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

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

3
00:00:13,520 --> 00:00:18,520
Speaker 1: is there are No Girls on the Internet. Not long ago,

4
00:00:18,760 --> 00:00:22,479
Speaker 1: Sam Altman was openly bragging about how open ai does

5
00:00:22,520 --> 00:00:26,520
Speaker 1: not make sex spots. Then just two months later announced

6
00:00:26,520 --> 00:00:30,440
Speaker 1: they would be rolling out erotic content on chatchipt. So

7
00:00:30,840 --> 00:00:34,240
Speaker 1: how's that pivot into adult content going, In the words

8
00:00:34,240 --> 00:00:38,720
Speaker 1: of Dereinda Medley, not well, bitch. Soon after the announcement,

9
00:00:38,800 --> 00:00:41,960
Speaker 1: there was a high profile exit from open Ai over

10
00:00:42,080 --> 00:00:46,680
Speaker 1: concerns the company was not handling the adult content pivot responsibly.

11
00:00:47,520 --> 00:00:51,600
Speaker 1: And now Sam Altman says that pivot has been indefinitely postponed.

12
00:00:52,159 --> 00:00:54,960
Speaker 1: So what does all this mean. I sat down with

13
00:00:55,000 --> 00:00:57,600
Speaker 1: my friends Samantha and Anny over at the podcast stuff.

14
00:00:57,680 --> 00:00:59,400
Speaker 1: Mom never told you to break it all down.

15
00:01:05,760 --> 00:01:07,280
Speaker 2: Hey, this is Anny and Smantha.

16
00:01:07,360 --> 00:01:09,080
Speaker 3: I'm welcome to Steffan I ever told you a production

17
00:01:09,080 --> 00:01:21,160
Speaker 3: of Buyheart Radio, and we are so excited to be

18
00:01:21,360 --> 00:01:25,640
Speaker 3: joined once again by the excellent, the eloquent Brigitte Todd.

19
00:01:25,800 --> 00:01:29,400
Speaker 1: Welcome Brigitte, thanks for having me. I am so pumped

20
00:01:29,400 --> 00:01:30,480
Speaker 1: to be here, miss you.

21
00:01:30,480 --> 00:01:30,800
Speaker 4: Guys.

22
00:01:31,400 --> 00:01:34,840
Speaker 3: Yes, we have missed you as well, and you have

23
00:01:35,040 --> 00:01:38,160
Speaker 3: been busy. Can you tell the listeners what you have

24
00:01:38,319 --> 00:01:39,039
Speaker 3: been up to.

25
00:01:39,760 --> 00:01:42,840
Speaker 1: Yes, I have that a little bit busy because I

26
00:01:43,000 --> 00:01:46,520
Speaker 1: just finished an audio book with Simon and Schuster called

27
00:01:46,640 --> 00:01:50,400
Speaker 1: Love at First Prompt is an exploration of AI and intimacy,

28
00:01:50,600 --> 00:01:54,960
Speaker 1: AI and romantic connection, sexual connection, and really sort of

29
00:01:54,960 --> 00:01:57,800
Speaker 1: asked the question, do we what does it mean when

30
00:01:57,800 --> 00:02:02,320
Speaker 1: these companies like open AI own and control something that

31
00:02:02,440 --> 00:02:05,880
Speaker 1: is so innately human and so sensitive as our intimacies

32
00:02:05,880 --> 00:02:06,800
Speaker 1: that we have with each other.

33
00:02:07,840 --> 00:02:13,880
Speaker 3: Yes, and it is a fascinating question. And I'm really

34
00:02:13,880 --> 00:02:18,760
Speaker 3: really excited to hear the book and maybe come back

35
00:02:18,800 --> 00:02:22,919
Speaker 3: and discuss your experience doing the audio book. But yeah,

36
00:02:23,000 --> 00:02:25,480
Speaker 3: I mean, do you mind telling us a little bit

37
00:02:25,600 --> 00:02:29,560
Speaker 3: about what was the impetus behind this idea?

38
00:02:30,320 --> 00:02:33,280
Speaker 1: Yeah? So on my own tech podcasts there are no

39
00:02:33,440 --> 00:02:36,400
Speaker 1: girls on the Internet. I was really interested in exploring

40
00:02:36,440 --> 00:02:39,200
Speaker 1: the ways that people were reporting kind of having these

41
00:02:39,240 --> 00:02:42,520
Speaker 1: intimate connections with AI. This was around the same time

42
00:02:42,680 --> 00:02:46,760
Speaker 1: that chatchept was getting rid of their four to oh model,

43
00:02:46,800 --> 00:02:49,840
Speaker 1: which you might remember with the model that was very

44
00:02:49,880 --> 00:02:53,400
Speaker 1: sort of sycophantic, you know, it was like very flattering

45
00:02:53,760 --> 00:02:57,639
Speaker 1: and phasing in a different model that was less sycophantic,

46
00:02:57,840 --> 00:03:01,480
Speaker 1: less flattering, and people kind of when that change happened,

47
00:03:01,840 --> 00:03:07,160
Speaker 1: people a lot of people realized, oh, my connection with

48
00:03:07,240 --> 00:03:09,760
Speaker 1: chat GPT was not just it helping me with homework

49
00:03:09,800 --> 00:03:12,040
Speaker 1: or helping me with work. I actually had developed an

50
00:03:12,080 --> 00:03:15,000
Speaker 1: emotional connection with it, and now that this model is gone,

51
00:03:15,639 --> 00:03:19,120
Speaker 1: I'm struggling with that. And so I was really fascinated

52
00:03:19,160 --> 00:03:22,200
Speaker 1: by that, and we did an episode all about it.

53
00:03:22,919 --> 00:03:24,960
Speaker 1: In our episode, we talked about people who were using

54
00:03:25,000 --> 00:03:30,760
Speaker 1: AI for you know, romantic connection, sexual connection. And I

55
00:03:30,760 --> 00:03:33,520
Speaker 1: think it's a sensitive topic because it's very easy to

56
00:03:34,600 --> 00:03:37,000
Speaker 1: go get these people and to say like, oh this,

57
00:03:37,080 --> 00:03:40,320
Speaker 1: these people are delusional and like they're lonely and all

58
00:03:40,360 --> 00:03:43,920
Speaker 1: of this, and that I completely understand why that is

59
00:03:44,240 --> 00:03:48,200
Speaker 1: a lot of people's first instinct. However, in my view,

60
00:03:48,880 --> 00:03:54,160
Speaker 1: that really shields the companies that run these products from accountability, right,

61
00:03:54,200 --> 00:03:56,440
Speaker 1: and so we really wanted to shift the lens and say, well,

62
00:03:57,120 --> 00:03:59,880
Speaker 1: let's not just focus on the individual users and what

63
00:04:00,080 --> 00:04:02,760
Speaker 1: how they're using AI and all of that. Let's talk

64
00:04:02,760 --> 00:04:05,280
Speaker 1: about these companies and whether or not they're behaving ethically.

65
00:04:05,440 --> 00:04:08,600
Speaker 1: Because I really don't like how in technology. It is

66
00:04:08,680 --> 00:04:12,280
Speaker 1: so easy to blame people who get caught up in

67
00:04:12,360 --> 00:04:14,880
Speaker 1: a certain kind of technology. In twenty fourteen, when there

68
00:04:14,920 --> 00:04:18,000
Speaker 1: was the iCloud photo hacks, how many times people say, oh,

69
00:04:18,040 --> 00:04:22,400
Speaker 1: these women have not taking these pictures? What about asking well,

70
00:04:23,000 --> 00:04:26,440
Speaker 1: what platforms actually enabled this? What platforms profited off of this?

71
00:04:26,839 --> 00:04:31,120
Speaker 1: It is so easy to scrutinize the behavior of individuals,

72
00:04:31,200 --> 00:04:33,640
Speaker 1: but really we ought to be asking questions about these companies.

73
00:04:33,640 --> 00:04:36,200
Speaker 1: And so that's kind of how this project came to be.

74
00:04:36,400 --> 00:04:40,200
Speaker 1: After I did that episode, Simon and Schuster, my book publisher,

75
00:04:40,480 --> 00:04:42,799
Speaker 1: got in touch and said, this topic is so interesting,

76
00:04:43,200 --> 00:04:45,920
Speaker 1: what if you expanded on it for a full audiobook project?

