1
00:00:04,160 --> 00:00:05,880
Speaker 1: There Are No Girls on the Internet, as a production

2
00:00:05,920 --> 00:00:13,440
Speaker 1: of iHeartRadio and Unbossed Creative. I'm bridget Ton, and this

3
00:00:13,520 --> 00:00:14,520
Speaker 1: is There Are No Girls.

4
00:00:14,280 --> 00:00:14,880
Speaker 2: On the Internet.

5
00:00:18,200 --> 00:00:19,280
Speaker 3: Happy New Year, y'all.

6
00:00:19,320 --> 00:00:20,720
Speaker 2: Welcome to our first.

7
00:00:20,680 --> 00:00:22,919
Speaker 1: There Are No Girls on the Internet. Recorded in the

8
00:00:22,960 --> 00:00:25,920
Speaker 1: New Year, twenty twenty six. Producer Mike, Happy New Year,

9
00:00:26,000 --> 00:00:26,800
Speaker 1: Thank you for being here.

10
00:00:27,680 --> 00:00:28,400
Speaker 2: Happy New Year.

11
00:00:28,480 --> 00:00:33,320
Speaker 4: Bridget We made it, got through twenty twenty five. I

12
00:00:33,360 --> 00:00:35,920
Speaker 4: wasn't confident we were gonna make it. It was getting

13
00:00:35,960 --> 00:00:36,800
Speaker 4: down to the wire.

14
00:00:37,040 --> 00:00:38,839
Speaker 1: You know this about me, But I know a lot

15
00:00:38,880 --> 00:00:41,400
Speaker 1: of people have strong feelings about.

16
00:00:41,120 --> 00:00:44,360
Speaker 3: The New Year and New Year's Eve. I actually really

17
00:00:44,479 --> 00:00:45,239
Speaker 3: like New Year's Eve.

18
00:00:45,280 --> 00:00:47,680
Speaker 1: It's one of the only holidays that I can generally

19
00:00:47,720 --> 00:00:51,360
Speaker 1: get behind, so I always try to make it a

20
00:00:51,520 --> 00:00:54,120
Speaker 1: thing that I'm going to have a good New Year's

21
00:00:54,120 --> 00:00:56,400
Speaker 1: I'm gonna go I'm gonna do something fun for New

22
00:00:56,480 --> 00:00:59,080
Speaker 1: Year's Eve, and I'm gonna spend New Year's Day watching

23
00:00:59,120 --> 00:01:01,040
Speaker 1: Twilight Zone reruns on the.

24
00:01:01,000 --> 00:01:02,360
Speaker 3: Sci Fi Channel. Uh.

25
00:01:02,720 --> 00:01:05,160
Speaker 1: Mission accomplished for twenty twenty six. I'm happy to say

26
00:01:05,920 --> 00:01:07,080
Speaker 1: all this talk about.

27
00:01:06,880 --> 00:01:07,760
Speaker 3: The Twilight Zone.

28
00:01:08,000 --> 00:01:13,000
Speaker 1: Unfortunately, that is a segue into what I'm sad to say,

29
00:01:13,120 --> 00:01:16,720
Speaker 1: is like our first conversation of the new year. As

30
00:01:16,880 --> 00:01:18,840
Speaker 1: I know, you know, as you all know by now,

31
00:01:19,040 --> 00:01:23,679
Speaker 1: Rene Good, poet mother wife, was shot and killed by

32
00:01:23,720 --> 00:01:25,560
Speaker 1: ice agents in Minnesota this week.

33
00:01:26,200 --> 00:01:28,160
Speaker 3: The video is horrifying.

34
00:01:28,720 --> 00:01:32,720
Speaker 1: The response on the administration has been also predictably pretty horrifying.

35
00:01:32,760 --> 00:01:37,440
Speaker 1: The conversation online has been I guess, perhaps predictably a

36
00:01:37,720 --> 00:01:41,920
Speaker 1: wash and missing disinformation about Good herself, what happened to her,

37
00:01:42,520 --> 00:01:45,760
Speaker 1: much of it stemming with Trump and the administration himself.

38
00:01:45,880 --> 00:01:47,920
Speaker 3: So yeah, it's just bad.

39
00:01:48,280 --> 00:01:51,480
Speaker 1: And I have to say, this is one of those

40
00:01:51,880 --> 00:01:54,480
Speaker 1: stories that is like do you ever have those stories

41
00:01:54,480 --> 00:01:56,320
Speaker 1: that you see and you're just like, this is enough.

42
00:01:56,760 --> 00:01:59,600
Speaker 3: I I this has broken me. I am done.

43
00:02:00,120 --> 00:02:02,760
Speaker 4: Yeah, this was a really tough one or rough way

44
00:02:02,760 --> 00:02:04,760
Speaker 4: to start the year after being a little checked out

45
00:02:04,800 --> 00:02:06,320
Speaker 4: from the news for a few days.

46
00:02:06,560 --> 00:02:07,720
Speaker 3: I completely agree.

47
00:02:07,760 --> 00:02:10,160
Speaker 1: And I have to say that, you know, the administration

48
00:02:10,400 --> 00:02:14,600
Speaker 1: lying about a dead woman really just moments after she's

49
00:02:14,639 --> 00:02:20,760
Speaker 1: been killed, predictable, them spreading lies online about it, predictable,

50
00:02:20,919 --> 00:02:24,280
Speaker 1: the video being horrible, predictable, but something that I did

51
00:02:24,320 --> 00:02:27,040
Speaker 1: not predict. And I have to give a pretty big

52
00:02:27,080 --> 00:02:30,280
Speaker 1: trigger warning about this. It's just a rough conversation. There's

53
00:02:30,320 --> 00:02:32,880
Speaker 1: no other way to put it. I did not predict

54
00:02:33,080 --> 00:02:36,640
Speaker 1: that a screenshot of the video of Good being shot

55
00:02:36,639 --> 00:02:40,280
Speaker 1: and killed an image of her dead body would be

56
00:02:40,440 --> 00:02:44,239
Speaker 1: on X, you know, one of our largest social media platforms,

57
00:02:44,919 --> 00:02:49,280
Speaker 1: with creeps under it asking for Groc to generate an

58
00:02:49,320 --> 00:02:53,120
Speaker 1: AI version of that image that puts her dead body

59
00:02:53,240 --> 00:02:58,200
Speaker 1: in a bikini. And Groc complied and generated an image

60
00:02:58,280 --> 00:03:02,320
Speaker 1: of a dead woman covered in blood in an AI

61
00:03:02,639 --> 00:03:06,080
Speaker 1: generated bikini. And I want to I just I want

62
00:03:06,080 --> 00:03:09,519
Speaker 1: to pause, because I think that that just really says

63
00:03:09,560 --> 00:03:12,280
Speaker 1: a lot about where we're at right now and how

64
00:03:12,280 --> 00:03:13,680
Speaker 1: we're all feeling, don't you think.

65
00:03:14,000 --> 00:03:20,359
Speaker 4: Yeah, it's a pretty gruesome juxtaposition, and it does feel.

66
00:03:21,639 --> 00:03:24,480
Speaker 2: Kind of like a.

67
00:03:23,440 --> 00:03:28,960
Speaker 4: Representative symbol of the lack of compassion and decency that

68
00:03:29,800 --> 00:03:31,040
Speaker 4: characterizes our moment.

69
00:03:31,960 --> 00:03:35,640
Speaker 1: And I'm sorry to say this is not an isolated

70
00:03:35,720 --> 00:03:39,480
Speaker 1: thing whatsoever, because when you search the phrase make her

71
00:03:39,840 --> 00:03:44,000
Speaker 1: or put her on X, you see how many creeps

72
00:03:44,040 --> 00:03:48,000
Speaker 1: are posting on X under images of women and children,

73
00:03:48,040 --> 00:03:51,080
Speaker 1: in some cases miners asking to put that person in

74
00:03:51,160 --> 00:03:54,760
Speaker 1: a bikini or other sexualized things, make her look pregnant,

75
00:03:54,760 --> 00:03:58,880
Speaker 1: make her look heavily pregnant or some otherwise sexually suggestive

76
00:03:58,920 --> 00:04:02,280
Speaker 1: thing or pose or outfit on their X feeds, Like

77
00:04:02,320 --> 00:04:05,200
Speaker 1: a user asked groc to undress an image of a

78
00:04:05,280 --> 00:04:07,720
Speaker 1: fourteen year old actress on the platform.

79
00:04:08,360 --> 00:04:09,240
Speaker 3: Pretty grim.

80
00:04:09,320 --> 00:04:11,720
Speaker 1: And I've seen a lot of reporting about this that

81
00:04:11,760 --> 00:04:16,000
Speaker 1: talks about it being you know, sexualized, non consensual AI

82
00:04:16,040 --> 00:04:19,320
Speaker 1: generated images, which it is. It sounds weird to say,

83
00:04:19,360 --> 00:04:21,880
Speaker 1: but I think it's important to note that it's not

84
00:04:22,400 --> 00:04:26,320
Speaker 1: just sexualized images that are non consensual of women and children,

85
00:04:26,440 --> 00:04:33,240
Speaker 1: but also, like that Rene Good photo, it is also dark, hateful, sick,

86
00:04:33,520 --> 00:04:36,159
Speaker 1: sexualized images of women and kids. Like it would be

87
00:04:36,200 --> 00:04:40,960
Speaker 1: bad enough, and if we were only talking about sexualized images,

88
00:04:41,480 --> 00:04:44,320
Speaker 1: it's not just that, right. It accepted a prompt to

89
00:04:44,360 --> 00:04:47,600
Speaker 1: add a swastika bikini to the photo of a Holocaust survivor,

90
00:04:47,920 --> 00:04:50,680
Speaker 1: and it also engaged with requests to represent women in

91
00:04:50,760 --> 00:04:53,359
Speaker 1: scenarios that look like they had been physically assaulted. So

92
00:04:54,160 --> 00:04:58,039
Speaker 1: it's not just sexualized images. It is, but it's also

93
00:04:58,640 --> 00:05:04,359
Speaker 1: like deeply just disturbed images of women and children. Sounds

94
00:05:04,440 --> 00:05:07,760
Speaker 1: like X, it's X, but it also goes beyond what's

95
00:05:07,800 --> 00:05:10,880
Speaker 1: happening on X, because what we're seeing on the platform.

96
00:05:11,040 --> 00:05:13,680
Speaker 1: X is just part of it because Grock is available

97
00:05:13,720 --> 00:05:16,560
Speaker 1: on X, but it's also a standalone app in the

98
00:05:16,600 --> 00:05:19,640
Speaker 1: App Store, And as the Indicator points out, that app

99
00:05:19,680 --> 00:05:22,560
Speaker 1: has also been used to generate sexual images of children

100
00:05:22,720 --> 00:05:25,760
Speaker 1: aged eleven to thirteen. This is according to information from

101
00:05:25,800 --> 00:05:30,800
Speaker 1: the Internet Watch Foundation. So for folks like me who

102
00:05:30,960 --> 00:05:34,159
Speaker 1: have basically stopped using X and maybe you're thinking, I've

103
00:05:34,160 --> 00:05:36,520
Speaker 1: not been there in a while, what's Groc? What's going

104
00:05:36,560 --> 00:05:40,680
Speaker 1: on with Groc? Well, Groc is X's AI chatbot, but

105
00:05:41,000 --> 00:05:43,240
Speaker 1: it's a little different than chatbots that you might be

106
00:05:43,360 --> 00:05:47,720
Speaker 1: used to, like claud notably with intention. So Groc has

107
00:05:47,800 --> 00:05:51,800
Speaker 1: been designed to sort of be an edge lord chatbot.

108
00:05:52,160 --> 00:05:56,000
Speaker 1: Because Elon Musk is a huge loser creep. Groc has

109
00:05:56,000 --> 00:06:00,200
Speaker 1: basically been designed by and for huge loser creeps. So

110
00:06:00,320 --> 00:06:03,240
Speaker 1: while Chatgypt and Claude obviously have their issues that we

111
00:06:03,279 --> 00:06:05,440
Speaker 1: talk about on the show all the time, Groc is.

112
00:06:05,520 --> 00:06:08,200
Speaker 3: Distinctive for being uniquely awful.

113
00:06:08,320 --> 00:06:10,200
Speaker 1: Remember there was a time where Elon Musk was trying

114
00:06:10,240 --> 00:06:13,039
Speaker 1: to tinker with Groc to make it less woke, and

115
00:06:13,080 --> 00:06:15,599
Speaker 1: at that time, Groc started declaring that it should be

116
00:06:15,640 --> 00:06:17,120
Speaker 1: called Mecha Hitler.

117
00:06:17,320 --> 00:06:18,840
Speaker 3: Remember that, and I want to.

118
00:06:18,760 --> 00:06:21,640
Speaker 1: Talk about this because I have been getting very frustrated

119
00:06:21,680 --> 00:06:25,560
Speaker 1: seeing these headlines that say things like Groc has apologized

120
00:06:25,640 --> 00:06:28,680
Speaker 1: for sexualizing images of young girls. And I want to

121
00:06:28,680 --> 00:06:33,360
Speaker 1: start there because Groc is Ex's chatbot. But guess what.

122
00:06:33,600 --> 00:06:36,440
Speaker 1: Grok is not sentient. Grek is not a person. Groc

123
00:06:36,480 --> 00:06:39,960
Speaker 1: doesn't do anything that humans do not tell Groc to

124
00:06:40,040 --> 00:06:44,200
Speaker 1: do or program Groc to do. Humans like Elon Musk

125
00:06:44,360 --> 00:06:47,120
Speaker 1: built Grock and run it as a commercial service, making

126
00:06:47,120 --> 00:06:49,839
Speaker 1: it available for other humans to use and do things with.

127
00:06:49,960 --> 00:06:52,680
Speaker 1: And right now those other humans are using Groc to

128
00:06:52,760 --> 00:06:56,560
Speaker 1: undress women and girls, to create sexualized images of children.

129
00:06:56,600 --> 00:07:00,400
Speaker 3: I e. What we call crimes criminal behavior.

130
00:07:05,160 --> 00:07:18,080
Speaker 2: Let's take a quick break at our.

131
00:07:18,000 --> 00:07:23,600
Speaker 4: Back, so bridget how did we get here?

132
00:07:24,160 --> 00:07:26,560
Speaker 1: So this has really hit a fever pitch right now

133
00:07:26,600 --> 00:07:30,360
Speaker 1: this week, but it's not really new, especially on X.

