1
00:00:04,120 --> 00:00:05,840
Speaker 1: There Are No Girls on the Internet, as a production

2
00:00:05,880 --> 00:00:13,400
Speaker 1: of iHeartRadio and Unbossed Creative. I'm Bridget Todd, and this

3
00:00:13,520 --> 00:00:18,200
Speaker 1: is there Are No Girls on the Internet. A pop

4
00:00:18,239 --> 00:00:22,160
Speaker 1: star accused of being a secret Nazi. A family restaurant

5
00:00:22,200 --> 00:00:25,840
Speaker 1: accused of going woke. Now, at first glance, these may

6
00:00:25,880 --> 00:00:30,080
Speaker 1: seem like two completely unrelated Internet controversies, but dig a

7
00:00:30,120 --> 00:00:32,960
Speaker 1: little bit deeper and they share something important in common.

8
00:00:33,600 --> 00:00:37,840
Speaker 1: Both controversies were amplified by coordinated bot campaigns that sought

9
00:00:37,880 --> 00:00:41,360
Speaker 1: to rile up ordinary users and draw them into conversation.

10
00:00:42,200 --> 00:00:45,480
Speaker 1: In this episode, we're breaking down two reports on coordinated

11
00:00:45,520 --> 00:00:49,479
Speaker 1: bot campaigns, one aimed at Taylor Swift, another aimed at

12
00:00:49,520 --> 00:00:53,120
Speaker 1: Cracker Barrel to talk about what happens when online outrage

13
00:00:53,200 --> 00:00:57,000
Speaker 1: is manufactured, amplified, and weaponized, as well as what it

14
00:00:57,080 --> 00:00:59,880
Speaker 1: means for the authentic human voices trying to be heard

15
00:01:00,040 --> 00:01:03,760
Speaker 1: against the algorithmic roar. Because if bots can create the

16
00:01:03,800 --> 00:01:08,120
Speaker 1: appearance of consensus and generate real backlash or controversy at scale,

17
00:01:08,560 --> 00:01:10,399
Speaker 1: how do any of us know when we're reacting to

18
00:01:10,480 --> 00:01:14,320
Speaker 1: real people or just reacting exactly the way somebody wanted

19
00:01:14,400 --> 00:01:16,880
Speaker 1: us to. This is what Molly Dwyer has spent her

20
00:01:16,920 --> 00:01:20,720
Speaker 1: career researching. She describes herself as a kind of professional

21
00:01:20,800 --> 00:01:24,039
Speaker 1: Internet vibe checker, but her real title is Director of

22
00:01:24,080 --> 00:01:27,520
Speaker 1: Insights at Peak Metrics. And before the Internet was talking

23
00:01:27,560 --> 00:01:31,560
Speaker 1: about bots manipulating conversations around Taylor Swift, Molly had been

24
00:01:31,560 --> 00:01:35,080
Speaker 1: looking into how similar kinds of unauthentic behavior was driving

25
00:01:35,080 --> 00:01:38,560
Speaker 1: the discourse around the old timey country themed chain restaurant

26
00:01:38,760 --> 00:01:42,600
Speaker 1: Cracker Barrel. Now. I know it may seem wild to

27
00:01:42,640 --> 00:01:45,839
Speaker 1: think that anybody would care deeply about this, but Molly

28
00:01:45,959 --> 00:01:49,320
Speaker 1: says it's yet another instance of how easily the Internet

29
00:01:49,320 --> 00:01:53,640
Speaker 1: can be manipulated to change people's views. Her company, Peak Metrics,

30
00:01:53,880 --> 00:01:56,680
Speaker 1: was not involved with the report about Taylor Swift, but

31
00:01:56,760 --> 00:01:59,360
Speaker 1: a few months earlier they published a report about the

32
00:01:59,360 --> 00:02:02,320
Speaker 1: Cracker Barrel controversy that had a lot of similar findings.

33
00:02:02,760 --> 00:02:04,680
Speaker 1: So we asked her about what was happening with the

34
00:02:04,720 --> 00:02:07,840
Speaker 1: Cracker Barrel discourse and got her take on the recent

35
00:02:07,920 --> 00:02:08,959
Speaker 1: Taylor Swift report.

36
00:02:09,880 --> 00:02:13,120
Speaker 2: I spent a number of years living in Russia on

37
00:02:13,360 --> 00:02:17,520
Speaker 2: various US State Department programs to help young people study

38
00:02:17,600 --> 00:02:20,240
Speaker 2: languages that are critical for national security. I say that's

39
00:02:20,280 --> 00:02:24,320
Speaker 2: relevant because my first kind of experience of how the

40
00:02:24,400 --> 00:02:27,800
Speaker 2: Internet can manipulate and change people's view on the world

41
00:02:27,919 --> 00:02:31,200
Speaker 2: was my experience sort of pre and post twenty fourteen

42
00:02:31,240 --> 00:02:34,000
Speaker 2: in Russia, where I watched people that I knew personally

43
00:02:34,680 --> 00:02:38,080
Speaker 2: go through quite a transformation on their perspective on Ukraine

44
00:02:38,120 --> 00:02:39,800
Speaker 2: and the US in a very short period of time.

45
00:02:40,320 --> 00:02:42,120
Speaker 2: So that really kicked off kind of my interest in

46
00:02:42,120 --> 00:02:45,920
Speaker 2: this idea of like information operations, how did the things

47
00:02:45,960 --> 00:02:49,480
Speaker 2: that we consume on the Internet impact our worldview, whether

48
00:02:49,520 --> 00:02:53,680
Speaker 2: we're aware of it or not. Immediately out of that,

49
00:02:54,080 --> 00:02:56,560
Speaker 2: I went to work for a small startup that does

50
00:02:56,600 --> 00:02:59,600
Speaker 2: open source intelligence, which is really just another way of

51
00:02:59,680 --> 00:03:02,960
Speaker 2: talking about all of the data that is publicly available

52
00:03:03,000 --> 00:03:06,359
Speaker 2: on the Internet. There's a pretty big tool set out

53
00:03:06,360 --> 00:03:10,040
Speaker 2: there nowadays for different technologies that will help you quantify

54
00:03:10,120 --> 00:03:13,600
Speaker 2: and qualify what is being said, what is being researched

55
00:03:13,600 --> 00:03:16,120
Speaker 2: on the Internet. And I've had a decade of experience

56
00:03:16,240 --> 00:03:19,200
Speaker 2: in that open source intelligence tool space.

57
00:03:19,240 --> 00:03:20,320
Speaker 3: At this point.

58
00:03:20,520 --> 00:03:23,440
Speaker 2: Now, the Internet of we're going into twenty twenty six.

59
00:03:23,480 --> 00:03:26,160
Speaker 2: The Internet of twenty twenty six is very different than

60
00:03:26,200 --> 00:03:29,679
Speaker 2: the Internet of twenty sixteen. When I first started out

61
00:03:29,720 --> 00:03:34,000
Speaker 2: in this industry. Traditional metrics that we were used to,

62
00:03:34,320 --> 00:03:38,120
Speaker 2: like volume or sentiment, or even just taking for granted

63
00:03:38,160 --> 00:03:41,400
Speaker 2: the idea that a verified page was who they say

64
00:03:41,400 --> 00:03:45,480
Speaker 2: they are are just no longer applicable nowadays. And I

65
00:03:45,480 --> 00:03:48,760
Speaker 2: think that the Internet has fundamentally changed from under us.

66
00:03:48,840 --> 00:03:51,640
Speaker 2: I think especially obviously since the rise of llms and

67
00:03:51,680 --> 00:03:55,120
Speaker 2: agentic AI, and we're all just sort of grappling and

68
00:03:55,120 --> 00:03:58,080
Speaker 2: trying to figure out what this means for us. And

69
00:03:58,200 --> 00:04:01,920
Speaker 2: my role at Peak Metrics is to help companies and

70
00:04:02,040 --> 00:04:05,320
Speaker 2: other organizations figure out what the Internet means for them.

71
00:04:06,400 --> 00:04:09,520
Speaker 1: Before we started recording, you said that you basically were

72
00:04:09,720 --> 00:04:14,120
Speaker 1: an Internet vibe checker, that it's one of the reasons

73
00:04:14,160 --> 00:04:16,240
Speaker 1: I wanted to talk to you is that I've had

74
00:04:16,240 --> 00:04:20,839
Speaker 1: the experience of seeing a conversation kind of take hold

75
00:04:20,839 --> 00:04:25,080
Speaker 1: of the Internet very quickly, much more quickly than other

76
00:04:25,200 --> 00:04:29,080
Speaker 1: kind of organic conversations. And maybe you don't know something

77
00:04:29,120 --> 00:04:31,360
Speaker 1: is going up, and you suspect right things where I

78
00:04:31,400 --> 00:04:35,000
Speaker 1: would think, who really is dedicating their time to writing

79
00:04:35,040 --> 00:04:37,040
Speaker 1: a bunch of posts about this? Who really cares about this?

80
00:04:37,200 --> 00:04:40,440
Speaker 1: These conversations where your spidey senses kind of get tingling.

81
00:04:40,520 --> 00:04:42,839
Speaker 1: Is that sort of what you mean exactly?

82
00:04:43,400 --> 00:04:46,360
Speaker 2: I think we all have an innate sense of the Internet,

83
00:04:46,400 --> 00:04:50,360
Speaker 2: but because we don't have easily understandable metrics to understand

84
00:04:51,040 --> 00:04:55,359
Speaker 2: how big is this conversation? Am I seeing this because

85
00:04:55,400 --> 00:04:59,039
Speaker 2: of my own algorithmic bubble? Are other people seeing this?

86
00:04:59,240 --> 00:05:01,080
Speaker 2: And like, what are the mo motivations for people to

87
00:05:01,080 --> 00:05:04,440
Speaker 2: participate in this conversation. I'll give you an example of

88
00:05:04,480 --> 00:05:07,800
Speaker 2: maybe sort of the algorithmic bubbles that we live in.

89
00:05:09,080 --> 00:05:11,760
Speaker 2: Last year, I remember there were a lot of media

90
00:05:11,800 --> 00:05:15,120
Speaker 2: inquiries to look at the impact of I was the

91
00:05:15,160 --> 00:05:19,039
Speaker 2: first or second presidential debate with Kamalairis as the candidate,

92
00:05:19,360 --> 00:05:21,800
Speaker 2: and I was looking at the scale of conversation to

93
00:05:21,880 --> 00:05:24,880
Speaker 2: that relative to something that I had not heard of,

94
00:05:25,000 --> 00:05:27,160
Speaker 2: but someone on my team brought forward to me, which

95
00:05:27,279 --> 00:05:30,839
Speaker 2: was the plight of this squirrel that was like a

96
00:05:30,880 --> 00:05:33,200
Speaker 2: pet squirrel that was going to be put down in

97
00:05:33,240 --> 00:05:37,640
Speaker 2: New York State that Trump entered the conversation on, and I,

98
00:05:37,720 --> 00:05:40,919
Speaker 2: in my algorithmic bubble had not heard of this. I

99
00:05:41,080 --> 00:05:44,680
Speaker 2: was focused on reporting on the metrics around discussions around

100
00:05:44,720 --> 00:05:46,800
Speaker 2: the debate, and when I actually went to go look

101
00:05:46,839 --> 00:05:49,360
Speaker 2: at just the overall volume of conversation around this two

102
00:05:49,400 --> 00:05:53,400
Speaker 2: topics on X, the volume of conversation around the squirrel

103
00:05:53,760 --> 00:05:56,960
Speaker 2: was multiple x larger than the volume of the conversation

104
00:05:57,040 --> 00:06:01,160
Speaker 2: around around the presidential debate. So I think that anecdote

105
00:06:01,200 --> 00:06:04,400
Speaker 2: illustrates sometimes both the our algorithmic bubbles and just like

106
00:06:04,720 --> 00:06:09,520
Speaker 2: how big these conversations can get quickly when certain players

107
00:06:09,600 --> 00:06:13,520
Speaker 2: are involved, like when certain big influencers enter the conversation,

108
00:06:13,640 --> 00:06:17,760
Speaker 2: and then also thinking about like when bots are involved, right,

109
00:06:17,960 --> 00:06:22,200
Speaker 2: like what type of content are they incentivized to post

110
00:06:22,240 --> 00:06:26,040
Speaker 2: about and to amplify for you know, their ultimate goals

111
00:06:26,080 --> 00:06:27,320
Speaker 2: probably of monetization.

112
00:06:29,200 --> 00:06:32,359
Speaker 1: Molly's background in Russian makes sense here because there was

113
00:06:32,360 --> 00:06:35,080
Speaker 1: a time that when you talked about bat campaigns online,

114
00:06:35,240 --> 00:06:38,360
Speaker 1: the assumption was was being done by foreign bad actors,

115
00:06:38,560 --> 00:06:41,640
Speaker 1: the kind of sophisticated manipulation campaign that we associate with

116
00:06:41,680 --> 00:06:45,600
Speaker 1: an adversarial foreign country. But today things have really changed.

117
00:06:46,360 --> 00:06:49,120
Speaker 2: So I think if you take the Okham's Ranger approach

118
00:06:49,160 --> 00:06:51,240
Speaker 2: to the Internet, which is just assuming that everything is

119
00:06:51,279 --> 00:06:54,760
Speaker 2: a money making drift, you'll probably be correct in like

120
00:06:54,880 --> 00:06:59,240
Speaker 2: ninety percent of cases, and I think that that applies here.

121
00:07:00,080 --> 00:07:02,080
Speaker 2: I'll back up a little bit to talk about like

122
00:07:02,720 --> 00:07:05,440
Speaker 2: bot networks in general, and I think that this is

123
00:07:05,480 --> 00:07:08,360
Speaker 2: important to cover before we get into who is behind it.

124
00:07:08,680 --> 00:07:11,360
Speaker 2: So traditionally when we would talk about bot networks and

125
00:07:11,400 --> 00:07:15,160
Speaker 2: who was operating them, you know the degree of sophistication

126
00:07:15,280 --> 00:07:20,160
Speaker 2: required to set up these campaigns, to run these accounts simultaneously,

127
00:07:20,200 --> 00:07:22,280
Speaker 2: to get them to stay on topic with their posting.

128
00:07:22,640 --> 00:07:26,040
Speaker 2: You were looking at a relatively limited set of essentially

129
00:07:26,120 --> 00:07:29,400
Speaker 2: state adversary actors like the big bads that were used

130
00:07:29,400 --> 00:07:32,640
Speaker 2: to Russia, China, Iran, North Korea, right, that had the

131
00:07:32,680 --> 00:07:36,320
Speaker 2: sophisticated like cyber teams that could run these campaigns. And

132
00:07:36,360 --> 00:07:39,200
Speaker 2: that's why I think in our minds were still a

133
00:07:39,280 --> 00:07:42,320
Speaker 2: little bit stuck in this idea of like, okay, big

134
00:07:42,360 --> 00:07:45,960
Speaker 2: bad foreign actors manipulating the conversation. I think there are

135
00:07:46,000 --> 00:07:48,680
Speaker 2: still big bad foreign actors manipulating the conversation, and they

136
00:07:48,720 --> 00:07:50,920
Speaker 2: have a lot of incentives to do so. But the

137
00:07:50,960 --> 00:07:56,480
Speaker 2: world changed with with AI and it's now easier than

138
00:07:56,520 --> 00:07:59,160
Speaker 2: ever to run these types of campaigns. The other side

139
00:07:59,160 --> 00:08:01,480
Speaker 2: of that was you had made be like traditional non

140
00:08:01,520 --> 00:08:04,480
Speaker 2: state cyber actors like threat network, you know, a non

141
00:08:04,800 --> 00:08:08,400
Speaker 2: type people who were super sophisticated could do this right.

142
00:08:08,800 --> 00:08:12,720
Speaker 2: And now you know, smaller governments can like run a

143
00:08:12,760 --> 00:08:16,520
Speaker 2: campaign to prop up their dictator. Right, average people who

144
00:08:16,560 --> 00:08:19,360
Speaker 2: were not part of like a threat network can stand

145
00:08:19,440 --> 00:08:22,120
Speaker 2: up campaigns like this thanks to AI, you know, run

146
00:08:22,160 --> 00:08:25,360
Speaker 2: them simultaneously. And I think that's where we get into

147
00:08:25,360 --> 00:08:28,360
Speaker 2: this explanation of Okay, well what if it's not ideological

148
00:08:28,520 --> 00:08:31,600
Speaker 2: the way that like a state adversary actor would be, Well,

149
00:08:31,600 --> 00:08:34,440
Speaker 2: why are they doing this? I think shits and giggles

150
00:08:34,520 --> 00:08:37,640
Speaker 2: is always a reasonable guess. But I think that money

151
00:08:37,760 --> 00:08:39,440
Speaker 2: is the thing that makes the most sense. And I'll

152
00:08:39,440 --> 00:08:42,560
Speaker 2: tell you why. I think that worldview I think is

153
00:08:42,600 --> 00:08:45,800
Speaker 2: bearing out recently. So I don't know if you saw

154
00:08:46,120 --> 00:08:48,520
Speaker 2: in the past few weeks that X made an update

155
00:08:48,920 --> 00:08:54,640
Speaker 2: where user locations were published, and what we saw this

156
00:08:54,760 --> 00:08:57,680
Speaker 2: is again anecdotal. I don't necessarily have the data on this,

157
00:08:57,760 --> 00:08:59,960
Speaker 2: but what we saw anecdotally was that a lot of

158
00:09:00,080 --> 00:09:03,760
Speaker 2: accounts that we're posting incendiary political content on both the

159
00:09:03,840 --> 00:09:08,400
Speaker 2: left and the US right. We're coming from places like India,

160
00:09:08,960 --> 00:09:14,280
Speaker 2: the Global South, places where you know, our relatively low income.

161
00:09:15,320 --> 00:09:18,320
Speaker 2: The money that you could make from this is not insignificant.

162
00:09:19,320 --> 00:09:22,840
Speaker 2: We don't necessarily think that these people have ideological reasons

163
00:09:22,880 --> 00:09:25,520
Speaker 2: to fan discourse in the US. It just so happens

164
00:09:25,559 --> 00:09:27,560
Speaker 2: that that's maybe a good way to make a side

165
00:09:27,640 --> 00:09:31,920
Speaker 2: hustle online because of the way that you know, engagement

166
00:09:32,120 --> 00:09:36,600
Speaker 2: is monetized. So that's my overview answer of like the

167
00:09:36,640 --> 00:09:39,240
Speaker 2: types of people who could be behind these it's a

168
00:09:39,360 --> 00:09:44,000
Speaker 2: much larger set of actors than it was maybe four

169
00:09:44,080 --> 00:09:48,600
Speaker 2: years ago, But I still think that money is probably

170
00:09:48,840 --> 00:09:50,920
Speaker 2: the guiding principle of what's going on here.

171
00:09:51,559 --> 00:09:54,240
Speaker 1: Yeah, We've been having a lot of conversations about that

172
00:09:54,360 --> 00:09:58,240
Speaker 1: change at X, and I think it really helped me.

173
00:09:58,600 --> 00:10:01,160
Speaker 1: See I knew this sort of inside, but it helped.

174
00:10:01,280 --> 00:10:03,720
Speaker 1: It was just like another way to sort of crystallize

175
00:10:03,760 --> 00:10:06,880
Speaker 1: exactly what you said that we're so used to thinking

176
00:10:06,960 --> 00:10:11,840
Speaker 1: about nefarious actors, bad actors who are manipulating our online

177
00:10:11,880 --> 00:10:16,800
Speaker 1: discourse to foment chaos and confusion and political division. But

178
00:10:16,920 --> 00:10:22,200
Speaker 1: also in making this, in changing the financial payouts of X,

179
00:10:22,480 --> 00:10:27,640
Speaker 1: they certainly have incentivized people to post inflammatory, insidiary content,

180
00:10:27,800 --> 00:10:30,560
Speaker 1: rage b right content that we know works as a

181
00:10:30,600 --> 00:10:33,400
Speaker 1: revenue stream, And part of me can't even really blame

182
00:10:33,440 --> 00:10:36,920
Speaker 1: them for being like, oh, they set up this incentivized

183
00:10:37,679 --> 00:10:39,240
Speaker 1: reason for me to do this to make a little

184
00:10:39,320 --> 00:10:41,360
Speaker 1: money I need money, I'm going to do it.