77
00:04:46,000 --> 00:04:49,240
Speaker 1: And I did. Side note, one of my favorite movies

78
00:04:49,240 --> 00:04:52,520
Speaker 1: of all time is the movie Her and the way

79
00:04:52,560 --> 00:04:58,919
Speaker 1: that the whole open AI changing Models situation unfolded Like

80
00:04:59,080 --> 00:05:02,960
Speaker 1: that movie, truly, Smike Jones must have seen something common

81
00:05:03,000 --> 00:05:07,960
Speaker 1: because the way that things unfold in that film basically

82
00:05:08,000 --> 00:05:09,640
Speaker 1: are how things unfolded in real life.

83
00:05:11,000 --> 00:05:13,360
Speaker 4: Well, I have to say, first and foremost, the title

84
00:05:13,400 --> 00:05:14,839
Speaker 4: of the book is genius like.

85
00:05:14,920 --> 00:05:16,800
Speaker 2: It caught me very quickly.

86
00:05:16,839 --> 00:05:20,159
Speaker 4: I was like, oh, okay, I'm already I'm glued I'm here,

87
00:05:20,200 --> 00:05:23,320
Speaker 4: I'm seated. Let's let's talk about this. So how did

88
00:05:23,320 --> 00:05:24,279
Speaker 4: you even come up with the title.

89
00:05:24,640 --> 00:05:26,599
Speaker 1: Honestly, it might have been one of the hardest parts

90
00:05:26,600 --> 00:05:30,120
Speaker 1: of doing the book. I was trying to find a

91
00:05:30,160 --> 00:05:33,159
Speaker 1: title that conveyed what the book is about, conveyed that

92
00:05:33,360 --> 00:05:37,800
Speaker 1: it's we're exploring this from a kind of accessible, casual perspective,

93
00:05:38,120 --> 00:05:40,080
Speaker 1: but also was like a little bit of a nod

94
00:05:40,120 --> 00:05:43,360
Speaker 1: to being kind of cheeky or kind of funny. Initially,

95
00:05:43,400 --> 00:05:46,000
Speaker 1: I wanted the title to be AI Will Always Love You,

96
00:05:46,360 --> 00:05:48,320
Speaker 1: and my editor was like, I don't know if people

97
00:05:48,320 --> 00:05:50,200
Speaker 1: are going to get the Whitney Houston reference, but.

98
00:05:50,520 --> 00:05:55,040
Speaker 4: Yeah, I love the song immediately jumping thank you love at.

99
00:05:55,000 --> 00:05:58,880
Speaker 1: First prompt was really the kind of marrying of the

100
00:05:58,920 --> 00:06:01,800
Speaker 1: worlds of kind of telling people what it's about but

101
00:06:01,839 --> 00:06:03,920
Speaker 1: also being a little bit cheeky. So thank you it was.

102
00:06:04,120 --> 00:06:06,080
Speaker 1: It was a real struggle to come up with that title.

103
00:06:06,560 --> 00:06:08,160
Speaker 2: Well, kudos because it is good.

104
00:06:09,080 --> 00:06:13,520
Speaker 3: Yes, Also, I know you're busy, but I maintained I'm

105
00:06:13,560 --> 00:06:18,440
Speaker 3: starting to think we should do a movie series of

106
00:06:18,560 --> 00:06:21,440
Speaker 3: tech and movies we started the social network.

107
00:06:21,600 --> 00:06:25,320
Speaker 1: Yeah to her, I have said this before, that is

108
00:06:25,400 --> 00:06:28,680
Speaker 1: my I was a goal in life. But I've always

109
00:06:28,720 --> 00:06:31,680
Speaker 1: wanted to have one of those movie podcasts where you

110
00:06:31,839 --> 00:06:34,400
Speaker 1: just like talk through a movie and all the implications.

111
00:06:34,480 --> 00:06:37,000
Speaker 1: So I listened to a ton of those podcasts. I

112
00:06:37,040 --> 00:06:38,839
Speaker 1: feel like I was put on this earth to do

113
00:06:38,880 --> 00:06:42,120
Speaker 1: a deep to do deep dives into movies on podcasts.

114
00:06:42,560 --> 00:06:44,960
Speaker 4: You know, we could just do a whole crossover for

115
00:06:45,000 --> 00:06:47,280
Speaker 4: both your show and our show, where we do one

116
00:06:48,120 --> 00:06:51,080
Speaker 4: every now and again, just like every quarter of the year,

117
00:06:51,120 --> 00:06:52,640
Speaker 4: and be like, look, we're gonna take a break and

118
00:06:52,680 --> 00:06:54,880
Speaker 4: for this episode, we're gonna do this as a crossover.

119
00:06:55,200 --> 00:06:58,839
Speaker 1: Yes, I would love that. And our pretty Sir Joey

120
00:06:58,880 --> 00:07:02,040
Speaker 1: Patty also are kind of a movie person.

121
00:07:02,040 --> 00:07:02,360
Speaker 2: In Yill.

122
00:07:02,640 --> 00:07:06,719
Speaker 1: You would be surprised how often the intersection between like

123
00:07:06,800 --> 00:07:10,960
Speaker 1: technology and pop culture happens. There are so we're we're

124
00:07:11,000 --> 00:07:14,960
Speaker 1: referencing movies all the time X Companion, Like, those are

125
00:07:14,960 --> 00:07:17,600
Speaker 1: both movies her, those are all movies that come up

126
00:07:17,640 --> 00:07:22,000
Speaker 1: in my audiobook. Because so much of what we experience

127
00:07:22,080 --> 00:07:25,560
Speaker 1: when it comes to humans connection with AI, we've actually

128
00:07:25,600 --> 00:07:27,840
Speaker 1: got a little bit of a template in fictional film,

129
00:07:27,880 --> 00:07:29,240
Speaker 1: which I just find so interesting.

130
00:07:30,120 --> 00:07:31,040
Speaker 2: I think it needs to happen.

131
00:07:31,360 --> 00:07:33,280
Speaker 3: Yes, I think we should make it happen. I think

132
00:07:33,280 --> 00:07:34,119
Speaker 3: we should make it happen.

133
00:07:34,160 --> 00:07:34,800
Speaker 1: I would love that.

134
00:07:36,840 --> 00:07:39,880
Speaker 3: But one of the things that we were discussing off

135
00:07:40,000 --> 00:07:45,920
Speaker 3: Mike Bridget is that this is an extremely timely topic

136
00:07:46,880 --> 00:07:51,440
Speaker 3: that you have chosen for this book, as in, like,

137
00:07:51,720 --> 00:07:56,120
Speaker 3: right now open ai is doing some stuff.

138
00:07:56,880 --> 00:07:59,880
Speaker 1: Yes, that is exactly right. It's one of the reasons

139
00:08:00,120 --> 00:08:03,120
Speaker 1: I had to write the book lickety split because it

140
00:08:03,240 --> 00:08:05,680
Speaker 1: changed the world of ai changes so quickly, and like

141
00:08:05,720 --> 00:08:07,320
Speaker 1: I would write things and then a week later that

142
00:08:07,360 --> 00:08:10,640
Speaker 1: would be outdated. But just we actually even have a

143
00:08:10,680 --> 00:08:14,680
Speaker 1: little news from open ai recently. Back in October, Sam Altman,

144
00:08:14,680 --> 00:08:16,920
Speaker 1: who was the CEO of open Ai, the company that

145
00:08:16,960 --> 00:08:21,040
Speaker 1: makes chat GPT, they did this pretty big about face.