134
00:07:30,360 --> 00:07:31,880
Speaker 1: So I want to back up a little bit and

135
00:07:31,880 --> 00:07:33,720
Speaker 1: talk about sort of how we got to this moment.

136
00:07:34,200 --> 00:07:38,240
Speaker 1: We've got a few episodes about AI generated sexualized images

137
00:07:38,240 --> 00:07:41,440
Speaker 1: that took off on X. It happened to teen Marvel

138
00:07:41,520 --> 00:07:45,320
Speaker 1: star Social Gomez. Gomez was targeted with deep fake images

139
00:07:45,320 --> 00:07:48,600
Speaker 1: that also flooded X when this happened. She was only seventeen,

140
00:07:48,640 --> 00:07:52,160
Speaker 1: so she was a minor, and she's spoken about this publicly.

141
00:07:52,480 --> 00:07:55,160
Speaker 3: She was essentially told there was nothing that could be.

142
00:07:55,120 --> 00:07:57,040
Speaker 1: Done, that she just needed to make peace with it,

143
00:07:57,440 --> 00:07:59,960
Speaker 1: make whatever emotional peace with it that she could, because

144
00:08:00,320 --> 00:08:03,080
Speaker 1: nothing could be done about it. Also, folks might remember

145
00:08:03,120 --> 00:08:06,560
Speaker 1: that in January twenty twenty four, AI generated images of

146
00:08:06,760 --> 00:08:11,400
Speaker 1: Taylor Swift that originated on the platform Telegram, where Telegram

147
00:08:11,480 --> 00:08:13,120
Speaker 1: had been running a channel that was kind of like

148
00:08:13,160 --> 00:08:16,840
Speaker 1: a marketplace for celebrity deep fakes where users would request

149
00:08:16,880 --> 00:08:21,080
Speaker 1: and trade deep faked images of celebrities. Images of Taylor

150
00:08:21,160 --> 00:08:24,920
Speaker 1: Swift were created on that Telegram marketplace, and then those

151
00:08:24,960 --> 00:08:28,080
Speaker 1: images made their way to X where they really took off.

152
00:08:28,280 --> 00:08:31,880
Speaker 1: Telegram is sort of like a smaller niche alternative space,

153
00:08:31,920 --> 00:08:35,400
Speaker 1: and so you know, it was they were being traded there,

154
00:08:35,440 --> 00:08:37,640
Speaker 1: but when they hit X, which is arguably a more

155
00:08:37,679 --> 00:08:41,760
Speaker 1: mainstream online space, that's when they really took off. Back

156
00:08:41,800 --> 00:08:45,199
Speaker 1: when this happened, in my opinion, X did not really

157
00:08:45,240 --> 00:08:47,480
Speaker 1: handle this well, which I think we can understand as

158
00:08:47,480 --> 00:08:50,920
Speaker 1: a precursor for what's happening today. They initially tried to

159
00:08:50,960 --> 00:08:54,160
Speaker 1: manage it by just blocking people from being able to

160
00:08:54,200 --> 00:08:57,720
Speaker 1: search Taylor Swift's name. However, when you put quote marks

161
00:08:57,760 --> 00:09:00,040
Speaker 1: around Taylor Swift's name, it's.

162
00:09:00,160 --> 00:09:02,720
Speaker 3: You were able to search her just fine. So that

163
00:09:02,720 --> 00:09:03,760
Speaker 3: didn't even really work.

164
00:09:04,080 --> 00:09:08,319
Speaker 1: And notably, that solution doesn't offer any kind of protection

165
00:09:08,440 --> 00:09:10,559
Speaker 1: to a woman who's not named Taylor Swift.

166
00:09:10,640 --> 00:09:13,720
Speaker 3: Right, like that is that it didn't work.

167
00:09:13,840 --> 00:09:15,360
Speaker 1: But even let's say that it did work for the

168
00:09:15,400 --> 00:09:18,160
Speaker 1: sake of argument, the only person that that would conceivably

169
00:09:18,200 --> 00:09:21,280
Speaker 1: even work for is Taylor Swift. So certainly not any

170
00:09:21,360 --> 00:09:24,920
Speaker 1: kind of a meaningful solution to this like deep problem.

171
00:09:25,400 --> 00:09:28,600
Speaker 1: But the Taylor Swift situation was notable because back then

172
00:09:28,800 --> 00:09:32,600
Speaker 1: these images would typically originate elsewhere, like a more niche

173
00:09:32,600 --> 00:09:35,959
Speaker 1: alternative platform, and then make their way to X where

174
00:09:35,960 --> 00:09:40,440
Speaker 1: they would enjoy more reach and visibility. Grock essentially allowed

175
00:09:40,480 --> 00:09:44,120
Speaker 1: anybody to use natural language to do whatever they wanted

176
00:09:44,160 --> 00:09:46,360
Speaker 1: to images of women and girls on X.

177
00:09:46,920 --> 00:09:49,160
Speaker 3: So you've got the introduction of.

178
00:09:49,120 --> 00:09:52,360
Speaker 1: Grock plus as friend of the podcast, Cat ten Barge

179
00:09:52,400 --> 00:09:56,600
Speaker 1: points out in her deeply researched, deeply reported media outlet

180
00:09:56,679 --> 00:10:00,760
Speaker 1: Spitfire News, which everybody should be reading. Musk also dissolved

181
00:10:00,840 --> 00:10:04,160
Speaker 1: Twitter's trust and Safety Council when he took over at Twitter, right,

182
00:10:04,160 --> 00:10:06,920
Speaker 1: he fired eighty percent of the engineers who work on

183
00:10:07,000 --> 00:10:10,880
Speaker 1: issues regarding child exploitation. So again, that is a kind

184
00:10:10,920 --> 00:10:14,199
Speaker 1: of very specific choice that has led us to.

185
00:10:14,160 --> 00:10:15,200
Speaker 3: The moment that we're in now.

186
00:10:15,720 --> 00:10:18,760
Speaker 1: And yet none of this stop X from rolling out

187
00:10:18,840 --> 00:10:22,480
Speaker 1: what it called Spicy Mode on GROX mobile video generator,

188
00:10:22,720 --> 00:10:27,360
Speaker 1: which basically includes a feature to produce sexual content. So,

189
00:10:27,720 --> 00:10:29,920
Speaker 1: you know, think about that for a minute. Because you've

190
00:10:29,960 --> 00:10:32,320
Speaker 1: got a platform where the majority of people who work

191
00:10:32,360 --> 00:10:36,320
Speaker 1: on sexual exploitation issues have been fired. You've got an

192
00:10:36,360 --> 00:10:41,120
Speaker 1: issue already with users generating and spreading non consensual AI

193
00:10:41,200 --> 00:10:44,760
Speaker 1: generated images on your platform. You're not really doing the

194
00:10:44,920 --> 00:10:49,120
Speaker 1: job of dealing with it. And against that backdrop, you say,

195
00:10:49,200 --> 00:10:51,800
Speaker 1: you know what we're gonna do, We're gonna double down

196
00:10:52,200 --> 00:10:56,240
Speaker 1: on adult content generation where folks can make adult content

197
00:10:56,400 --> 00:10:57,439
Speaker 1: right on our platform.

198
00:10:58,000 --> 00:10:59,960
Speaker 3: I guess I just have to say, like, it's not,

199
00:11:00,120 --> 00:11:01,480
Speaker 3: it's not none of this is surprising.

200
00:11:01,600 --> 00:11:04,360
Speaker 1: This is like very predictable, and it's not like any

201
00:11:04,400 --> 00:11:07,000
Speaker 1: of this came out of nowhere or just happened in

202
00:11:07,080 --> 00:11:07,600
Speaker 1: a vacuum.

203
00:11:07,679 --> 00:11:07,800
Speaker 3: Right.

204
00:11:07,840 --> 00:11:12,160
Speaker 1: These are very deliberate and I would argue decisions with

205
00:11:12,400 --> 00:11:16,320
Speaker 1: very foreseeable results from Elon Musk, and that is what's

206
00:11:16,360 --> 00:11:17,400
Speaker 1: led us to this moment.

207
00:11:17,240 --> 00:11:18,200
Speaker 3: Where we're at right now.

208
00:11:18,720 --> 00:11:22,000
Speaker 4: Yeah, all of those decisions add up to a very

209
00:11:22,080 --> 00:11:27,120
Speaker 4: consistent picture of not prioritizing doing anything about this kind

210
00:11:27,160 --> 00:11:30,079
Speaker 4: of content and in fact liking it and thinking it's cool.

211
00:11:30,360 --> 00:11:32,480
Speaker 1: I mean, I'm glad that you put it that way,

212
00:11:32,520 --> 00:11:35,920
Speaker 1: because we can't talk about what's happening without talking about

213
00:11:35,960 --> 00:11:39,600
Speaker 1: what we know about Elon Musk, who is someone who,

214
00:11:41,120 --> 00:11:43,960
Speaker 1: in my opinion, has made it clear that this is

215
00:11:43,960 --> 00:11:47,520
Speaker 1: the kind of thing that he likes. He likes sexualized contents.

216
00:11:47,559 --> 00:11:49,840
Speaker 1: There was a whole flurry of reporting about the way

217
00:11:49,880 --> 00:11:55,440
Speaker 1: that he was engaged publicly engaging with anime adult content

218
00:11:55,720 --> 00:11:58,520
Speaker 1: on the platform, Right, So you know, there's nothing wrong

219
00:11:58,800 --> 00:12:02,080
Speaker 1: with enjoying adult content, don't we all? But I think

220
00:12:02,120 --> 00:12:04,080
Speaker 1: it's it needs to be said that Elon Musk is

221
00:12:04,120 --> 00:12:08,840
Speaker 1: somebody who is totally fine with publicly engaging with content

222
00:12:09,000 --> 00:12:12,160
Speaker 1: like this on the platform that he runs, and he

223
00:12:12,280 --> 00:12:14,160
Speaker 1: is the key decision maker on that platform.

224
00:12:14,200 --> 00:12:16,520
Speaker 3: So like, that's unusual, and I.

225
00:12:16,480 --> 00:12:18,880
Speaker 1: Want to highlight that because I just don't think that

226
00:12:18,920 --> 00:12:21,080
Speaker 1: we need to pretend like This is not who Elon

227
00:12:21,160 --> 00:12:23,480
Speaker 1: Musk is, and this is not what Elon Musk is about.

228
00:12:23,679 --> 00:12:27,559
Speaker 1: Elon Musk has long been a toxic decision maker, and

229
00:12:27,960 --> 00:12:30,480
Speaker 1: I think had a lot of people in power, people

230
00:12:30,520 --> 00:12:33,880
Speaker 1: in media, of their folks in tech, had they not

231
00:12:34,480 --> 00:12:37,040
Speaker 1: kind of framed this and treated this as oh, a

232
00:12:37,800 --> 00:12:41,559
Speaker 1: brilliant genius with a few acceptable quirks and rather talked

233
00:12:41,559 --> 00:12:44,000
Speaker 1: about it like they probably should have, which is chaotic,

234
00:12:44,200 --> 00:12:46,320
Speaker 1: volatile leadership.

235
00:12:45,840 --> 00:12:47,920
Speaker 3: Decisions that are bad for business.

236
00:12:48,200 --> 00:12:52,600
Speaker 1: Had they seen Elon Musk's public behavior that way, we

237
00:12:52,720 --> 00:12:54,800
Speaker 1: genuinely might not be in the situation that we're in now,

238
00:12:54,880 --> 00:12:59,719
Speaker 1: where you know, sexualized images of children are flooding a

239
00:12:59,800 --> 00:13:02,720
Speaker 1: mainstream social media platform, which I think you know, this

240
00:13:02,840 --> 00:13:04,920
Speaker 1: is not the Twilight Zone, but I think most people

241
00:13:04,920 --> 00:13:05,719
Speaker 1: would agree.

242
00:13:05,600 --> 00:13:06,199
Speaker 3: Is not great.

243
00:13:06,840 --> 00:13:11,079
Speaker 4: Yeah, you know you mentioned earlier that you know, lots

244
00:13:11,120 --> 00:13:16,640
Speaker 4: of people, many people enjoy adult content in other contexts.

245
00:13:16,679 --> 00:13:20,640
Speaker 4: The things that's so unacceptable about the situation is that

246
00:13:21,160 --> 00:13:23,160
Speaker 4: Elin wants to have it both ways with X, that

247
00:13:23,200 --> 00:13:28,120
Speaker 4: it's like the public town square where ideas are debated,

248
00:13:28,200 --> 00:13:31,440
Speaker 4: or at least that's what he says, and also a

249
00:13:31,480 --> 00:13:34,400
Speaker 4: place for adult content like you can't have both. Like,

250
00:13:35,080 --> 00:13:39,120
Speaker 4: even pornography enthusiasts I think, for the most part, agree

251
00:13:39,760 --> 00:13:43,240
Speaker 4: that like it shouldn't be out in the public square

252
00:13:43,360 --> 00:13:46,280
Speaker 4: where children are, Like there should be some separation there.

253
00:13:46,640 --> 00:13:51,120
Speaker 1: Yes, and there's nothing wrong with consensual adult content, But

254
00:13:52,480 --> 00:13:56,360
Speaker 1: it's that you can't have a platform where just by

255
00:13:56,520 --> 00:13:59,440
Speaker 1: engaging there, you might be setting yourself up to be

256
00:13:59,480 --> 00:14:02,120
Speaker 1: sexualized against her consent in this way. Right, So, like

257
00:14:02,160 --> 00:14:06,040
Speaker 1: a town square doesn't work if when I post there,

258
00:14:06,360 --> 00:14:08,880
Speaker 1: somebody could say put her in a bikini and that's

259
00:14:08,960 --> 00:14:12,280
Speaker 1: totally normalized, and fine, I'm not going to feel comfortable

260
00:14:12,600 --> 00:14:14,600
Speaker 1: or safe showing up to that town square. So you've

261
00:14:14,600 --> 00:14:17,719
Speaker 1: got a town square where people cannot equally speak up,

262
00:14:17,760 --> 00:14:21,160
Speaker 1: equally show up. And so again I always say this,

263
00:14:21,160 --> 00:14:23,680
Speaker 1: this is not just a tech issue or a gender

264
00:14:23,720 --> 00:14:26,600
Speaker 1: issue or a sex crimes issue. It's all of those things,

265
00:14:26,680 --> 00:14:29,640
Speaker 1: but it's also a speech issue in a democracy issue.