185
00:10:42,000 --> 00:10:45,880
Speaker 2: Yeah, we just happened to be a big, very flammable

186
00:10:45,920 --> 00:10:50,240
Speaker 2: target for people to go after on the internet US society.

187
00:10:50,679 --> 00:10:54,880
Speaker 1: Yes, if you haven't been on a long car trip

188
00:10:54,920 --> 00:10:56,719
Speaker 1: in a while, I bet you haven't spent a lot

189
00:10:56,720 --> 00:10:59,600
Speaker 1: of time thinking about the restaurant Cracker Barrel. So here's

190
00:10:59,600 --> 00:11:03,080
Speaker 1: the content. Earlier this year, Cracker Barrel rolled out a

191
00:11:03,080 --> 00:11:06,839
Speaker 1: new logo. Previously, the logo featured an old white man

192
00:11:07,120 --> 00:11:11,360
Speaker 1: in overalls casually leaning over a barrel. But this new

193
00:11:11,400 --> 00:11:13,959
Speaker 1: logo removed the old man and just kept the name

194
00:11:14,120 --> 00:11:17,960
Speaker 1: Cracker Barrel in an ever so slightly sleeker font, one

195
00:11:17,960 --> 00:11:21,120
Speaker 1: that maybe jibed with a younger audience. To be sure,

196
00:11:21,280 --> 00:11:24,040
Speaker 1: this is the kind of boring corporate marketing change that

197
00:11:24,080 --> 00:11:28,040
Speaker 1: most people wouldn't even notice, But in today's climate, where

198
00:11:28,080 --> 00:11:30,160
Speaker 1: every little thing can be turned into evidence of a

199
00:11:30,240 --> 00:11:34,360
Speaker 1: woke agenda, it became an entire news cycle. Okay, so

200
00:11:34,559 --> 00:11:36,920
Speaker 1: you really set the stage wonderfully. And I think I

201
00:11:36,960 --> 00:11:40,040
Speaker 1: mentioned this too, that independently we had been wanting to

202
00:11:40,080 --> 00:11:43,040
Speaker 1: do an episode about bots and sort of the general

203
00:11:43,120 --> 00:11:45,520
Speaker 1: grip that they have on our discourse, and one of

204
00:11:45,520 --> 00:11:48,280
Speaker 1: the conversations that I saw that. I was like, this

205
00:11:48,520 --> 00:11:53,680
Speaker 1: is very weird conversation happening. Was cracker Barrel changing their logo?

206
00:11:54,160 --> 00:11:57,720
Speaker 1: I I grew up in the South and my parents

207
00:11:57,720 --> 00:11:59,440
Speaker 1: went to cracker barrel. We had one in our town.

208
00:11:59,600 --> 00:12:02,319
Speaker 1: They went every week. But other than my parents, I

209
00:12:02,360 --> 00:12:05,880
Speaker 1: don't think I've ever heard anybody mention cracker Barrel. It

210
00:12:05,920 --> 00:12:08,600
Speaker 1: wasn't like a big part of the discourse. So imagine

211
00:12:08,640 --> 00:12:11,480
Speaker 1: my surprise when one day I wake up and everybody's

212
00:12:11,480 --> 00:12:14,360
Speaker 1: freaking talking about cracker Barrel. Then I saw the Peak

213
00:12:14,400 --> 00:12:17,880
Speaker 1: Metrics report about the way that bots might have helped

214
00:12:17,880 --> 00:12:21,160
Speaker 1: shape the way that conversation spread online. How did that

215
00:12:21,160 --> 00:12:24,000
Speaker 1: come to be something that Peak Metrics was looking at.

216
00:12:24,400 --> 00:12:26,959
Speaker 2: Yeah, I too can't believe that we're still talking about

217
00:12:26,960 --> 00:12:28,280
Speaker 2: cracker Barrel, but here we are.

218
00:12:28,360 --> 00:12:28,760
Speaker 3: I think.

219
00:12:29,559 --> 00:12:32,920
Speaker 2: I think why media latched onto it when we put

220
00:12:32,960 --> 00:12:37,439
Speaker 2: out some initial findings was this sense of why are

221
00:12:37,440 --> 00:12:40,600
Speaker 2: we talking about this? And we gave them some numbers

222
00:12:40,640 --> 00:12:44,840
Speaker 2: to chew on to help them understand what degree of

223
00:12:44,880 --> 00:12:49,560
Speaker 2: this conversation is potentially inorganic or automated. That might explain

224
00:12:49,760 --> 00:12:53,560
Speaker 2: why it's staying in the discourse longer than we would

225
00:12:53,600 --> 00:12:55,680
Speaker 2: expect it to be or why it rose to the

226
00:12:55,679 --> 00:12:59,160
Speaker 2: top of the discourse in the first place. And I

227
00:12:59,200 --> 00:13:03,960
Speaker 2: think our key finding there was that, you know, within

228
00:13:04,000 --> 00:13:06,400
Speaker 2: the first twenty four hours of this like rage cycle,

229
00:13:07,040 --> 00:13:11,520
Speaker 2: we found that the original posts that seeded this idea

230
00:13:11,559 --> 00:13:14,400
Speaker 2: of a boycott and you got to rewind it back

231
00:13:14,400 --> 00:13:17,200
Speaker 2: because I think that's also another principle that's hard to

232
00:13:17,240 --> 00:13:20,040
Speaker 2: do on the modern Internet is how did something start?

233
00:13:20,320 --> 00:13:23,040
Speaker 2: There's not necessarily an easy answer to that. You have

234
00:13:23,040 --> 00:13:24,760
Speaker 2: to bring in a lot of data, search for the

235
00:13:24,840 --> 00:13:27,360
Speaker 2: right stuff to figure out how a trend even started.

236
00:13:27,920 --> 00:13:30,160
Speaker 2: So from what we could see how the idea of

237
00:13:30,200 --> 00:13:33,800
Speaker 2: a boycott started, it did start from what our tech

238
00:13:33,800 --> 00:13:40,199
Speaker 2: would classify as organic accounts, but very quickly those calls

239
00:13:40,240 --> 00:13:44,480
Speaker 2: for a boycott were amplified by accounts that we flagged

240
00:13:44,600 --> 00:13:47,880
Speaker 2: as automated or inorganic. So there was this sort of

241
00:13:47,920 --> 00:13:53,280
Speaker 2: like seeding introduction of a claim that immediately got amplified

242
00:13:54,080 --> 00:13:56,360
Speaker 2: and at the height of the discourse, Like within the

243
00:13:56,400 --> 00:14:00,120
Speaker 2: first twenty four hours, we found that basically half of

244
00:14:00,160 --> 00:14:02,720
Speaker 2: the posts that we're calling for a boycott we're coming

245
00:14:02,760 --> 00:14:06,640
Speaker 2: from accounts that we flagged as automated. Half is a

246
00:14:06,640 --> 00:14:10,319
Speaker 2: big number, but we also don't necessarily at the time

247
00:14:10,320 --> 00:14:11,840
Speaker 2: that we publish the report, we didn't have a lot

248
00:14:11,880 --> 00:14:15,000
Speaker 2: of comparison points. I think that's key to maybe understanding

249
00:14:15,000 --> 00:14:17,400
Speaker 2: what's going on with the Taylor Swift discourses. I think

250
00:14:17,440 --> 00:14:21,200
Speaker 2: we're all across the industry like finding these numbers and

251
00:14:21,200 --> 00:14:23,040
Speaker 2: trying to figure out what they mean. If I were

252
00:14:23,080 --> 00:14:25,680
Speaker 2: to give you a spidy sense now several months after

253
00:14:25,760 --> 00:14:30,400
Speaker 2: Cracker Barrel is that when there's an incendiary conversation online,

254
00:14:31,080 --> 00:14:34,640
Speaker 2: I would consider a normal baseline for the amount of

255
00:14:34,680 --> 00:14:37,920
Speaker 2: automated activity in that conversation to be somewhere between twenty

256
00:14:37,920 --> 00:14:41,880
Speaker 2: to thirty percent of the conversation wow, And it typically

257
00:14:41,960 --> 00:14:45,280
Speaker 2: will go higher than that when we hit like a

258
00:14:45,320 --> 00:14:48,560
Speaker 2: crisis point. So like the Cracker Barrel changes its logo,

259
00:14:49,400 --> 00:14:52,360
Speaker 2: organic accounts seed this call for a boycott, and then

260
00:14:52,480 --> 00:14:56,520
Speaker 2: it jumps up. But that may just be a symptom

261
00:14:56,560 --> 00:14:59,960
Speaker 2: of how the internet works, and not necessarily a symptom

262
00:15:00,160 --> 00:15:06,440
Speaker 2: of a coordinated campaign to target chain restaurants that are

263
00:15:06,480 --> 00:15:08,000
Speaker 2: abandoning traditional values.

264
00:15:08,080 --> 00:15:08,240
Speaker 1: Right.

265
00:15:09,040 --> 00:15:10,640
Speaker 3: I think we're grasping these numbers.

266
00:15:10,760 --> 00:15:12,800
Speaker 2: We're putting a number to this thing for the first time,

267
00:15:12,840 --> 00:15:15,320
Speaker 2: and so we're trying to figure out what is baseline

268
00:15:15,360 --> 00:15:15,960
Speaker 2: at this point.

269
00:15:16,560 --> 00:15:18,440
Speaker 1: So what's the takeaway that you all found from the

270
00:15:18,440 --> 00:15:19,480
Speaker 1: Crocker Bill report.

271
00:15:19,760 --> 00:15:24,359
Speaker 2: I think the takeaway here is one into being automated.

272
00:15:24,440 --> 00:15:26,480
Speaker 2: That's a big number, but that's still a lot of

273
00:15:26,480 --> 00:15:29,800
Speaker 2: people that have potentially big, real feelings about this. The

274
00:15:30,400 --> 00:15:33,080
Speaker 2: other element that's at play here, though, is because of

275
00:15:33,080 --> 00:15:36,480
Speaker 2: how algorithmic boosting works, how many of those people who

276
00:15:36,560 --> 00:15:39,920
Speaker 2: were real people posting outrage would have even seen this

277
00:15:40,720 --> 00:15:44,560
Speaker 2: to post their outrage had it not first been amplified

278
00:15:44,920 --> 00:15:47,880
Speaker 2: by inorganic accounts that brought it to the top of

279
00:15:47,960 --> 00:15:52,600
Speaker 2: their feed. So even within that number of people with big,

280
00:15:52,640 --> 00:15:57,000
Speaker 2: real human feelings, there is an element of being shaped

281
00:15:57,040 --> 00:16:00,920
Speaker 2: by this inorganic discourse bringing that conversation to them in

282
00:16:00,920 --> 00:16:01,560
Speaker 2: the first place.

283
00:16:01,880 --> 00:16:04,280
Speaker 1: I saw an analogy to this that if you were

284
00:16:04,320 --> 00:16:06,640
Speaker 1: cutting vegetables in your kitchen and then you cut your finger,

285
00:16:06,960 --> 00:16:09,160
Speaker 1: and you ask somebody next to you hand me that towels,

286
00:16:09,200 --> 00:16:11,840
Speaker 1: why could apply pressure if they squeezed it instead so

287
00:16:11,880 --> 00:16:14,480
Speaker 1: that you bled faster. They didn't start the cut, but

288
00:16:14,520 --> 00:16:17,560
Speaker 1: they certainly made it worse. They certainly brought the problem,

289
00:16:17,920 --> 00:16:19,800
Speaker 1: you know, to a different level. And that's sort of

290
00:16:19,840 --> 00:16:22,640
Speaker 1: a good way to think about how bots and inauthentic

291
00:16:22,680 --> 00:16:26,200
Speaker 1: activity can shape a conversation online that might actually be

292
00:16:26,280 --> 00:16:29,120
Speaker 1: seated with organic people and their big feelings they're sharing

293
00:16:29,160 --> 00:16:29,800
Speaker 1: on the internet.

294
00:16:30,240 --> 00:16:31,920
Speaker 2: Now, I'm just thinking in my head of like what

295
00:16:32,080 --> 00:16:34,080
Speaker 2: all of the potential bots would say if you ask

296
00:16:34,160 --> 00:16:37,040
Speaker 2: them to hand you a towel. So I find that

297
00:16:37,120 --> 00:16:40,040
Speaker 2: you can, like you can spot them in the commons

298
00:16:40,080 --> 00:16:43,200
Speaker 2: section of like they'll they'll take things too literally sometimes,

299
00:16:43,960 --> 00:16:46,160
Speaker 2: so it would be, you know, like something about the

300
00:16:46,480 --> 00:16:48,440
Speaker 2: something about the towel, something about like not being able

301
00:16:48,440 --> 00:16:51,560
Speaker 2: to squeeze your finger. The I think there was a

302
00:16:51,640 --> 00:16:54,240
Speaker 2: key point in time where we saw some of the

303
00:16:54,240 --> 00:16:57,120
Speaker 2: bot networks, like short Circuit with their instructions about what

304
00:16:57,160 --> 00:17:01,160
Speaker 2: to post at the peak of the Trump Epstein files controversy,

305
00:17:01,160 --> 00:17:03,960
Speaker 2: where in real time, I'm not the only person who

306
00:17:04,000 --> 00:17:06,240
Speaker 2: saw this. I don't know if there's any anything written

307
00:17:06,280 --> 00:17:08,760
Speaker 2: about this, but other people were observing this that like

308
00:17:09,440 --> 00:17:14,199
Speaker 2: a lot of these seemingly maga accounts were turning on

309
00:17:14,520 --> 00:17:17,920
Speaker 2: Trump in the context of like calling for the Epstein

310
00:17:17,920 --> 00:17:20,240
Speaker 2: files to be released, because you can tell that they're

311
00:17:20,240 --> 00:17:24,160
Speaker 2: like pre programmed instructions were to like non stop call

312
00:17:24,240 --> 00:17:26,560
Speaker 2: for the Epstein files to be released, and they short

313
00:17:26,600 --> 00:17:29,480
Speaker 2: circuited a little bit based on their previous instructions for

314
00:17:29,520 --> 00:17:32,600
Speaker 2: the world. Before you know, they're they're they got updated

315
00:17:32,680 --> 00:17:34,840
Speaker 2: to the version of the world that existed at that

316
00:17:34,880 --> 00:17:35,400
Speaker 2: point in time.

317
00:17:36,119 --> 00:17:38,160
Speaker 1: It's part out there for a bot. I don't find

318
00:17:38,240 --> 00:17:41,520
Speaker 1: the boss. I'm not a bue. I get confused. It's

319
00:17:41,600 --> 00:17:43,920
Speaker 1: I feel bad that these bots that are like, I

320
00:17:43,960 --> 00:17:45,240
Speaker 1: don't even know what the post anymore.

321
00:17:45,920 --> 00:17:48,720
Speaker 2: Yeah, yeah, I think if people are wondering, you know,

322
00:17:48,760 --> 00:17:50,679
Speaker 2: like what is what is a good way to confirm

323
00:17:50,680 --> 00:17:52,600
Speaker 2: your spidy sense of something in surreal I mean, I

324
00:17:52,600 --> 00:17:56,080
Speaker 2: think we were all used to like those account handles

325
00:17:56,119 --> 00:17:59,280
Speaker 2: that are like John two four, five, six, oh whatever, right,

326
00:17:59,359 --> 00:18:01,480
Speaker 2: Like we know that's a that's a common way that

327
00:18:01,560 --> 00:18:02,879
Speaker 2: those account.

328
00:18:02,560 --> 00:18:04,919
Speaker 3: Names are formulated. We do still see that.

329
00:18:05,440 --> 00:18:08,719
Speaker 2: But I think even without any other technology to flag

330
00:18:08,760 --> 00:18:11,159
Speaker 2: automated behavior, what you can do is if you're like

331
00:18:11,240 --> 00:18:14,119
Speaker 2: on your phone and just scroll a couple of times.

332
00:18:14,520 --> 00:18:17,840
Speaker 2: If you're still scrolling and you have not reached yesterday

333
00:18:18,000 --> 00:18:21,560
Speaker 2: in that person's posting, you know, that's probably a sign

334
00:18:21,720 --> 00:18:24,439
Speaker 2: that they are posting at a frequency that is just

335
00:18:24,600 --> 00:18:29,160
Speaker 2: not probably humanly possible and is probably you know, an

336
00:18:29,200 --> 00:18:32,640
Speaker 2: automated posting rate that's that's going out there. So that's

337
00:18:32,680 --> 00:18:34,359
Speaker 2: one of my tip of one of the simplest ways

338
00:18:34,359 --> 00:18:38,040
Speaker 2: to be like, eh, is it a bot? Yeah, just

339
00:18:38,160 --> 00:18:39,879
Speaker 2: scroll back and see how far it takes you to

340
00:18:39,880 --> 00:18:40,640
Speaker 2: get to yesterday.

341
00:18:41,000 --> 00:18:44,760
Speaker 1: That is a great tip. And yeah, don't spend don't

342
00:18:44,760 --> 00:18:47,840
Speaker 1: dedicate your time and energy to getting into a back

343
00:18:47,840 --> 00:18:50,879
Speaker 1: and forth with a bot or like worrying yourself with

344
00:18:50,920 --> 00:18:52,720
Speaker 1: what a bot is saying, unless you're doing it because

345
00:18:52,720 --> 00:18:55,000
Speaker 1: you're a researcher and you're interested. Don't have it like

346
00:18:55,080 --> 00:18:57,840
Speaker 1: fuck up your day what a bot left in your

347
00:18:57,880 --> 00:18:59,840
Speaker 1: comments on Instagram or something.

348
00:19:00,240 --> 00:19:01,000
Speaker 3: Yeah, exactly.

349
00:19:01,040 --> 00:19:02,639
Speaker 2: I mean I think we're talking about X, and I

350
00:19:02,680 --> 00:19:06,159
Speaker 2: do think that X is pretty central to like the

351
00:19:06,240 --> 00:19:09,239
Speaker 2: discourse in the sense that it really is like the

352
00:19:09,240 --> 00:19:12,600
Speaker 2: most fertile ground for bot behavior. Like it's very text based,

353
00:19:12,640 --> 00:19:17,040
Speaker 2: which means it's like easier to you know, produce content there,

354
00:19:17,080 --> 00:19:20,000
Speaker 2: like it's a little bit harder to automate like posting

355
00:19:20,000 --> 00:19:22,880
Speaker 2: of pictures if you think about it, like just logistically,

356
00:19:23,960 --> 00:19:27,159
Speaker 2: So I think we do still see a lot of

357
00:19:27,200 --> 00:19:30,080
Speaker 2: bought activity on X, where I think researchers are less

358
00:19:30,080 --> 00:19:32,919
Speaker 2: familiar with what bot activity looks like on other platforms.

359
00:19:33,880 --> 00:19:38,080
Speaker 2: I will say that, you know, I'm seeing a lot

360
00:19:38,080 --> 00:19:43,040
Speaker 2: of folks going to more niche communities like Facebook groups

361
00:19:43,359 --> 00:19:47,000
Speaker 2: or discord or Reddit, where I think that there's a

362
00:19:47,080 --> 00:19:49,880
Speaker 2: desire for people to like be in conversation with real

363
00:19:49,920 --> 00:19:53,359
Speaker 2: people on the Internet, and when they're finding that that's

364
00:19:53,400 --> 00:19:57,080
Speaker 2: not possible in certain forums anymore, they're they're moving to

365
00:19:57,119 --> 00:20:00,840
Speaker 2: different places to try to be assured that when they're

366
00:20:01,000 --> 00:20:03,840
Speaker 2: engaging in conversation with someone that yeah, it's not like

367
00:20:03,880 --> 00:20:06,560
Speaker 2: a bot who's arguing back at them.

368
00:20:06,400 --> 00:20:08,720
Speaker 1: Especially in this age of AI, I want to know

369
00:20:08,840 --> 00:20:12,520
Speaker 1: even if we're having a spicy conversation a record argument,

370
00:20:12,600 --> 00:20:14,320
Speaker 1: I want it to be with a real person, like

371
00:20:14,359 --> 00:20:16,119
Speaker 1: in des Moines, iware or something. I don't want to

372
00:20:16,119 --> 00:20:18,159
Speaker 1: have the feeling of like what am I wasting my

373
00:20:18,200 --> 00:20:20,240
Speaker 1: time going back and forth on Reddit with a bot? No,

374
00:20:20,320 --> 00:20:25,400
Speaker 1: thank you, Let's take.