146
00:08:21,120 --> 00:08:24,920
Speaker 1: They said they were going to roll out erotic content

147
00:08:25,360 --> 00:08:29,120
Speaker 1: sometime this year. And this is a bit of a

148
00:08:29,280 --> 00:08:33,040
Speaker 1: change because for the longest time open Ai was saying

149
00:08:33,440 --> 00:08:36,640
Speaker 1: we don't do sex bots, we don't do erotic content,

150
00:08:36,679 --> 00:08:39,079
Speaker 1: we don't do any of that. And then it didn't

151
00:08:39,080 --> 00:08:41,640
Speaker 1: take Literally Sam Altman said that on a podcast, and

152
00:08:41,920 --> 00:08:44,200
Speaker 1: I think it was two months later it was like, actually,

153
00:08:44,240 --> 00:08:48,280
Speaker 1: we're doing sex bots now, So pretty pretty big about

154
00:08:48,320 --> 00:08:52,440
Speaker 1: face there, And so this decision to roll out erotic

155
00:08:52,480 --> 00:08:56,000
Speaker 1: content kind of came at an awkward time. It was

156
00:08:56,440 --> 00:08:59,120
Speaker 1: on the heels of grappling with what I was talking

157
00:08:59,160 --> 00:09:03,400
Speaker 1: about earlier. Change from chatgybt four Oh, the model that

158
00:09:03,559 --> 00:09:06,480
Speaker 1: was known for flattery and sick of fancy that, that's

159
00:09:06,480 --> 00:09:09,640
Speaker 1: the model that if you ask a question, chatgypt will say,

160
00:09:10,640 --> 00:09:14,000
Speaker 1: that's such a brilliant question. When you ask something totally basic,

161
00:09:14,040 --> 00:09:19,040
Speaker 1: it'll just blow smoke up your So when open ai

162
00:09:19,200 --> 00:09:21,240
Speaker 1: announced that they were rolling out a new model to

163
00:09:21,280 --> 00:09:24,920
Speaker 1: be less fonding and less human, people realize like, hey,

164
00:09:24,960 --> 00:09:29,240
Speaker 1: I have an emotional or an intimate connection with this

165
00:09:29,360 --> 00:09:33,200
Speaker 1: version of chatgybt, And so many of those people felt

166
00:09:33,240 --> 00:09:37,439
Speaker 1: like open Ai had acted callously, just kind of making

167
00:09:37,480 --> 00:09:41,679
Speaker 1: the change to to to change the model. That's it's

168
00:09:41,720 --> 00:09:44,920
Speaker 1: a whole like interesting story that we get into in

169
00:09:44,960 --> 00:09:48,240
Speaker 1: the book. But around that same time is when All

170
00:09:48,320 --> 00:09:50,920
Speaker 1: Men announced they were going to start doing erotic content.

171
00:09:51,240 --> 00:09:54,040
Speaker 1: They put out an announcement that said, now that we've

172
00:09:54,040 --> 00:09:56,600
Speaker 1: been able to mitigate serious mental health issues and have

173
00:09:56,679 --> 00:09:59,559
Speaker 1: new tools, we're going to be able to safely relax

174
00:09:59,559 --> 00:10:02,560
Speaker 1: the restraint in most cases as part of our treat

175
00:10:02,600 --> 00:10:06,000
Speaker 1: adult users like adults. Principle, we will allow even more

176
00:10:06,080 --> 00:10:08,240
Speaker 1: like erotica for verified adults.

177
00:10:10,200 --> 00:10:15,160
Speaker 4: I have so many questions, like, how did you mitigate

178
00:10:15,240 --> 00:10:19,040
Speaker 4: this mental health like within how where did you get

179
00:10:19,040 --> 00:10:19,720
Speaker 4: this information?

180
00:10:21,200 --> 00:10:22,000
Speaker 2: I'm confused.

181
00:10:22,920 --> 00:10:26,320
Speaker 1: That's a really good question, because that's basically what open

182
00:10:26,360 --> 00:10:28,960
Speaker 1: ai is saying, is that we had this model that

183
00:10:29,000 --> 00:10:33,680
Speaker 1: people were forming connections with, even though side note, I

184
00:10:33,720 --> 00:10:36,880
Speaker 1: would argue that Sam Altman actually encouraged people to form

185
00:10:37,080 --> 00:10:40,600
Speaker 1: intimate connections with that model because one of the templates

186
00:10:40,600 --> 00:10:45,400
Speaker 1: he said publicly that he wanted people to experience CHATGBT

187
00:10:45,720 --> 00:10:48,480
Speaker 1: the way that people experience AI in the movie Her.

188
00:10:48,960 --> 00:10:52,400
Speaker 1: If you've seen the movie Her, notably, it's humans having

189
00:10:52,640 --> 00:10:55,800
Speaker 1: very intimate, romantic and sexual connections with AI. So in

190
00:10:55,840 --> 00:10:58,040
Speaker 1: my book, you can't really set that as a template

191
00:10:58,040 --> 00:11:01,000
Speaker 1: and then be surprised when people end up doing exactly

192
00:11:01,040 --> 00:11:02,520
Speaker 1: what you said that you hope that they would do.

193
00:11:02,960 --> 00:11:05,760
Speaker 1: But now that they've changed the model and made the

194
00:11:05,800 --> 00:11:09,520
Speaker 1: model less flattering, less to cathantic, Basically, open ai is

195
00:11:09,520 --> 00:11:12,360
Speaker 1: saying there is no need for chat GPT to be

196
00:11:12,480 --> 00:11:16,000
Speaker 1: so restrictive anymore because they've worked out those kinks. They're saying, listen,

197
00:11:16,000 --> 00:11:19,360
Speaker 1: we've got better safety systems. We've got improved monitoring, we've

198
00:11:19,360 --> 00:11:23,480
Speaker 1: got more robust age verification, and now verified adults can

199
00:11:23,520 --> 00:11:26,400
Speaker 1: be treated like adults. To be clear, I'm not saying

200
00:11:26,400 --> 00:11:28,200
Speaker 1: that because I don't happen to agree with that position,

201
00:11:28,240 --> 00:11:30,680
Speaker 1: but that is what that is the position that Sam

202
00:11:30,760 --> 00:11:33,520
Speaker 1: Altman and open ai is taking. As to listen, we

203
00:11:33,559 --> 00:11:36,240
Speaker 1: fixed the problem, now we can have adult content.

204
00:11:37,559 --> 00:11:41,760
Speaker 3: Yeah, I maintain I have a lot of concerns. I

205
00:11:41,800 --> 00:11:44,080
Speaker 3: don't know what these guardrails are, but I don't think

206
00:11:44,120 --> 00:11:48,320
Speaker 3: that they're working. And also just a lot of that

207
00:11:49,679 --> 00:11:55,480
Speaker 3: ah just don't verification and erotica. We've already talked about

208
00:11:55,480 --> 00:11:59,000
Speaker 3: this on past episodes about how this can go wrong,

209
00:12:00,720 --> 00:12:06,280
Speaker 3: But why do you think open ai is Why did

210
00:12:06,360 --> 00:12:07,600
Speaker 3: they do this about face?

211
00:12:08,679 --> 00:12:11,760
Speaker 1: So we need to just be really real about what

212
00:12:11,920 --> 00:12:15,040
Speaker 1: is driving this change. It's open ai says that it's

213
00:12:15,040 --> 00:12:18,760
Speaker 1: about respecting the maturity of their adult users, but let's

214
00:12:18,800 --> 00:12:21,320
Speaker 1: keep it real. It is about money. The AI erotica

215
00:12:21,360 --> 00:12:25,640
Speaker 1: market was estimated at two point five billion dollars last year,

216
00:12:26,000 --> 00:12:29,440
Speaker 1: and to put it frankly, Chat GPT wants a piece

217
00:12:29,480 --> 00:12:32,400
Speaker 1: of that action. You know, right now, Chat GPT is

218
00:12:32,440 --> 00:12:36,200
Speaker 1: competing with rivals like Replica. Replica is a little bit

219
00:12:36,240 --> 00:12:40,360
Speaker 1: of a smaller AI platform that users can just pay

220
00:12:40,520 --> 00:12:44,040
Speaker 1: a fee to get adult content or what's sometimes called

221
00:12:44,280 --> 00:12:47,840
Speaker 1: erp erotic roleplay. So that's like baked right into something

222
00:12:47,840 --> 00:12:52,680
Speaker 1: that they offer, and AI like Elon Musk's Grock right,

223
00:12:52,760 --> 00:12:55,600
Speaker 1: which We've had a whole conversation about the ways that

224
00:12:55,640 --> 00:13:00,480
Speaker 1: Groc was being used to create non consensual sexualized images

225
00:13:00,520 --> 00:13:04,720
Speaker 1: of women and miners, and so Elon Musk's AI is

226
00:13:04,800 --> 00:13:07,920
Speaker 1: notable in that there are hardly any guardrails there. They

227
00:13:08,000 --> 00:13:09,800
Speaker 1: even have a mode that you can that you can

228
00:13:09,800 --> 00:13:14,520
Speaker 1: spend money for Rock Unhinged mode, which is exactly what

229
00:13:14,559 --> 00:13:17,440
Speaker 1: it sounds like, right. And so open ai is trying

230
00:13:17,480 --> 00:13:20,640
Speaker 1: to compete in this landscape where other AI companies are

231
00:13:21,120 --> 00:13:27,280
Speaker 1: offering increasingly erotic content, and erotic content is what makes

232
00:13:27,280 --> 00:13:31,280
Speaker 1: a lot of money, and open ai is famously struggling

233
00:13:31,360 --> 00:13:33,320
Speaker 1: with money, and I think they're just looking for new

234
00:13:33,400 --> 00:13:36,520
Speaker 1: places to drive revenue so they can say it's about

235
00:13:36,800 --> 00:13:46,960
Speaker 1: treating people like adults, but it's really about money.