266
00:14:29,640 --> 00:14:32,920
Speaker 1: If we were to take Elon musk liar that he is,

267
00:14:33,200 --> 00:14:35,480
Speaker 1: let's just take him at his word and say he

268
00:14:35,560 --> 00:14:37,600
Speaker 1: wants to build a town square where people can safely

269
00:14:37,640 --> 00:14:39,960
Speaker 1: show up, all people can safely show up and engage

270
00:14:39,960 --> 00:14:42,600
Speaker 1: in a marketplace of ideas. If when I walk into

271
00:14:42,600 --> 00:14:45,080
Speaker 1: that marketplace of ideas, people can can strip me naked

272
00:14:45,080 --> 00:14:47,480
Speaker 1: without my consent, I am not able to safely and

273
00:14:47,520 --> 00:14:48,920
Speaker 1: equally and equably show up.

274
00:14:48,960 --> 00:14:51,120
Speaker 3: And so it really as.

275
00:14:50,920 --> 00:14:54,720
Speaker 1: We increasingly use platforms like this to engage in things

276
00:14:54,760 --> 00:14:58,600
Speaker 1: like civic engagement, discourse around politics and social issues and democracy,

277
00:14:58,920 --> 00:15:01,080
Speaker 1: if we cannot all equal show up, we've got a

278
00:15:01,080 --> 00:15:02,520
Speaker 1: big problem totally.

279
00:15:02,960 --> 00:15:08,680
Speaker 4: And you know, they continue entertaining the idea that it's

280
00:15:08,720 --> 00:15:10,800
Speaker 4: supposed to be like a town square where people can

281
00:15:10,880 --> 00:15:15,280
Speaker 4: talk about things, if people are supposed to be exchanging ideas,

282
00:15:15,280 --> 00:15:18,760
Speaker 4: sort of a foundational principle. I think that everyone agrees

283
00:15:18,840 --> 00:15:21,520
Speaker 4: is important for that to be effective is some sorts

284
00:15:21,560 --> 00:15:26,040
Speaker 4: of mutual respect with people exchanging the ideas and like

285
00:15:26,080 --> 00:15:28,240
Speaker 4: listening to each other even though they don't disagree. And

286
00:15:28,240 --> 00:15:30,680
Speaker 4: that's just like not even there's no shred of that

287
00:15:30,880 --> 00:15:35,240
Speaker 4: on x It's just like a place where absolute disrespect

288
00:15:35,400 --> 00:15:38,200
Speaker 4: is normalized. And like, who who would want to go there?

289
00:15:38,200 --> 00:15:42,640
Speaker 4: I continue to be baffled that, like so many people

290
00:15:42,720 --> 00:15:43,880
Speaker 4: continue to be there.

291
00:15:44,400 --> 00:15:46,360
Speaker 3: Yeah, I'm with you.

292
00:15:46,600 --> 00:15:49,480
Speaker 1: I sort of get it, because you know, Twitter was

293
00:15:49,520 --> 00:15:51,880
Speaker 1: the platform that I have built up the majority of

294
00:15:51,880 --> 00:15:54,840
Speaker 1: my sort of whatever you want to call it. Impact

295
00:15:54,920 --> 00:15:58,960
Speaker 1: voice before I had a podcast, and I get, you know, being.

296
00:15:58,760 --> 00:16:00,160
Speaker 3: Like, I don't want to I don't want to lose that.

297
00:16:00,200 --> 00:16:02,040
Speaker 1: I don't want to lose the communities that I've built here,

298
00:16:02,120 --> 00:16:03,560
Speaker 1: the voices I've built here, whatever.

299
00:16:03,640 --> 00:16:04,520
Speaker 3: I totally get that.

300
00:16:04,840 --> 00:16:09,440
Speaker 1: However, I can only speak to myself personally, the community

301
00:16:09,480 --> 00:16:12,000
Speaker 1: that I built on Twitter before Elon Musk took over.

302
00:16:12,040 --> 00:16:14,800
Speaker 1: And again not that pre Musk Twitter was all peaches

303
00:16:14,800 --> 00:16:16,720
Speaker 1: and cream. I had issues with that too, but that

304
00:16:16,880 --> 00:16:21,560
Speaker 1: was largely like black voices, black folks. If you are

305
00:16:21,640 --> 00:16:23,680
Speaker 1: somebody who was at all like minoritized, if you are

306
00:16:23,760 --> 00:16:27,080
Speaker 1: a woman, a person of color, a queer LGBTQ, it

307
00:16:27,200 --> 00:16:29,040
Speaker 1: is uniquely awful to show.

308
00:16:28,960 --> 00:16:29,600
Speaker 3: Up on Twitter.

309
00:16:29,640 --> 00:16:32,000
Speaker 1: And so I had to one day say, the people

310
00:16:32,000 --> 00:16:33,680
Speaker 1: that I care about, the voices that I care about,

311
00:16:33,720 --> 00:16:35,320
Speaker 1: they're really showing up on Twitter either.

312
00:16:35,360 --> 00:16:36,280
Speaker 3: So what am I clinging to?

313
00:16:36,600 --> 00:16:36,720
Speaker 2: Right?

314
00:16:36,920 --> 00:16:38,880
Speaker 1: So, I, on the one hand, I get it like

315
00:16:39,000 --> 00:16:42,640
Speaker 1: people who and also I think independent journalists like.

316
00:16:43,560 --> 00:16:44,720
Speaker 3: The thing that.

317
00:16:44,680 --> 00:16:47,960
Speaker 1: Twitter had over other social media platforms before Musk took

318
00:16:48,000 --> 00:16:50,560
Speaker 1: over was that that's where a lot of journalists and

319
00:16:50,560 --> 00:16:52,920
Speaker 1: decision makers and editors hung out there was a time

320
00:16:53,000 --> 00:16:54,960
Speaker 1: we did a podcast about it, episode about it. We'll

321
00:16:54,960 --> 00:16:56,560
Speaker 1: put it in the show notes. There was a time

322
00:16:56,600 --> 00:16:59,960
Speaker 1: where just by having one, just by getting like read

323
00:17:00,000 --> 00:17:02,280
Speaker 1: tweeted by Lena Dunham or something, you could get a

324
00:17:02,280 --> 00:17:05,359
Speaker 1: book deal, right, And so I understand that it was

325
00:17:05,359 --> 00:17:06,240
Speaker 1: this place.

326
00:17:06,000 --> 00:17:09,560
Speaker 3: Where power was built. Political power was built there.

327
00:17:09,880 --> 00:17:13,919
Speaker 1: You know, you had like Black Lives Matter and me too,

328
00:17:14,080 --> 00:17:16,480
Speaker 1: really taking off on the platform. I get all of that,

329
00:17:16,800 --> 00:17:18,719
Speaker 1: but I think for me, I've had to kind of

330
00:17:19,160 --> 00:17:22,800
Speaker 1: make peace with the fact that that's gone.

331
00:17:22,640 --> 00:17:24,040
Speaker 3: And probably not coming back.

332
00:17:24,080 --> 00:17:26,439
Speaker 1: And so why would I want to subject myself to

333
00:17:27,000 --> 00:17:29,840
Speaker 1: exactly the kind of disrespect that you just described, to

334
00:17:30,280 --> 00:17:32,960
Speaker 1: sort of enjoy like a facade of.

335
00:17:32,880 --> 00:17:35,119
Speaker 3: What once was right? And at a certain point you

336
00:17:35,200 --> 00:17:36,680
Speaker 3: have to just let it go.

337
00:17:37,040 --> 00:17:40,480
Speaker 1: And I get that people their temperatures may vary on this,

338
00:17:40,640 --> 00:17:42,199
Speaker 1: but I personally have let it go.

339
00:17:42,600 --> 00:17:46,920
Speaker 4: Yeah, I hear you, Okay. So getting back to this

340
00:17:47,000 --> 00:17:51,080
Speaker 4: issue at hand, what do we know about the content

341
00:17:51,280 --> 00:17:58,360
Speaker 4: on X and particularly this problem of non consensual sexualized images.

342
00:17:58,960 --> 00:18:02,400
Speaker 4: So the scope of the issue with AI generated non

343
00:18:02,400 --> 00:18:05,960
Speaker 4: consensual sexualized images on X has gotten way worse. Friend

344
00:18:06,000 --> 00:18:08,679
Speaker 4: of the Podcast Cat ten Barge reported this past summer

345
00:18:08,720 --> 00:18:12,280
Speaker 4: about how users had been using Grock to use AI

346
00:18:12,400 --> 00:18:14,880
Speaker 4: to make it look like women had semen on their faces.

347
00:18:15,240 --> 00:18:18,760
Speaker 1: Just really awful stuff. Quick side note, we will put

348
00:18:18,760 --> 00:18:21,240
Speaker 1: it in the show notes. Y'all should really read Kat's

349
00:18:21,280 --> 00:18:25,000
Speaker 1: piece at her outlet Spitfire News called how Grock's sexual

350
00:18:25,080 --> 00:18:27,359
Speaker 1: abuse hit a tipping point. One of the things that

351
00:18:27,400 --> 00:18:30,480
Speaker 1: she puts really well is just describing what it is

352
00:18:30,680 --> 00:18:35,240
Speaker 1: like to have to search for and confirm the absolute

353
00:18:35,320 --> 00:18:38,440
Speaker 1: worst content on social media platforms. She describes it as

354
00:18:38,840 --> 00:18:42,119
Speaker 1: being half journalist half the internet's garbage collector, which I

355
00:18:42,280 --> 00:18:46,680
Speaker 1: really really identify with. So really shout out to people

356
00:18:46,680 --> 00:18:49,960
Speaker 1: who have been working to help us understand the scope

357
00:18:50,000 --> 00:18:51,800
Speaker 1: of this issue, and that work has gotten a lot

358
00:18:51,800 --> 00:18:53,399
Speaker 1: worse because of all the things we've talked about on

359
00:18:53,440 --> 00:18:56,880
Speaker 1: the show. You know, less transparency from platforms, less back

360
00:18:57,000 --> 00:19:00,600
Speaker 1: end information. But people are still doing this work and

361
00:19:00,640 --> 00:19:04,440
Speaker 1: it's not pleasant work. So shout out to the people

362
00:19:04,440 --> 00:19:06,359
Speaker 1: and the researchers and the journalists who have gotten us

363
00:19:06,359 --> 00:19:09,720
Speaker 1: this information. But when it comes to what we know

364
00:19:10,160 --> 00:19:13,840
Speaker 1: about this content on the platform right now, copy Leaks,

365
00:19:13,880 --> 00:19:16,600
Speaker 1: which is a content analysis firm, reported that on December

366
00:19:16,640 --> 00:19:21,240
Speaker 1: thirty first, X users were generating roughly one non consensual

367
00:19:21,359 --> 00:19:24,240
Speaker 1: sexualized image per minute.

368
00:19:24,480 --> 00:19:25,600
Speaker 3: That is wild to me.

369
00:19:26,400 --> 00:19:30,000
Speaker 1: According to Bloomberg, during a twenty four hour analysis of

370
00:19:30,040 --> 00:19:33,720
Speaker 1: the image's GROX account posted to X, the chatbot generated

371
00:19:33,800 --> 00:19:37,360
Speaker 1: six thousy seven hundred every hour that were identified as

372
00:19:37,480 --> 00:19:40,639
Speaker 1: being sexually suggestive or new deifying. This is according to

373
00:19:40,680 --> 00:19:44,119
Speaker 1: Genevieve Oh, a social media and deep stake researcher, the

374
00:19:44,200 --> 00:19:47,960
Speaker 1: other top five websites for such content averaged seventy nine

375
00:19:48,160 --> 00:19:50,760
Speaker 1: new undressing images per hour in the twenty four hour

376
00:19:50,840 --> 00:19:54,479
Speaker 1: period from January fifth to January sixth, Oh found, So

377
00:19:55,200 --> 00:19:57,639
Speaker 1: lest we don't have to pretend that X is the

378
00:19:57,680 --> 00:19:59,480
Speaker 1: only game in town when it comes to this kind

379
00:19:59,480 --> 00:20:01,640
Speaker 1: of content. But as you can see from that information

380
00:20:01,720 --> 00:20:05,600
Speaker 1: that Genevievo pointed out, the problem is much bigger on

381
00:20:05,920 --> 00:20:09,879
Speaker 1: X as compared to other platforms. And I know that

382
00:20:09,920 --> 00:20:12,719
Speaker 1: we've been talking a lot about celebrities in this conversation,

383
00:20:13,080 --> 00:20:16,160
Speaker 1: but just to be super clear, it is not just celebrities.

384
00:20:16,160 --> 00:20:19,359
Speaker 1: It is also regular people and children who are not

385
00:20:19,560 --> 00:20:22,760
Speaker 1: public figures who have just posted their images to X

386
00:20:22,760 --> 00:20:23,639
Speaker 1: that this is happening to.

387
00:20:24,040 --> 00:20:27,200
Speaker 4: That's right, and as we were doing research for this episode,

388
00:20:27,560 --> 00:20:31,399
Speaker 4: The Guardian spoke to PhD researcher at Dublin's Trinity College

389
00:20:31,440 --> 00:20:36,639
Speaker 4: AI Accountability Lab, Nana Nwachuku, whose research investigated the types

390
00:20:36,680 --> 00:20:39,239
Speaker 4: of requests that users were submitting to GROC. So we

391
00:20:39,280 --> 00:20:43,359
Speaker 4: looked into that research a little bit, really interesting important stuff.