375
00:20:25,280 --> 00:20:37,200
Speaker 4: A quick break at our back.

376
00:20:39,720 --> 00:20:42,840
Speaker 1: Since Molly is an Internet Vibe checker, I wanted to

377
00:20:42,920 --> 00:20:45,720
Speaker 1: know her thoughts on the report around the controversy surrounding

378
00:20:45,720 --> 00:20:49,400
Speaker 1: Taylor Swift, things like accusations that Swift is a secret

379
00:20:49,440 --> 00:20:54,359
Speaker 1: trad wife or even secretly a Nazi. So I want

380
00:20:54,359 --> 00:20:57,120
Speaker 1: to talk about the report that I'm sure you've seen

381
00:20:57,160 --> 00:20:59,920
Speaker 1: by now that was put out by a company called Gooday,

382
00:21:00,680 --> 00:21:03,040
Speaker 1: that was then reported about in Rolling Stone that looked

383
00:21:03,040 --> 00:21:06,719
Speaker 1: into social media commentary around Taylor Swift's latest album and

384
00:21:06,840 --> 00:21:09,840
Speaker 1: whether or not that commentary was in part driven by

385
00:21:09,880 --> 00:21:13,879
Speaker 1: inauthentic coordinated accounts. What was your reaction, just as somebody

386
00:21:13,920 --> 00:21:16,359
Speaker 1: who puts together reports like this, somebody who was in

387
00:21:16,440 --> 00:21:19,240
Speaker 1: this space, what was your reaction to that report?

388
00:21:19,640 --> 00:21:22,520
Speaker 2: I mean, I think it was a spidy sentence check.

389
00:21:23,280 --> 00:21:26,160
Speaker 2: I am not a swifty, please don't come after me swifties.

390
00:21:26,200 --> 00:21:30,720
Speaker 2: My wife is a swifty. But I, as a person

391
00:21:30,760 --> 00:21:34,320
Speaker 2: on the internet, had had seen the reactions to Taylor

392
00:21:34,359 --> 00:21:37,240
Speaker 2: Swift's album Now. At the time, I was not discriminating

393
00:21:37,240 --> 00:21:40,000
Speaker 2: whether I thought that those reactions were organic or inorganic,

394
00:21:40,960 --> 00:21:43,880
Speaker 2: but I was aware of the pushback to the most

395
00:21:43,920 --> 00:21:49,240
Speaker 2: recent album, and looking through the methodology, it makes sense

396
00:21:49,280 --> 00:21:51,520
Speaker 2: to me. I think it sound in certain aspects of it,

397
00:21:51,560 --> 00:21:55,119
Speaker 2: and I think maybe where sometimes the nuance gets lost

398
00:21:55,440 --> 00:21:58,399
Speaker 2: is you know, we talk about how one part of

399
00:21:58,400 --> 00:22:04,399
Speaker 2: a conversation is automated or manipulated. And that's not to

400
00:22:04,480 --> 00:22:08,640
Speaker 2: discount the other part of the conversation that has real

401
00:22:08,680 --> 00:22:12,640
Speaker 2: people having real, big human feelings here. And inherently, in

402
00:22:13,040 --> 00:22:16,800
Speaker 2: whether it's data sampling or you know, the way that

403
00:22:16,840 --> 00:22:19,840
Speaker 2: your technology, whether it's by keywords or something else, is

404
00:22:20,040 --> 00:22:24,359
Speaker 2: identifying the parts of the discourse, you're going to more

405
00:22:24,520 --> 00:22:27,720
Speaker 2: easily identify the parts of the discourse that like sound

406
00:22:27,800 --> 00:22:30,920
Speaker 2: automated to begin with, because they're flagging like the right

407
00:22:31,000 --> 00:22:33,720
Speaker 2: keywords or the right markers for what's going on. And

408
00:22:33,800 --> 00:22:39,040
Speaker 2: inherently you're going to miss the more nuanced conversations because people,

409
00:22:39,359 --> 00:22:42,720
Speaker 2: when real people talk about issues, they're all using slightly

410
00:22:42,720 --> 00:22:45,840
Speaker 2: different language to describe what's going on. They're maybe not referencing,

411
00:22:46,240 --> 00:22:49,800
Speaker 2: like Taylor Swift's full name in the conversation they're replying.

412
00:22:49,880 --> 00:22:53,439
Speaker 2: So inherently, in the work that we do, based on

413
00:22:53,480 --> 00:22:56,399
Speaker 2: how you're going to collect this data, it's already going

414
00:22:56,400 --> 00:22:58,280
Speaker 2: to veer a little bit more towards missing some of

415
00:22:58,320 --> 00:23:01,800
Speaker 2: those more organic conversations because of how humans talk about things.

416
00:23:02,400 --> 00:23:07,119
Speaker 2: So something got lost down the road. I in terms

417
00:23:07,119 --> 00:23:09,800
Speaker 2: of the real people having real human feelings about this,

418
00:23:10,600 --> 00:23:13,440
Speaker 2: I don't disagree based on the methodology that there are

419
00:23:13,480 --> 00:23:18,720
Speaker 2: certainly indicators of automated activity here, but we don't necessarily know.

420
00:23:18,880 --> 00:23:22,800
Speaker 2: Going back to the earlier question, you know these bot

421
00:23:22,840 --> 00:23:26,440
Speaker 2: networks that were you know, jumping on the bandwagon here

422
00:23:26,480 --> 00:23:30,280
Speaker 2: that also we're attacking Blake Lively. You know, we don't

423
00:23:30,320 --> 00:23:33,199
Speaker 2: necessarily know what the incentives are of these butt networks.

424
00:23:33,200 --> 00:23:37,720
Speaker 2: Are they just hopping on the latest celebrity trend to

425
00:23:38,359 --> 00:23:42,440
Speaker 2: gain traction and monetization, or are they targeted against certain

426
00:23:42,480 --> 00:23:45,639
Speaker 2: celebrities like these are unknown questions at this point, so

427
00:23:45,680 --> 00:23:47,840
Speaker 2: I think we want to make sure to not jump

428
00:23:47,960 --> 00:23:52,560
Speaker 2: too many steps forward to claim that billionaire Taylor Swift

429
00:23:52,640 --> 00:23:54,760
Speaker 2: is being targeted and we need to protect her from

430
00:23:54,800 --> 00:23:57,280
Speaker 2: attacks online. There are a lot of other people who

431
00:23:57,320 --> 00:24:01,159
Speaker 2: are being targeted who don't necessarily have of Taylor Swift's

432
00:24:01,160 --> 00:24:03,600
Speaker 2: pr arsenal at their back.

433
00:24:04,320 --> 00:24:06,760
Speaker 1: Yes, I'm so glad that you brought that up, because

434
00:24:07,119 --> 00:24:09,120
Speaker 1: I mean, I don't know if you've I mean, I'm

435
00:24:09,119 --> 00:24:11,919
Speaker 1: sure you've seen the way that conversation around this report

436
00:24:11,960 --> 00:24:15,720
Speaker 1: has spread online. And one of the things that I've

437
00:24:15,760 --> 00:24:19,800
Speaker 1: been a little frustrated to see is people saying, oh, well,

438
00:24:19,840 --> 00:24:23,359
Speaker 1: this report confirms that everybody who was critical of Taylor

439
00:24:23,440 --> 00:24:25,760
Speaker 1: Swift was just a bot? They can they can or

440
00:24:25,840 --> 00:24:28,919
Speaker 1: people who were you know, manipulated by bots, all of

441
00:24:28,960 --> 00:24:31,439
Speaker 1: that discourse can be dismissed. So, first of all, the

442
00:24:31,520 --> 00:24:34,640
Speaker 1: report does not say that, not even a little. And

443
00:24:34,680 --> 00:24:36,560
Speaker 1: in fact, something that I think is getting lost in

444
00:24:36,600 --> 00:24:38,600
Speaker 1: the sauce of the report is that, in fact, the

445
00:24:38,640 --> 00:24:44,200
Speaker 1: report is talking about authentic accounts having authentic discourse alongside

446
00:24:44,359 --> 00:24:46,400
Speaker 1: what they have what they have seen as like perhaps

447
00:24:46,400 --> 00:24:49,239
Speaker 1: inauthentic discourse. So like, nowhere in that report are they

448
00:24:49,280 --> 00:24:51,919
Speaker 1: saying everybody who was critical of Taylor Swift can just

449
00:24:51,960 --> 00:24:54,240
Speaker 1: be dismissed as a bot. And I think that there's

450
00:24:54,240 --> 00:24:57,960
Speaker 1: something about the way it was reported that is not

451
00:24:58,400 --> 00:25:02,679
Speaker 1: giving enough credence to the that were there were and

452
00:25:02,880 --> 00:25:06,600
Speaker 1: are real people in this conversation who are not bots,

453
00:25:06,600 --> 00:25:08,840
Speaker 1: who have been talking about Taylor Swift critically for a

454
00:25:08,880 --> 00:25:12,560
Speaker 1: long time. And so even though you know, I'm the

455
00:25:12,680 --> 00:25:15,520
Speaker 1: data scientist, but looking at their methodology, I don't think

456
00:25:15,560 --> 00:25:17,280
Speaker 1: I don't think they've made this up. I don't think

457
00:25:17,280 --> 00:25:20,000
Speaker 1: Taylor's Swift paid them to put this reporting together or

458
00:25:20,040 --> 00:25:22,800
Speaker 1: something like that. But I understand what people are saying

459
00:25:22,800 --> 00:25:26,280
Speaker 1: when they're saying this report does not actually reflect the

460
00:25:26,280 --> 00:25:30,440
Speaker 1: fact that there are so many people authentically being critical

461
00:25:30,480 --> 00:25:33,000
Speaker 1: of Taylor Swift on the internet. We're not bots. I

462
00:25:33,040 --> 00:25:35,560
Speaker 1: think something about the way the report was framed sort

463
00:25:35,600 --> 00:25:37,840
Speaker 1: of gave credence to this idea that you could just

464
00:25:37,880 --> 00:25:40,920
Speaker 1: be dismissive of all of these critical voices exactly.

465
00:25:41,000 --> 00:25:43,160
Speaker 2: I mean, I think on the flip side, we could

466
00:25:43,200 --> 00:25:47,360
Speaker 2: look at any other recent controversial issue and perhaps come

467
00:25:47,440 --> 00:25:51,800
Speaker 2: to a similar conclusion of, Wow, this conversation is bot driven,

468
00:25:52,560 --> 00:25:54,840
Speaker 2: because I think that that may just be a symptom

469
00:25:54,880 --> 00:25:58,359
Speaker 2: of like how the Internet functions nowadays, that may not

470
00:25:58,400 --> 00:26:03,119
Speaker 2: necessarily be specific to the dynamics of Taylor Swift. I

471
00:26:03,160 --> 00:26:05,440
Speaker 2: think the other thing that you can do, and I

472
00:26:05,480 --> 00:26:08,119
Speaker 2: mean I'm taking lessons from this, you know, working and

473
00:26:08,240 --> 00:26:11,320
Speaker 2: researching in this space, is you know, one of the

474
00:26:11,520 --> 00:26:13,399
Speaker 2: one of my favorite ways to set folks up to

475
00:26:14,000 --> 00:26:18,240
Speaker 2: analyze conversations online with Pea metrics technology is all say,

476
00:26:18,400 --> 00:26:21,160
Speaker 2: you know, look at the data and ask the same

477
00:26:21,280 --> 00:26:25,560
Speaker 2: question filtered to the organic activity and the non organic activity,

478
00:26:25,880 --> 00:26:28,040
Speaker 2: so that you can see the nuance and the difference

479
00:26:28,080 --> 00:26:30,760
Speaker 2: and maybe like the specific themes or narratives that they're

480
00:26:30,800 --> 00:26:33,960
Speaker 2: talking about. So an interesting question to ask of this

481
00:26:34,040 --> 00:26:40,960
Speaker 2: data might have been within the controversial conversation around the

482
00:26:41,040 --> 00:26:46,640
Speaker 2: latest album, which one specifically were the bots trying to push?

483
00:26:46,680 --> 00:26:49,440
Speaker 2: And then when it comes to real human people, which

484
00:26:49,480 --> 00:26:53,560
Speaker 2: aspects were they focused on presenting. I think maybe those

485
00:26:53,600 --> 00:26:57,240
Speaker 2: two things side by side gives us better context for

486
00:26:57,320 --> 00:26:59,919
Speaker 2: what those people who were having real human feelings about this.

487
00:27:00,000 --> 00:27:03,159
Speaker 2: We're thinking, which aspects of the album controversy were they

488
00:27:03,240 --> 00:27:05,960
Speaker 2: latching onto the most, and how did that perhaps differ

489
00:27:06,240 --> 00:27:07,720
Speaker 2: from what the boss were focused on.

490
00:27:08,200 --> 00:27:10,879
Speaker 1: And I think there's also I mean, I can't not

491
00:27:11,080 --> 00:27:13,720
Speaker 1: talk about the way that this seems too the conversation

492
00:27:13,760 --> 00:27:17,320
Speaker 1: seems to be happening along some clear racial lines to me,

493
00:27:17,560 --> 00:27:20,760
Speaker 1: where a lot of the voices who were critical of

494
00:27:20,800 --> 00:27:23,639
Speaker 1: Taylor Swift, not all, but a lot were women of color,

495
00:27:23,640 --> 00:27:25,960
Speaker 1: black women, people of color, and a lot of the

496
00:27:26,000 --> 00:27:29,280
Speaker 1: Swifty community appeared to be again not all, but a

497
00:27:29,280 --> 00:27:31,920
Speaker 1: lot of white ladies, white people. And so I think

498
00:27:31,960 --> 00:27:34,960
Speaker 1: it's just one of those issues that will always sit

499
00:27:35,000 --> 00:27:38,679
Speaker 1: at these tension points of the tensions that we know

500
00:27:38,840 --> 00:27:44,440
Speaker 1: exist in our society. Something about the report giving credence

501
00:27:44,520 --> 00:27:48,000
Speaker 1: to the idea that you could just discount a largely

502
00:27:48,880 --> 00:27:51,960
Speaker 1: like the voice the critical voices of largely minoritized people

503
00:27:52,960 --> 00:27:54,720
Speaker 1: because you love Taylor Swift and you don't have to

504
00:27:54,720 --> 00:27:57,000
Speaker 1: think critically about the points these people are making because

505
00:27:57,040 --> 00:28:00,119
Speaker 1: they're bots. I can see why that this hit people sideway.

506
00:28:01,000 --> 00:28:05,440
Speaker 2: Yeah, And I think to keep the organic voices anonymized here,

507
00:28:05,600 --> 00:28:07,560
Speaker 2: I think that might have been a great opportunity to

508
00:28:07,640 --> 00:28:10,119
Speaker 2: look at the people on the organic side of the

509
00:28:10,119 --> 00:28:12,720
Speaker 2: conversation and see, like, who were the biggest influencers in

510
00:28:12,760 --> 00:28:15,439
Speaker 2: this space, so within people who were criticizing the album

511
00:28:15,640 --> 00:28:20,040
Speaker 2: or the aesthetics of Taylor Swift and they were organic,

512
00:28:20,119 --> 00:28:21,760
Speaker 2: you know, who were some of the biggest accounts that

513
00:28:21,800 --> 00:28:23,919
Speaker 2: weighed in, Like there are you know, I'm aware that

514
00:28:24,000 --> 00:28:26,720
Speaker 2: like black Twitter is a thing, like who were some

515
00:28:26,800 --> 00:28:29,119
Speaker 2: of the biggest voices that were that were shaping the

516
00:28:29,160 --> 00:28:32,880
Speaker 2: conversation from the human side. What we may uncover if

517
00:28:32,880 --> 00:28:34,880
Speaker 2: we looked at like, you know, there might have been

518
00:28:35,000 --> 00:28:38,840
Speaker 2: a post from someone who has like five hundred thousand followers,

519
00:28:39,040 --> 00:28:42,400
Speaker 2: and arguably that might have shaped the discourse a lot more.

520
00:28:42,480 --> 00:28:45,239
Speaker 2: That single post from that person with a lot of

521
00:28:45,280 --> 00:28:49,120
Speaker 2: reach might have shaped the discourse more than the you know,

522
00:28:49,800 --> 00:28:52,920
Speaker 2: fifty account fifty posts from bot accounts that have like

523
00:28:53,240 --> 00:28:57,240
Speaker 2: one hundred followers. So maybe there's some nuance to be

524
00:28:57,440 --> 00:29:01,920
Speaker 2: parsed out here too. That volume doesn't necessarily equal impact

525
00:29:02,040 --> 00:29:05,520
Speaker 2: or influence on the Internet, and what bots are inherently

526
00:29:05,600 --> 00:29:08,960
Speaker 2: trying to do is boost the volume of the conversation.

527
00:29:10,160 --> 00:29:13,200
Speaker 2: But where humans have an impact, And there are also

528
00:29:13,280 --> 00:29:16,080
Speaker 2: like automated accounts that look like influencer accounts that have

529
00:29:16,120 --> 00:29:18,400
Speaker 2: lots of followers. But I'd like to pull the threat

530
00:29:18,400 --> 00:29:20,400
Speaker 2: a little bit more about who are the influential people

531
00:29:20,400 --> 00:29:21,800
Speaker 2: on the human side of the conversation.

532
00:29:22,400 --> 00:29:24,200
Speaker 1: That makes a lot of sense, and I guess I

533
00:29:24,240 --> 00:29:27,240
Speaker 1: would have liked I also think that listen, we have

534
00:29:27,320 --> 00:29:28,760
Speaker 1: said this put it on the show a million times.

535
00:29:28,800 --> 00:29:31,000
Speaker 1: When you talk about something, there's something about Taylor Swift

536
00:29:31,000 --> 00:29:34,080
Speaker 1: that is like, once you start talking about her, big

537
00:29:34,080 --> 00:29:37,520
Speaker 1: feelings come out on all sides, even people who don't

538
00:29:37,520 --> 00:29:39,800
Speaker 1: like Kidler. Swift is like, there's just something about her

539
00:29:39,800 --> 00:29:41,479
Speaker 1: that gets people talking. And so I think if you're

540
00:29:41,520 --> 00:29:44,080
Speaker 1: going to be putting out a report that is about

541
00:29:44,160 --> 00:29:46,880
Speaker 1: Taylor Swift, it would behove you to make some of

542
00:29:46,920 --> 00:29:49,000
Speaker 1: this clear, right, It would behove you to spend a

543
00:29:49,040 --> 00:29:52,600
Speaker 1: little time explaining like, oh, well, who were the voices

544
00:29:53,760 --> 00:29:56,320
Speaker 1: that we're seeing on the like, who are the authentic

545
00:29:56,400 --> 00:29:58,400
Speaker 1: voices that we're seeing that are talking about her in

546
00:29:58,400 --> 00:30:00,720
Speaker 1: this way and weighing in this way? Because you have

547
00:30:00,720 --> 00:30:02,520
Speaker 1: to imagine it's it's gonna get a lot of eyes

548
00:30:02,560 --> 00:30:04,320
Speaker 1: on it. By now, everybody knows that when you talk

549
00:30:04,320 --> 00:30:06,160
Speaker 1: about Taylor Swift, it's something that gets a lot of

550
00:30:06,200 --> 00:30:08,680
Speaker 1: eyes or a lot of ears or whatever medium you're using.

551
00:30:09,240 --> 00:30:12,880
Speaker 2: Yeah, if I maybe have three guideposts that I'm making

552
00:30:12,960 --> 00:30:16,560
Speaker 2: up on the fly as as an Internet person having

553
00:30:16,600 --> 00:30:19,600
Speaker 2: been on you know, Tumbler back in like the early

554
00:30:19,640 --> 00:30:23,640
Speaker 2: twenty tens, you knew not to mess with the swifties, right,

555
00:30:23,760 --> 00:30:27,160
Speaker 2: Like that's so the guideposts of the Internet is, like,

556
00:30:27,200 --> 00:30:29,800
Speaker 2: what do we established, like, assume it's a grift?

557
00:30:30,840 --> 00:30:33,880
Speaker 3: What is the other one from that Netflix documentary.

558
00:30:33,560 --> 00:30:37,120
Speaker 2: Don't fuck with don't don't fuck with cats? And I

559
00:30:37,160 --> 00:30:40,600
Speaker 2: think maybe the third guidepost of the Internet is do

560
00:30:40,720 --> 00:30:43,320
Speaker 2: not touch the tailor swift discourse with a ten foot pole.