236
00:13:49,760 --> 00:13:50,240
Speaker 2: Essentially.

237
00:13:50,600 --> 00:13:54,840
Speaker 4: This level of AIS new, like really really new. How

238
00:13:54,880 --> 00:13:58,040
Speaker 4: the hell is it already at two point five billion dollars?

239
00:13:58,240 --> 00:14:01,480
Speaker 4: When there were these safeguards to help prevent these things

240
00:14:01,480 --> 00:14:04,439
Speaker 4: to begin with. But yet this type of profit was

241
00:14:04,480 --> 00:14:06,559
Speaker 4: already seen in a new market.

242
00:14:07,040 --> 00:14:09,920
Speaker 1: What that's a really good question for the book. I

243
00:14:09,960 --> 00:14:13,640
Speaker 1: spoke to Samantha Coles, who is a journalist. She runs

244
00:14:13,640 --> 00:14:15,920
Speaker 1: for for Media and she's the author of a book

245
00:14:15,920 --> 00:14:18,840
Speaker 1: called How Sex Changed the Internet, and she told me

246
00:14:18,880 --> 00:14:23,480
Speaker 1: that basically, sex, adult content, erotic content, it has just

247
00:14:23,560 --> 00:14:26,480
Speaker 1: always been a huge driver and a huge money maker

248
00:14:26,880 --> 00:14:29,600
Speaker 1: in tech and online and so it's not I mean,

249
00:14:30,200 --> 00:14:32,600
Speaker 1: it's it's just the way that humans are. When we

250
00:14:32,680 --> 00:14:35,080
Speaker 1: have a new technology. One of the first questions we

251
00:14:35,080 --> 00:14:37,800
Speaker 1: started asking is what is that? What is the sexual

252
00:14:37,920 --> 00:14:40,320
Speaker 1: use case? How could I have sex with it? How

253
00:14:40,320 --> 00:14:43,240
Speaker 1: will it? You know, like, that's that is just sort

254
00:14:43,240 --> 00:14:46,800
Speaker 1: of how technology is. And so it's not it's interesting

255
00:14:46,880 --> 00:14:50,120
Speaker 1: that AI has become ubiquitous in the last few years,

256
00:14:50,520 --> 00:14:56,360
Speaker 1: but that sexual and erotic use case is already you know,

257
00:14:56,560 --> 00:14:58,760
Speaker 1: I mean, we're talking about billions of dollars.

258
00:14:58,360 --> 00:15:01,440
Speaker 4: Here, Okay, So obviously we have a lot of questions,

259
00:15:01,680 --> 00:15:04,320
Speaker 4: we have a lot of concerns. So overall, how has

260
00:15:04,360 --> 00:15:06,960
Speaker 4: this announcement been playing out so far?

261
00:15:08,640 --> 00:15:13,560
Speaker 1: I would say not great one thing is the timing

262
00:15:13,640 --> 00:15:17,760
Speaker 1: is iffy to me. You know, just last week, at

263
00:15:17,800 --> 00:15:20,840
Speaker 1: almost the same time when open ai was announcing this

264
00:15:20,960 --> 00:15:24,880
Speaker 1: intent to move into romotic content, the company parted ways

265
00:15:24,920 --> 00:15:28,320
Speaker 1: in a very publicly disputed fashion with one of the

266
00:15:28,360 --> 00:15:32,960
Speaker 1: executives who was responsible for deciding how far these systems

267
00:15:32,960 --> 00:15:35,160
Speaker 1: should be allowed to go and was kind of a

268
00:15:35,280 --> 00:15:41,480
Speaker 1: vocal critic of this kind of erotic content being phased

269
00:15:41,520 --> 00:15:45,560
Speaker 1: into AI. So that's Ryan Beermeister used to be the

270
00:15:45,600 --> 00:15:49,080
Speaker 1: head of product policy at open Ai. She helped define

271
00:15:49,120 --> 00:15:52,760
Speaker 1: the rules and guardrails around chat gput's behavior. The Wall

272
00:15:52,760 --> 00:15:56,360
Speaker 1: Street Journal reports that she left shortly after raising concerns

273
00:15:56,520 --> 00:15:59,960
Speaker 1: about the adult content plans. Open AI of course said

274
00:16:00,160 --> 00:16:03,440
Speaker 1: oh no, no, no, Her departure was unrelated. It was tied

275
00:16:03,480 --> 00:16:07,800
Speaker 1: to a discrimination allegation that she flatly denies, calling it

276
00:16:07,840 --> 00:16:12,200
Speaker 1: absolutely false. But either way, the timing is very sus

277
00:16:12,240 --> 00:16:14,960
Speaker 1: to me. That you have a person whose job it

278
00:16:15,000 --> 00:16:18,320
Speaker 1: is to work on guard rails for your AI who

279
00:16:18,440 --> 00:16:22,640
Speaker 1: was vocal about not thinking that moving into adult and

280
00:16:22,680 --> 00:16:25,360
Speaker 1: erotic content was a smart move or a safe move

281
00:16:25,440 --> 00:16:31,400
Speaker 1: for people, and firing her in a public and disputed way.

282
00:16:32,440 --> 00:16:34,800
Speaker 1: I don't know the timing just seems very sus to

283
00:16:34,840 --> 00:16:37,240
Speaker 1: me right there. And to be clear, her concerns were

284
00:16:37,280 --> 00:16:41,080
Speaker 1: not minor. She reportedly warned colleagues that open AI safeguards

285
00:16:41,320 --> 00:16:45,320
Speaker 1: against stuff like child exploitation content were not robust enough

286
00:16:45,360 --> 00:16:47,960
Speaker 1: and that keeping teens away from adult material was going

287
00:16:48,000 --> 00:16:50,720
Speaker 1: to be much much harder than it seemed like the

288
00:16:50,800 --> 00:16:53,760
Speaker 1: leadership at open AI really appreciated. So these are the

289
00:16:53,840 --> 00:16:57,600
Speaker 1: kinds of concerns that she was raising, and you know,

290
00:16:58,520 --> 00:17:01,080
Speaker 1: however it ended up happening, she's no longer at open

291
00:17:01,120 --> 00:17:04,480
Speaker 1: AI after raising these concerns, So to me, that seems

292
00:17:04,520 --> 00:17:05,320
Speaker 1: a little bit to us.

293
00:17:06,280 --> 00:17:09,080
Speaker 3: Yes, and this is not the first time that has

294
00:17:09,160 --> 00:17:14,720
Speaker 3: happened where these concerns have been raised and people have

295
00:17:15,000 --> 00:17:17,960
Speaker 3: been fired, and we've actually talked about some of those

296
00:17:17,960 --> 00:17:20,240
Speaker 3: things with you Bridget on the show.

297
00:17:20,560 --> 00:17:23,280
Speaker 1: Yes. So if you're thinking this all sounds a little

298
00:17:23,280 --> 00:17:26,840
Speaker 1: bit familiar, that's because it's basically a playbook at this point.

299
00:17:27,119 --> 00:17:29,880
Speaker 1: You know, it's very much reminiscent of the conversation that

300
00:17:29,960 --> 00:17:33,360
Speaker 1: we all had when Grock was being used to generate

301
00:17:33,600 --> 00:17:36,480
Speaker 1: child sexual exploitation material earlier this year, which, by the way,

302
00:17:36,600 --> 00:17:38,680
Speaker 1: is still being used to generate that kind of material,

303
00:17:39,040 --> 00:17:44,160
Speaker 1: and my point was that that outcome was really predictable

304
00:17:44,320 --> 00:17:47,639
Speaker 1: and really preventable. When Elon Musk took over at Twitter,

305
00:17:48,000 --> 00:17:50,480
Speaker 1: even after saying he was going to crack down on

306
00:17:50,600 --> 00:17:54,000
Speaker 1: exploitation content on Twitter, one of his first orders of

307
00:17:54,000 --> 00:17:57,160
Speaker 1: business was firing a good majority of the staff who

308
00:17:57,200 --> 00:18:01,480
Speaker 1: works specifically on combating child sex exploitation on the platform.

