet, it 

1002
00:46:04,200 --> 00:46:08,100
turns out, not even by a long 
shot, and there's another large 

1003
00:46:08,100 --> 00:46:11,100
company. 
I'm sure smoogle that had lots 

1004
00:46:11,100 --> 00:46:15,100
of stuff going on that quietly 
found, it never told the public 

1005
00:46:15,200 --> 00:46:17,300
never told molars team never 
told anybody because they're not

1006
00:46:17,300 --> 00:46:19,100
legally required to what you 
think. 

1007
00:46:19,100 --> 00:46:21,900
Google is worse than Facebook 
and Russian interference in the 

1008
00:46:21,908 --> 00:46:25,500
2016 election. 
I don't I don't know because we 

1009
00:46:25,500 --> 00:46:30,900
don't know anything and And 
since then their playbook has 

1010
00:46:30,900 --> 00:46:33,300
been to not share anything, 
right? 

1011
00:46:33,300 --> 00:46:35,800
So like Facebook has crowd 
tangle which might get killed. 

1012
00:46:35,800 --> 00:46:39,100
So I like crowd angles is 
fantastic platform. 

1013
00:46:39,400 --> 00:46:42,800
That is super useful and was 
incredibly key to our 2020 work 

1014
00:46:43,200 --> 00:46:45,500
and all the signs look, like 
Facebook's going to kill it. 

1015
00:46:45,500 --> 00:46:48,100
But as of right now, Facebook 
has incredible transparency. 

1016
00:46:48,300 --> 00:46:49,800
Twitter has incredible 
transparency. 

1017
00:46:50,000 --> 00:46:54,200
YouTube is almost impossible to 
study as an outside researcher. 

1018
00:46:54,200 --> 00:46:56,600
Tick-Tock is almost impossible 
to say, he's outside researcher,

1019
00:46:56,600 --> 00:46:59,100
and so, I think we're going to 
need to have mandated. 

1020
00:46:59,300 --> 00:47:01,100
Here. 
Because what's happened so far, 

1021
00:47:01,100 --> 00:47:02,200
is that the companies have 
opened up. 

1022
00:47:02,200 --> 00:47:04,100
Have had a huge amount of 
criticism and there's never any 

1023
00:47:04,100 --> 00:47:06,500
criticism the companies that 
have decided not to open up. 

1024
00:47:06,700 --> 00:47:09,300
And as long as that Dynamic 
continues to exist, I think 

1025
00:47:09,300 --> 00:47:11,700
we're going to have this 
lopsided coverage but also the 

1026
00:47:11,700 --> 00:47:13,100
inability to fix some of these 
problems. 

1027
00:47:13,600 --> 00:47:15,000
So guys, I'm sorry. 
I have the time out. 

1028
00:47:15,000 --> 00:47:16,600
My daughter, schools about to 
call her the rubble. 

1029
00:47:16,600 --> 00:47:18,800
So we're about 15 minutes so I 
can continue this conversation. 

1030
00:47:18,800 --> 00:47:20,400
15 minutes or we've got to wrap 
it up. 

1031
00:47:20,400 --> 00:47:23,000
I'm so I think we've gone long. 
Sure I give it like a little 

1032
00:47:23,000 --> 00:47:25,000
goodbye. 
See ya. 

1033
00:47:25,000 --> 00:47:26,400
Yes thanks so much for having 
me. 

1034
00:47:26,900 --> 00:47:30,600
All right. 
We're like what's a concluding 

1035
00:47:30,600 --> 00:47:32,900
Point earlier that's not to say?
Well Alex I think we've all 

1036
00:47:32,900 --> 00:47:36,000
agreed that Facebook and the 
media companies are all the same

1037
00:47:36,100 --> 00:47:39,600
that we all need to get in the 
same ship together and and and 

1038
00:47:39,600 --> 00:47:41,100
figure it all out. 
They're not all the same. 

1039
00:47:41,200 --> 00:47:45,100
I think what we have write my 
conclusion would be we have 

1040
00:47:45,100 --> 00:47:48,700
fundamental weaknesses in our 
media environment as well as 

1041
00:47:48,700 --> 00:47:51,000
like the psychology of Americans
and we have to look at the big 

1042
00:47:51,000 --> 00:47:52,800
picture if we're going to 
understand or deal with any of 

1043
00:47:52,800 --> 00:47:54,700
these problems. 
Thank you so much for having me.

1044
00:47:54,700 --> 00:47:55,600
I'll send you guys the 
recording. 

1045
00:47:55,600 --> 00:47:57,500
Sounds good, thank you. 
Thank you so much Alex. 

1046
00:47:57,500 --> 00:47:59,600
We really appreciate it. 
Goodbye, 

1047
00:48:09,700 --> 00:48:12,400
goodbye. 
Goodbye, goodbye, goodbye, 

1048
00:48:12,500 --> 00:48:13,400
goodbye. 
Goodbye.