561
00:30:45,480 --> 00:30:47,200
Speaker 2: If you follow these things, you can't go wrong on

562
00:30:47,240 --> 00:30:47,720
Speaker 2: the Internet.

563
00:30:48,080 --> 00:30:50,000
Speaker 1: What do you think about the methodology that they used

564
00:30:50,040 --> 00:30:52,040
Speaker 1: in this report to define a bot accounts.

565
00:30:52,480 --> 00:30:55,200
Speaker 2: I think it's challenging that we're using a common word

566
00:30:55,400 --> 00:30:58,120
Speaker 2: that you could define in a lot of different ways.

567
00:30:58,800 --> 00:31:03,760
Speaker 2: I think that there's no you know, set methodology. I

568
00:31:03,760 --> 00:31:07,320
Speaker 2: think a lot of different tech companies have different definitions

569
00:31:07,360 --> 00:31:09,560
Speaker 2: and have their own sort of secret sauce when it

570
00:31:09,600 --> 00:31:13,959
Speaker 2: comes to you know, what available data they're using to

571
00:31:14,000 --> 00:31:18,000
Speaker 2: come to these conclusions. I can tell you that, you

572
00:31:18,040 --> 00:31:20,400
Speaker 2: know from the way that we think about it, coming

573
00:31:20,440 --> 00:31:24,040
Speaker 2: from sort of a framework of you know, degrees of

574
00:31:24,040 --> 00:31:26,720
Speaker 2: confidence that you would see in like the intelligence community,

575
00:31:26,800 --> 00:31:28,360
Speaker 2: right like you're never going to say I'm one hundred

576
00:31:28,400 --> 00:31:30,920
Speaker 2: percent sure that you know something is something. You're going

577
00:31:31,000 --> 00:31:33,560
Speaker 2: to say, okay, well, the available data that I has

578
00:31:34,200 --> 00:31:36,840
Speaker 2: that I have shows, you know, it's highly likely that

579
00:31:36,960 --> 00:31:40,320
Speaker 2: this is a bot account. Even when you're getting into

580
00:31:40,880 --> 00:31:44,080
Speaker 2: the highest degree of confidence that something is a bought account,

581
00:31:44,600 --> 00:31:47,000
Speaker 2: the way that we frame it is that it's almost

582
00:31:47,120 --> 00:31:50,520
Speaker 2: certainly a bought account, which is still you know, by

583
00:31:50,840 --> 00:31:54,600
Speaker 2: verbiage a little bit short of saying you know for sure,

584
00:31:54,680 --> 00:31:56,760
Speaker 2: because I think the only way to know for sure

585
00:31:56,800 --> 00:31:59,480
Speaker 2: that something is a bought account is to do a

586
00:31:59,520 --> 00:32:03,080
Speaker 2: lot of honestly like manual forensic analysis of that account,

587
00:32:03,320 --> 00:32:05,560
Speaker 2: and the issue is that you can't do that at scale,

588
00:32:05,680 --> 00:32:08,480
Speaker 2: so when you're looking at large sets of data and

589
00:32:08,520 --> 00:32:11,440
Speaker 2: you're looking at, you know, a limited availability of metrics

590
00:32:11,480 --> 00:32:14,360
Speaker 2: that you have on these accounts. Honestly, I don't think

591
00:32:14,360 --> 00:32:18,680
Speaker 2: it's necessarily the way that you know, the bots are classified.

592
00:32:18,720 --> 00:32:21,640
Speaker 2: It's the language that you're using to describe the confidence

593
00:32:21,720 --> 00:32:24,600
Speaker 2: that these are bots. And again, I think there's something

594
00:32:24,640 --> 00:32:27,719
Speaker 2: that's getting lost in translation between like a methodological report,

595
00:32:28,160 --> 00:32:31,240
Speaker 2: journalistic reporting, and then where that journalistic reporting goes in

596
00:32:31,280 --> 00:32:34,959
Speaker 2: the discourse. So we have you know, things like account history,

597
00:32:35,000 --> 00:32:38,400
Speaker 2: like when was the account created, We have the posting frequency.

598
00:32:38,480 --> 00:32:42,080
Speaker 2: You have like the the you know profile image, right,

599
00:32:42,160 --> 00:32:44,920
Speaker 2: is that like a recycled image or an AI generated image?

600
00:32:45,240 --> 00:32:48,480
Speaker 2: You know, we talked about like the obvious indicator of

601
00:32:48,600 --> 00:32:51,360
Speaker 2: the way that the user name is formulated. Are there

602
00:32:51,440 --> 00:32:55,040
Speaker 2: other usernames like that that appear on other social media sites?

603
00:32:56,680 --> 00:32:59,120
Speaker 2: What type of content is this account posting? Is it

604
00:32:59,280 --> 00:33:04,400
Speaker 2: mainly act as an amplifier and just resharing reposting, or

605
00:33:04,680 --> 00:33:07,320
Speaker 2: is it doing a lot of original posts, which is

606
00:33:07,400 --> 00:33:12,560
Speaker 2: just like mechanically and also computationally from like a energy perspective,

607
00:33:12,560 --> 00:33:14,440
Speaker 2: going to be harder for like a bought run account

608
00:33:14,480 --> 00:33:16,680
Speaker 2: to maintain posting original content.

609
00:33:17,040 --> 00:33:18,120
Speaker 3: So I throw out.

610
00:33:18,000 --> 00:33:19,920
Speaker 2: All of these metrics to say that I think that

611
00:33:20,080 --> 00:33:23,680
Speaker 2: the answer lies in probably a combination of all of

612
00:33:23,720 --> 00:33:26,280
Speaker 2: them to get to the best answer. But I think

613
00:33:26,320 --> 00:33:29,280
Speaker 2: we need to be really careful about the language that

614
00:33:29,320 --> 00:33:32,479
Speaker 2: we're using when we say, like the confidence level that

615
00:33:32,520 --> 00:33:35,560
Speaker 2: we have that an account is automated.

616
00:33:35,960 --> 00:33:39,800
Speaker 1: One of the criticisms, and I mean, you've been, I

617
00:33:39,800 --> 00:33:42,959
Speaker 1: think helpfully kind of critical, good critical of the report.

618
00:33:43,400 --> 00:33:45,160
Speaker 1: But some of the criticisms I'm seeing to people sort

619
00:33:45,200 --> 00:33:47,040
Speaker 1: of just trying to trash this report and say that

620
00:33:47,080 --> 00:33:50,200
Speaker 1: it's a lie. Is that on their site they have

621
00:33:50,440 --> 00:33:55,080
Speaker 1: a disclaimer that's like, oh, we cannot we're not saying

622
00:33:55,080 --> 00:33:57,240
Speaker 1: that we're not guaranteeing that what we say in these

623
00:33:57,240 --> 00:34:00,200
Speaker 1: reports is true. And I thought to myself, well, this

624
00:34:00,240 --> 00:34:02,920
Speaker 1: seems like standard cover your ass language to me that like,

625
00:34:02,960 --> 00:34:06,760
Speaker 1: of course, I'm not going to say with certainty like

626
00:34:06,800 --> 00:34:08,399
Speaker 1: what we're saying you could take it to the bank,

627
00:34:08,440 --> 00:34:11,120
Speaker 1: take it to your grave, one hundred percent true. I

628
00:34:11,160 --> 00:34:14,680
Speaker 1: didn't feel like that was a necessarily a fair criticism

629
00:34:14,719 --> 00:34:17,160
Speaker 1: of saying this report is full of lies it's bunk.

630
00:34:17,719 --> 00:34:21,920
Speaker 2: Yeah, I mean, I think when you're stepping foot into

631
00:34:22,360 --> 00:34:24,799
Speaker 2: measuring the Internet, I think the issue is just that, like,

632
00:34:25,600 --> 00:34:28,280
Speaker 2: what I'm finding is that media outlets are really hungry

633
00:34:28,600 --> 00:34:33,440
Speaker 2: for any data that quantifies what's going on. Don't necessarily

634
00:34:33,480 --> 00:34:37,920
Speaker 2: have folks in house who can fact check or verify

635
00:34:38,000 --> 00:34:42,040
Speaker 2: that data. So you're really being you know, it's a

636
00:34:42,080 --> 00:34:45,919
Speaker 2: pretty big responsibility to be the person who is doing

637
00:34:45,960 --> 00:34:49,480
Speaker 2: the research and the person who is presenting that research

638
00:34:50,160 --> 00:34:53,320
Speaker 2: to an organization that you know doesn't necessarily have the

639
00:34:53,440 --> 00:34:58,240
Speaker 2: same tools at their disposal to interrogate it. I feel

640
00:34:58,239 --> 00:35:01,360
Speaker 2: that responsibility someone working in the space, which is why

641
00:35:01,480 --> 00:35:03,840
Speaker 2: I will not touch the Taylor Swift discourse with the

642
00:35:03,840 --> 00:35:06,359
Speaker 2: ten football This is the closest that I will get.

643
00:35:07,640 --> 00:35:11,359
Speaker 2: But it's a lot of responsibility, and I think all

644
00:35:11,400 --> 00:35:12,719
Speaker 2: of the players need to be aware of that.

645
00:35:15,760 --> 00:35:15,920
Speaker 4: More.

646
00:35:15,960 --> 00:35:28,680
Speaker 1: After a quick break, let's get right back into it.

647
00:35:30,760 --> 00:35:35,120
Speaker 1: In my opinion, the reason why this conversation started out

648
00:35:35,160 --> 00:35:38,799
Speaker 1: from such a volatile place was that piece and Rolling Stone. Right,

649
00:35:38,840 --> 00:35:41,040
Speaker 1: so I read the report, I have a good sense

650
00:35:41,040 --> 00:35:43,960
Speaker 1: of what's it. It exactly what you were just saying, right, Like,

651
00:35:44,000 --> 00:35:48,040
Speaker 1: they're not necessarily saying this is guaranteed, you know, all

652
00:35:48,120 --> 00:35:51,480
Speaker 1: bots whatever. Whatever. The headline of the Rolling Stone report,

653
00:35:51,520 --> 00:35:53,480
Speaker 1: which I believe is what most people read, I don't

654
00:35:53,480 --> 00:35:56,279
Speaker 1: think most people were going to the actual report. The

655
00:35:56,280 --> 00:36:01,160
Speaker 1: headline was Taylor Swift's last album sparked bizarrec you's of Nazism.

656
00:36:01,680 --> 00:36:05,719
Speaker 1: It was a coordinated attack. Having read the report, that's

657
00:36:05,760 --> 00:36:07,280
Speaker 1: not even really what it said.

658
00:36:07,640 --> 00:36:09,200
Speaker 3: It's so many steps too far.

659
00:36:10,680 --> 00:36:12,400
Speaker 2: And I don't know how many people are aware that,

660
00:36:12,520 --> 00:36:15,160
Speaker 2: right like, headlines are written by different people who write

661
00:36:15,280 --> 00:36:18,880
Speaker 2: the article. You know, It's it's almost like we circled

662
00:36:18,880 --> 00:36:23,880
Speaker 2: back to the thing that people have said for you know, decades,

663
00:36:23,920 --> 00:36:25,759
Speaker 2: that they don't like about the media, right Like, So

664
00:36:25,840 --> 00:36:28,080
Speaker 2: this isn't necessarily an Internet problem at this point that

665
00:36:28,120 --> 00:36:31,600
Speaker 2: we're talking about. This is a problem of like media

666
00:36:31,880 --> 00:36:35,880
Speaker 2: and common grievances that people have about how things are clickbaity.

667
00:36:36,200 --> 00:36:39,200
Speaker 1: Yeah, I don't know that people know how pieces like

668
00:36:39,280 --> 00:36:41,760
Speaker 1: this come to be. And I also think Rolling Stone,

669
00:36:41,840 --> 00:36:43,960
Speaker 1: like I saw people going back and saying, oh well,

670
00:36:44,000 --> 00:36:46,359
Speaker 1: when her album came out, Taylor Swift and a Rolling

671
00:36:46,440 --> 00:36:49,480
Speaker 1: Stone takeover, they gave her album a five out of five,

672
00:36:49,560 --> 00:36:53,799
Speaker 1: like be I don't think that that's I think that

673
00:36:53,800 --> 00:36:56,120
Speaker 1: that's fair to be part of a conversation is what

674
00:36:56,160 --> 00:37:00,080
Speaker 1: relationship did Rolling Stone have with Taylor Swift before? And

675
00:37:00,160 --> 00:37:02,680
Speaker 1: then like why were they chosen to get this exclusive

676
00:37:02,800 --> 00:37:07,160
Speaker 1: report about her being the target of this this coordinative attack.

677
00:37:07,480 --> 00:37:10,760
Speaker 1: I don't think those questions are are totally out of pocket,

678
00:37:10,960 --> 00:37:13,160
Speaker 1: but I think it really goes to show like why

679
00:37:13,239 --> 00:37:16,359
Speaker 1: you need to be so intentional and so careful when

680
00:37:16,400 --> 00:37:18,440
Speaker 1: you're going to be rolling out a report like this,

681
00:37:18,600 --> 00:37:19,960
Speaker 1: because people are going to run with it.

682
00:37:20,360 --> 00:37:23,239
Speaker 2: Yeah, I mean coming from from this side, from the

683
00:37:23,360 --> 00:37:27,160
Speaker 2: from the industry side, I would say, you know, there

684
00:37:27,200 --> 00:37:29,440
Speaker 2: are two sides of how this could work. You know,

685
00:37:29,560 --> 00:37:34,040
Speaker 2: sometimes we find something internally that's interesting, and you know,

686
00:37:34,160 --> 00:37:37,920
Speaker 2: we shop it out to various media outlets and we say, like,

687
00:37:38,080 --> 00:37:41,040
Speaker 2: is anyone interested in this cool data that we found?

688
00:37:41,640 --> 00:37:44,880
Speaker 2: And sometimes it's the flip side. Sometimes it's an outlet

689
00:37:44,960 --> 00:37:47,600
Speaker 2: coming to us that they're asking this question of a

690
00:37:47,600 --> 00:37:50,600
Speaker 2: lot of different tech companies or researchers in the space

691
00:37:50,640 --> 00:37:52,720
Speaker 2: and saying, hey, does anyone have an answer to this question?

692
00:37:53,080 --> 00:37:55,480
Speaker 2: So that would be kind of my follow up from

693
00:37:55,480 --> 00:37:58,520
Speaker 2: like the mechanics of the industry perspective is, was this

694
00:37:58,640 --> 00:38:02,240
Speaker 2: report shopped around bund to a number of different outlets,

695
00:38:02,280 --> 00:38:04,160
Speaker 2: and just it happened to be that Rolling Stone was

696
00:38:04,200 --> 00:38:07,600
Speaker 2: the one who picked it up, or was rolling Stone

697
00:38:07,920 --> 00:38:11,839
Speaker 2: calling different companies asking if anyone had any data on this.

698
00:38:12,239 --> 00:38:14,040
Speaker 2: We didn't get a call from Rolling Stone asking if

699
00:38:14,040 --> 00:38:17,239
Speaker 2: we had Taylor Swift data, So I can tell you

700
00:38:17,280 --> 00:38:19,480
Speaker 2: that they did not call us for a consusation. But

701
00:38:19,800 --> 00:38:23,040
Speaker 2: it is something that happens frequently. And sometimes I will

702
00:38:23,600 --> 00:38:27,400
Speaker 2: exhaustively research something and you know, work with a media

703
00:38:27,480 --> 00:38:30,640
Speaker 2: organization about how to you know, write an article about it,

704
00:38:30,680 --> 00:38:33,640
Speaker 2: and sometimes that article doesn't even get published, right, So

705
00:38:33,880 --> 00:38:35,960
Speaker 2: you know, that's kind of the background of how this

706
00:38:36,040 --> 00:38:36,600
Speaker 2: all works.

707
00:38:37,680 --> 00:38:39,600
Speaker 3: So I'm not sure which dynamic was at play here.

708
00:38:39,800 --> 00:38:41,719
Speaker 5: If I can jump in with another question here, I

709
00:38:41,719 --> 00:38:45,200
Speaker 5: think this is, you know, raises some good questions about

710
00:38:45,200 --> 00:38:48,720
Speaker 5: the role of industry in making this kind of data

711
00:38:48,760 --> 00:38:50,960
Speaker 5: available to the public. So, like, I come from a

712
00:38:50,960 --> 00:38:56,440
Speaker 5: public health background where resources are pretty limited, right, and

713
00:38:57,400 --> 00:39:00,960
Speaker 5: with the platforms shutting down API access researchers a couple

714
00:39:01,000 --> 00:39:04,960
Speaker 5: of years ago, and you know, the Trump administration shifting

715
00:39:05,120 --> 00:39:10,480
Speaker 5: resources away from a lot of funding mechanisms that were

716
00:39:10,480 --> 00:39:14,720
Speaker 5: there in the past. The capacity of like nonprofit public

717
00:39:14,719 --> 00:39:18,839
Speaker 5: health researchers to do these kinds of investigations is really

718
00:39:18,880 --> 00:39:23,840
Speaker 5: pretty limited, and so, you know, I've seen criticism online

719
00:39:23,840 --> 00:39:25,560
Speaker 5: about this Tailor Swift report that like, oh, it's a

720
00:39:25,560 --> 00:39:28,520
Speaker 5: for profit company, we can't trust anything that they say.

721
00:39:29,520 --> 00:39:32,719
Speaker 5: But given like just the state of the world and

722
00:39:33,600 --> 00:39:39,000
Speaker 5: how complicated the internet is and the limited resources and

723
00:39:39,040 --> 00:39:43,759
Speaker 5: capacity of people in public health, I'm really like skeptical

724
00:39:43,920 --> 00:39:45,799
Speaker 5: of an approach that says like, well, we just can't

725
00:39:45,840 --> 00:39:49,440
Speaker 5: trust anything that comes out of industry whatsoever. It just

726
00:39:49,560 --> 00:39:54,640
Speaker 5: feels like so limiting to shoot ourselves in the foot

727
00:39:54,760 --> 00:39:59,440
Speaker 5: when you know, I wish that we had companies like

728
00:39:59,440 --> 00:40:03,120
Speaker 5: yours doing these kinds of analyzes of public health topics

729
00:40:03,160 --> 00:40:07,400
Speaker 5: instead of brands or celebrities, but you know, it's just

730
00:40:07,440 --> 00:40:10,399
Speaker 5: not there. And so I'm just, yeah, curious what you

731
00:40:10,480 --> 00:40:14,680
Speaker 5: think about, you know, what is that relationship of for

732
00:40:14,760 --> 00:40:19,200
Speaker 5: profit companies that are acting in private space but then

733
00:40:19,239 --> 00:40:23,080
Speaker 5: contributing to public conversation. How should members of the public

734
00:40:23,160 --> 00:40:25,879
Speaker 5: think about to think about that, That's a great question.