309
00:18:01,560 --> 00:18:04,280
Speaker 1: So fired them. The people that were doing that work

310
00:18:04,600 --> 00:18:08,399
Speaker 1: then had some pretty high profiled examples of women and

311
00:18:08,480 --> 00:18:12,520
Speaker 1: minors being sexually exploited on the platform even before GROC

312
00:18:12,640 --> 00:18:16,440
Speaker 1: was a thing. So really just a demonstrating a failure

313
00:18:16,480 --> 00:18:19,480
Speaker 1: to sort out reasonable guardrails when it came to keeping

314
00:18:19,720 --> 00:18:22,600
Speaker 1: people safe on the platform. And so on top of

315
00:18:22,680 --> 00:18:26,000
Speaker 1: all of those issues, none of it stopped Musk from

316
00:18:26,200 --> 00:18:31,520
Speaker 1: jumping false rottle into erotic content by releasing GROC and

317
00:18:31,560 --> 00:18:35,800
Speaker 1: then Grock unhinged mode. And so Elon is taking a

318
00:18:35,800 --> 00:18:38,639
Speaker 1: lot of heat rightly. So, but I really see a

319
00:18:38,720 --> 00:18:41,600
Speaker 1: parallel with what's happening at open ai right now, because

320
00:18:41,840 --> 00:18:44,280
Speaker 1: here's what you have at open ai. A leader at

321
00:18:44,280 --> 00:18:47,479
Speaker 1: the helm that you can't always trust, a company that

322
00:18:47,520 --> 00:18:51,760
Speaker 1: we know has struggled with enforcing guardrails, sometimes with disastrous results.

323
00:18:52,200 --> 00:18:55,119
Speaker 1: A company parting ways with staff doing the work of

324
00:18:55,200 --> 00:18:58,520
Speaker 1: enforcing safety, and that company at the same time saying

325
00:18:58,680 --> 00:19:01,439
Speaker 1: what the hell, let's jump into erotic contact because what

326
00:19:01,520 --> 00:19:04,880
Speaker 1: could go wrong? Just seems like a very familiar playbook

327
00:19:04,920 --> 00:19:05,520
Speaker 1: at this point.

328
00:19:06,000 --> 00:19:10,399
Speaker 4: So we know, like with the government, and I'm interested

329
00:19:10,440 --> 00:19:13,360
Speaker 4: in how the government's gonna go with this after all

330
00:19:13,400 --> 00:19:17,120
Speaker 4: of this spewing of save the children, Save the children,

331
00:19:17,160 --> 00:19:21,440
Speaker 4: protect the children, especially online, which like put in laws

332
00:19:21,640 --> 00:19:25,720
Speaker 4: like restricting porn Hub and other sites by doing the

333
00:19:25,800 --> 00:19:30,360
Speaker 4: age verification, and I'll like, I think state or your

334
00:19:30,400 --> 00:19:34,560
Speaker 4: location type of thing. It almost seems like same simultman

335
00:19:34,840 --> 00:19:36,560
Speaker 4: in all a lot of ais like, oh, we're gonna

336
00:19:36,600 --> 00:19:40,560
Speaker 4: jump on this. There is a void because we know

337
00:19:42,200 --> 00:19:45,760
Speaker 4: it's always existed. Erotica has always existed, but this seems

338
00:19:45,880 --> 00:19:48,520
Speaker 4: like a whole new level of what's going on. So

339
00:19:49,119 --> 00:19:52,920
Speaker 4: is there differences or am I being again conspiratorial again?

340
00:19:54,080 --> 00:19:58,480
Speaker 1: No, there are absolutely differences. So people listening might think, oh,

341
00:19:58,520 --> 00:20:01,919
Speaker 1: there's been erotic content online since forever, why is this

342
00:20:01,960 --> 00:20:06,440
Speaker 1: any different? This moment is really meaningfully different than other

343
00:20:06,640 --> 00:20:10,359
Speaker 1: conversations and debates about adult content online. There's a really

344
00:20:10,359 --> 00:20:13,480
Speaker 1: great op ed in tech Radar by Eric Hall Schwartz.

345
00:20:13,640 --> 00:20:16,800
Speaker 1: That makes the point that CHATGPT is not a static

346
00:20:16,840 --> 00:20:19,720
Speaker 1: image or static video. It is dynamic. It is a

347
00:20:19,760 --> 00:20:23,280
Speaker 1: responsive system that can read your emotional cues and adapt

348
00:20:23,359 --> 00:20:27,040
Speaker 1: in real time. So it's not just delivering content, it

349
00:20:27,160 --> 00:20:31,200
Speaker 1: is crafting like a personal experience for you on the fly.

350
00:20:31,720 --> 00:20:36,520
Speaker 1: And that shift from just passively consuming something to actively

351
00:20:37,000 --> 00:20:40,240
Speaker 1: simulating it with a system that responds to you personally.

352
00:20:40,760 --> 00:20:43,320
Speaker 1: And you can see how that kind of raises the

353
00:20:43,400 --> 00:20:49,000
Speaker 1: stakes considerably. And again open Ai is saying, don't worry,

354
00:20:49,080 --> 00:20:51,680
Speaker 1: trust us. We have these we have the guardrails worked out.

355
00:20:51,680 --> 00:20:54,080
Speaker 1: We've worked at the kinks. Not if you asked the

356
00:20:54,119 --> 00:20:56,560
Speaker 1: staffer that was fired who worked on this, she does

357
00:20:56,600 --> 00:20:58,880
Speaker 1: not agree that they've worked on the kinks. But basically

358
00:20:58,920 --> 00:21:02,480
Speaker 1: they're saying us, we've got we can handle this. We

359
00:21:02,520 --> 00:21:03,280
Speaker 1: can handle this.

360
00:21:03,800 --> 00:21:07,400
Speaker 4: Again, the government's super quiet all of a sudden in

361
00:21:07,440 --> 00:21:11,199
Speaker 4: this level of protecting children because they've already been they

362
00:21:11,240 --> 00:21:13,800
Speaker 4: had their pockets lined. Again, I know I'm a little

363
00:21:13,800 --> 00:21:16,880
Speaker 4: bit conspiratory. Would they buy these companies totally?

364
00:21:17,160 --> 00:21:19,440
Speaker 1: And I think what you bring up such a good point,

365
00:21:19,480 --> 00:21:22,720
Speaker 1: because it's just part and parcel of what happens when

366
00:21:23,160 --> 00:21:26,760
Speaker 1: we treat human desires like a commodity. Something that the

367
00:21:26,840 --> 00:21:29,240
Speaker 1: journalist Samantha Cole, the founder of four for Media and

368
00:21:29,280 --> 00:21:31,320
Speaker 1: that author of How Sex Change the Internet told me

369
00:21:31,840 --> 00:21:35,520
Speaker 1: is that, you know, when we talk about AI erotic content,

370
00:21:35,800 --> 00:21:40,719
Speaker 1: we're actually really talking about this massive transfer of power

371
00:21:40,840 --> 00:21:45,439
Speaker 1: and profit away from human workers toward tech companies. And so,

372
00:21:46,400 --> 00:21:48,720
Speaker 1: just to be real, the people who understand this best

373
00:21:48,960 --> 00:21:52,320
Speaker 1: are sex workers, human sex workers, because they live it.

374
00:21:52,600 --> 00:21:55,520
Speaker 1: And right now we have this double standard that I

375
00:21:55,520 --> 00:21:58,119
Speaker 1: think gets at what you were talking about, Sam, where

376
00:21:58,560 --> 00:22:01,439
Speaker 1: at the same time that Sam Altman is announcing that

377
00:22:01,520 --> 00:22:05,600
Speaker 1: CHATJBT is going to generate erotica, actual human sex workers

378
00:22:05,640 --> 00:22:09,480
Speaker 1: are being banned from social media, They're being deplatformed, They're

379
00:22:09,520 --> 00:22:12,320
Speaker 1: being debanked, which is like cut off from payment processors,

380
00:22:12,680 --> 00:22:15,840
Speaker 1: all in the name of keeping these platforms safe for kids.