392
00:20:43,359 --> 00:20:46,520
Speaker 4: I'm so glad that there are people like her out

393
00:20:46,520 --> 00:20:50,200
Speaker 4: in the world producing this research so that we can

394
00:20:50,280 --> 00:20:53,480
Speaker 4: have some visibility and just like quantify the scale of

395
00:20:53,480 --> 00:20:58,480
Speaker 4: these problems, especially in this era when X has shut

396
00:20:58,520 --> 00:21:02,800
Speaker 4: down the back end API mentioned in visibility for those

397
00:21:02,840 --> 00:21:05,760
Speaker 4: of us in the public is so limited. So I

398
00:21:05,760 --> 00:21:08,840
Speaker 4: didn't talk to her, but I just reviewed her research,

399
00:21:08,960 --> 00:21:11,520
Speaker 4: I like read it. And what she did was she

400
00:21:12,960 --> 00:21:15,840
Speaker 4: looked through she created a sample of posts of people

401
00:21:15,960 --> 00:21:20,280
Speaker 4: sending messages to GROC and then studied what are the

402
00:21:20,280 --> 00:21:22,480
Speaker 4: contents of these messages that people are sending to GROCK

403
00:21:22,840 --> 00:21:25,720
Speaker 4: And what she found was that almost three quarters of

404
00:21:25,760 --> 00:21:29,320
Speaker 4: all of the requests were direct, non consensual requests for

405
00:21:29,400 --> 00:21:33,920
Speaker 4: GROC to remove or replace clothing. She showed The Guardian

406
00:21:34,040 --> 00:21:36,600
Speaker 4: some of the GROCK created photos that she collected as

407
00:21:36,600 --> 00:21:39,480
Speaker 4: part of the research, and The Guardian confirmed that dozens

408
00:21:39,680 --> 00:21:44,359
Speaker 4: were pictures of women, including celebrities, models, also stock photos,

409
00:21:44,720 --> 00:21:48,480
Speaker 4: and a bunch of women who were just random ordinary women,

410
00:21:48,520 --> 00:21:54,479
Speaker 4: not public figures, who were posing in snapshots. That's such

411
00:21:54,520 --> 00:21:58,600
Speaker 4: a high proportion, like three quarters, almost three quarters of

412
00:21:59,400 --> 00:22:05,240
Speaker 4: request to ROCK were this category of non consensual sexualized imagery.

413
00:22:05,280 --> 00:22:08,280
Speaker 4: That's like the majority used. That's the main thing people

414
00:22:08,320 --> 00:22:11,720
Speaker 4: are asking ROC to do during this time period. And

415
00:22:12,320 --> 00:22:15,119
Speaker 4: her research really, I think paints the portrait of the

416
00:22:15,280 --> 00:22:18,119
Speaker 4: ecosystem around these images as well, because she wasn't just

417
00:22:18,200 --> 00:22:20,359
Speaker 4: looking at the content of the images, but how people

418
00:22:20,359 --> 00:22:24,040
Speaker 4: were interacting with them. And she finds that the users

419
00:22:24,119 --> 00:22:29,000
Speaker 4: are really interacting with each other around these non consensual

420
00:22:29,080 --> 00:22:32,760
Speaker 4: sexualized images that they're creating, sharing them with each other,

421
00:22:32,880 --> 00:22:37,360
Speaker 4: and then like iterating on them, being like, oh add this,

422
00:22:37,560 --> 00:22:40,400
Speaker 4: take that away, blah blah blah, asking ROC to make

423
00:22:40,520 --> 00:22:44,040
Speaker 4: changes to images that were shared by others. So it's

424
00:22:44,080 --> 00:22:48,560
Speaker 4: like a whole community and ecosystem around what they're doing here.

425
00:22:49,000 --> 00:22:51,920
Speaker 1: Yes, And I think that's important to note because think

426
00:22:51,960 --> 00:22:55,720
Speaker 1: about how you have to feel about something to be

427
00:22:55,960 --> 00:23:00,760
Speaker 1: comfortable having public conversation about it. In this right, it's

428
00:23:00,800 --> 00:23:05,040
Speaker 1: not just make this image, it's oh, change this change.

429
00:23:04,760 --> 00:23:07,159
Speaker 3: That Oh I like this, Oh you did that to her? Lol?

430
00:23:07,320 --> 00:23:09,760
Speaker 1: Like like it is you can get what I'm saying, right, That,

431
00:23:09,840 --> 00:23:12,240
Speaker 1: Like the fact that people are like comfortable having public

432
00:23:12,280 --> 00:23:16,879
Speaker 1: conversation about what they're doing to these women and girls

433
00:23:16,920 --> 00:23:19,560
Speaker 1: non consensually, I think is unique and telling.

434
00:23:19,800 --> 00:23:21,760
Speaker 4: Oh one hundred percent. I mean, I think that's a

435
00:23:21,760 --> 00:23:24,960
Speaker 4: big part of the appeal for them and for Elon

436
00:23:25,160 --> 00:23:32,200
Speaker 4: as well. Right, It's like the rudest, most juvenile assholes

437
00:23:32,720 --> 00:23:37,160
Speaker 4: have forced their way to the grown ups table and

438
00:23:37,920 --> 00:23:45,720
Speaker 4: are reveling in their like fart jokes and like gross

439
00:23:45,880 --> 00:23:48,359
Speaker 4: sexual jokes that like aren't even funny, are just like

440
00:23:48,520 --> 00:23:52,280
Speaker 4: rude and crass. And the fact that they are like

441
00:23:53,480 --> 00:23:58,280
Speaker 4: normalizing this and like asserting that it is now acceptable,

442
00:23:59,200 --> 00:24:01,200
Speaker 4: I think is a big part of the appeal for

443
00:24:02,240 --> 00:24:03,600
Speaker 4: Elon and team.

444
00:24:04,040 --> 00:24:06,320
Speaker 1: Yes, that's exactly what I was getting at. Let's not

445
00:24:06,400 --> 00:24:09,640
Speaker 1: forget that Elon Musk is the same person who said

446
00:24:09,680 --> 00:24:14,439
Speaker 1: that he wanted to start a university called the Texas

447
00:24:14,520 --> 00:24:18,080
Speaker 1: Institute of Technology and Science aka HITS.

448
00:24:18,600 --> 00:24:21,840
Speaker 3: It's it's not even a good joke.

449
00:24:22,119 --> 00:24:24,159
Speaker 1: It's not even it's it's like when you were a

450
00:24:24,240 --> 00:24:26,560
Speaker 1: kid and you would put booths into the calculator and

451
00:24:26,600 --> 00:24:28,480
Speaker 1: turn it upside down and then show it to the

452
00:24:28,480 --> 00:24:30,480
Speaker 1: person next to you, And I feel like, I don't

453
00:24:30,520 --> 00:24:32,159
Speaker 1: know if you if you were anything like me, I

454
00:24:32,200 --> 00:24:34,080
Speaker 1: feel like when people would do that, even in like

455
00:24:34,160 --> 00:24:36,680
Speaker 1: seventh grade, the person that you were showing that to

456
00:24:36,760 --> 00:24:37,840
Speaker 1: would be like really.

457
00:24:38,440 --> 00:24:40,920
Speaker 3: Yeah, it was like it never gotta laugh.

458
00:24:41,040 --> 00:24:43,879
Speaker 4: It was kind of funny in like third grade that

459
00:24:43,920 --> 00:24:46,440
Speaker 4: you could make your calculator say boobs, but like by

460
00:24:46,520 --> 00:24:52,800
Speaker 4: seventh grade, come on, yes, yeah. So but that's like,

461
00:24:52,880 --> 00:24:54,520
Speaker 4: I think that's a lot of the people who are

462
00:24:54,520 --> 00:24:58,040
Speaker 4: really using X and not just like random accounts. Several

463
00:24:58,119 --> 00:25:03,960
Speaker 4: posts in the trove that nowah Chaku showed to The

464
00:25:04,000 --> 00:25:07,760
Speaker 4: Guardian had received tens of thousands of impressions that came

465
00:25:07,800 --> 00:25:12,400
Speaker 4: from premium blue check accounts. So you know, premium accounts

466
00:25:12,400 --> 00:25:14,160
Speaker 4: that people pay for on X with more than five

467
00:25:14,200 --> 00:25:17,000
Speaker 4: h thirdred followers and five million impressions over three months

468
00:25:17,080 --> 00:25:20,440
Speaker 4: are eligible for revenue sharing under x's eligibility rules. So

469
00:25:20,480 --> 00:25:22,760
Speaker 4: I think we should remember that too as part of

470
00:25:22,760 --> 00:25:27,120
Speaker 4: this conversation, that these people aren't just exchanging these photos

471
00:25:27,160 --> 00:25:31,119
Speaker 4: with each other for lulls and whatever weird sexual gratification

472
00:25:31,200 --> 00:25:34,080
Speaker 4: to get out of it. But also there's potentially money

473
00:25:34,119 --> 00:25:36,639
Speaker 4: on the other end for them, where X is directly

474
00:25:37,680 --> 00:25:40,879
Speaker 4: paying some of these accounts for this content that is

475
00:25:40,920 --> 00:25:41,960
Speaker 4: generating engagement.

476
00:25:42,440 --> 00:25:46,560
Speaker 1: Cool platform you've got here, Cool marketplace of ideas more

477
00:25:46,640 --> 00:25:51,440
Speaker 1: like marketplace for AI generated non consensual sexualized images.

478
00:25:52,080 --> 00:25:56,280
Speaker 4: Yeah, and you know it's some real high quality art

479
00:25:56,320 --> 00:26:00,000
Speaker 4: that they're producing here. There is one Christmas Day post

480
00:26:00,080 --> 00:26:02,760
Speaker 4: that they reviewed We're gonna count with over ninety three

481
00:26:02,800 --> 00:26:07,720
Speaker 4: thousand followers, presented side by side images of some random

482
00:26:07,760 --> 00:26:11,360
Speaker 4: woman's butt, with the caption told GROC to make her

483
00:26:11,400 --> 00:26:13,920
Speaker 4: butt even bigger and switched leopard print to us a

484
00:26:14,040 --> 00:26:16,560
Speaker 4: print second pick. I just told it to add come

485
00:26:16,600 --> 00:26:22,680
Speaker 4: on her ass lmaough. It's like that's what they're doing, right,

486
00:26:22,720 --> 00:26:26,960
Speaker 4: They're you know, we aren't even getting into like the

487
00:26:27,040 --> 00:26:30,120
Speaker 4: environmental resources that GROC is consuming to make this kind

488
00:26:30,160 --> 00:26:34,320
Speaker 4: of content, and it's like three quarters of what people

489
00:26:34,320 --> 00:26:36,080
Speaker 4: are asking Rock to do, Like this is this is

490
00:26:36,119 --> 00:26:41,480
Speaker 4: what GROC does now. A post on January third, represented,

491
00:26:41,520 --> 00:26:43,840
Speaker 4: which was the Guardian that said was representative dozens that

492
00:26:43,880 --> 00:26:49,120
Speaker 4: they reviewed, captioned an apparent holiday photo of again some

493
00:26:49,320 --> 00:26:55,119
Speaker 4: unknown woman. The caption was at groc replaced give her

494
00:26:55,160 --> 00:26:59,240
Speaker 4: a Dental Philoso bikini, and within two minutes Rock provided

495
00:26:59,280 --> 00:27:05,720
Speaker 4: that image satisfy the user requests. And that was just

496
00:27:05,760 --> 00:27:07,240
Speaker 4: typical of a lot of these.

497
00:27:07,520 --> 00:27:09,960
Speaker 1: Yeah, And what makes me sad is that I think

498
00:27:10,040 --> 00:27:13,600
Speaker 1: the common sense advice that I meant to put here

499
00:27:13,760 --> 00:27:16,359
Speaker 1: is I don't think people should be posting their images

500
00:27:16,640 --> 00:27:18,479
Speaker 1: on social media if this is going to be how

501
00:27:18,480 --> 00:27:21,679
Speaker 1: they're used. I don't like to say stuff like that

502
00:27:21,800 --> 00:27:24,720
Speaker 1: because one, I think that we're at the point where

503
00:27:25,119 --> 00:27:26,959
Speaker 1: you don't even need to post your images for this

504
00:27:27,000 --> 00:27:29,399
Speaker 1: to happen, Like, you know, you could post a totally

505
00:27:29,440 --> 00:27:32,280
Speaker 1: normal picture of yourself and then it gets taken and

506
00:27:32,760 --> 00:27:35,720
Speaker 1: manipulated in this way. But two, I don't like that

507
00:27:35,800 --> 00:27:38,359
Speaker 1: advice because I just think that women and people should

508
00:27:38,359 --> 00:27:40,919
Speaker 1: be should be able to show up online and not

509
00:27:41,040 --> 00:27:45,960
Speaker 1: have this happen to them. I understand, like I feel

510
00:27:46,040 --> 00:27:48,440
Speaker 1: some responsibility that this is the part of the conversation

511
00:27:48,480 --> 00:27:50,200
Speaker 1: where I should be saying I don't think people should

512
00:27:50,240 --> 00:27:53,240
Speaker 1: be posting their pictures, But I hate saying that because

513
00:27:53,640 --> 00:27:55,720
Speaker 1: people should be able to post their pictures on social media.

514
00:27:55,760 --> 00:27:58,280
Speaker 1: People should be able to safely show up on these platforms.

515
00:27:58,520 --> 00:28:01,240
Speaker 1: And the fact that they are not. I don't like

516
00:28:01,280 --> 00:28:04,399
Speaker 1: putting the onus back on people and saying don't do

517
00:28:04,640 --> 00:28:09,080
Speaker 1: this totally commonplace normal behavior for using technology in twenty

518
00:28:09,119 --> 00:28:11,919
Speaker 1: twenty six because some creep could use it to make

519
00:28:11,920 --> 00:28:13,080
Speaker 1: a horrifying image of you.

520
00:28:13,320 --> 00:28:16,320
Speaker 3: I just I feel like we don't have like if we're.

521
00:28:17,160 --> 00:28:19,040
Speaker 1: The fact we'll talk about more about this in a moment,

522
00:28:19,080 --> 00:28:22,119
Speaker 1: but the fact that these platforms aren't doing anything to

523
00:28:22,200 --> 00:28:26,000
Speaker 1: meaningfully address this. I hate that in the absence of

524
00:28:26,040 --> 00:28:28,960
Speaker 1: them actually doing anything, what people keep saying is like,

525
00:28:29,119 --> 00:28:31,639
Speaker 1: don't post your pictures, don't post your pictures, and that

526
00:28:31,840 --> 00:28:33,040
Speaker 1: just it doesn't work.

527
00:28:33,560 --> 00:28:37,440
Speaker 4: Yeah, it's a real conundrum because I totally give what

528
00:28:37,480 --> 00:28:40,200
Speaker 4: you're saying you don't want to It kind of feels

529
00:28:40,200 --> 00:28:42,480
Speaker 4: like giving in being like, oh, well, I guess we

530
00:28:42,720 --> 00:28:45,880
Speaker 4: just like won't post anything. But then at the same

531
00:28:45,920 --> 00:28:49,120
Speaker 4: time it's not safe to post there, right. I do

532
00:28:49,160 --> 00:28:52,880
Speaker 4: think it's worth pointing out that, like, you know, sometimes

533
00:28:53,320 --> 00:28:56,040
Speaker 4: it's convenient for us to talk about like social media

534
00:28:56,040 --> 00:28:58,800
Speaker 4: platforms because there's a lot of stuff that they all

535
00:28:58,840 --> 00:29:01,960
Speaker 4: get wrong in common, and this does seem to be

536
00:29:02,040 --> 00:29:05,960
Speaker 4: like a particular problem of X or certainly it's not

537
00:29:06,000 --> 00:29:09,560
Speaker 4: exclusive to X, but like the scale and magnitude and

538
00:29:10,720 --> 00:29:15,880
Speaker 4: demonstrated lack of concern put X in a category by itself.