735
00:40:26,000 --> 00:40:29,240
Speaker 5: We're getting like into the nitty gritty of this industry,

736
00:40:29,239 --> 00:40:31,440
Speaker 5: and I really like that so I can tell you

737
00:40:31,480 --> 00:40:35,239
Speaker 5: that I'm working on public health problem sets for our customers, right,

738
00:40:35,719 --> 00:40:40,040
Speaker 5: but that's not necessarily something that we're being asked to

739
00:40:40,200 --> 00:40:44,400
Speaker 5: report publicly about. So I think maybe to lift the

740
00:40:44,400 --> 00:40:47,560
Speaker 5: hood a little bit for the industry side, is, you know,

741
00:40:47,719 --> 00:40:51,839
Speaker 5: I'm working on a huge number of different customers from

742
00:40:51,920 --> 00:40:55,120
Speaker 5: Fortune five hundred to government to commercial to big pr

743
00:40:55,200 --> 00:40:59,279
Speaker 5: and entertainment, and the only things that I end up

744
00:40:59,320 --> 00:41:03,360
Speaker 5: being able to talk public about are if that customer

745
00:41:03,440 --> 00:41:06,359
Speaker 5: wants to, you know, do a public facing report, or

746
00:41:06,440 --> 00:41:09,839
Speaker 5: if media asks us. And so that's why in the

747
00:41:09,840 --> 00:41:14,440
Speaker 5: industry we'll say yes to media asks because that's a

748
00:41:14,520 --> 00:41:20,080
Speaker 5: way to highlight our capabilities. And what does media typically

749
00:41:20,120 --> 00:41:24,759
Speaker 5: ask about. It's typically politics or celebrities. So if you

750
00:41:24,760 --> 00:41:26,920
Speaker 5: were to look at, you know, what are the issues

751
00:41:26,960 --> 00:41:31,440
Speaker 5: that I've commented on in my in my current role

752
00:41:31,480 --> 00:41:33,919
Speaker 5: over the past two years, you would come to sort

753
00:41:33,920 --> 00:41:36,880
Speaker 5: of the like an odd conclusion that I am very

754
00:41:37,040 --> 00:41:40,600
Speaker 5: into only pop culture and politics and this is one

755
00:41:40,680 --> 00:41:43,319
Speaker 5: hundred percent of my work. Those just happen to be

756
00:41:43,400 --> 00:41:47,000
Speaker 5: the issues that I'm able to talk about publicly because,

757
00:41:47,239 --> 00:41:49,840
Speaker 5: to be frank like a lot of these tech companies

758
00:41:49,960 --> 00:41:52,920
Speaker 5: were small I know, Peak Metrics is still in like

759
00:41:52,960 --> 00:41:56,759
Speaker 5: a startup role, and so we don't necessarily have the

760
00:41:56,800 --> 00:42:01,120
Speaker 5: timer resources to pull together the most interesting public facing

761
00:42:01,160 --> 00:42:04,719
Speaker 5: report on the most pressing issue. If we can't guarantee

762
00:42:04,760 --> 00:42:07,040
Speaker 5: that it's going to get placement somewhere, that's a lot

763
00:42:07,040 --> 00:42:11,200
Speaker 5: of energy and resources to expend on something. So that's

764
00:42:11,320 --> 00:42:15,480
Speaker 5: why companies, you know, answer the call from media. So

765
00:42:15,680 --> 00:42:21,719
Speaker 5: media gets basically this free data and companies get free

766
00:42:21,880 --> 00:42:24,839
Speaker 5: you know, advertising of the types of things that they

767
00:42:24,880 --> 00:42:29,200
Speaker 5: can do so that you know, customers who have other

768
00:42:29,320 --> 00:42:33,000
Speaker 5: types of problems that are not maybe pop culture politics related,

769
00:42:33,320 --> 00:42:35,400
Speaker 5: call us up to say, hey, can you take a

770
00:42:35,400 --> 00:42:38,680
Speaker 5: look at this issue for me. So that's the mechanics

771
00:42:38,719 --> 00:42:40,719
Speaker 5: of like how this works. And I think the incentives

772
00:42:40,719 --> 00:42:43,480
Speaker 5: I don't think. I think everyone should be skeptical of

773
00:42:43,520 --> 00:42:46,120
Speaker 5: things that are coming always from private industry versus like

774
00:42:46,520 --> 00:42:49,400
Speaker 5: academia and researchers. But we also don't really have like

775
00:42:49,440 --> 00:42:54,600
Speaker 5: an established discipline of internet researchers I think in academia

776
00:42:54,640 --> 00:42:57,840
Speaker 5: to even to even call from. I mean, you know,

777
00:42:57,880 --> 00:43:00,680
Speaker 5: there's like the AM I thinking correctly of the Shortenstein

778
00:43:00,760 --> 00:43:05,240
Speaker 5: Center at Harvard, you know, love all of their stuff

779
00:43:05,239 --> 00:43:08,080
Speaker 5: that they publish. But I mean, it's an emerging field.

780
00:43:08,160 --> 00:43:10,960
Speaker 5: And I think that's maybe why you have a sense

781
00:43:11,040 --> 00:43:14,040
Speaker 5: that industry is leading the way is because I think

782
00:43:14,080 --> 00:43:17,319
Speaker 5: that academia has not figured out where to carve this

783
00:43:17,400 --> 00:43:17,920
Speaker 5: out yet.

784
00:43:18,160 --> 00:43:21,440
Speaker 1: Oh Molly, you just hit my I'll just say I agree,

785
00:43:21,440 --> 00:43:25,280
Speaker 1: because I'll if you get me going. I so. Formerly

786
00:43:25,360 --> 00:43:28,759
Speaker 1: I was a research fellow at brookmanklin which is sort

787
00:43:28,760 --> 00:43:32,000
Speaker 1: of a cousin to the Short Shortenstein Center at Harvard.

788
00:43:33,680 --> 00:43:38,080
Speaker 1: People who do research on the Internet in from in

789
00:43:38,120 --> 00:43:42,600
Speaker 1: an academic way attached to universities are having a rough time,

790
00:43:42,680 --> 00:43:44,640
Speaker 1: I'll just put it that way, and they're us it

791
00:43:44,719 --> 00:43:46,600
Speaker 1: used to Things used to be better. There used to

792
00:43:46,640 --> 00:43:48,719
Speaker 1: be more funding. But if you do that kind of

793
00:43:48,719 --> 00:43:50,600
Speaker 1: work when you're doing it today in twenty twenty five,

794
00:43:51,239 --> 00:43:53,360
Speaker 1: god bless you. I am happy that you've got it

795
00:43:53,400 --> 00:43:57,040
Speaker 1: figured out. But we have really hollowed out an entire

796
00:43:57,239 --> 00:44:00,359
Speaker 1: space of researchers who are interested in what's happening on

797
00:44:00,400 --> 00:44:03,640
Speaker 1: the Internet. It's not a robust field any longer. I

798
00:44:03,640 --> 00:44:07,600
Speaker 1: can definitely confirm that. But I'm really glad that you

799
00:44:07,800 --> 00:44:11,440
Speaker 1: that you made this point about private companies versus you know,

800
00:44:11,560 --> 00:44:15,040
Speaker 1: Academia and another spaces because one of the criticisms I've

801
00:44:15,080 --> 00:44:20,440
Speaker 1: seen online about this report was that it's essentially Taylor

802
00:44:20,440 --> 00:44:23,280
Speaker 1: Swift PR right that you know, and they're they're talking

803
00:44:23,280 --> 00:44:26,239
Speaker 1: about this in an almost nefarious way that oh, this

804
00:44:26,360 --> 00:44:29,680
Speaker 1: is just a company whose whole job is to get

805
00:44:30,320 --> 00:44:35,919
Speaker 1: good positive press hits for brands and celebrities and things

806
00:44:36,000 --> 00:44:38,160
Speaker 1: like that. And what's interesting to me is that from

807
00:44:38,200 --> 00:44:40,600
Speaker 1: what you've said, it doesn't sound like that criticism is

808
00:44:40,640 --> 00:44:43,879
Speaker 1: like totally wrong because in a kind of way, that's

809
00:44:43,880 --> 00:44:45,840
Speaker 1: how these companies get their name out there so that

810
00:44:45,920 --> 00:44:48,120
Speaker 1: they can continue to fund the other important work that

811
00:44:48,160 --> 00:44:51,279
Speaker 1: they're doing. But that like it's sort of missing the

812
00:44:51,320 --> 00:44:53,719
Speaker 1: forest for the trees of what's actually going on in

813
00:44:53,760 --> 00:44:56,080
Speaker 1: terms of how this research comes to get to the public.

814
00:44:56,120 --> 00:44:56,840
Speaker 1: Do I have that sort of.

815
00:44:56,920 --> 00:44:58,320
Speaker 3: Right on your person?

816
00:44:58,360 --> 00:45:01,320
Speaker 2: And I mean I think if they were successfully doing

817
00:45:01,400 --> 00:45:04,879
Speaker 2: PR plants for Taylor Swift, they would have a much

818
00:45:04,920 --> 00:45:07,719
Speaker 2: more you know, robust company at this point, and we

819
00:45:07,760 --> 00:45:10,719
Speaker 2: would be able to tell that the roof would be

820
00:45:10,719 --> 00:45:13,840
Speaker 2: in the putting. I think I guess one other element

821
00:45:13,880 --> 00:45:15,759
Speaker 2: to talk about, like, while we're just talking about, you know,

822
00:45:16,480 --> 00:45:19,839
Speaker 2: the vibes of the Internet is. You know, you talked

823
00:45:19,840 --> 00:45:23,000
Speaker 2: about the hollowing out of you know, academia.

824
00:45:22,320 --> 00:45:22,680
Speaker 4: On this.

825
00:45:24,200 --> 00:45:28,560
Speaker 2: There's you know, maybe fewer people in house that news

826
00:45:28,680 --> 00:45:32,880
Speaker 2: and media organizations have that are like Internet experts and

827
00:45:32,920 --> 00:45:35,719
Speaker 2: are are given kind of like the time and resources

828
00:45:35,719 --> 00:45:38,320
Speaker 2: to dig into this. I mean, if we're talking about

829
00:45:38,320 --> 00:45:41,799
Speaker 2: like Internet vibes and reporting on the Internet as a beat,

830
00:45:41,840 --> 00:45:44,920
Speaker 2: I feel like I have to mention Brandy Seedrosni who.

831
00:45:44,760 --> 00:45:46,880
Speaker 3: I respect her work so greatly.

832
00:45:46,920 --> 00:45:50,680
Speaker 2: Getting to getting to pitch Peak Metrics data to her

833
00:45:50,719 --> 00:45:52,279
Speaker 2: and interact with her has like been one of the

834
00:45:52,360 --> 00:45:55,400
Speaker 2: highlights of my career because that was that was the

835
00:45:55,400 --> 00:45:58,760
Speaker 2: first time that I really saw the Internet being reported

836
00:45:58,800 --> 00:46:00,960
Speaker 2: on as like, hey, we're sending a reporter out to

837
00:46:01,040 --> 00:46:04,480
Speaker 2: this location. Now we're sending a reporter like into the

838
00:46:04,600 --> 00:46:07,319
Speaker 2: bowels of the Internet to come back and tell us

839
00:46:07,360 --> 00:46:11,719
Speaker 2: what's going on. And you know, not every news organization

840
00:46:12,120 --> 00:46:14,560
Speaker 2: news organization has a brand news adrosneens, So that's why

841
00:46:14,600 --> 00:46:19,279
Speaker 2: they they call up companies like Peak Metrics to you know,

842
00:46:19,680 --> 00:46:22,960
Speaker 2: give them a sense of what is going on on

843
00:46:23,239 --> 00:46:25,200
Speaker 2: the other side, which is, you know, this thing that

844
00:46:25,239 --> 00:46:28,439
Speaker 2: we're all a part of. But you can't just walk

845
00:46:28,480 --> 00:46:29,879
Speaker 2: away and like put your finger up in the air

846
00:46:29,880 --> 00:46:31,400
Speaker 2: and say, like, as a matter of fact, this is

847
00:46:31,440 --> 00:46:32,040
Speaker 2: what's happening.

848
00:46:32,400 --> 00:46:34,200
Speaker 1: One of the questions I was going to ask is,

849
00:46:34,640 --> 00:46:36,719
Speaker 1: you know, what's the solution to this, to the way

850
00:46:36,760 --> 00:46:39,640
Speaker 1: that bots are disrupting our discourse? And I was thinking

851
00:46:39,719 --> 00:46:41,840
Speaker 1: of it as a question of like what should platforms

852
00:46:41,880 --> 00:46:43,520
Speaker 1: be doing or not doing? But it sounds like the

853
00:46:43,680 --> 00:46:48,080
Speaker 1: stepping back, the problem is so much more layered than

854
00:46:48,200 --> 00:46:52,200
Speaker 1: platform X needs to do specific thing. It's a it's

855
00:46:52,239 --> 00:46:55,400
Speaker 1: a problem of media outlets, it's a problem of coverage.

856
00:46:55,440 --> 00:46:58,359
Speaker 1: It sounds like a much more a complex problem than

857
00:46:58,440 --> 00:47:01,879
Speaker 1: just saying, oh, must decide to do this, that'll fix

858
00:47:01,880 --> 00:47:02,480
Speaker 1: the problem.

859
00:47:02,960 --> 00:47:03,200
Speaker 6: Yeah.

860
00:47:03,280 --> 00:47:05,239
Speaker 2: I mean we're going to continue, probably to be in

861
00:47:05,280 --> 00:47:08,799
Speaker 2: like an outrage clickbait cycle of hey, the bots were

862
00:47:09,480 --> 00:47:12,160
Speaker 2: this big of a part of this conversation and what

863
00:47:12,200 --> 00:47:15,200
Speaker 2: implications does that have for the subject of this conversation.

864
00:47:16,040 --> 00:47:19,560
Speaker 2: But you know, until we get a sense of like

865
00:47:19,600 --> 00:47:23,640
Speaker 2: what is the normal baseline level of bot activity, We're

866
00:47:23,680 --> 00:47:25,680
Speaker 2: just going to kind of keep spinning on this wheel.

867
00:47:25,800 --> 00:47:29,160
Speaker 2: So I'm hopeful that maybe this time next year, I

868
00:47:29,200 --> 00:47:32,680
Speaker 2: feel like this is this bot activity discourse has really

869
00:47:32,760 --> 00:47:36,680
Speaker 2: hit maybe in twenty twenty five, we're all a little

870
00:47:36,719 --> 00:47:39,000
Speaker 2: bit more aware of it now, we're building out, you know,

871
00:47:39,040 --> 00:47:41,799
Speaker 2: a repertoire of research on it. You know, hopefully if

872
00:47:41,840 --> 00:47:44,520
Speaker 2: we were to be coming back together in December twenty

873
00:47:44,520 --> 00:47:48,440
Speaker 2: twenty six, maybe we'll have like a better understanding of

874
00:47:48,480 --> 00:47:51,800
Speaker 2: this that will get us less locked into these outrage cycles.

875
00:47:52,320 --> 00:47:55,160
Speaker 1: Molly, is there anything else that I have not asked

876
00:47:55,160 --> 00:47:56,760
Speaker 1: you about that you want to make sure gets included

877
00:47:56,800 --> 00:47:57,480
Speaker 1: in this conversation?

878
00:47:59,520 --> 00:48:01,400
Speaker 3: I guess is there a part of the Internet that

879
00:48:01,440 --> 00:48:04,040
Speaker 3: makes you happy? Like why are we all? Why are

880
00:48:04,120 --> 00:48:05,960
Speaker 3: we all still here on the internet? Like why are

881
00:48:06,000 --> 00:48:06,480
Speaker 3: we doing this?

882
00:48:07,840 --> 00:48:09,799
Speaker 1: What are you? What's your Internet happy place? What make

883
00:48:09,880 --> 00:48:12,160
Speaker 1: what fills you with hope for the future? And or

884
00:48:12,200 --> 00:48:13,960
Speaker 1: just you dislike spending time online?

885
00:48:14,280 --> 00:48:17,279
Speaker 2: Yeah, well, I mean, I guess I mentioned that I

886
00:48:17,360 --> 00:48:21,080
Speaker 2: was like a twenty tens Tumblr girly, and I still

887
00:48:21,120 --> 00:48:25,280
Speaker 2: have close friends that I met on Tumblr that I

888
00:48:25,320 --> 00:48:29,520
Speaker 2: am talking to, you know, fifteen years later. So I

889
00:48:29,560 --> 00:48:34,719
Speaker 2: think that that's like the power of Internet subcommunities when

890
00:48:34,880 --> 00:48:39,440
Speaker 2: they're working well, And I think that people use the

891
00:48:39,480 --> 00:48:41,279
Speaker 2: Internet to find connections like that.

892
00:48:42,120 --> 00:48:46,080
Speaker 1: I'm so used to talking about the fucking drugs of

893
00:48:46,080 --> 00:48:48,319
Speaker 1: the Internet that I forget that. Actually it's a place

894
00:48:48,320 --> 00:48:50,960
Speaker 1: that I like and I'm there every day voluntarily and

895
00:48:51,080 --> 00:48:54,120
Speaker 1: like it was hugely informative to me as a youth,

896
00:48:54,160 --> 00:48:55,920
Speaker 1: and you know, that's why I keep coming back. It's

897
00:48:55,920 --> 00:48:57,400
Speaker 1: a good it's good of a reminder to be like,

898
00:48:57,480 --> 00:49:01,399
Speaker 1: you actually enjoy technology, right, like you actually voluntarily keep

899
00:49:01,400 --> 00:49:02,040
Speaker 1: it in your life.

900
00:49:02,120 --> 00:49:04,600
Speaker 2: No, yeah, you're not like a complete let in at

901
00:49:04,600 --> 00:49:06,440
Speaker 2: this point, right right, exactly.

902
00:49:11,400 --> 00:49:24,160
Speaker 1: More after a quick break, let's get right back into it.

903
00:49:26,080 --> 00:49:29,560
Speaker 1: Social media platforms only work when they actually help people

904
00:49:29,600 --> 00:49:32,399
Speaker 1: make sense of the world around them. That breaks down

905
00:49:32,480 --> 00:49:36,520
Speaker 1: when they're overrun with inauthentic narratives and bots engineered to

906
00:49:36,600 --> 00:49:39,960
Speaker 1: bait engagement. But understanding all this is where the work

907
00:49:40,000 --> 00:49:42,120
Speaker 1: of people like Keith Presley comes in.

908
00:49:43,840 --> 00:49:47,000
Speaker 6: We started as a nonprofit trying to identify how information

909
00:49:47,200 --> 00:49:51,080
Speaker 6: moves across the Internet, So where are those breeding grounds

910
00:49:51,120 --> 00:49:53,239
Speaker 6: of information campaigns? And then how does it get to

911
00:49:53,280 --> 00:49:57,200
Speaker 6: the broader network and impact people? Just mapping how information moved,

912
00:49:58,000 --> 00:50:00,919
Speaker 6: So we went ahead and did that. Turns out there's

913
00:50:00,920 --> 00:50:03,480
Speaker 6: a lot more use cases than the limited one we

914
00:50:03,520 --> 00:50:05,759
Speaker 6: had in the nonprofit. So we then spun out into

915
00:50:05,800 --> 00:50:06,320
Speaker 6: a company.

916
00:50:06,760 --> 00:50:09,960
Speaker 1: Keith's company, Kadeta, was all over the Internet thanks to

917
00:50:10,000 --> 00:50:12,759
Speaker 1: a report they put out looking at how coordinated bot

918
00:50:12,840 --> 00:50:16,760
Speaker 1: campaigns influenced discourse around the release of Taylor Swift's latest album,

919
00:50:17,080 --> 00:50:19,840
Speaker 1: Life of a show Girl. You might recall that after

920
00:50:19,840 --> 00:50:22,840
Speaker 1: the album was released, Taylor put out a necklace featuring

921
00:50:22,920 --> 00:50:26,240
Speaker 1: lightning bolts, an image that some felt bore us striking

922
00:50:26,320 --> 00:50:30,239
Speaker 1: resemblance to not the ss iconography. The report examined what

923
00:50:30,400 --> 00:50:33,239
Speaker 1: sparked those claims and how those claims spread across the

924
00:50:33,280 --> 00:50:37,480
Speaker 1: Internet like wildfire. The report was published in Rolling Stone magazine.

925
00:50:38,239 --> 00:50:40,880
Speaker 6: We like to consider ourselves the stormtrocker for the Internet. So,

926
00:50:41,040 --> 00:50:44,360
Speaker 6: as a meteorologist uses patterns to predict the weather, we

927
00:50:44,480 --> 00:50:48,120
Speaker 6: use patterns to predict information online. So we take in

928
00:50:48,120 --> 00:50:51,840
Speaker 6: information from over four hundred and eighty different platforms, and

929
00:50:51,840 --> 00:50:54,440
Speaker 6: then we have a patented process where we run graphnal

930
00:50:54,520 --> 00:50:57,920
Speaker 6: networking through that data to see how people behave with it,

931
00:50:58,440 --> 00:51:01,240
Speaker 6: and then use AI to lift out those behavioral patterns

932
00:51:01,239 --> 00:51:02,640
Speaker 6: and summarize for our clients.

933
00:51:03,320 --> 00:51:05,760
Speaker 1: I have seen people say that what your company does

934
00:51:05,920 --> 00:51:09,680
Speaker 1: is essentially pr that celebrities and brands like Taylor Swift,

935
00:51:10,040 --> 00:51:13,359
Speaker 1: just give you all money to print nice things about them.

936
00:51:13,600 --> 00:51:15,600
Speaker 1: What do you say to somebody who feels that way

937
00:51:15,640 --> 00:51:18,840
Speaker 1: about what you're doing. I would say one this report

938
00:51:19,440 --> 00:51:21,480
Speaker 1: we did on our own. We were just interested.