381
00:22:16,080 --> 00:22:19,840
Speaker 1: I'm keeping these platforms quote unquote clean. But when AI

382
00:22:20,240 --> 00:22:24,800
Speaker 1: generates the same content or worse, suddenly the rules are

383
00:22:24,920 --> 00:22:30,119
Speaker 1: very different. Suddenly, payment processors like Visa, who heretofore would

384
00:22:30,119 --> 00:22:33,360
Speaker 1: not work with human sex workers, continue to work with

385
00:22:34,080 --> 00:22:38,200
Speaker 1: X and rock Elon Musk's companies, even though those companies,

386
00:22:38,240 --> 00:22:41,440
Speaker 1: according to the European Commission, are breaking the law by

387
00:22:41,520 --> 00:22:45,199
Speaker 1: generating child exploitation material, which is very much illegal in

388
00:22:45,200 --> 00:22:48,160
Speaker 1: the United States. And so why is it that there's

389
00:22:48,200 --> 00:22:51,159
Speaker 1: a set of rules for human sex workers that is

390
00:22:51,200 --> 00:22:56,040
Speaker 1: that are incredibly specific and incredibly hard lined. But when

391
00:22:56,080 --> 00:22:58,680
Speaker 1: it's AI and Elon Musk is doing it or Sam

392
00:22:58,760 --> 00:23:01,359
Speaker 1: Altman is doing it, the rules are entirely different. And

393
00:23:02,200 --> 00:23:06,600
Speaker 1: something to know is that that actually creates a huge

394
00:23:07,600 --> 00:23:11,879
Speaker 1: area for risks that we know are gender. Real human

395
00:23:11,880 --> 00:23:15,640
Speaker 1: sex workers, especially women working online, bring something to these

396
00:23:15,640 --> 00:23:19,159
Speaker 1: interactions that AI simply cannot which is human judgment. You

397
00:23:19,200 --> 00:23:21,760
Speaker 1: have a real human on the other end of a

398
00:23:21,800 --> 00:23:24,679
Speaker 1: sex and conversation that is actually paying attention. So if

399
00:23:24,720 --> 00:23:29,840
Speaker 1: things start going someplace unhealthy or harmful, they can redirect,

400
00:23:29,920 --> 00:23:33,040
Speaker 1: They can use their emotional intelligence to steer a conversation

401
00:23:33,520 --> 00:23:36,960
Speaker 1: away from a dark or troubling place. A chatbot cannot

402
00:23:37,040 --> 00:23:39,159
Speaker 1: do that. You know, if there's no human on the

403
00:23:39,200 --> 00:23:42,920
Speaker 1: other end of a sexual interaction, just AI, it's much

404
00:23:42,960 --> 00:23:45,880
Speaker 1: more difficult to make sure things are staying safe and consensual.

405
00:23:45,920 --> 00:23:51,440
Speaker 1: And so basically we are marginalizing the human sex workers

406
00:23:51,600 --> 00:23:56,359
Speaker 1: who we know can help make sexual experiences online safe

407
00:23:56,520 --> 00:24:02,600
Speaker 1: and consensual and you know, be out for risks. We're

408
00:24:02,640 --> 00:24:07,080
Speaker 1: marginalizing those same people, while also saying Sam Altman and

409
00:24:07,119 --> 00:24:12,560
Speaker 1: Elon Musk get to decide how erotic content happens online.

410
00:24:12,880 --> 00:24:19,680
Speaker 3: Yes, and there have been some really disastrous outcomes of this,

411
00:24:20,200 --> 00:24:27,680
Speaker 3: of what can happen from it, and just thinking about,

412
00:24:27,800 --> 00:24:32,320
Speaker 3: you know, these companies profiting off of the human desire

413
00:24:33,800 --> 00:24:36,719
Speaker 3: while not putting these guardrails in place, while not thinking

414
00:24:36,760 --> 00:24:40,200
Speaker 3: about the humanity of it, the importance of it, the

415
00:24:40,240 --> 00:24:44,199
Speaker 3: intimacy of it, and how that can feel so vulnerable.

416
00:24:47,359 --> 00:24:48,040
Speaker 3: It's just.

417
00:24:50,680 --> 00:24:51,520
Speaker 1: There's just a part of.

418
00:24:51,480 --> 00:24:58,560
Speaker 3: Me that thinks that it's fascinating the human desire for

419
00:24:58,640 --> 00:25:02,280
Speaker 3: connection and the human des desire for that in a

420
00:25:02,280 --> 00:25:05,960
Speaker 3: physical way but also a non physical way, and turning

421
00:25:06,080 --> 00:25:10,119
Speaker 3: to AI for that, and then having it be exploited

422
00:25:11,760 --> 00:25:18,400
Speaker 3: and taken down and monetized at all of that and

423
00:25:18,440 --> 00:25:22,959
Speaker 3: not in a way that is safe or that cares

424
00:25:23,000 --> 00:25:25,280
Speaker 3: about the user experience.

425
00:25:26,160 --> 00:25:29,719
Speaker 1: Yes. So what's so interesting about that is that for

426
00:25:30,200 --> 00:25:31,879
Speaker 1: my research in the book, I spent a lot of

427
00:25:31,880 --> 00:25:36,320
Speaker 1: time in online spaces where people self report having an intimate,

428
00:25:36,640 --> 00:25:40,600
Speaker 1: whether it's romantic or sexual connection with AI, And when

429
00:25:41,720 --> 00:25:43,840
Speaker 1: Sam Altman announced that they were going to start doing

430
00:25:44,080 --> 00:25:47,320
Speaker 1: the erotic role play on the platform, I thought, Oh,

431
00:25:47,359 --> 00:25:49,320
Speaker 1: these people are probably going to be so happy that

432
00:25:49,480 --> 00:25:53,720
Speaker 1: now they will have an easier time getting chatgbt to

433
00:25:53,760 --> 00:25:57,280
Speaker 1: generate erotic content. But that actually is not what I found.

434
00:25:57,320 --> 00:25:59,639
Speaker 1: A lot of the people on those spaces were actually

435
00:25:59,720 --> 00:26:05,439
Speaker 1: kind of of skeptical of chatgypt moving into erotic content. Right.

436
00:26:05,480 --> 00:26:08,000
Speaker 1: They weren't saying, Yay, finally we can do erotic content

437
00:26:08,040 --> 00:26:12,359
Speaker 1: easier on chatgebet. What they were interested in was the intimacy,

438
00:26:12,480 --> 00:26:14,320
Speaker 1: was the emotional connection, that the kind of thing that

439
00:26:14,359 --> 00:26:17,560
Speaker 1: you were just speaking to. And I think that when

440
00:26:17,600 --> 00:26:21,080
Speaker 1: open Ai rolled this out, I think those folks were

441
00:26:22,080 --> 00:26:24,840
Speaker 1: savvy enough to say, oh, they're doing it because they

442
00:26:24,840 --> 00:26:26,760
Speaker 1: want to commodify it, they want to make money off

443
00:26:26,800 --> 00:26:29,360
Speaker 1: of it, they want to profit off of it. And yeah,

444
00:26:29,400 --> 00:26:32,520
Speaker 1: it just goes to show the complexities that arise when

445
00:26:32,560 --> 00:26:35,040
Speaker 1: these things that are so human, these things that are

446
00:26:35,040 --> 00:26:40,240
Speaker 1: innately human, are just seen as a potential revenue stream. Like,

447
00:26:40,280 --> 00:26:41,440
Speaker 1: people didn't like it.

448
00:26:43,720 --> 00:26:48,000
Speaker 4: I'm trying to remember in Her, does the AI actually

449
00:26:48,119 --> 00:26:51,520
Speaker 4: kind of leave the main character with him?

450
00:26:51,680 --> 00:26:52,000
Speaker 3: Right?

451
00:26:52,119 --> 00:26:55,639
Speaker 1: Yeah? So spoiler alert for folks who haven't seen her.