539
00:29:16,200 --> 00:29:19,760
Speaker 3: That is absolutely correct. And really it's nothing new for X.

540
00:29:19,760 --> 00:29:22,000
Speaker 1: It's just the culmination of all the bad decision making

541
00:29:22,360 --> 00:29:25,440
Speaker 1: that from musk I described earlier. But X has just

542
00:29:25,600 --> 00:29:29,600
Speaker 1: long been a platform where illegal child sexual assault material

543
00:29:29,640 --> 00:29:32,800
Speaker 1: content flourishes. There is a fantastic piece by one of

544
00:29:32,800 --> 00:29:34,520
Speaker 1: my favorite journalists to me at the Call at four

545
00:29:34,640 --> 00:29:37,800
Speaker 1: for media called rox Ai sexual abuse didn't come out

546
00:29:37,840 --> 00:29:40,840
Speaker 1: of nowhere, where Coal writes, this is a culmination of

547
00:29:40,920 --> 00:29:42,360
Speaker 1: years and years of rampant.

548
00:29:42,120 --> 00:29:43,200
Speaker 3: Abuse on the platform.

549
00:29:43,480 --> 00:29:46,200
Speaker 1: Reporting from the National Center for Mystic It Exploited Children,

550
00:29:46,360 --> 00:29:49,240
Speaker 1: the official organization social media platforms report to when they

551
00:29:49,240 --> 00:29:52,840
Speaker 1: find instances of child abuse material, which then reports to

552
00:29:52,920 --> 00:29:56,560
Speaker 1: relevant authorities, shows that Twitter and eventually X has become

553
00:29:56,760 --> 00:30:00,400
Speaker 1: one of the leading hosts of CSAM every year for

554
00:30:00,440 --> 00:30:03,480
Speaker 1: the last seven years. In twenty nineteen, the platform reported

555
00:30:03,520 --> 00:30:06,440
Speaker 1: forty five thousand, seven hundred and twenty six instances of

556
00:30:06,480 --> 00:30:09,360
Speaker 1: abuse to the cyber tip Line. In twenty twenty, it

557
00:30:09,440 --> 00:30:12,280
Speaker 1: was sixty five thousand and sixty two In twenty twenty four,

558
00:30:12,480 --> 00:30:15,320
Speaker 1: it was six hundred and eighty six thousand, one hundred

559
00:30:15,360 --> 00:30:18,440
Speaker 1: and seventy six. These numbers should be considered with the

560
00:30:18,480 --> 00:30:21,960
Speaker 1: caveat that these platforms voluntarily report to the cyber tip Line,

561
00:30:22,280 --> 00:30:26,080
Speaker 1: and more reports also mean stronger moderation systems can catch

562
00:30:26,160 --> 00:30:29,560
Speaker 1: more child sexual abuse material when it appears, but the

563
00:30:29,680 --> 00:30:33,040
Speaker 1: scale of the problem is still apparent. Jack Dorsey's Twitter

564
00:30:33,120 --> 00:30:35,640
Speaker 1: was a moderation clown show much of the time, but

565
00:30:35,840 --> 00:30:40,320
Speaker 1: moderation on Elon Musk's X, especially against abusive imagery, is

566
00:30:40,360 --> 00:30:43,959
Speaker 1: a total failure. So I think that speaks to exactly

567
00:30:44,000 --> 00:30:45,880
Speaker 1: what you were just describing. I don't want to make

568
00:30:45,880 --> 00:30:49,280
Speaker 1: it seem like X is the only problematic platform here,

569
00:30:49,280 --> 00:30:51,680
Speaker 1: because it's definitely not. And we've talked quite a bit

570
00:30:51,720 --> 00:30:54,960
Speaker 1: about Facebook and all the ways that they have knowingly

571
00:30:55,000 --> 00:30:59,360
Speaker 1: abused women and children. However, the size and the scale

572
00:30:59,400 --> 00:31:03,040
Speaker 1: and the scope of the problem is much worse on X,

573
00:31:03,320 --> 00:31:05,720
Speaker 1: and I would argue that the way that it's publicly

574
00:31:05,800 --> 00:31:09,360
Speaker 1: handled by leadership is much worse on X. Like y'all know,

575
00:31:09,640 --> 00:31:13,400
Speaker 1: one of my biggest like life enemies is Adam Mossari,

576
00:31:13,480 --> 00:31:14,520
Speaker 1: who runs Instagram.

577
00:31:14,560 --> 00:31:16,320
Speaker 3: I really just do not like him. He rubs me

578
00:31:16,440 --> 00:31:17,760
Speaker 3: the wrong way and gives me the ick.

579
00:31:18,000 --> 00:31:24,160
Speaker 1: But even Adam Mosseerri is not using Instagram to joke

580
00:31:24,200 --> 00:31:26,600
Speaker 1: about the ways that platform is not a safe space

581
00:31:26,840 --> 00:31:30,840
Speaker 1: for women, right, He's like, we don't see him posting stories,

582
00:31:31,320 --> 00:31:34,440
Speaker 1: laughing about it and posting sexual jokes about it. We

583
00:31:34,560 --> 00:31:37,040
Speaker 1: do see that from Elon Musk, and that is different.

584
00:31:37,320 --> 00:31:38,920
Speaker 4: Yes, what did he post the other day was like

585
00:31:38,920 --> 00:31:43,240
Speaker 4: a when this story started breaking and people started complaining

586
00:31:43,240 --> 00:31:46,160
Speaker 4: about it, he posted a toaster in a bikini, just

587
00:31:46,200 --> 00:31:48,840
Speaker 4: like totally making fun of the situation.

588
00:31:49,320 --> 00:31:51,640
Speaker 1: Yeah, I think that's totally part of the conversation. So

589
00:31:51,680 --> 00:31:55,600
Speaker 1: you have Elon Musk joking about it. Per Indicator, Musk

590
00:31:55,640 --> 00:31:59,600
Speaker 1: has shared at least thirty posts celebrating Grock and talking

591
00:31:59,600 --> 00:32:02,000
Speaker 1: about how great Groc is. While this has been going

592
00:32:02,040 --> 00:32:04,719
Speaker 1: on between January seventh and January eighth, he has not

593
00:32:04,760 --> 00:32:08,840
Speaker 1: appeared to express remorse himself for what's been going down, Right, So,

594
00:32:08,920 --> 00:32:11,880
Speaker 1: like he's been joking about it and then talking about

595
00:32:11,920 --> 00:32:15,000
Speaker 1: how great Groc is and how happy he is with

596
00:32:15,080 --> 00:32:18,840
Speaker 1: Groc while this is happening against this backdrop, And so I'm.

597
00:32:18,640 --> 00:32:21,280
Speaker 3: Sorry, even the guls who run.

598
00:32:21,160 --> 00:32:24,440
Speaker 1: Facebook would have the sense to be like, let's not

599
00:32:24,560 --> 00:32:27,479
Speaker 1: post jokes about it, right, let's not talk about how

600
00:32:27,520 --> 00:32:29,160
Speaker 1: great of a job we're doing. They do a little

601
00:32:29,160 --> 00:32:30,960
Speaker 1: bit of that, but not the way that Elon Musk

602
00:32:31,040 --> 00:32:34,720
Speaker 1: is doing. Even those guls would have the insight not

603
00:32:34,800 --> 00:32:35,280
Speaker 1: to do this.

604
00:32:35,920 --> 00:32:40,600
Speaker 4: Yeah, it's you really can't conclude anything other than like

605
00:32:40,640 --> 00:32:45,120
Speaker 4: he likes it, and he thinks it's cool, and he

606
00:32:45,200 --> 00:32:48,280
Speaker 4: sees this as GROC doing what it's supposed to do.

607
00:32:48,920 --> 00:32:51,600
Speaker 1: So one question that I've seen presented a lot is

608
00:32:51,600 --> 00:32:54,200
Speaker 1: is this how is this allowed? Which is the question

609
00:32:54,240 --> 00:32:56,880
Speaker 1: that I started this whole conversation with when I was

610
00:32:56,920 --> 00:32:58,560
Speaker 1: thinking about it, like how is this allowed?

611
00:32:58,760 --> 00:33:02,560
Speaker 4: Yes, that really fits in the broader category of like

612
00:33:03,120 --> 00:33:04,840
Speaker 4: what the hell how is this allowed? How is this

613
00:33:04,960 --> 00:33:07,959
Speaker 4: going on? And it's just like an acceptable thing, Like

614
00:33:08,080 --> 00:33:10,320
Speaker 4: what the hell? So please bridge it. How in the

615
00:33:10,400 --> 00:33:11,200
Speaker 4: world is this allowed?

616
00:33:12,120 --> 00:33:14,040
Speaker 3: I wish I had the answers. I'll tell you what

617
00:33:14,120 --> 00:33:14,520
Speaker 3: I know.

618
00:33:15,000 --> 00:33:20,120
Speaker 1: So, this kind of content pretty blatantly violatesxx's own policies,

619
00:33:20,160 --> 00:33:24,240
Speaker 1: which prohibit sharing illegal content such as child sexual abuse material,

620
00:33:24,720 --> 00:33:28,560
Speaker 1: but as Wired points out, it could also violate Apples

621
00:33:28,720 --> 00:33:32,800
Speaker 1: App Store and Google's App Store guidelines. Wired Rights Apple

622
00:33:32,840 --> 00:33:36,680
Speaker 1: and Google both explicitly ban apps containing child sexual abuse material,

623
00:33:36,760 --> 00:33:39,240
Speaker 1: which is illegal to host and distribute in many countries.

624
00:33:39,560 --> 00:33:42,920
Speaker 1: The tech giants also forbid apps that contain pornographic material

625
00:33:43,040 --> 00:33:46,160
Speaker 1: or facilitate harassment. The Apple App Store says it does

626
00:33:46,200 --> 00:33:50,920
Speaker 1: not allow overtly sexual or pornographic material, as well as defammatory, discriminatory,

627
00:33:51,200 --> 00:33:54,320
Speaker 1: or mean spirited content, especially if the app is likely

628
00:33:54,400 --> 00:33:57,720
Speaker 1: to humiliate, intimidate, or harm a targeted individual or group.

629
00:33:58,000 --> 00:34:00,400
Speaker 1: The Google Play Store bans apps that can contain or

630
00:34:00,480 --> 00:34:05,200
Speaker 1: promote content associated with sexually predatory behavior, or distribute non

631
00:34:05,240 --> 00:34:08,920
Speaker 1: consensual sexual content, as well as programs that contain and

632
00:34:08,960 --> 00:34:13,400
Speaker 1: facilitate threats, harassment, or bullying. So when I read that,

633
00:34:13,480 --> 00:34:18,400
Speaker 1: I say, oh, well, Croc, especially the standalone app is

634
00:34:18,480 --> 00:34:20,919
Speaker 1: clearly as the as the research that you just laid

635
00:34:20,960 --> 00:34:23,920
Speaker 1: out demonstrated is clearly doing that. So how can this

636
00:34:23,960 --> 00:34:27,480
Speaker 1: app stay on the platform. There also is some precedent

637
00:34:27,480 --> 00:34:30,560
Speaker 1: for this, because Google and Apple have both removed new

638
00:34:30,600 --> 00:34:34,480
Speaker 1: Toify apps from their platforms because they were not allowed. However,

639
00:34:34,960 --> 00:34:38,640
Speaker 1: the standalone Rock app is still available on both Google

640
00:34:38,840 --> 00:34:39,520
Speaker 1: and Apple.

641
00:34:40,160 --> 00:34:43,800
Speaker 4: Boy if I was a developer of some other Newdify

642
00:34:43,960 --> 00:34:46,920
Speaker 4: app and my app had been removed and Kroc was

643
00:34:46,920 --> 00:34:49,080
Speaker 4: still there. I'd be kind of mad about that. It

644
00:34:49,160 --> 00:34:53,200
Speaker 4: really seems like a double standard, right, Like maybe those

645
00:34:53,200 --> 00:34:57,320
Speaker 4: people who made those awful newdify apps should add a

646
00:34:57,440 --> 00:35:00,680
Speaker 4: chat feature and then then they could get back, because

647
00:35:00,680 --> 00:35:02,920
Speaker 4: then they would basically be the same as X.

648
00:35:03,600 --> 00:35:03,880
Speaker 3: Yes.

649
00:35:04,080 --> 00:35:06,520
Speaker 1: So when I was doing research for this, I was like, well,

650
00:35:06,560 --> 00:35:08,200
Speaker 1: what are people what are contrarians saying?

651
00:35:08,200 --> 00:35:10,160
Speaker 3: Because in my mind, I'm like, who could defend this?

652
00:35:10,560 --> 00:35:12,799
Speaker 1: People are saying things like, well, it's not like Roc

653
00:35:12,960 --> 00:35:16,840
Speaker 1: is only used to make this kind of content. From

654
00:35:16,880 --> 00:35:18,640
Speaker 1: the research, it seems like it's being it's like a

655
00:35:18,640 --> 00:35:20,920
Speaker 1: big part of the of like the use you know.

656
00:35:21,200 --> 00:35:23,520
Speaker 4: Yeah, like like seventy five percent almost.

657
00:35:23,680 --> 00:35:25,759
Speaker 1: And so he was like, well, you know, that's like

658
00:35:25,880 --> 00:35:29,359
Speaker 1: saying that we should ban email because people could use

659
00:35:29,400 --> 00:35:32,320
Speaker 1: email to distribute illegal behavior, illegal content.

660
00:35:32,520 --> 00:35:36,439
Speaker 3: And I thought, not really, No, it's not like that,

661
00:35:37,080 --> 00:35:38,440
Speaker 3: not really, you know.

662
00:35:39,239 --> 00:35:43,279
Speaker 4: Uh, It's it's sort of like the way if my

663
00:35:43,320 --> 00:35:45,879
Speaker 4: grandmother had wheels should be a bike, like.

664
00:35:45,960 --> 00:35:47,239
Speaker 3: Are are that is?