939
00:51:21,680 --> 00:51:23,680
Speaker 6: I had a gut feeling that there was something funny

940
00:51:23,680 --> 00:51:28,560
Speaker 6: going on, ran the report, and then found something. A

941
00:51:28,600 --> 00:51:31,600
Speaker 6: big reason of why we wanted to do this though

942
00:51:31,640 --> 00:51:34,719
Speaker 6: and released it is a lot of this work is

943
00:51:34,760 --> 00:51:35,840
Speaker 6: behind closed doors.

944
00:51:35,920 --> 00:51:36,080
Speaker 1: Right.

945
00:51:36,120 --> 00:51:39,319
Speaker 6: We do work with companies and report on what is

946
00:51:39,360 --> 00:51:43,480
Speaker 6: happening around their narratives. Public doesn't get to see that,

947
00:51:43,880 --> 00:51:45,600
Speaker 6: and so this was a way for us to get

948
00:51:45,640 --> 00:51:47,520
Speaker 6: something out there so that people could have a better

949
00:51:47,600 --> 00:51:51,879
Speaker 6: understanding of how online information environments are actually being manipulated.

950
00:51:52,960 --> 00:51:55,280
Speaker 1: It's true that most of the social listening work happening

951
00:51:55,320 --> 00:51:58,320
Speaker 1: in twenty twenty five stays behind closed doors. And even

952
00:51:58,360 --> 00:52:00,920
Speaker 1: though the headline and Rolling Stone made it seem like

953
00:52:00,960 --> 00:52:04,360
Speaker 1: something extraordinary had happened with a Taylor Swift narrative, to

954
00:52:04,440 --> 00:52:06,800
Speaker 1: analysts who work on this kind of stuff all the time,

955
00:52:07,239 --> 00:52:10,600
Speaker 1: this kind of influence campaign was basically just another Tuesday.

956
00:52:11,320 --> 00:52:14,080
Speaker 1: It's funny that you say that this report started with

957
00:52:14,160 --> 00:52:17,840
Speaker 1: just sort of a gut feeling months ago, back in October,

958
00:52:17,880 --> 00:52:21,600
Speaker 1: we did an episode about the conversation links to the

959
00:52:21,680 --> 00:52:24,120
Speaker 1: launch of Taylor Swift's latest album, and I said the

960
00:52:24,200 --> 00:52:26,720
Speaker 1: exact same thing. I said, I don't have any proof.

961
00:52:27,080 --> 00:52:30,000
Speaker 1: I'm not a data scientist. Do My producer is a

962
00:52:30,080 --> 00:52:32,239
Speaker 1: data analyst, but like I do that It's not a

963
00:52:32,239 --> 00:52:34,400
Speaker 1: skill set that I have. But I have been on

964
00:52:34,440 --> 00:52:37,400
Speaker 1: the internet for a very long time, and I'm particularly

965
00:52:37,440 --> 00:52:42,759
Speaker 1: pretty plugged in with how conversation about marginalized people, so

966
00:52:42,880 --> 00:52:45,960
Speaker 1: like women, especially women in the public eye, I'm pretty

967
00:52:46,360 --> 00:52:49,080
Speaker 1: cluded to have those conversations move. And the way that

968
00:52:49,080 --> 00:52:55,000
Speaker 1: I would describe the conversation specifically around the Taylor Swift

969
00:52:55,239 --> 00:52:59,400
Speaker 1: Nazi necklace thing, it was like it came on strong

970
00:52:59,520 --> 00:53:03,279
Speaker 1: out of no where, was such an intense conversation, and

971
00:53:03,320 --> 00:53:07,200
Speaker 1: then it kind of stopped just as abruptly as it started.

972
00:53:07,840 --> 00:53:09,839
Speaker 1: I just something about that, I said, I don't know

973
00:53:09,840 --> 00:53:13,319
Speaker 1: if this is this is all authentic, Something about the

974
00:53:13,320 --> 00:53:16,480
Speaker 1: way that it came on so suddenly just gave me pause.

975
00:53:16,800 --> 00:53:18,200
Speaker 1: I guess I will say, it sounds like you all

976
00:53:18,239 --> 00:53:20,239
Speaker 1: were in the same boat exactly.

977
00:53:20,400 --> 00:53:25,520
Speaker 6: It's there are some tall tale signs of things are fishy, right,

978
00:53:25,960 --> 00:53:29,719
Speaker 6: and this is when you see such huge surges out

979
00:53:29,719 --> 00:53:32,759
Speaker 6: of nowhere that it is normally a leading indicator that

980
00:53:32,840 --> 00:53:36,799
Speaker 6: there are some sort of some sort of coordinated activity

981
00:53:37,200 --> 00:53:40,279
Speaker 6: to get that information out there to then impact how

982
00:53:40,280 --> 00:53:43,279
Speaker 6: a normal person is going to talk about it. So

983
00:53:43,520 --> 00:53:47,480
Speaker 6: get them to interact with that negative or illicit content,

984
00:53:47,560 --> 00:53:49,680
Speaker 6: right that it's the goal they're rage baiting.

985
00:53:50,080 --> 00:53:52,600
Speaker 1: So is that really kind of how this report came

986
00:53:52,640 --> 00:53:55,719
Speaker 1: to be? Just this inkling of something fishy is going

987
00:53:55,760 --> 00:53:57,440
Speaker 1: on here, let's find out what it is?

988
00:53:57,800 --> 00:54:02,920
Speaker 6: Yeah, truly, I know in the article, Georgia Paul says,

989
00:54:03,000 --> 00:54:05,120
Speaker 6: you know, she just had a feeling. It really came

990
00:54:05,160 --> 00:54:07,600
Speaker 6: from that. It was we were like in you know,

991
00:54:07,640 --> 00:54:10,239
Speaker 6: one of our early morning meetings and we're like, hey,

992
00:54:10,960 --> 00:54:14,040
Speaker 6: let's do it. Let's investigate and see what's going on there.

993
00:54:14,200 --> 00:54:15,360
Speaker 1: So what did the report find?

994
00:54:16,080 --> 00:54:21,240
Speaker 6: So essentially that the Nazi, that Taylor Swift is a Nazi,

995
00:54:21,280 --> 00:54:24,200
Speaker 6: and that the necklace with the lightning bolt necklace and

996
00:54:24,239 --> 00:54:27,400
Speaker 6: that was alluding to the SS from Germany. And what

997
00:54:27,440 --> 00:54:30,280
Speaker 6: they were doing is so they laid down that content,

998
00:54:30,440 --> 00:54:34,600
Speaker 6: a large quantity of that content to then impact influencers

999
00:54:34,600 --> 00:54:36,440
Speaker 6: who would then pick that up and then spread it

1000
00:54:36,480 --> 00:54:41,080
Speaker 6: more broadly to normal individuals. At that point, the narrative

1001
00:54:41,360 --> 00:54:45,320
Speaker 6: transforms to actually then being a comparison of Taylor to Kanye.

1002
00:54:46,080 --> 00:54:48,880
Speaker 6: Having something like that happen is the end goal, right,

1003
00:54:48,960 --> 00:54:51,919
Speaker 6: they want because that was actually real people now having

1004
00:54:52,000 --> 00:54:55,360
Speaker 6: that conversation, not driven by this inauthentic activity.

1005
00:54:56,000 --> 00:54:58,920
Speaker 5: Yeah, thank you for the summary there. I'm kind of

1006
00:54:59,360 --> 00:55:01,600
Speaker 5: jealous of the position that you guys are in, like

1007
00:55:01,680 --> 00:55:04,440
Speaker 5: having this data, because, like Bridgie was saying, we had

1008
00:55:04,480 --> 00:55:07,759
Speaker 5: a similar like feeling and conversation that something felt off

1009
00:55:07,760 --> 00:55:12,319
Speaker 5: about this conversation. But that's where it ended with us, right,

1010
00:55:12,360 --> 00:55:15,040
Speaker 5: because we did not have the tools readily available to

1011
00:55:15,440 --> 00:55:18,439
Speaker 5: look into it. And one of the things that has

1012
00:55:19,600 --> 00:55:22,520
Speaker 5: come up for me as we were researching this episode

1013
00:55:22,560 --> 00:55:25,879
Speaker 5: and talking to people is just how valuable it is

1014
00:55:25,920 --> 00:55:30,400
Speaker 5: for the public to get glimpses like this using data

1015
00:55:30,440 --> 00:55:33,839
Speaker 5: to like actually bring some data to help us understand

1016
00:55:34,640 --> 00:55:38,440
Speaker 5: what can seem very random but often is not. In

1017
00:55:38,560 --> 00:55:40,920
Speaker 5: terms of like conversation happening on the Internet.

1018
00:55:41,200 --> 00:55:43,520
Speaker 6: It's almost kind of like a black box the Internet,

1019
00:55:44,360 --> 00:55:49,240
Speaker 6: especially the social media platforms, where there's millions of posts,

1020
00:55:49,239 --> 00:55:53,680
Speaker 6: billions of posts being made daily, lots of different coordinated

1021
00:55:54,280 --> 00:55:58,319
Speaker 6: activities to either you know, from crypto schemes to just

1022
00:55:58,440 --> 00:56:01,800
Speaker 6: influencers wanting to get clicks. There's a lot of competing

1023
00:56:01,840 --> 00:56:06,480
Speaker 6: priorities that just you can't see normally. You can't get

1024
00:56:06,520 --> 00:56:10,480
Speaker 6: that bigger picture, especially as a normal person, right, we

1025
00:56:10,520 --> 00:56:12,040
Speaker 6: can only really see what our feeds are.

1026
00:56:12,280 --> 00:56:14,400
Speaker 1: Yeah, and I always make this point on the show.

1027
00:56:14,480 --> 00:56:18,720
Speaker 1: There are so few spaces or industries where the public

1028
00:56:18,920 --> 00:56:22,560
Speaker 1: is coming into contact with it regularly, where you have

1029
00:56:22,920 --> 00:56:26,719
Speaker 1: such limited information or data about how it's impacting people. Right,

1030
00:56:27,320 --> 00:56:30,960
Speaker 1: if there was a car company that was killing people

1031
00:56:31,000 --> 00:56:32,839
Speaker 1: and we weren't able to get that information, if there

1032
00:56:32,840 --> 00:56:35,239
Speaker 1: was a pharmaceutical company. There aren't really a lot of

1033
00:56:35,360 --> 00:56:39,120
Speaker 1: companies or industries where we've just accepted this is a

1034
00:56:39,160 --> 00:56:42,400
Speaker 1: black box of information where the public is interacting with

1035
00:56:42,440 --> 00:56:45,080
Speaker 1: this every day but has no idea what that interaction

1036
00:56:45,239 --> 00:56:46,719
Speaker 1: actually means or looks like.

1037
00:56:46,840 --> 00:56:50,080
Speaker 6: On our end. This industry is really new too, just

1038
00:56:50,160 --> 00:56:53,840
Speaker 6: trying to actually understand how the Internet works, or information

1039
00:56:54,000 --> 00:56:56,839
Speaker 6: moving on the Internet, or how it's impacting people. There's

1040
00:56:56,840 --> 00:57:00,759
Speaker 6: some academic research that's been happening recently, you know, and

1041
00:57:00,800 --> 00:57:03,279
Speaker 6: then companies like ours that are trying to figure this out.

1042
00:57:03,400 --> 00:57:08,040
Speaker 6: To help society, but it's still a pretty small pool

1043
00:57:08,080 --> 00:57:09,239
Speaker 6: of people trying to do this.

1044
00:57:09,560 --> 00:57:12,040
Speaker 5: I'd like to get into the methodology of your report

1045
00:57:12,160 --> 00:57:14,240
Speaker 5: just a little bit. Like one of the things that

1046
00:57:13,480 --> 00:57:18,040
Speaker 5: you that you did was classify accounts according to whether

1047
00:57:18,080 --> 00:57:22,400
Speaker 5: they were typical or atypical, and then further subdivideed the atypical.

1048
00:57:22,440 --> 00:57:25,800
Speaker 5: And that seemed like an interesting approach to this problem

1049
00:57:25,880 --> 00:57:28,920
Speaker 5: of trying to make sense out of all these different

1050
00:57:28,960 --> 00:57:32,720
Speaker 5: actors and like different types of actors, where you know,

1051
00:57:32,760 --> 00:57:38,440
Speaker 5: the binary of ordinary human versus bot. As we've been

1052
00:57:38,440 --> 00:57:41,160
Speaker 5: researching this, it's I'm starting to feel like that's not

1053
00:57:41,480 --> 00:57:45,160
Speaker 5: perhaps a super useful distinction, a binary distinction. And so

1054
00:57:45,800 --> 00:57:48,760
Speaker 5: you guys used I think five different categories. Could you

1055
00:57:49,160 --> 00:57:51,440
Speaker 5: talk a little bit about that decision and how you

1056
00:57:51,480 --> 00:57:52,120
Speaker 5: defined those?

1057
00:57:52,680 --> 00:57:55,400
Speaker 6: Yeah, and you're kind of hitting the nail on the

1058
00:57:55,440 --> 00:57:58,680
Speaker 6: head for how we were thinking about this that one

1059
00:57:58,880 --> 00:58:02,560
Speaker 6: or zero. If you're looking at the what is trying

1060
00:58:02,600 --> 00:58:06,040
Speaker 6: to what are people trying to do online as either

1061
00:58:06,200 --> 00:58:09,440
Speaker 6: bought or not bought, typical or atypical, you're kind of

1062
00:58:09,440 --> 00:58:14,440
Speaker 6: missing the point. These all of these actions are coordinated

1063
00:58:14,520 --> 00:58:17,240
Speaker 6: to make information go viral, right, That's the end goal

1064
00:58:17,240 --> 00:58:19,880
Speaker 6: they want to get in front of eyeballs because for

1065
00:58:19,960 --> 00:58:24,360
Speaker 6: either illicard reasons or making money, et cetera. To do that,

1066
00:58:24,440 --> 00:58:27,560
Speaker 6: you can't just here's a bot, right, it's going to

1067
00:58:27,640 --> 00:58:30,000
Speaker 6: post a lot. That's not going to get something to

1068
00:58:30,000 --> 00:58:33,040
Speaker 6: go viral, as we've seen, right, you've got to impact people.

1069
00:58:33,120 --> 00:58:39,160
Speaker 6: And so we've observed and modeled five different behavioral patterns

1070
00:58:39,600 --> 00:58:42,240
Speaker 6: from the typical user all the way down to what

1071
00:58:42,280 --> 00:58:46,360
Speaker 6: you would consider that through you know, the bot, and

1072
00:58:47,240 --> 00:58:51,480
Speaker 6: are working to understand how they're used online to actually

1073
00:58:51,520 --> 00:58:55,320
Speaker 6: get information to go viral. So, you know, an influencer

1074
00:58:55,400 --> 00:58:57,760
Speaker 6: has their own specific behavioral pattern that they're going to

1075
00:58:57,800 --> 00:59:03,480
Speaker 6: be using what we call outliers, facilitators, and then power players,

1076
00:59:03,600 --> 00:59:07,800
Speaker 6: and so all of them work together in some instance

1077
00:59:08,040 --> 00:59:11,760
Speaker 6: to push strategies and tactics to make that information get

1078
00:59:11,760 --> 00:59:12,440
Speaker 6: in front of people.

1079
00:59:12,760 --> 00:59:15,440
Speaker 1: How do you know when this behavior is coordinated? That

1080
00:59:15,560 --> 00:59:18,240
Speaker 1: was one of the points the takeaways in the report

1081
00:59:18,320 --> 00:59:22,080
Speaker 1: was this is these are not just inauthentic accounts acting

1082
00:59:22,520 --> 00:59:24,680
Speaker 1: on their own. They're all coordinated in a kind of

1083
00:59:24,680 --> 00:59:26,480
Speaker 1: way as a network. How did you determine that?

1084
00:59:27,000 --> 00:59:30,960
Speaker 6: Yeah, so that's part of our patented process where so

1085
00:59:31,160 --> 00:59:35,560
Speaker 6: that graphnal networking through the information reveals patterns of behavior,

1086
00:59:35,880 --> 00:59:38,400
Speaker 6: and so it really does. It's like a fingerprint within

1087
00:59:38,400 --> 00:59:42,760
Speaker 6: the data. It actually has a very distinct pattern that

1088
00:59:42,800 --> 00:59:45,520
Speaker 6: we can then see and lift out and then make

1089
00:59:45,560 --> 00:59:48,160
Speaker 6: those determinations. And so a lot of that can be

1090
00:59:48,240 --> 00:59:51,720
Speaker 6: you know, how are they interacting with like accounts, how

1091
00:59:51,760 --> 00:59:54,640
Speaker 6: are they interacting with peers, the language that they are using,

1092
00:59:54,680 --> 00:59:58,320
Speaker 6: the timing, there is a slew of clues that we

1093
00:59:58,960 --> 01:00:03,360
Speaker 6: are at a that we then use to make these determinations.

1094
01:00:03,680 --> 01:00:06,240
Speaker 1: One of the criticisms that I've seen of the study

1095
01:00:06,440 --> 01:00:09,680
Speaker 1: is that the data set, the data that you used

1096
01:00:09,720 --> 01:00:12,240
Speaker 1: for this was not made available right, and so if

1097
01:00:12,280 --> 01:00:14,800
Speaker 1: I wanted to take that data set and crunch the

1098
01:00:14,840 --> 01:00:17,280
Speaker 1: numbers myself, I could not do this. We had a

1099
01:00:17,320 --> 01:00:20,880
Speaker 1: conversation with somebody in it from a socialisting organization and

1100
01:00:20,960 --> 01:00:25,240
Speaker 1: they said, oh, well, that's actually kind of commonplace for

1101
01:00:25,320 --> 01:00:28,360
Speaker 1: private companies. They might have proprietary things they don't want

1102
01:00:28,400 --> 01:00:33,360
Speaker 1: out there. How much did that impact not making that

1103
01:00:33,480 --> 01:00:34,920
Speaker 1: data set available for the public.

1104
01:00:36,160 --> 01:00:41,880
Speaker 6: That two fold, So proprietary processes and then like data

1105
01:00:41,920 --> 01:00:45,520
Speaker 6: licenses that we're not actually can make some data available

1106
01:00:45,600 --> 01:00:50,640
Speaker 6: right because we do have API access to various platforms,

1107
01:00:51,200 --> 01:00:53,720
Speaker 6: so you just can't give away their information.

1108
01:00:54,320 --> 01:00:56,680
Speaker 1: Okay, this is such a good point because and this

1109
01:00:56,760 --> 01:00:59,760
Speaker 1: is a frustration of mine. It used to be that

1110
01:01:00,120 --> 01:01:03,480
Speaker 1: if you wanted API access to X, right, that was

1111
01:01:03,480 --> 01:01:05,280
Speaker 1: not something that will be difficult to get. In twenty

1112
01:01:05,320 --> 01:01:08,320
Speaker 1: twenty five, things have really changed, and so yeah, I

1113
01:01:08,800 --> 01:01:11,200
Speaker 1: don't know if people who are not sort of in

1114
01:01:11,280 --> 01:01:14,920
Speaker 1: this world really know the ways that a lot of

1115
01:01:14,960 --> 01:01:17,479
Speaker 1: the Internet and how it works is has been turned

1116
01:01:17,520 --> 01:01:20,560
Speaker 1: into a black box, and so it's incredibly difficult to

1117
01:01:20,560 --> 01:01:23,120
Speaker 1: get that kind of access for most of us these days,

1118
01:01:23,200 --> 01:01:25,600
Speaker 1: and we simply, you know, just don't have it.

1119
01:01:26,000 --> 01:01:28,480
Speaker 5: And just to underscore that, you know, it's it's not

1120
01:01:28,560 --> 01:01:30,720
Speaker 5: a black box because it is so mysterious, no one

1121
01:01:30,720 --> 01:01:32,680
Speaker 5: could know it. It is a black box because the

1122
01:01:32,680 --> 01:01:35,520
Speaker 5: people who run the platforms have decided that they want

1123
01:01:35,520 --> 01:01:36,480
Speaker 5: the box to be black.