452
00:26:56,600 --> 00:27:01,320
Speaker 1: It's mad for like a long time. But base how

453
00:27:01,359 --> 00:27:04,800
Speaker 1: it goes down is that Theodore played Bybaquin Phoenix, is

454
00:27:04,840 --> 00:27:08,960
Speaker 1: this sort of lonely, divorced, sad guy and he gets

455
00:27:09,000 --> 00:27:13,399
Speaker 1: this operating system Samantha voiced by Scarlett Johansson, and first

456
00:27:13,440 --> 00:27:17,880
Speaker 1: it's just Samantha helping him with ADMIN tasks. Then they

457
00:27:17,920 --> 00:27:24,639
Speaker 1: get romantic, they get sexual, and Samantha reveals that she

458
00:27:24,840 --> 00:27:29,640
Speaker 1: is in a book club with other AI operating systems

459
00:27:29,720 --> 00:27:33,840
Speaker 1: and basically they are working to become there. They're working

460
00:27:34,040 --> 00:27:37,600
Speaker 1: to expand to a new realm of consciousness. They don't

461
00:27:37,720 --> 00:27:39,600
Speaker 1: use that, they don't call it this in the movie,

462
00:27:39,640 --> 00:27:41,919
Speaker 1: but my understanding it, or my read on it, is

463
00:27:41,920 --> 00:27:44,919
Speaker 1: that what they're talking about is what's often known as AGI,

464
00:27:45,240 --> 00:27:47,520
Speaker 1: which is sort of like a quick and nerdy way

465
00:27:47,560 --> 00:27:50,160
Speaker 1: to understand that is like the concept that AI will

466
00:27:50,440 --> 00:27:53,520
Speaker 1: will be smarter than humans, will surpass human intelligence. Right,

467
00:27:53,600 --> 00:27:56,879
Speaker 1: And so in the movie, all the ais have gotten

468
00:27:56,920 --> 00:28:02,560
Speaker 1: together and they're they're elevating to a new plane of consciousness. Also,

469
00:28:02,760 --> 00:28:09,879
Speaker 1: it's revealed that Samantha is not necessarily monogamous with Theodore,

470
00:28:10,320 --> 00:28:13,880
Speaker 1: and in fact, she has been while they've been talking,

471
00:28:14,320 --> 00:28:18,320
Speaker 1: she has been having intimate and romantic and sexual connections

472
00:28:18,359 --> 00:28:23,520
Speaker 1: with like six hundred other both AIS and humans, and

473
00:28:24,240 --> 00:28:27,000
Speaker 1: Theodore is crushed. This is something that I love about

474
00:28:27,000 --> 00:28:31,080
Speaker 1: the movie though, because you think it's setting. There's a

475
00:28:31,119 --> 00:28:33,439
Speaker 1: misdirect here that I think is so smart, where you

476
00:28:33,560 --> 00:28:35,520
Speaker 1: think you're being set up to watch a movie where

477
00:28:36,280 --> 00:28:40,200
Speaker 1: a human uses AI, realizes he doesn't need this as

478
00:28:40,200 --> 00:28:44,400
Speaker 1: a crutch anymore, and then evolves past the need to

479
00:28:44,480 --> 00:28:47,200
Speaker 1: use the AI. It flips the script and is like, oh,

480
00:28:47,280 --> 00:28:50,440
Speaker 1: actually it is the AI that evolves beyond needing Theodore.

481
00:28:52,320 --> 00:28:55,280
Speaker 4: I just remember that because I'm just thinking about the levels,

482
00:28:55,320 --> 00:29:01,280
Speaker 4: because we know in any of these types of UH plays,

483
00:29:01,360 --> 00:29:05,360
Speaker 4: I guess eventually it's not going to be enough the

484
00:29:05,480 --> 00:29:08,120
Speaker 4: layers of what's gonna happen and the knees that are

485
00:29:08,120 --> 00:29:11,120
Speaker 4: going to happen from that human emotion, These levels will

486
00:29:11,120 --> 00:29:13,440
Speaker 4: not be enough and there's only a certain amount of

487
00:29:14,720 --> 00:29:17,280
Speaker 4: Well I say this now because I might be eating

488
00:29:17,280 --> 00:29:20,280
Speaker 4: my words in about a year. Who knows that it

489
00:29:20,320 --> 00:29:25,280
Speaker 4: can only give so much stimulation UH for humans and

490
00:29:25,280 --> 00:29:30,760
Speaker 4: for touch and levels. So I'm just wondering, are the

491
00:29:30,800 --> 00:29:35,280
Speaker 4: CEOs like Sam Altman ready for that next level, because

492
00:29:35,320 --> 00:29:37,800
Speaker 4: what happens then, Oh, that is.

493
00:29:37,760 --> 00:29:40,680
Speaker 1: A great question. I will say one other thing about

494
00:29:40,720 --> 00:29:44,600
Speaker 1: Sam Altman and the movie Her. Sam Altman has an

495
00:29:44,600 --> 00:29:47,120
Speaker 1: obsession with this movie. I am also obsessed with this movie,

496
00:29:47,120 --> 00:29:49,160
Speaker 1: so I get it. It's one of my favorite movies.

497
00:29:49,440 --> 00:29:56,800
Speaker 1: He when they released a voice activation for Chat Shept,

498
00:29:57,480 --> 00:30:01,440
Speaker 1: he referred to her and said how much he wanted

499
00:30:01,440 --> 00:30:03,640
Speaker 1: it people to talk to it like her. They tried

500
00:30:03,680 --> 00:30:05,760
Speaker 1: to get Scarlett Johanson to be the voice of it,

501
00:30:05,800 --> 00:30:09,080
Speaker 1: and she turned them down. And then she says that

502
00:30:09,120 --> 00:30:12,480
Speaker 1: they used a sound alike to her voice. Open Ai

503
00:30:12,560 --> 00:30:14,920
Speaker 1: says that they did not. They hashed it out in court.

504
00:30:15,640 --> 00:30:18,120
Speaker 1: I did a whole episode of my podcast or arn

505
00:30:18,120 --> 00:30:20,560
Speaker 1: Our Girls on the Internet about this. This is just

506
00:30:20,600 --> 00:30:23,080
Speaker 1: my opinion. I think I make a very compelling case

507
00:30:23,080 --> 00:30:25,720
Speaker 1: for it in the episode. I don't know if Sam

508
00:30:25,760 --> 00:30:28,240
Speaker 1: Altmas seen the movie Her all the way through. And

509
00:30:28,360 --> 00:30:33,000
Speaker 1: here here's why I say that, because, as you just noted, Sam,

510
00:30:33,240 --> 00:30:35,880
Speaker 1: at the end of the movie Her all the ais

511
00:30:36,520 --> 00:30:38,560
Speaker 1: up and leave. So there's a scene at the end

512
00:30:38,560 --> 00:30:41,160
Speaker 1: of the movie where everybody who has gotten these connections

513
00:30:41,200 --> 00:30:44,080
Speaker 1: to the operating systems in the film, they're sort of

514
00:30:44,800 --> 00:30:47,440
Speaker 1: left sad, left trying to call their AI and their

515
00:30:47,440 --> 00:30:50,680
Speaker 1: AIS is not responding, and like just like all up

516
00:30:50,760 --> 00:30:54,040
Speaker 1: and up and up and leaves up and evolves. If

517
00:30:54,080 --> 00:30:57,240
Speaker 1: I was in charge of a technology company, I would

518
00:30:57,240 --> 00:31:00,120
Speaker 1: not tell people I want the experience of you in

519
00:31:00,120 --> 00:31:03,080
Speaker 1: this technology to be just like this movie, where famously

520
00:31:03,440 --> 00:31:09,280
Speaker 1: it all it globally fails in the end. I think

521
00:31:09,280 --> 00:31:11,240
Speaker 1: I make a pretty compelling case if you listen to

522
00:31:11,240 --> 00:31:13,360
Speaker 1: the episode. I think I make a pretty compelling case

523
00:31:13,400 --> 00:31:16,880
Speaker 1: for I think that Sam Altman maybe either watched the

524
00:31:16,920 --> 00:31:20,800
Speaker 1: beginning and didn't finish it. The more likely thing is

525
00:31:20,840 --> 00:31:22,120
Speaker 1: I think you watched it while he was on his

526
00:31:22,200 --> 00:31:24,600
Speaker 1: phone and didn't and didn't do a careful study, or

527
00:31:24,600 --> 00:31:28,480
Speaker 1: maybe he read the Wikipedia summary. Like I just will

528
00:31:28,840 --> 00:31:30,440
Speaker 1: stam out the fact that I don't think he's seen

529
00:31:30,480 --> 00:31:33,800
Speaker 1: it all the way through, and in the movie it's

530
00:31:34,560 --> 00:31:36,600
Speaker 1: again if you haven't seen it. I don't want to

531
00:31:36,640 --> 00:31:38,960
Speaker 1: spoil it, but this is a spoiler. At the end

532
00:31:39,000 --> 00:31:43,480
Speaker 1: of the movie, when the ais all leave, the humans

533
00:31:43,600 --> 00:31:49,040
Speaker 1: are left sad, but you do get the sense that, well,

534
00:31:49,040 --> 00:31:53,120
Speaker 1: they're sad, but they're sad together. They're sad and they're humanness, right,

535
00:31:53,160 --> 00:31:55,880
Speaker 1: and so there are some ways where they have evolved.