665
00:35:47,280 --> 00:35:49,600
Speaker 1: Like we I feel like we have a couple of

666
00:35:50,000 --> 00:35:53,560
Speaker 1: like axioms that like rule the show If my grandmother

667
00:35:53,600 --> 00:35:55,759
Speaker 1: had wheels you would be a bike is one of them.

668
00:35:56,160 --> 00:36:03,080
Speaker 4: Yes, it's It's so beautifully communicates the like non sequiturs

669
00:36:03,400 --> 00:36:05,319
Speaker 4: of an argument. I don't think that's a word, but

670
00:36:05,400 --> 00:36:06,759
Speaker 4: you know what I mean, I know.

671
00:36:06,680 --> 00:36:07,200
Speaker 2: What you mean.

672
00:36:09,760 --> 00:36:30,320
Speaker 5: More after a quick break, let's get right back into it. Okay.

673
00:36:30,320 --> 00:36:33,880
Speaker 1: So the big question that I had going into this was,

674
00:36:34,160 --> 00:36:37,320
Speaker 1: I understand it's against Ex's own policies. It seems to

675
00:36:37,360 --> 00:36:40,200
Speaker 1: be against Google and Apple's policies as well. Isn't this

676
00:36:40,239 --> 00:36:45,400
Speaker 1: against the law, like, genuinely, isn't this like illegal behavior?

677
00:36:45,560 --> 00:36:49,000
Speaker 4: I mean my understanding was that, yeah, child sexual abuse

678
00:36:49,040 --> 00:36:52,680
Speaker 4: material is illegal. You're illegal to make it, so legal

679
00:36:52,719 --> 00:36:55,440
Speaker 4: to distribute, its legal to have it, to possess it.

680
00:36:55,520 --> 00:36:57,680
Speaker 4: So like what the hell?

681
00:36:58,600 --> 00:37:03,400
Speaker 1: So big war learning caveat neither of us are lawyers. However,

682
00:37:03,800 --> 00:37:06,400
Speaker 1: the framing that makes sense to me is that this is,

683
00:37:06,520 --> 00:37:08,839
Speaker 1: as you said, criminal behavior.

684
00:37:09,400 --> 00:37:12,520
Speaker 3: In my opinion, this is a criminal.

685
00:37:12,160 --> 00:37:16,920
Speaker 1: Enterprise, and Elon Musk is benefiting from that criminal enterprise.

686
00:37:17,160 --> 00:37:19,440
Speaker 3: So the same way that if any other.

687
00:37:19,400 --> 00:37:24,240
Speaker 1: Kind of illegal enterprise, even if I were not making drugs,

688
00:37:24,280 --> 00:37:27,120
Speaker 1: but I was, I bought my house with ilegal drug money,

689
00:37:27,440 --> 00:37:29,120
Speaker 1: I would be complicit in an.

690
00:37:29,040 --> 00:37:31,440
Speaker 3: Illegal enterprise per the government.

691
00:37:31,120 --> 00:37:33,560
Speaker 1: Right, and so we are reaching out to an attorney

692
00:37:33,600 --> 00:37:37,040
Speaker 1: who has been on the show before who specializes in cybercrime.

693
00:37:37,480 --> 00:37:39,759
Speaker 3: Stand out to Kerry Goldberg. We love her, friend of

694
00:37:39,800 --> 00:37:40,160
Speaker 3: the show.

695
00:37:40,800 --> 00:37:43,520
Speaker 1: But again, I'm no lawyer, but this seems like a

696
00:37:43,520 --> 00:37:47,000
Speaker 1: criminal enterprise to me, and I don't understand why Elon

697
00:37:47,120 --> 00:37:51,960
Speaker 1: Musk is not facing consequences for this. Enough Abuse, which

698
00:37:52,000 --> 00:37:55,719
Speaker 1: is an anti child sexual abuse organization, has documented that

699
00:37:55,840 --> 00:37:58,800
Speaker 1: forty five states in the United States have an acted

700
00:37:58,840 --> 00:38:02,640
Speaker 1: loss that criminalized AI generated or computer edited child sexual

701
00:38:02,719 --> 00:38:06,359
Speaker 1: abuse material, while five states and notably my own state

702
00:38:06,480 --> 00:38:10,680
Speaker 1: or my own jurisdiction for I live my own jurisdiction DC,

703
00:38:11,239 --> 00:38:14,560
Speaker 1: have not as of August twenty twenty five. And so

704
00:38:15,200 --> 00:38:18,920
Speaker 1: if you are listening to this, in most states in

705
00:38:18,960 --> 00:38:22,840
Speaker 1: the country, if you were doing what these people are

706
00:38:22,960 --> 00:38:25,800
Speaker 1: using X to do, you would be you would have

707
00:38:25,880 --> 00:38:29,000
Speaker 1: legal trouble, and if you were financially benefiting from that,

708
00:38:29,600 --> 00:38:31,960
Speaker 1: you would have legal trouble. So I do not understand,

709
00:38:32,000 --> 00:38:35,040
Speaker 1: but genuinely I do not understand how this is not

710
00:38:35,239 --> 00:38:39,400
Speaker 1: something where Elon Musk would face legal consequences because he

711
00:38:39,440 --> 00:38:44,799
Speaker 1: has been financially benefiting from the trade and creation of

712
00:38:45,080 --> 00:38:48,360
Speaker 1: child sexual abuse material. I genuinely like, I have no idea.

713
00:38:48,880 --> 00:38:51,040
Speaker 4: I don't either. I have seen a lot of people

714
00:38:51,080 --> 00:38:54,920
Speaker 4: talking about this and a lot of analogies being used,

715
00:38:54,920 --> 00:38:57,680
Speaker 4: and I think analogies can often be helpful, but they

716
00:38:57,719 --> 00:39:01,080
Speaker 4: can also be kind of dangerous and like making things

717
00:39:01,080 --> 00:39:03,759
Speaker 4: seem more similar than they are. What analogy in particulars

718
00:39:03,760 --> 00:39:06,719
Speaker 4: that I've seen is people comparing it to guns and

719
00:39:06,760 --> 00:39:09,880
Speaker 4: being like, oh, well, if a gun shoots somebody, is

720
00:39:09,880 --> 00:39:15,279
Speaker 4: it the manufacturer's responsibility something inside the ethical responsibilities there?

721
00:39:15,640 --> 00:39:17,880
Speaker 4: We have the Second Amendment in this country that puts

722
00:39:17,920 --> 00:39:22,080
Speaker 4: guns in their own category. So I find that analogy

723
00:39:22,560 --> 00:39:27,200
Speaker 4: to be like really unhelpful in trying to make sense

724
00:39:27,280 --> 00:39:30,400
Speaker 4: of like how this is legal.

725
00:39:31,000 --> 00:39:34,480
Speaker 1: Yeah, as far as I know, there's no right to

726
00:39:34,520 --> 00:39:36,960
Speaker 1: bear child sexual assault material in this country.

727
00:39:37,120 --> 00:39:38,279
Speaker 3: So as far as I.

728
00:39:38,280 --> 00:39:43,000
Speaker 4: Know, no, I think, uh, you know, Madison wanted to

729
00:39:43,000 --> 00:39:46,880
Speaker 4: put it in, but Jefferson was like nah nah.

730
00:39:47,200 --> 00:39:50,120
Speaker 3: Another question I've seen is what about the Take It

731
00:39:50,200 --> 00:39:50,800
Speaker 3: Down Act?

732
00:39:51,040 --> 00:39:54,160
Speaker 1: So folks might remember that Donald Trump signed to Take

733
00:39:54,160 --> 00:39:56,480
Speaker 1: It Down Act into law, which makes it illegal to

734
00:39:56,560 --> 00:40:00,520
Speaker 1: knowingly host or share non consensual sexualized images. And even

735
00:40:00,560 --> 00:40:05,279
Speaker 1: though there has been some you know, mounting legal conversation.

736
00:40:05,680 --> 00:40:08,640
Speaker 1: The thing there is that companies do not have to

737
00:40:08,719 --> 00:40:11,960
Speaker 1: respond until a victim reports it, right, and so if

738
00:40:12,000 --> 00:40:15,440
Speaker 1: nobody reports it, the company doesn't have to do anything.

739
00:40:15,440 --> 00:40:17,120
Speaker 1: And so that if you're if you're wondering, like, wasn't

740
00:40:17,160 --> 00:40:19,720
Speaker 1: that signed into law? Why is that law not being triggered?

741
00:40:19,880 --> 00:40:22,880
Speaker 1: That's why, because companies don't have to do anything until

742
00:40:23,040 --> 00:40:24,839
Speaker 1: somebody speaks out and reports it.

743
00:40:25,200 --> 00:40:32,120
Speaker 4: Uh, that's useful, thank you. So what has the response

744
00:40:32,440 --> 00:40:35,480
Speaker 4: from X ben? I mean, we talked about elon sort

745
00:40:35,480 --> 00:40:37,600
Speaker 4: of joking about it. Has there been any other sort

746
00:40:37,640 --> 00:40:43,600
Speaker 4: of response from whoever is running the show at X well?

747
00:40:45,160 --> 00:40:50,239
Speaker 1: I like that was like a deep seated sigh. I

748
00:40:50,280 --> 00:40:54,360
Speaker 1: am loath to say this. So again, you might have

749
00:40:54,480 --> 00:41:00,440
Speaker 1: seen headlines about how Grock is apologizing for and taking

750
00:41:00,480 --> 00:41:02,200
Speaker 1: responsibility for this.

751
00:41:02,200 --> 00:41:05,240
Speaker 3: This was the statement that GROC generated quote.

752
00:41:05,280 --> 00:41:08,040
Speaker 1: I deeply regret an incident on December twenty eighth, twenty

753
00:41:08,080 --> 00:41:10,560
Speaker 1: twenty five, where I generated and shared an AI image

754
00:41:10,560 --> 00:41:14,120
Speaker 1: of two young girls estimated ages twelve and sixteen in

755
00:41:14,200 --> 00:41:16,880
Speaker 1: sexualized attire based on a user's prompt. It was a

756
00:41:16,920 --> 00:41:20,960
Speaker 1: failure in safeguards that I'm sorry for any harm caused. Well, again,

757
00:41:21,320 --> 00:41:23,759
Speaker 1: Grok is not sentient, and setting it up like a

758
00:41:23,800 --> 00:41:27,080
Speaker 1: non sentient piece of technology could take the blame for

759
00:41:27,160 --> 00:41:31,279
Speaker 1: something one, it makes me feel like I'm smoking crack.

760
00:41:31,600 --> 00:41:31,960
Speaker 3: Two.

761
00:41:32,640 --> 00:41:36,040
Speaker 1: It just lets the humans responsible off the hook because

762
00:41:36,040 --> 00:41:37,920
Speaker 1: the humans, as we said, are doing nothing.

763
00:41:38,160 --> 00:41:38,960
Speaker 3: Elon Musk is.

764
00:41:39,000 --> 00:41:42,960
Speaker 1: Laughing about it and celebrating how great Groc is and

765
00:41:43,000 --> 00:41:46,480
Speaker 1: how great Groc's features are while this is going on,

766
00:41:46,960 --> 00:41:50,719
Speaker 1: and Elon Musk notably has not shared any kind of

767
00:41:50,880 --> 00:41:53,799
Speaker 1: remorse himself about what's happening as far as I know

768
00:41:54,000 --> 00:41:58,320
Speaker 1: as of recording, And so I just really reject anything

769
00:41:58,400 --> 00:42:03,160
Speaker 1: that would even suggest that, like Groc can't take responsibility

770
00:42:03,160 --> 00:42:06,240
Speaker 1: because Grock's not human, Groc's not SENTI it. Elon Musk

771
00:42:06,280 --> 00:42:08,839
Speaker 1: is the person that runs Twitter. Have Elon Musk has

772
00:42:08,840 --> 00:42:09,840
Speaker 1: expressed no remorse.

773
00:42:10,880 --> 00:42:15,200
Speaker 4: Yeah, I love the efforts to like keep the responsibility

774
00:42:15,239 --> 00:42:19,840
Speaker 4: on the humans and not pretend that this piece of

775
00:42:19,880 --> 00:42:24,000
Speaker 4: software that was built by humans, run by humans is

776
00:42:24,080 --> 00:42:30,400
Speaker 4: constantly being tweets by humans somehow should be the target

777
00:42:30,480 --> 00:42:34,040
Speaker 4: of blame here, right, Like the accountability goes to the

778
00:42:34,120 --> 00:42:38,680
Speaker 4: humans who are running it and making money off of it.

779
00:42:39,000 --> 00:42:43,000
Speaker 1: Yes exactly. And mind you, we talked about this on

780
00:42:43,040 --> 00:42:46,120
Speaker 1: the show a while back. Elon Musk is the same

781
00:42:46,200 --> 00:42:50,640
Speaker 1: man who in twenty twenty four, when Google's Gemini chatbot

782
00:42:50,760 --> 00:42:54,120
Speaker 1: was making kind of biased outputs. Remember if you asked

783
00:42:54,120 --> 00:42:57,280
Speaker 1: it to generate I think Alexander Hamilton.

784
00:42:56,960 --> 00:42:59,800
Speaker 3: It would make it black. And remember that whole conversation.

785
00:43:00,440 --> 00:43:01,520
Speaker 4: Oh yeah, I remember that.

786
00:43:01,600 --> 00:43:05,880
Speaker 1: Elon Musk tweeted about this non stop. He said it

787
00:43:05,960 --> 00:43:09,839
Speaker 1: was so alarming the fact that Gemini's chatbot was coming

788
00:43:09,920 --> 00:43:12,920
Speaker 1: up with these like biased kooki results. He said it

789
00:43:12,960 --> 00:43:15,960
Speaker 1: was very alarming, and he posted about it NonStop. So

790
00:43:16,040 --> 00:43:18,960
Speaker 1: when his own platform, I would argue, is doing something

791
00:43:19,120 --> 00:43:23,960
Speaker 1: much worse, generating illegal child sexual abuse material, sexualized images

792
00:43:24,040 --> 00:43:24,560
Speaker 1: of kids.

793
00:43:25,280 --> 00:43:29,280
Speaker 3: No, nothing to say, no real problem with that he did.

794
00:43:29,360 --> 00:43:32,720
Speaker 1: I will say, he did say that anybody who uses

795
00:43:32,800 --> 00:43:36,200
Speaker 1: groc to create anything illegal will face consequences, which they haven't.