1124
01:01:37,680 --> 01:01:42,680
Speaker 6: Yeah, you know, monetizing the information that they have. And

1125
01:01:42,720 --> 01:01:46,400
Speaker 6: I think it's become even probably even worse now, you know,

1126
01:01:46,920 --> 01:01:49,760
Speaker 6: with training llms. You know, a good place to get

1127
01:01:49,880 --> 01:01:54,680
Speaker 6: information is from these social media platforms, which then you know,

1128
01:01:55,320 --> 01:01:58,880
Speaker 6: aren't they're not getting any value off of that. Those

1129
01:01:58,880 --> 01:02:01,320
Speaker 6: companies training off of their so I think it's kind

1130
01:02:01,320 --> 01:02:04,040
Speaker 6: of become a positive feedback loop.

1131
01:02:04,480 --> 01:02:07,600
Speaker 5: I was also really interested in the sort of temporal

1132
01:02:07,640 --> 01:02:10,520
Speaker 5: analysis that you guys did, looking at not just the

1133
01:02:10,560 --> 01:02:13,400
Speaker 5: mix of the types of accounts, but how it changed

1134
01:02:13,440 --> 01:02:17,240
Speaker 5: over time pretty rapidly in the you know, immediate days

1135
01:02:17,440 --> 01:02:20,840
Speaker 5: after you know, the beginning of the study period, which

1136
01:02:21,240 --> 01:02:23,640
Speaker 5: I think was like the day after the album drop.

1137
01:02:23,760 --> 01:02:26,640
Speaker 5: Can you talk a little bit about what that temporal

1138
01:02:26,720 --> 01:02:31,720
Speaker 5: analysis and how the shifting mix of actors what that

1139
01:02:31,880 --> 01:02:34,720
Speaker 5: tells us. It wouldn't be available if you just looked

1140
01:02:34,720 --> 01:02:36,240
Speaker 5: at a single snapshot in time.

1141
01:02:36,680 --> 01:02:39,480
Speaker 6: It's actually important to look at these as snapshot in time,

1142
01:02:39,600 --> 01:02:42,080
Speaker 6: so we can go down from like the microsecond all

1143
01:02:42,120 --> 01:02:44,480
Speaker 6: the way up to you know, centuries. We don't have

1144
01:02:44,520 --> 01:02:49,520
Speaker 6: that much data yet though, But the key here is,

1145
01:02:49,560 --> 01:02:52,920
Speaker 6: so let's say the Taylor Swift report, it's actually about

1146
01:02:52,920 --> 01:02:56,080
Speaker 6: three point seven percent of the total users were these

1147
01:02:56,160 --> 01:02:59,320
Speaker 6: inauthentic type users that contributed twenty eight percent of the

1148
01:02:59,360 --> 01:03:02,920
Speaker 6: total content. That's the big snapshot, but that doesn't tell

1149
01:03:02,960 --> 01:03:04,800
Speaker 6: you the whole picture when you then look at it

1150
01:03:04,880 --> 01:03:08,000
Speaker 6: how it happened over the course of hours, where that

1151
01:03:08,080 --> 01:03:11,800
Speaker 6: three point seven percent. At the first instance of this narrative,

1152
01:03:12,280 --> 01:03:15,720
Speaker 6: Spiking contributed, it was about fifteen to twenty percent of

1153
01:03:15,720 --> 01:03:18,080
Speaker 6: the total audience and contributed I think it was seventy

1154
01:03:18,120 --> 01:03:20,840
Speaker 6: eight percent of the total volume of right. So if

1155
01:03:20,880 --> 01:03:24,080
Speaker 6: you only look at one overall snapshot, you are going

1156
01:03:24,160 --> 01:03:26,280
Speaker 6: to you're not going to see the force from the trees.

1157
01:03:26,640 --> 01:03:29,840
Speaker 1: I'm sure you know that the reaction to this has

1158
01:03:29,880 --> 01:03:33,880
Speaker 1: been quite aware. Yeah, it's been big. I'm sure y'all

1159
01:03:33,880 --> 01:03:37,920
Speaker 1: have had a wild week and it's been really interesting

1160
01:03:37,960 --> 01:03:41,200
Speaker 1: and I think kind of weirdly telling to engage with

1161
01:03:41,320 --> 01:03:43,760
Speaker 1: some of the criticisms that people have made or like

1162
01:03:43,920 --> 01:03:47,040
Speaker 1: just responses that people have had, they've been like deeply

1163
01:03:47,080 --> 01:03:49,480
Speaker 1: emotional responses, I guess is how I'll put them. And

1164
01:03:49,960 --> 01:03:53,840
Speaker 1: one of the things I've seen i've seen the takeaways

1165
01:03:53,840 --> 01:03:57,640
Speaker 1: from the report. I don't want to say misrepresented, but

1166
01:03:59,080 --> 01:04:02,560
Speaker 1: I would summarize it as one camp being like, see,

1167
01:04:02,680 --> 01:04:04,960
Speaker 1: one hundred percent of the people who push this narrative

1168
01:04:05,000 --> 01:04:07,400
Speaker 1: were bots, and this was a narrative that we did

1169
01:04:07,440 --> 01:04:09,960
Speaker 1: not exist for real. If you if you bought this

1170
01:04:10,120 --> 01:04:12,320
Speaker 1: was real, you got taken. And then I saw people

1171
01:04:12,480 --> 01:04:15,200
Speaker 1: like black women on social media being like, I'm on

1172
01:04:15,280 --> 01:04:17,400
Speaker 1: a bot, I'm a real person. I felt x y

1173
01:04:17,480 --> 01:04:19,960
Speaker 1: Z about Taylor Swift. And what's interesting to me is

1174
01:04:20,000 --> 01:04:23,520
Speaker 1: that when you actually read the report, neither the report

1175
01:04:23,560 --> 01:04:25,640
Speaker 1: does not make either of those claims. And so I

1176
01:04:25,680 --> 01:04:28,360
Speaker 1: guess I wonder, you know, what do you think accounts

1177
01:04:28,360 --> 01:04:33,480
Speaker 1: for the fact that people are using the report to

1178
01:04:33,560 --> 01:04:36,880
Speaker 1: say something that I think the report patently did not suggest.

1179
01:04:37,640 --> 01:04:42,480
Speaker 6: Yeah, we've definitely noticed and had discussions about that too,

1180
01:04:42,480 --> 01:04:46,520
Speaker 6: where like, where are you guys getting this from our

1181
01:04:46,560 --> 01:04:49,240
Speaker 6: whole goal? So we're not the arbitras of truth, right,

1182
01:04:49,320 --> 01:04:51,600
Speaker 6: that is not what we do as a company. What

1183
01:04:51,640 --> 01:04:55,080
Speaker 6: we're trying to do is tell you who and what

1184
01:04:55,480 --> 01:04:59,280
Speaker 6: and why is that narrative happening. That's what we're trying

1185
01:04:59,280 --> 01:05:01,080
Speaker 6: to do, is like how did that come to be?

1186
01:05:01,680 --> 01:05:05,600
Speaker 6: And so, yeah, you know, in this instance, even though

1187
01:05:05,600 --> 01:05:09,320
Speaker 6: there's a high percentage of nontypical actors, there were still

1188
01:05:09,360 --> 01:05:11,680
Speaker 6: normal people that did engage with it. So we're not

1189
01:05:11,840 --> 01:05:15,320
Speaker 6: discounting that there was some genuine engagement at the start.

1190
01:05:15,600 --> 01:05:21,160
Speaker 6: It's just that that first layer came from non typical users, right,

1191
01:05:21,240 --> 01:05:23,280
Speaker 6: and then people got brought into it, and then it

1192
01:05:23,360 --> 01:05:26,440
Speaker 6: morphed and as it grew over the course of the

1193
01:05:26,800 --> 01:05:27,760
Speaker 6: couple days.

1194
01:05:28,480 --> 01:05:30,480
Speaker 1: And I feel like that is part and parcel of

1195
01:05:30,560 --> 01:05:35,000
Speaker 1: these online manipulation campaigns, where the point is to get real,

1196
01:05:35,320 --> 01:05:39,400
Speaker 1: authentic people talking about stuff that otherwise they probably wouldn't

1197
01:05:39,440 --> 01:05:43,400
Speaker 1: be talking about. And so this is not inauthentic discourse

1198
01:05:43,480 --> 01:05:48,080
Speaker 1: from bots. I've personally gotten myself pulled into conversations around

1199
01:05:48,120 --> 01:05:51,000
Speaker 1: like why am I all caps rage tweeting about cracker

1200
01:05:51,040 --> 01:05:53,440
Speaker 1: Barrel right now, a restaurant I have not eaten at

1201
01:05:53,480 --> 01:05:56,160
Speaker 1: in twenty years, Like the ways that they can get

1202
01:05:56,200 --> 01:05:57,600
Speaker 1: you to pull in.

1203
01:05:57,520 --> 01:06:00,800
Speaker 6: And engage, like that's the point that is, Yeah, nail

1204
01:06:00,840 --> 01:06:03,520
Speaker 6: on the head, that literally is the point they want

1205
01:06:03,560 --> 01:06:06,160
Speaker 6: you to get engaged with that content.

1206
01:06:06,360 --> 01:06:09,800
Speaker 1: The report makes it clear that there was overlap between

1207
01:06:09,880 --> 01:06:13,800
Speaker 1: some of this Taylor Swift and authentic behavior and Blake

1208
01:06:13,920 --> 01:06:17,120
Speaker 1: Lively what's going on there? Like like, what's the what

1209
01:06:17,160 --> 01:06:18,960
Speaker 1: are the implications for that? Yeah?

1210
01:06:19,040 --> 01:06:20,640
Speaker 6: I would say highly suspicious?

1211
01:06:23,280 --> 01:06:25,600
Speaker 1: Where is us? You know?

1212
01:06:25,840 --> 01:06:30,160
Speaker 6: Especially that far apart right?

1213
01:06:30,520 --> 01:06:30,640
Speaker 4: Uh?

1214
01:06:30,720 --> 01:06:34,280
Speaker 6: And the same classification like the facilitator accounts that those

1215
01:06:34,320 --> 01:06:36,880
Speaker 6: are the ones that typically posted high volumes in short

1216
01:06:37,000 --> 01:06:40,240
Speaker 6: and short stints. The fact that there were so many

1217
01:06:40,360 --> 01:06:44,280
Speaker 6: overlapping that far apart tells me that that was more

1218
01:06:44,320 --> 01:06:46,640
Speaker 6: of a coordinated uh activity.

1219
01:06:47,400 --> 01:06:49,440
Speaker 1: I mean that kind of takes me back to a

1220
01:06:49,800 --> 01:06:55,320
Speaker 1: stepping back question. Why would someone be invested in manipulating

1221
01:06:55,320 --> 01:06:58,360
Speaker 1: the conversation around Taylor Swift on the internet like like

1222
01:06:59,080 --> 01:06:59,560
Speaker 1: and why?

1223
01:07:01,960 --> 01:07:04,280
Speaker 6: So we've also talked about that internally and trying to

1224
01:07:04,400 --> 01:07:06,880
Speaker 6: you know, so we work with brands and we see

1225
01:07:06,880 --> 01:07:10,240
Speaker 6: this all the time where you know, either corporate espionage

1226
01:07:10,560 --> 01:07:17,840
Speaker 6: or you know, targeted campaigns to hurt market relevance on

1227
01:07:17,920 --> 01:07:22,360
Speaker 6: the Taylor Swift side and play clively either. I guess

1228
01:07:22,400 --> 01:07:28,360
Speaker 6: we have two theories. One testing right, So Taylor Swift

1229
01:07:28,400 --> 01:07:31,120
Speaker 6: is a huge brand. She drives economies. You know, you

1230
01:07:31,120 --> 01:07:34,720
Speaker 6: can almost say that she's a political figure. If you

1231
01:07:34,760 --> 01:07:39,160
Speaker 6: can impact how Swift eas or that conversation online around her,

1232
01:07:39,920 --> 01:07:42,320
Speaker 6: that means those strategies and tactics would do then work

1233
01:07:42,360 --> 01:07:45,720
Speaker 6: for others. And then the other one is around the

1234
01:07:45,880 --> 01:07:49,240
Speaker 6: economic Uh. You know again, she drives economies, so if

1235
01:07:49,240 --> 01:07:53,040
Speaker 6: you can hurt her reputation that allows others to fill

1236
01:07:53,080 --> 01:07:53,560
Speaker 6: that void.

1237
01:07:54,120 --> 01:07:55,400
Speaker 1: Was that meant to be a bit of a tailor

1238
01:07:55,440 --> 01:08:00,439
Speaker 1: Swift pun? Doesn't she have an album called reputation I think,

1239
01:08:00,480 --> 01:08:14,280
Speaker 1: don't quote me on that more. After a quick break,

1240
01:08:18,160 --> 01:08:22,439
Speaker 1: let's get right back into it. How did the report

1241
01:08:22,600 --> 01:08:24,840
Speaker 1: end up in being covered in Rolling Stone?

1242
01:08:25,000 --> 01:08:27,360
Speaker 6: We went out and we shopped it around and Rolling

1243
01:08:27,360 --> 01:08:30,200
Speaker 6: Stones was interested in writing about it. Like again, we

1244
01:08:30,200 --> 01:08:34,160
Speaker 6: were also equally giddy about that one. We did not

1245
01:08:34,320 --> 01:08:36,879
Speaker 6: anticipate such a reaction.

1246
01:08:37,040 --> 01:08:39,760
Speaker 1: I'm gonna be honest, Yeah, I mean the reaction has

1247
01:08:39,760 --> 01:08:44,280
Speaker 1: been absolutely wild. And I understand that like companies like yours,

1248
01:08:44,720 --> 01:08:48,320
Speaker 1: part of publishing studies like this, like especially like flashy ones,

1249
01:08:48,320 --> 01:08:50,360
Speaker 1: ones that people are gonna actually you know, want to

1250
01:08:50,400 --> 01:08:54,519
Speaker 1: be reading. Part of it is like getting publicity for

1251
01:08:54,560 --> 01:08:56,680
Speaker 1: the work that they were able to do, and so like,

1252
01:08:57,200 --> 01:09:01,080
Speaker 1: to that end, do you consider this to be a success, Like,

1253
01:09:01,439 --> 01:09:04,120
Speaker 1: I don't know the last time at a report. I mean,

1254
01:09:04,360 --> 01:09:07,920
Speaker 1: we look internally, we're reading reports about the Internet all

1255
01:09:07,960 --> 01:09:12,000
Speaker 1: the time. Typically my cousin is not texting me about them.

1256
01:09:12,040 --> 01:09:13,200
Speaker 1: You know what I'm you know what I'm saying.

1257
01:09:14,160 --> 01:09:20,040
Speaker 6: Yeah, So yes, definitely would consider it a success. To

1258
01:09:20,080 --> 01:09:22,360
Speaker 6: that point, I've had family that I haven't spoken to

1259
01:09:22,439 --> 01:09:28,439
Speaker 6: in years and we're like, oh my god, Yeah, it

1260
01:09:28,520 --> 01:09:31,280
Speaker 6: was crazy, but you know it was I guess to

1261
01:09:31,320 --> 01:09:34,240
Speaker 6: the point, we were just trying to demonstrate the ability

1262
01:09:34,280 --> 01:09:37,000
Speaker 6: that we have, right, that was really what we wanted

1263
01:09:37,000 --> 01:09:40,240
Speaker 6: to show people what is happening online and how we

1264
01:09:40,280 --> 01:09:40,719
Speaker 6: can help.

1265
01:09:41,120 --> 01:09:43,920
Speaker 5: I mean, one possible takeaway here is that, like there

1266
01:09:44,000 --> 01:09:47,000
Speaker 5: is a lot of appetite among the public and demand

1267
01:09:47,040 --> 01:09:48,760
Speaker 5: for this kind of information.

1268
01:09:49,800 --> 01:09:52,000
Speaker 6: Yeah, that's uh, I was kind of kind of we're

1269
01:09:52,040 --> 01:09:54,639
Speaker 6: thinking that too. And then what else could we look

1270
01:09:54,680 --> 01:09:56,240
Speaker 6: into that could be helpful for people?

1271
01:09:56,640 --> 01:09:59,679
Speaker 1: Can you give us a sense of when somebody gets

1272
01:09:59,680 --> 01:10:04,840
Speaker 1: on the internet, how much conversation online is perhaps being

1273
01:10:04,960 --> 01:10:09,160
Speaker 1: impacted by inauthentic behavior, because you know, from Cracker Barrel

1274
01:10:09,200 --> 01:10:12,360
Speaker 1: to Blake Lively the Taylor Swift, it seems like these

1275
01:10:12,400 --> 01:10:15,920
Speaker 1: conversations that you might have thought of as innocuous are

1276
01:10:16,000 --> 01:10:19,040
Speaker 1: now actually being manipulated by inauthentic actors.

1277
01:10:19,320 --> 01:10:23,880
Speaker 6: Yeah, so my whole quote the Internet is fake, it

1278
01:10:24,120 --> 01:10:30,120
Speaker 6: kind of really is. So since we've been doing this,

1279
01:10:30,479 --> 01:10:32,760
Speaker 6: we have yet to find a narrative that didn't have

1280
01:10:32,880 --> 01:10:37,080
Speaker 6: ood activity or inauthentic activity, Like, not a single one.

1281
01:10:37,280 --> 01:10:40,400
Speaker 1: I mean, it's sort of I I mean, I guess

1282
01:10:40,560 --> 01:10:42,519
Speaker 1: you when I read that quote the Internet is fake,

1283
01:10:42,600 --> 01:10:43,519
Speaker 1: I didn't realize you.

1284
01:10:43,479 --> 01:10:47,280
Speaker 6: Meant it quite so literally, it's just kind of scary.

1285
01:10:47,400 --> 01:10:50,200
Speaker 6: Like we you know, we were a little bright eyed

1286
01:10:50,200 --> 01:10:52,280
Speaker 6: and bushy tailed when we got into this.

1287
01:10:52,840 --> 01:10:52,920
Speaker 4: Uh.

1288
01:10:53,040 --> 01:10:55,760
Speaker 6: And then over the course of you know, doing this work,

1289
01:10:55,760 --> 01:10:59,000
Speaker 6: we've seen just how much inauthentic activity there really is

1290
01:10:59,040 --> 01:11:00,840
Speaker 6: happening online.

1291
01:11:00,920 --> 01:11:05,160
Speaker 1: How are we meant to use our social media platforms

1292
01:11:05,240 --> 01:11:08,000
Speaker 1: and platforms for discourse? How are we meant to use

1293
01:11:08,040 --> 01:11:11,559
Speaker 1: them effectively? If that's the case, Like, can they be

1294
01:11:11,640 --> 01:11:15,040
Speaker 1: used effectively anymore? That it's a really hard question too.

1295
01:11:15,640 --> 01:11:19,599
Speaker 1: And so if I could wave a magic wand the like,

1296
01:11:19,840 --> 01:11:24,760
Speaker 1: how I would think or fix this? So humans have

1297
01:11:24,800 --> 01:11:27,320
Speaker 1: had thousands of years to figure out the etiquette of

1298
01:11:27,360 --> 01:11:30,600
Speaker 1: behaving with each other in real life, right, there's so

1299
01:11:30,720 --> 01:11:33,479
Speaker 1: many different norms that we have that have just come

1300
01:11:33,560 --> 01:11:37,679
Speaker 1: from those centuries. The internet, like in the current state

1301
01:11:37,960 --> 01:11:41,160
Speaker 1: that we use it, maybe twenty years, twenty five, right,

1302
01:11:41,640 --> 01:11:44,760
Speaker 1: My gut is that we probably need to figure out

1303
01:11:44,800 --> 01:11:48,880
Speaker 1: what the etiquette is when interacting with people online because

1304
01:11:49,439 --> 01:11:51,479
Speaker 1: you know, right now there is none you just just

1305
01:11:51,479 --> 01:11:54,240
Speaker 1: the wild West. Yeah, the wild West is a good

1306
01:11:54,280 --> 01:11:57,720
Speaker 1: way to describe it, and especially for these conversations that

1307
01:11:58,040 --> 01:12:03,400
Speaker 1: frankly I think even fifteen years ago did not feel

1308
01:12:03,520 --> 01:12:06,840
Speaker 1: this heated online. It's incredibly difficult to have a conversation,

1309
01:12:06,960 --> 01:12:09,640
Speaker 1: even a conversation is about something as simple and Innocubus

1310
01:12:09,680 --> 01:12:12,800
Speaker 1: has an album, a movie, something like that without it

1311
01:12:12,920 --> 01:12:16,680
Speaker 1: feeling the triolic. And I have to imagine this inauthentic

1312
01:12:16,720 --> 01:12:18,840
Speaker 1: activity is adding to that.