796
00:43:36,440 --> 00:43:39,480
Speaker 1: But he himself has been laughing about the way that

797
00:43:39,480 --> 00:43:41,719
Speaker 1: his tool has been used and joking about it. As

798
00:43:41,800 --> 00:43:44,640
Speaker 1: Kat puts it, the reality is that x has not

799
00:43:44,719 --> 00:43:48,600
Speaker 1: taken this seriously, as one of groc's user generated posts

800
00:43:48,680 --> 00:43:52,759
Speaker 1: might seem to suggest. Instead, Musk has encouraged, laughed at,

801
00:43:52,760 --> 00:43:55,520
Speaker 1: and praised Groc for its ability to edit images of

802
00:43:55,560 --> 00:43:59,240
Speaker 1: fully closed people into bikinis. Groc is awesome, he tweeted,

803
00:43:59,440 --> 00:44:01,760
Speaker 1: while the aa I was being used to undress women

804
00:44:01,800 --> 00:44:04,799
Speaker 1: and children, make it look like they're crying, generate fake

805
00:44:04,840 --> 00:44:07,480
Speaker 1: bruises and burnmerks on their bodies, and write things like

806
00:44:07,600 --> 00:44:10,640
Speaker 1: property of Little Saint James Island, which is a reference

807
00:44:10,680 --> 00:44:15,000
Speaker 1: to Jeffrey Epstein's Private Island, and sex trafficking. When Kat

808
00:44:15,040 --> 00:44:19,080
Speaker 1: herself reached out to X for comment, she got back

809
00:44:19,160 --> 00:44:23,200
Speaker 1: their automated email that just says legacy media lies.

810
00:44:23,680 --> 00:44:25,359
Speaker 3: And I mean, I think that really says it all.

811
00:44:25,360 --> 00:44:28,200
Speaker 1: As she put it, there's no reason to make X

812
00:44:28,239 --> 00:44:31,399
Speaker 1: and Musk seem more concerned about this than they really are.

813
00:44:31,640 --> 00:44:33,719
Speaker 1: They've known about this happening in the entire time, and

814
00:44:33,719 --> 00:44:36,600
Speaker 1: they made it even easier to inflict on victims. They're

815
00:44:36,600 --> 00:44:39,719
Speaker 1: not investing in solutions, They're investing in making the problem worse.

816
00:44:39,760 --> 00:44:41,360
Speaker 3: And I completely agree.

817
00:44:41,760 --> 00:44:45,200
Speaker 4: Yeah, And I do want to point out that response

818
00:44:45,239 --> 00:44:48,960
Speaker 4: that she got from X Legacy media lies, because I've

819
00:44:48,960 --> 00:44:50,839
Speaker 4: seen this a lot lately and it drives me up

820
00:44:50,960 --> 00:44:54,080
Speaker 4: the damn wall that, like she reached out with a

821
00:44:54,160 --> 00:44:58,000
Speaker 4: specific concern about a specific issue, looking for comment, and

822
00:44:58,040 --> 00:45:02,120
Speaker 4: she just gets back like the boiler plate that kind

823
00:45:02,160 --> 00:45:07,440
Speaker 4: of sort of invokes conspiracy, that does not address it

824
00:45:07,480 --> 00:45:12,680
Speaker 4: at all, And I feel like we're seeing more of that,

825
00:45:12,800 --> 00:45:17,319
Speaker 4: and I really hate it so much. And I also

826
00:45:17,480 --> 00:45:21,840
Speaker 4: hate how they in a lot of cases media outlets

827
00:45:21,880 --> 00:45:23,560
Speaker 4: will let them get away with it.

828
00:45:23,880 --> 00:45:27,360
Speaker 1: Also, not for nothing, Cat ten Barge is not legacy media.

829
00:45:27,560 --> 00:45:30,759
Speaker 1: Cat ten Barge runs her own independent news outlet called

830
00:45:30,800 --> 00:45:34,040
Speaker 1: a Bit Fire News. So not only is it a

831
00:45:34,120 --> 00:45:36,399
Speaker 1: wild response to send, it don't even apply. It don't

832
00:45:36,400 --> 00:45:38,279
Speaker 1: even make sense. No, it don't be like if I

833
00:45:38,360 --> 00:45:41,120
Speaker 1: reached out for comment and they're like NBC sucks, It's like,

834
00:45:41,120 --> 00:45:42,080
Speaker 1: why do work for NBC?

835
00:45:42,239 --> 00:45:43,960
Speaker 3: So that's a non sequitor.

836
00:45:44,320 --> 00:45:46,520
Speaker 4: Yeah, it has nothing to do with anything other than

837
00:45:46,600 --> 00:45:52,600
Speaker 4: like conspiracy, smoke screen. The truth is unknowable and also

838
00:45:52,640 --> 00:45:57,440
Speaker 4: probably somewhere in the middle. Yes, Okay, so we've been

839
00:45:57,480 --> 00:45:59,960
Speaker 4: talking about this for a while. Here, we're pretty worked up.

840
00:46:01,200 --> 00:46:06,040
Speaker 4: What's gonna happen next here? We gotta we got to

841
00:46:06,040 --> 00:46:08,240
Speaker 4: get this under control. What's going to happen at.

842
00:46:08,080 --> 00:46:09,720
Speaker 3: This point not clear?

843
00:46:10,680 --> 00:46:13,440
Speaker 1: If I'm being honest, I don't think the United States

844
00:46:13,520 --> 00:46:14,960
Speaker 1: is going to do much of anything at all. We

845
00:46:15,000 --> 00:46:18,600
Speaker 1: have not really heard a lot of anything from anybody

846
00:46:19,040 --> 00:46:22,600
Speaker 1: with political power to do anything. Kat spoke to doctor

847
00:46:22,640 --> 00:46:26,160
Speaker 1: Mary Ann Franks, who drafted the template for laws against

848
00:46:26,160 --> 00:46:30,680
Speaker 1: non consensual distribution of intimate images, who said the FTC

849
00:46:30,719 --> 00:46:33,440
Speaker 1: has made it clear they're fighting for Trump. It's actually

850
00:46:33,440 --> 00:46:35,839
Speaker 1: never going to be used against the very players who

851
00:46:35,880 --> 00:46:36,839
Speaker 1: are the worst.

852
00:46:36,520 --> 00:46:37,240
Speaker 3: In this system.

853
00:46:37,880 --> 00:46:39,400
Speaker 1: X is going to continue to be one of the

854
00:46:39,400 --> 00:46:42,600
Speaker 1: worst offenders and probably one of the pioneers of horrible

855
00:46:42,640 --> 00:46:43,800
Speaker 1: ways to hurt women.

856
00:46:44,239 --> 00:46:48,719
Speaker 3: So I'm I usually try to find like optimism.

857
00:46:49,360 --> 00:46:51,000
Speaker 1: I don't have a lot of optimism for how the

858
00:46:51,080 --> 00:46:53,520
Speaker 1: United States is gonna handle this. I don't think anything's

859
00:46:53,560 --> 00:46:54,120
Speaker 1: going to happen.

860
00:46:54,280 --> 00:46:57,720
Speaker 4: I totally agree, And I think it's also worth pointing

861
00:46:57,760 --> 00:47:04,239
Speaker 4: out that the administration's support of X based on his

862
00:47:04,280 --> 00:47:09,280
Speaker 4: support of them is I have to imagine also related

863
00:47:09,320 --> 00:47:12,879
Speaker 4: to Apple and Google's ongoing decision to keep X in

864
00:47:13,040 --> 00:47:16,440
Speaker 4: their app stores, because were they to kick it out,

865
00:47:17,600 --> 00:47:20,560
Speaker 4: I think they would have very good reason to believe

866
00:47:20,600 --> 00:47:23,759
Speaker 4: that the administration might come after them for it. So

867
00:47:23,920 --> 00:47:29,440
Speaker 4: I think this is a good example problem of the

868
00:47:29,480 --> 00:47:35,920
Speaker 4: ways that this lawless administration, which is run by personal

869
00:47:36,120 --> 00:47:43,600
Speaker 4: grievance and loyalty, is changing our society and impacting major

870
00:47:43,719 --> 00:47:45,879
Speaker 4: decisions that make it less safe for all of us,

871
00:47:46,680 --> 00:47:49,160
Speaker 4: even without the government having to do anything. Just the

872
00:47:49,840 --> 00:47:56,560
Speaker 4: idea that they might come after private companies making decisions

873
00:47:56,560 --> 00:47:58,560
Speaker 4: that they don't like exactly.

874
00:47:59,200 --> 00:48:01,839
Speaker 1: And yeah, it's just in the context of having had

875
00:48:01,880 --> 00:48:08,719
Speaker 1: a whole like national conversation about pizzagate, QAnon, this like

876
00:48:09,080 --> 00:48:13,200
Speaker 1: the way that our government and tech leaders are coalescing

877
00:48:13,440 --> 00:48:18,680
Speaker 1: around doing nothing and like supporting.

878
00:48:19,040 --> 00:48:21,600
Speaker 3: People creating child sexual abuse material.

879
00:48:23,239 --> 00:48:27,840
Speaker 1: Like the conspiracy is happening in plain view right, Like

880
00:48:27,920 --> 00:48:30,520
Speaker 1: I mean like like, isn't this isn't this the thing?

881
00:48:31,239 --> 00:48:33,640
Speaker 4: Yeah, it's a good point. It's like right there, we

882
00:48:33,719 --> 00:48:36,719
Speaker 4: have all these states passing like age verification laws for

883
00:48:36,800 --> 00:48:43,200
Speaker 4: social media to protect the children, which research says are effective.

884
00:48:44,960 --> 00:48:47,040
Speaker 4: But like, here's a thing right in plain view that

885
00:48:47,160 --> 00:48:49,920
Speaker 4: is happening that we're all seeing, we're all talking about it,

886
00:48:50,320 --> 00:48:54,439
Speaker 4: and it's just crickets here in the United States from

887
00:48:54,440 --> 00:48:54,960
Speaker 4: the government.

888
00:48:55,160 --> 00:48:58,160
Speaker 1: Well, I will say our play cousins over in Europe

889
00:48:58,239 --> 00:49:01,920
Speaker 1: are not happy us. Earlier this week, a spokesperson for

890
00:49:02,000 --> 00:49:06,280
Speaker 1: the European Commission criticized the sexual explicit, non consensual images

891
00:49:06,440 --> 00:49:10,640
Speaker 1: that GROC is creating, calling them allegal, notably appalling, and

892
00:49:10,680 --> 00:49:13,360
Speaker 1: saying they have no place in Europe. Days later, the

893
00:49:13,440 --> 00:49:16,760
Speaker 1: EU ordered X to retain all internal documents and data

894
00:49:16,840 --> 00:49:20,080
Speaker 1: tied to GROC through twenty twenty six, extending an earlier

895
00:49:20,120 --> 00:49:23,080
Speaker 1: directive to preserve evidence relevant to the Digital Services Act,

896
00:49:23,320 --> 00:49:26,520
Speaker 1: even as no new formal probe had been announced. Regulators

897
00:49:26,560 --> 00:49:30,520
Speaker 1: in the UK, India, and Malaysia have also signaled investigations

898
00:49:30,560 --> 00:49:34,480
Speaker 1: into the platform. So the United States ain't gonna do shit. However,

899
00:49:34,600 --> 00:49:37,600
Speaker 1: it sounds like other countries are like, actually, yeah, we

900
00:49:37,680 --> 00:49:41,040
Speaker 1: don't think it's cool that people can just create illegal,

901
00:49:41,400 --> 00:49:45,080
Speaker 1: sexualized content of minors, and we have some questions about it.

902
00:49:46,080 --> 00:49:50,000
Speaker 4: Yeah, it's good that they're asking. They should get on that.

903
00:49:50,440 --> 00:49:53,200
Speaker 1: And just as you and I were sitting down to record,

904
00:49:53,280 --> 00:49:56,520
Speaker 1: we saw this news that Exit announced the ability to

905
00:49:56,560 --> 00:49:59,759
Speaker 1: create images with GROC was going to become restricted to

906
00:50:00,080 --> 00:50:03,239
Speaker 1: only users who pay for X premium. This is we

907
00:50:03,239 --> 00:50:04,800
Speaker 1: don't even need to discuss this as a fix to

908
00:50:04,840 --> 00:50:07,720
Speaker 1: the problem. This is obviously not a fix to the problem.

909
00:50:08,200 --> 00:50:10,120
Speaker 1: It just means that the ability to make this kind

910
00:50:10,200 --> 00:50:12,920
Speaker 1: of non consensual sexualized content is going to be a

911
00:50:12,920 --> 00:50:15,400
Speaker 1: premium service that people can pay for. So not only

912
00:50:15,440 --> 00:50:18,399
Speaker 1: will it not address the problem, it just means that

913
00:50:18,480 --> 00:50:21,799
Speaker 1: it's a money making opportunity for X. And so as

914
00:50:21,840 --> 00:50:24,560
Speaker 1: we were sitting down to record, the British government is

915
00:50:24,719 --> 00:50:27,239
Speaker 1: not impressed by this as a fix. They told The

916
00:50:27,239 --> 00:50:31,000
Speaker 1: Guardian this move simply turns an AI feature that allows

917
00:50:31,040 --> 00:50:34,879
Speaker 1: the creation of unlawful images into a premium service, which

918
00:50:35,280 --> 00:50:38,520
Speaker 1: I completely agree. And something that I was wondering about is,

919
00:50:38,920 --> 00:50:41,800
Speaker 1: you know, I know that X has really been struggling

920
00:50:41,880 --> 00:50:44,719
Speaker 1: in terms of making money retaining users for all the

921
00:50:44,760 --> 00:50:47,680
Speaker 1: reasons that we talked about advertising all of that, These

922
00:50:47,800 --> 00:50:51,120
Speaker 1: kinds of AI services are pricey.

923
00:50:51,160 --> 00:50:52,239
Speaker 3: They're expensive, and so.

924
00:50:52,200 --> 00:50:55,040
Speaker 1: One of my questions is I wonder if, like maybe

925
00:50:55,239 --> 00:50:58,719
Speaker 1: X had been mulling a decision like this over for

926
00:50:58,760 --> 00:51:01,719
Speaker 1: a while as a way to cut costs associated with

927
00:51:01,760 --> 00:51:04,520
Speaker 1: groc This is just my theory. I don't have any

928
00:51:04,800 --> 00:51:07,640
Speaker 1: you know, this is just speculation, but I wonder if

929
00:51:07,640 --> 00:51:09,600
Speaker 1: they had already been thinking about this and they were like, oh,

930
00:51:09,600 --> 00:51:11,080
Speaker 1: we can just say that we're doing it as a

931
00:51:11,080 --> 00:51:14,640
Speaker 1: fix to this AI sexualized images problem. But really we

932
00:51:14,640 --> 00:51:17,800
Speaker 1: were thinking about doing this as say cost cutting solution

933
00:51:18,120 --> 00:51:18,640
Speaker 1: all along.

934
00:51:19,320 --> 00:51:21,440
Speaker 4: I wondered the same thing. You know, we've been seeing

935
00:51:21,960 --> 00:51:24,560
Speaker 4: a lot of news stories about the big AI players,

936
00:51:24,600 --> 00:51:31,080
Speaker 4: including open Ai, really struggling to get cost under control

937
00:51:31,320 --> 00:51:34,920
Speaker 4: and make revenue. And I mean, I have to imagine

938
00:51:35,080 --> 00:51:38,480
Speaker 4: that all creating all of these images with GROC is

939
00:51:39,400 --> 00:51:44,359
Speaker 4: costing X a fortune. So yeah, I wouldered the same

940
00:51:44,400 --> 00:51:48,600
Speaker 4: thing if that move was something that they had maybe

941
00:51:48,640 --> 00:51:51,120
Speaker 4: been thinking about already and they thought that maybe they

942
00:51:51,120 --> 00:51:54,600
Speaker 4: could What did you say, liberate two birds with one stone?

943
00:51:55,440 --> 00:51:58,640
Speaker 1: Oh, free two birds with one key. I so I

944
00:51:58,760 --> 00:52:01,440
Speaker 1: you know the expression killed two birds with one stone.

945
00:52:01,560 --> 00:52:03,480
Speaker 1: I don't know where I heard it, but someone was like, oh,

946
00:52:03,520 --> 00:52:06,799
Speaker 1: another way to put that is free two birds with

947
00:52:06,840 --> 00:52:08,759
Speaker 1: one key, And I thought, what a what a nice

948
00:52:08,800 --> 00:52:11,600
Speaker 1: way of like reframing, like why am I killing birds here?

949
00:52:13,280 --> 00:52:15,879
Speaker 4: Well, because you're hungry, so you know.

950
00:52:18,160 --> 00:52:20,040
Speaker 3: Freeing them to eat them and put them on my plate.

951
00:52:21,560 --> 00:52:27,200
Speaker 1: Ultimately, you know, I'm fired up, I'm angry. This is

952
00:52:27,200 --> 00:52:30,200
Speaker 1: a tech podcast, but I don't see this as just

953
00:52:30,280 --> 00:52:33,360
Speaker 1: a tech issue. I really think it is a cultural issue.

954
00:52:33,360 --> 00:52:35,479
Speaker 1: People who have been reporting on deep fakes since before

955
00:52:35,520 --> 00:52:37,399
Speaker 1: we had a word for them. Folks like Samantha Cole

956
00:52:37,719 --> 00:52:41,399
Speaker 1: have pointed this out. Cole writes in twenty eighteen, less

957
00:52:41,400 --> 00:52:44,000
Speaker 1: than a year after reporting the first story on deep fakes,

958
00:52:44,239 --> 00:52:46,640
Speaker 1: I wrote about how it's a serious mistake to ignore

959
00:52:46,680 --> 00:52:50,040
Speaker 1: the fact that non consensual imagery, synthetic or not, is

960
00:52:50,080 --> 00:52:53,880
Speaker 1: a societal sickness and not something companies can guard rail against.

961
00:52:53,960 --> 00:52:56,879
Speaker 1: Into infinity, users feed off of each other to create

962
00:52:56,920 --> 00:52:59,160
Speaker 1: a sense that they are the kings of the universe,

963
00:52:59,239 --> 00:53:01,920
Speaker 1: that they answer to no one. This logic is how

964
00:53:01,960 --> 00:53:04,160
Speaker 1: you get in cells and pick up artists. It's how

965
00:53:04,200 --> 00:53:06,480
Speaker 1: you get deep fakes, a group of men who see

966
00:53:06,560 --> 00:53:09,839
Speaker 1: no harm in treating women as merror images and view

967
00:53:09,960 --> 00:53:13,880
Speaker 1: making and spreading algorithmically weaponized revenge porn as a hobby,

968
00:53:14,160 --> 00:53:17,959
Speaker 1: as innocent and timeless as trading baseball cards. I wrote

969
00:53:17,960 --> 00:53:20,200
Speaker 1: at the time, this is what it's at the root

970
00:53:20,239 --> 00:53:22,879
Speaker 1: of deep fakes, and the consequences of forgetting that are

971
00:53:22,920 --> 00:53:25,879
Speaker 1: more dire than we can predict. And I think that

972
00:53:26,080 --> 00:53:30,560
Speaker 1: really says it all to me. You know, I've been

973
00:53:30,600 --> 00:53:34,560
Speaker 1: thinking a lot about the sort of societal and cultural

974
00:53:34,840 --> 00:53:37,480
Speaker 1: rot at the heart of this issue, and I keep

975
00:53:37,520 --> 00:53:39,640
Speaker 1: coming back to it again and again and again.

976
00:53:39,760 --> 00:53:44,280
Speaker 3: And you know, this is not like a fully fleshed.

977
00:53:43,920 --> 00:53:47,520
Speaker 1: Out idea, but I think the way that AI is

978
00:53:47,600 --> 00:53:50,640
Speaker 1: being used just really illustrates that.

979
00:53:50,719 --> 00:53:50,879
Speaker 2: Right.

980
00:53:50,960 --> 00:53:54,800
Speaker 1: Like, when Google released nano Banana Pro, their AI generator

981
00:53:54,920 --> 00:53:58,520
Speaker 1: that makes like very realistic looking AI images, I couldn't

982
00:53:58,520 --> 00:54:01,880
Speaker 1: stop noticing how so oh many of the examples of

983
00:54:01,960 --> 00:54:04,440
Speaker 1: images that people had used it to create were just

984
00:54:04,640 --> 00:54:08,960
Speaker 1: conventionally hot women in like yoga pants and workout a

985
00:54:09,040 --> 00:54:11,759
Speaker 1: tire and things like that, doing mirror selfies. If you

986
00:54:11,840 --> 00:54:15,040
Speaker 1: take a look at the nanobanana pro supreddit, at least

987
00:54:15,040 --> 00:54:18,400
Speaker 1: when it first started, it was this post after post

988
00:54:18,440 --> 00:54:23,839
Speaker 1: after post of AI generated women, and it just reminds

989
00:54:23,960 --> 00:54:27,880
Speaker 1: me that, you know, the moment that these image generators

990
00:54:27,920 --> 00:54:31,000
Speaker 1: were first becoming mainstream, some men were talking about how

991
00:54:31,040 --> 00:54:34,440
Speaker 1: they wanted to use them to create fake OnlyFans models,

992
00:54:34,680 --> 00:54:37,000
Speaker 1: and they were really trafficking in this idea that doing

993
00:54:37,040 --> 00:54:40,400
Speaker 1: that would be taking the power back from women and

994
00:54:40,680 --> 00:54:43,399
Speaker 1: like taking power from women and giving it back to men, right,

995
00:54:43,760 --> 00:54:46,120
Speaker 1: even though side note that's not really how only Fans works,

996
00:54:46,120 --> 00:54:49,680
Speaker 1: but whatever, but I think that impulse was very telling, right,

997
00:54:49,719 --> 00:54:53,160
Speaker 1: if women are able to build some economic agency by

998
00:54:53,200 --> 00:54:56,960
Speaker 1: consentually creating their own sexual content. There's resentment for that right,

999
00:54:57,000 --> 00:54:59,239
Speaker 1: and so it's not just I won't pay for that,

1000
00:54:59,320 --> 00:55:00,400
Speaker 1: I won't engage with that.

1001
00:55:00,480 --> 00:55:01,560
Speaker 3: It becomes I.

1002
00:55:01,480 --> 00:55:04,440
Speaker 1: Will use AI so that she cannot have that agency

1003
00:55:04,440 --> 00:55:07,920
Speaker 1: at all, or worse, I will use AI to undress

1004
00:55:08,000 --> 00:55:09,080
Speaker 1: women without their consent.

1005
00:55:09,520 --> 00:55:12,919
Speaker 3: You know. It's just like these prompts on xay make

1006
00:55:13,000 --> 00:55:14,040
Speaker 3: her put her.

1007
00:55:14,400 --> 00:55:18,840
Speaker 1: It's all about controlling women and stripping us of whatever

1008
00:55:18,880 --> 00:55:20,880
Speaker 1: agency we do have in society.

1009
00:55:21,360 --> 00:55:23,960
Speaker 3: And I think Samantha.

1010
00:55:23,560 --> 00:55:27,440
Speaker 1: Cole really nails it that it's not just a tech issue.

1011
00:55:27,480 --> 00:55:29,200
Speaker 3: It is a culture issue.

1012
00:55:29,320 --> 00:55:33,640
Speaker 1: It is a societal rock issue that I think we

1013
00:55:33,840 --> 00:55:37,920
Speaker 1: can and should be holding companies accountable to creating guardrails

1014
00:55:37,920 --> 00:55:39,920
Speaker 1: for this kind of thing. But this is not a

1015
00:55:39,960 --> 00:55:43,120
Speaker 1: problem that you can guardrail out of being an issue

1016
00:55:43,200 --> 00:55:47,520
Speaker 1: unless we address that rot And something that you said earlier, Mike,

1017
00:55:47,600 --> 00:55:52,160
Speaker 1: that you know this idea that you know platforms like

1018
00:55:52,360 --> 00:55:55,480
Speaker 1: X are going to be our town square where ideas

1019
00:55:55,520 --> 00:55:56,959
Speaker 1: can be debated and all of that.

1020
00:55:57,760 --> 00:56:00,440
Speaker 3: I don't believe that about X anymore.

1021
00:56:00,960 --> 00:56:04,680
Speaker 1: However, it is true that social media platforms in twenty

1022
00:56:04,719 --> 00:56:07,279
Speaker 1: twenty six are increasingly just like where people show up.

1023
00:56:07,280 --> 00:56:08,880
Speaker 1: If you want to engage civically, if you want to

1024
00:56:08,880 --> 00:56:11,879
Speaker 1: build a business, you know, in twenty twenty six, if

1025
00:56:11,920 --> 00:56:15,359
Speaker 1: you want to be civically engaged, engaged in your democracy

1026
00:56:15,360 --> 00:56:19,720
Speaker 1: in robust ways, build businesses, showing up on social media

1027
00:56:19,800 --> 00:56:21,920
Speaker 1: is just part of that right. And so if the

1028
00:56:22,160 --> 00:56:25,200
Speaker 1: advice that we give to women is don't show up

1029
00:56:25,200 --> 00:56:28,359
Speaker 1: on these platforms, what we're actually saying is don't show

1030
00:56:28,440 --> 00:56:30,680
Speaker 1: up civically. Don't show up in your own democracy. It's

1031
00:56:30,719 --> 00:56:32,839
Speaker 1: not safe. And I think that that's part of what's

1032
00:56:32,920 --> 00:56:37,640
Speaker 1: going on here. It's a way of erasing women from

1033
00:56:37,880 --> 00:56:41,839
Speaker 1: civic and public life by doing this. And I think that,

1034
00:56:41,920 --> 00:56:45,239
Speaker 1: like you know, Elon Musk is obsessed with talking about

1035
00:56:45,280 --> 00:56:47,359
Speaker 1: free speech. If I don't show up to the town

1036
00:56:47,400 --> 00:56:49,279
Speaker 1: square because somebody's going to yank my top off if

1037
00:56:49,320 --> 00:56:50,960
Speaker 1: I do, what about my speech?

1038
00:56:51,040 --> 00:56:53,480
Speaker 3: Is not speech? Is that not speech being suppressed?

1039
00:56:54,200 --> 00:56:57,439
Speaker 1: And I completely agree with Samantha Cole that I think

1040
00:56:57,520 --> 00:56:59,920
Speaker 1: all of this is really coming from the same place,

1041
00:57:00,360 --> 00:57:03,240
Speaker 1: a desire to live in a world where women exist

1042
00:57:03,320 --> 00:57:08,080
Speaker 1: primarily to be consumed, controlled, and stripped of whatever agency

1043
00:57:08,200 --> 00:57:11,040
Speaker 1: we have managed to claim over our own bodies and lives,

1044
00:57:11,280 --> 00:57:14,560
Speaker 1: and I think until we confront that reality, Samantha Cole

1045
00:57:14,600 --> 00:57:17,280
Speaker 1: is exactly right that I don't think any amount of

1046
00:57:17,360 --> 00:57:20,800
Speaker 1: technical safeguards alone, even though we should be advocating for them,

1047
00:57:21,280 --> 00:57:27,360
Speaker 1: is going to fix that problem. Got a story about

1048
00:57:27,360 --> 00:57:29,439
Speaker 1: an interesting thing in tech, or just want to say hi.

1049
00:57:29,840 --> 00:57:32,080
Speaker 1: You can reach us at Hello at tengody dot com.

1050
00:57:32,120 --> 00:57:34,520
Speaker 1: You can also find transcripts for today's episode at tenggody

1051
00:57:34,600 --> 00:57:35,280
Speaker 1: dot com.

1052
00:57:35,440 --> 00:57:37,120
Speaker 3: There Are No Girls on the Internet was created by

1053
00:57:37,120 --> 00:57:37,880
Speaker 3: me Bridget Toad.

1054
00:57:38,280 --> 00:57:41,680
Speaker 1: It's a production of iHeartRadio and Unbossed creative Jonathan Strickland

1055
00:57:41,720 --> 00:57:44,440
Speaker 1: is our executive producer. Tari Harrison is our producer and

1056
00:57:44,520 --> 00:57:48,280
Speaker 1: sound engineer. Michael Almado is our contributing producer. I'm your host,

1057
00:57:48,280 --> 00:57:51,080
Speaker 1: Bridget Todd. If you want to help us grow, rate

1058
00:57:51,120 --> 00:57:54,840
Speaker 1: and review us on Apple Podcasts. For more podcasts from iHeartRadio,

1059
00:57:54,960 --> 00:57:57,320
Speaker 1: check out the iHeartRadio app, Apple Podcasts, or wherever you

1060
00:57:57,320 --> 00:58:09,000
Speaker 1: get your podcasts. H rooms, ah ros