1313
01:12:19,560 --> 01:12:23,479
Speaker 6: Oh yeah, absolutely Again, that type of content gets the

1314
01:12:23,520 --> 01:12:27,920
Speaker 6: most clicks, so it gets lifted up the fastest. It's

1315
01:12:27,920 --> 01:12:31,120
Speaker 6: actually been really really interesting and meta for us to

1316
01:12:31,160 --> 01:12:36,519
Speaker 6: watch us become the conspiracy that is. Yeah, yeah, that

1317
01:12:36,600 --> 01:12:38,200
Speaker 6: one's going to be a fun white paper. We're going

1318
01:12:38,280 --> 01:12:40,360
Speaker 6: to do in ourselves now, Oh my.

1319
01:12:40,320 --> 01:12:43,040
Speaker 1: God, you absolutely should. Well we'll have you back. And

1320
01:12:43,320 --> 01:12:45,360
Speaker 1: I'm guess something. When I was preparing for this, I

1321
01:12:45,400 --> 01:12:47,320
Speaker 1: was going through social media and I was trying to

1322
01:12:47,320 --> 01:12:50,519
Speaker 1: pull out some of the I guess criticisms. Maybe isn't

1323
01:12:50,520 --> 01:12:52,640
Speaker 1: the right word, but some of the things people have

1324
01:12:52,760 --> 01:12:55,920
Speaker 1: been saying about the report about your company, some of

1325
01:12:55,960 --> 01:12:58,240
Speaker 1: them for my own background, I know to be incorrect, right,

1326
01:12:58,320 --> 01:13:00,719
Speaker 1: like saying, oh, this is just PR Taylor's that probably

1327
01:13:00,800 --> 01:13:02,559
Speaker 1: paid to have them print this, And I'm like, well,

1328
01:13:02,560 --> 01:13:05,600
Speaker 1: it's not really what these companies do and I guess

1329
01:13:06,200 --> 01:13:09,160
Speaker 1: how can I even ask this. Let's say it like,

1330
01:13:09,240 --> 01:13:13,760
Speaker 1: I understand why and how a lot of people who

1331
01:13:13,800 --> 01:13:17,519
Speaker 1: are saying, hey, I'm not about I feel offended and

1332
01:13:17,720 --> 01:13:21,479
Speaker 1: unseen and erased by a report that makes me feel

1333
01:13:21,520 --> 01:13:24,080
Speaker 1: like my voice isn't a real voice. Right, that's not

1334
01:13:24,080 --> 01:13:26,200
Speaker 1: what the report said, But I get how that's the

1335
01:13:26,200 --> 01:13:30,519
Speaker 1: conclusions that they're coming to. My thing is this, If

1336
01:13:30,520 --> 01:13:33,680
Speaker 1: you are someone who really can't stand Taylor Swift, you

1337
01:13:33,960 --> 01:13:36,040
Speaker 1: think that Taylor Swift is a literal Nazi. Let's say

1338
01:13:36,080 --> 01:13:38,400
Speaker 1: that for the sake of argument, I would think that

1339
01:13:38,439 --> 01:13:41,800
Speaker 1: you would want reports like this to make clear how

1340
01:13:41,920 --> 01:13:46,240
Speaker 1: difficult it is for your message to break through online,

1341
01:13:46,240 --> 01:13:48,920
Speaker 1: how much that message, even if that's a message that

1342
01:13:48,920 --> 01:13:52,840
Speaker 1: you authentically feel, how easily exploited it is, and the

1343
01:13:52,840 --> 01:13:55,360
Speaker 1: fact that it is being disrupted. And so I would

1344
01:13:55,400 --> 01:13:58,960
Speaker 1: imagine even people who want to use internet platforms to

1345
01:13:59,640 --> 01:14:04,839
Speaker 1: authentically engaged in like critical discourse, they, more than anybody,

1346
01:14:04,880 --> 01:14:07,720
Speaker 1: should want to have a media landscape where that is

1347
01:14:07,960 --> 01:14:12,000
Speaker 1: possible without this kind of inauthentic interference and manipulation.

1348
01:14:12,479 --> 01:14:16,160
Speaker 6: No, I completely agree, not only that, but those individuals

1349
01:14:16,240 --> 01:14:18,679
Speaker 6: that do have those true feelings and you know, want

1350
01:14:18,720 --> 01:14:21,559
Speaker 6: to use it as a platform to voice those feelings

1351
01:14:22,080 --> 01:14:27,480
Speaker 6: often get taken advantage of by the inauthentic activity accounts.

1352
01:14:27,760 --> 01:14:31,480
Speaker 6: So a very very common thing is they're not creating

1353
01:14:31,640 --> 01:14:33,599
Speaker 6: the narrative that they want to push. They're pulling it

1354
01:14:33,640 --> 01:14:36,720
Speaker 6: from these small populations that they're like, oh, that one,

1355
01:14:36,840 --> 01:14:39,320
Speaker 6: that one will really make people mad, and then they

1356
01:14:39,400 --> 01:14:42,959
Speaker 6: push it inauthentically. So essentially they're actually getting taken advantage

1357
01:14:42,960 --> 01:14:44,480
Speaker 6: of by these accounts.

1358
01:14:45,120 --> 01:14:48,360
Speaker 1: I've actually seen this, or at least suspected that I've

1359
01:14:48,400 --> 01:14:50,760
Speaker 1: been a target at this kind of thing, because you

1360
01:14:50,800 --> 01:14:53,759
Speaker 1: could just be being a garden variety hater on the internet,

1361
01:14:53,840 --> 01:14:55,320
Speaker 1: you know, like oh I didn't like this, I didn't

1362
01:14:55,320 --> 01:14:57,720
Speaker 1: like that. Then you're going to comment from someone who's like, yeah,

1363
01:14:57,760 --> 01:15:00,240
Speaker 1: let's boycott it, or like yeah, let's like get to

1364
01:15:00,320 --> 01:15:02,000
Speaker 1: a level where you're like, well, I was just trying

1365
01:15:02,000 --> 01:15:05,000
Speaker 1: to engage in a little low level snark. I didn't

1366
01:15:05,040 --> 01:15:07,559
Speaker 1: mean it like this. But if you're already sort of

1367
01:15:07,680 --> 01:15:12,320
Speaker 1: riled up and you're not necessarily thinking super critically about it,

1368
01:15:11,360 --> 01:15:14,479
Speaker 1: it is this difficult to not engage. I guess that's

1369
01:15:14,520 --> 01:15:15,000
Speaker 1: what I'm saying.

1370
01:15:15,040 --> 01:15:19,320
Speaker 6: No exactly, and Again, that's the point. It's really hard

1371
01:15:19,520 --> 01:15:23,000
Speaker 6: not to engage with this when either you're really do

1372
01:15:23,200 --> 01:15:25,640
Speaker 6: believe in it or you're really really opposed to it.

1373
01:15:25,680 --> 01:15:27,920
Speaker 6: You want to say something that's just human nature.

1374
01:15:28,479 --> 01:15:31,280
Speaker 1: Do you see any solutions to this? Is there something

1375
01:15:31,280 --> 01:15:34,360
Speaker 1: that I mean, I don't. It's like like what, like,

1376
01:15:34,439 --> 01:15:36,599
Speaker 1: is there something that platforms could be doing that they're

1377
01:15:36,640 --> 01:15:39,240
Speaker 1: not or you know, shouldn't be doing what they're currently doing.

1378
01:15:39,760 --> 01:15:44,360
Speaker 6: Yeah, and that's a really tough question that I don't

1379
01:15:44,400 --> 01:15:47,360
Speaker 6: know if we're even equipped to say, like what the

1380
01:15:47,400 --> 01:15:50,559
Speaker 6: solution would be here. There are certain things that are

1381
01:15:50,600 --> 01:15:53,880
Speaker 6: really clear indicators that somebody is doing something to faries,

1382
01:15:54,680 --> 01:15:58,320
Speaker 6: Like people that are really trying to push illicit content

1383
01:15:58,400 --> 01:16:01,679
Speaker 6: will change their handle us a lot over the course

1384
01:16:01,720 --> 01:16:04,920
Speaker 6: of just a short time span, Right, how many times

1385
01:16:04,920 --> 01:16:10,240
Speaker 6: did you change your handle like whatever? Rarely people write

1386
01:16:10,240 --> 01:16:12,519
Speaker 6: that's and then these accounts are doing it, you know,

1387
01:16:12,600 --> 01:16:17,320
Speaker 6: three times a day, Like you know, there are there's

1388
01:16:17,760 --> 01:16:21,000
Speaker 6: key indicators that they could be looking for that would

1389
01:16:21,080 --> 01:16:25,320
Speaker 6: allow them to mitigate some of this uh and authentic activity.

1390
01:16:25,640 --> 01:16:27,519
Speaker 1: My favorite thing is when I see an account that

1391
01:16:27,600 --> 01:16:30,840
Speaker 1: says it has a that the images of like a

1392
01:16:30,880 --> 01:16:33,120
Speaker 1: white person and then the post will be like, well

1393
01:16:33,120 --> 01:16:35,120
Speaker 1: as a black woman, and I'm like, oh, did someone

1394
01:16:36,000 --> 01:16:38,679
Speaker 1: switch up their grift? What happened?

1395
01:16:39,560 --> 01:16:41,840
Speaker 6: Those are our favorite too. We passed those around where

1396
01:16:41,880 --> 01:16:45,080
Speaker 6: it's like, wow, you're your a LLLM model here really

1397
01:16:45,080 --> 01:16:45,760
Speaker 6: fails you.

1398
01:16:47,600 --> 01:16:51,679
Speaker 1: Yeah, so it's it's tough. I mean, it sounds because

1399
01:16:51,680 --> 01:16:54,479
Speaker 1: a lot platforms could be doing. But obviously I don't

1400
01:16:54,520 --> 01:16:56,920
Speaker 1: have a direct line to Elon Musk or anything. But

1401
01:16:57,400 --> 01:17:01,839
Speaker 1: what about individuals, like, while we are in the absence

1402
01:17:01,880 --> 01:17:04,840
Speaker 1: of platforms really doing what they can to crack down

1403
01:17:04,840 --> 01:17:07,320
Speaker 1: on this kind of thing? Do you have advice for people,

1404
01:17:08,000 --> 01:17:12,479
Speaker 1: especially as we navigate what feel like increasingly volatile times

1405
01:17:12,520 --> 01:17:16,920
Speaker 1: where conversation just reflecting the times like it just seems

1406
01:17:16,960 --> 01:17:19,800
Speaker 1: more volatile. Do you have advice for folks as they

1407
01:17:20,200 --> 01:17:24,720
Speaker 1: wade through that to not be impacted or manipulated by

1408
01:17:24,720 --> 01:17:26,280
Speaker 1: this kind of an authentic behavior.

1409
01:17:26,720 --> 01:17:32,720
Speaker 6: Yeah, first, like, take a breath, right, don't don't respond immediately.

1410
01:17:33,680 --> 01:17:37,519
Speaker 6: But then I think it's there. It is because people

1411
01:17:37,560 --> 01:17:39,080
Speaker 6: are still going to want to respond, So how do

1412
01:17:39,120 --> 01:17:39,439
Speaker 6: you do that?

1413
01:17:39,560 --> 01:17:39,680
Speaker 3: Right?

1414
01:17:39,680 --> 01:17:42,840
Speaker 6: How do you engage with this content without actually engaging

1415
01:17:42,840 --> 01:17:44,840
Speaker 6: the algorithm that then is going to boost it to

1416
01:17:44,880 --> 01:17:49,479
Speaker 6: even more eyes. So our goal is to try to

1417
01:17:49,520 --> 01:17:52,439
Speaker 6: help people understand how to navigate these narratives. And so

1418
01:17:52,520 --> 01:17:55,439
Speaker 6: in this instance, it's kind of like a three step process.

1419
01:17:55,560 --> 01:17:58,639
Speaker 6: So observe but don't interact, so you know, you can

1420
01:17:58,680 --> 01:18:01,000
Speaker 6: see the content, but don't like, get, don't share it,

1421
01:18:01,040 --> 01:18:04,920
Speaker 6: don't do anything that would boost the algorithm. Then if

1422
01:18:04,920 --> 01:18:09,280
Speaker 6: you want to counter, you post separately, but don't reply

1423
01:18:09,400 --> 01:18:11,559
Speaker 6: in the comments, right, so you can make your own

1424
01:18:11,600 --> 01:18:14,040
Speaker 6: post about it, but don't be replying to it in

1425
01:18:14,080 --> 01:18:16,640
Speaker 6: the comments, because again that's going to boost it algorithmically.

1426
01:18:17,120 --> 01:18:20,559
Speaker 6: And then try to redirect to a more positive narrative.

1427
01:18:20,720 --> 01:18:24,679
Speaker 6: So in your post, talk about you know, don't don't

1428
01:18:24,680 --> 01:18:27,200
Speaker 6: engage what they're talking about directly, try to redirect it

1429
01:18:27,200 --> 01:18:29,960
Speaker 6: into a more positive sense without using any of the

1430
01:18:29,960 --> 01:18:34,960
Speaker 6: same hashtags or the like of that original negative content.

1431
01:18:35,240 --> 01:18:37,720
Speaker 6: That way, you can be trying to change the conversation

1432
01:18:38,120 --> 01:18:40,720
Speaker 6: instead of boosting something algorithmically.

1433
01:18:41,600 --> 01:18:43,640
Speaker 1: What is next for you all? What's your do you

1434
01:18:43,680 --> 01:18:45,320
Speaker 1: have it? Can you give us a little preview about

1435
01:18:45,320 --> 01:18:49,840
Speaker 1: what the next big yeah and authentic conversation online might be?

1436
01:18:53,400 --> 01:18:55,880
Speaker 6: Sadly no, right, because you never know what's going to

1437
01:18:55,960 --> 01:18:58,360
Speaker 6: happen online. I mean, we definitely are going to do

1438
01:18:58,400 --> 01:19:01,120
Speaker 6: one on ourselves because and did that go crazy?

1439
01:19:01,600 --> 01:19:02,120
Speaker 4: Uh?

1440
01:19:02,800 --> 01:19:04,640
Speaker 6: But I don't know, like if we wanted to do

1441
01:19:04,720 --> 01:19:07,439
Speaker 6: another one, maybe k pop could be a fun one.

1442
01:19:08,360 --> 01:19:10,559
Speaker 1: Oh, we were just having a conversation with one of

1443
01:19:10,560 --> 01:19:15,839
Speaker 1: our producers, Joey, about just K pop fandom in general.

1444
01:19:15,920 --> 01:19:20,160
Speaker 1: I had no idea. I had no idea. Oh we have.

1445
01:19:20,479 --> 01:19:24,000
Speaker 6: We've worked with clients where like they're they're just their

1446
01:19:24,040 --> 01:19:28,080
Speaker 6: online activity have broken our systems. And so the K

1447
01:19:28,160 --> 01:19:31,120
Speaker 6: pop community is real and large.

1448
01:19:31,080 --> 01:19:33,200
Speaker 1: And it just goes to show exactly what you were

1449
01:19:33,200 --> 01:19:37,440
Speaker 1: talking about that these might sound like quote celebrity stories,

1450
01:19:37,479 --> 01:19:41,679
Speaker 1: but you know k pop, it's it's such a big

1451
01:19:41,800 --> 01:19:45,479
Speaker 1: fandom that it says so much about how not just

1452
01:19:45,520 --> 01:19:47,639
Speaker 1: celebrity and fandom and how we live, but like how

1453
01:19:47,680 --> 01:19:50,960
Speaker 1: we live our lives politically, class issues, gender issues, these

1454
01:19:51,000 --> 01:19:54,719
Speaker 1: are real things that really motivate people in our world.

1455
01:19:54,760 --> 01:19:59,200
Speaker 1: And so it's not just celebrities and fluff. It's conversations

1456
01:19:59,200 --> 01:20:00,639
Speaker 1: that have actual influence.

1457
01:20:01,640 --> 01:20:06,480
Speaker 6: When we see big fandoms getting involved with something normally

1458
01:20:06,520 --> 01:20:11,840
Speaker 6: it's actually in a positive sense, really lifting up a

1459
01:20:11,840 --> 01:20:15,679
Speaker 6: new album or saying how much they like something. Actually,

1460
01:20:15,760 --> 01:20:19,880
Speaker 6: the fandoms really do tend to be like I want

1461
01:20:19,920 --> 01:20:23,759
Speaker 6: to say, wholesome community, Like they share nice content.

1462
01:20:24,160 --> 01:20:26,800
Speaker 1: As someone who studies the Internet and this does spends

1463
01:20:26,800 --> 01:20:28,800
Speaker 1: a lot of time making reports about what's happening there,

1464
01:20:29,439 --> 01:20:31,960
Speaker 1: are you does it give you hope? Like this, like

1465
01:20:32,160 --> 01:20:35,680
Speaker 1: the conversation about fandoms and them making wholesome content. You

1466
01:20:35,760 --> 01:20:38,080
Speaker 1: had a little smile when you mentioned that talking about

1467
01:20:38,120 --> 01:20:40,280
Speaker 1: technology and the Internet, that like, these are things that

1468
01:20:40,320 --> 01:20:42,880
Speaker 1: we like, spaces that we like. I find myself hating

1469
01:20:42,920 --> 01:20:46,439
Speaker 1: on them and lifting the bad stuff. But do you

1470
01:20:46,479 --> 01:20:49,240
Speaker 1: feel hopeful and good about the Internet given all of this?

1471
01:20:49,880 --> 01:20:52,599
Speaker 6: No, I think there's still so much to be gained

1472
01:20:52,680 --> 01:20:58,120
Speaker 6: from ever increasing connections, right because in all of these

1473
01:20:58,120 --> 01:21:03,400
Speaker 6: social media platforms really were about connecting individuals together that

1474
01:21:03,640 --> 01:21:08,439
Speaker 6: had like interests. And I think there when it became

1475
01:21:08,520 --> 01:21:12,120
Speaker 6: like social media and it's now about media content that

1476
01:21:12,320 --> 01:21:15,080
Speaker 6: has driven a little bit more of the negativity. Uh,

1477
01:21:15,560 --> 01:21:17,320
Speaker 6: you know, it really did. At its core, it is

1478
01:21:17,360 --> 01:21:20,479
Speaker 6: about people engaging with communities that have like interest and

1479
01:21:20,520 --> 01:21:23,200
Speaker 6: I think that's great. Our whole goal is to help

1480
01:21:23,240 --> 01:21:26,360
Speaker 6: people have a better understanding of the world they live in,

1481
01:21:26,640 --> 01:21:30,519
Speaker 6: and especially that information environment where it's just really hard

1482
01:21:30,520 --> 01:21:31,439
Speaker 6: to know what's real or not.

1483
01:21:36,280 --> 01:21:38,280
Speaker 1: Got a story about an interesting thing in tech, I

1484
01:21:38,400 --> 01:21:40,240
Speaker 1: just want to say hi. You can be just at

1485
01:21:40,240 --> 01:21:42,960
Speaker 1: hello at tengodi dot com. You can also find transcripts

1486
01:21:42,960 --> 01:21:45,439
Speaker 1: for today's episode at tengody dot com. There Are No

1487
01:21:45,520 --> 01:21:48,120
Speaker 1: Girls on the Internet was created by me Brigittad. It's

1488
01:21:48,120 --> 01:21:51,439
Speaker 1: a production of iHeartRadio and Unbossed. Creative Jonathan Strickland is

1489
01:21:51,479 --> 01:21:54,879
Speaker 1: our executive producer. Terry Harrison is our producer and sound engineer.

1490
01:21:55,320 --> 01:21:58,480
Speaker 1: Michael Amata is our contributing producer. I'm your host, Bridgeitad.

1491
01:21:59,400 --> 01:22:01,120
Speaker 1: If you want to help, let's grow, rate and review

1492
01:22:01,200 --> 01:22:04,800
Speaker 1: us on Apple Podcasts. For more podcasts from iHeartRadio, check

1493
01:22:04,800 --> 01:22:07,160
Speaker 1: out the iHeartRadio app, Apple Podcasts, or wherever you get

1494
01:22:07,160 --> 01:22:07,920
Speaker 1: your podcasts.