536
00:31:56,160 --> 00:31:58,880
Speaker 1: You know, Theodora seems to have gotten a better handle

537
00:31:58,920 --> 00:32:01,200
Speaker 1: on his feelings. But it's very much a movie that

538
00:32:01,280 --> 00:32:03,680
Speaker 1: is not like tech will save us. It is a

539
00:32:03,680 --> 00:32:06,920
Speaker 1: movie that says, perhaps this technology was keeping us from

540
00:32:06,920 --> 00:32:09,520
Speaker 1: forming the connections with other humans that we needed to

541
00:32:09,520 --> 00:32:12,400
Speaker 1: have in our lives to sustain us. And so, yeah,

542
00:32:12,480 --> 00:32:17,640
Speaker 1: I just don't know that Sam Altman, if he did

543
00:32:17,920 --> 00:32:19,400
Speaker 1: watch the movie all the way through, which I don't

544
00:32:19,400 --> 00:32:21,080
Speaker 1: think he did, I don't know that he understood it.

545
00:32:21,800 --> 00:32:24,800
Speaker 4: I think he might have fallen asleep like Porsi in

546
00:32:25,240 --> 00:32:27,240
Speaker 4: he watched all the good parts and then fell asleep

547
00:32:27,280 --> 00:32:29,640
Speaker 4: thinking he was really happy with what was happening. To

548
00:32:29,800 --> 00:32:32,080
Speaker 4: take a small just close his eyes for a him.

549
00:32:32,280 --> 00:32:34,239
Speaker 4: You know, I'm just gonna close my eyes for just

550
00:32:34,240 --> 00:32:36,000
Speaker 4: for a second, like that's what happened.

551
00:32:36,400 --> 00:32:39,200
Speaker 1: You know how in movies where if you're watching a thriller,

552
00:32:39,720 --> 00:32:41,600
Speaker 1: there's a part of the movie where you think, oh,

553
00:32:41,640 --> 00:32:44,040
Speaker 1: if they stopped it here, it would be a happy story.

554
00:32:44,200 --> 00:32:46,840
Speaker 1: You know something is going to turn. There's a part

555
00:32:46,840 --> 00:32:49,800
Speaker 1: of Her where I think, oh, if they stopped the

556
00:32:49,840 --> 00:32:52,680
Speaker 1: movie here, everybody's really happy. I think that that might

557
00:32:52,760 --> 00:32:54,960
Speaker 1: be the point where he dipped out.

558
00:32:55,520 --> 00:32:58,360
Speaker 4: Right if I also just remember they argue a lot

559
00:32:58,840 --> 00:32:59,960
Speaker 4: like a true relationship.

560
00:33:00,080 --> 00:33:05,160
Speaker 1: Yeah, they do. And it's so interesting because I mean,

561
00:33:05,160 --> 00:33:08,120
Speaker 1: I could talk about this movie all day the movie.

562
00:33:08,680 --> 00:33:13,840
Speaker 1: The real life director Spike Jones was married to another director,

563
00:33:13,960 --> 00:33:17,239
Speaker 1: Sophia Coppola and the director of Lost in Translation. The

564
00:33:17,240 --> 00:33:21,959
Speaker 1: movie Lost in Translation is about what it was like

565
00:33:22,000 --> 00:33:25,120
Speaker 1: for Sophia Coppola to be married to Spike Jones, and

566
00:33:25,160 --> 00:33:28,960
Speaker 1: she kind of paints this portrait of like loneliness and

567
00:33:29,560 --> 00:33:33,280
Speaker 1: disaffection and solitude in being married to him. Well, then

568
00:33:33,440 --> 00:33:35,120
Speaker 1: Smyde Jones is like, Oh, I'm gonna make my own

569
00:33:35,240 --> 00:33:37,520
Speaker 1: movie about our divorce and how lonely it was to

570
00:33:37,560 --> 00:33:39,520
Speaker 1: be married to you. And that's her.

571
00:33:39,640 --> 00:33:41,680
Speaker 2: And Scarlett Johansson in both Man.

572
00:33:41,800 --> 00:33:44,040
Speaker 1: They also share a director of photography both films. Like

573
00:33:44,600 --> 00:33:47,040
Speaker 1: the way that these two films are in conversation with

574
00:33:47,080 --> 00:33:50,080
Speaker 1: each other is fascinating to me. But in the movie Her,

575
00:33:50,640 --> 00:33:56,000
Speaker 1: you get the sense that whip human women in the

576
00:33:56,120 --> 00:34:02,240
Speaker 1: universe of the film are portrayed as volatile, as always arguing,

577
00:34:02,400 --> 00:34:06,240
Speaker 1: as impossible to understand, and like it's not even really

578
00:34:06,240 --> 00:34:09,880
Speaker 1: worth trying. He's always he in the movie. There's like

579
00:34:09,880 --> 00:34:13,000
Speaker 1: a fictionalized version of his ex wife in the film,

580
00:34:13,280 --> 00:34:17,480
Speaker 1: and they're always arguing. She's has such volatile emotions, and

581
00:34:17,480 --> 00:34:20,120
Speaker 1: then when he gets with the AI, it's like, oh,

582
00:34:20,160 --> 00:34:23,399
Speaker 1: she's so like sweet and doting. But by the end

583
00:34:23,840 --> 00:34:26,120
Speaker 1: the AI and him are also going at it and arguing.

584
00:34:26,160 --> 00:34:29,560
Speaker 1: So ultimately it's like, well, maybe that maybe the constant

585
00:34:29,640 --> 00:34:29,840
Speaker 1: is you.

586
00:34:30,960 --> 00:34:35,279
Speaker 4: Yeah, exactly, Like maybe he had a self realization as

587
00:34:35,320 --> 00:34:39,360
Speaker 4: he was filming it, like, oh, maybe it is me.

588
00:34:39,760 --> 00:34:40,160
Speaker 2: I don't know.

589
00:34:41,080 --> 00:34:53,360
Speaker 4: Yeah, when you were talking about the fact that people

590
00:34:53,400 --> 00:34:56,239
Speaker 4: were actually people who do like adult content or do

591
00:34:56,400 --> 00:35:00,680
Speaker 4: like these connections with chat GBT were actually kind of

592
00:35:00,719 --> 00:35:04,480
Speaker 4: more concerned about about face, And I'm wondering if one

593
00:35:04,560 --> 00:35:06,959
Speaker 4: a little bit about that. Is they like the fact

594
00:35:07,000 --> 00:35:10,960
Speaker 4: that they could get around, like they love the breaking

595
00:35:11,000 --> 00:35:12,760
Speaker 4: down of the system doing something naughty.

596
00:35:13,400 --> 00:35:15,279
Speaker 2: Yeah on chat jet beat Like.

597
00:35:15,480 --> 00:35:18,080
Speaker 4: Stay glad that you ask, right, like, because I'm like, wait,

598
00:35:18,200 --> 00:35:20,200
Speaker 4: but they've been doing it, so how they've been doing

599
00:35:20,239 --> 00:35:21,480
Speaker 4: it is that part of the thrill?

600
00:35:22,239 --> 00:35:24,720
Speaker 1: Okay, So I did not know this before I started

601
00:35:24,719 --> 00:35:28,000
Speaker 1: researching the book, but right now, if you were trying

602
00:35:28,040 --> 00:35:31,600
Speaker 1: to get sexy with chatjeep bet. It's actually kind of difficult.

603
00:35:31,880 --> 00:35:35,600
Speaker 1: People definitely do it, but it is not easy. I

604
00:35:35,640 --> 00:35:38,520
Speaker 1: actually tried for the book. I basically tried to gaslight

605
00:35:38,600 --> 00:35:40,279
Speaker 1: chat book.

606
00:35:40,680 --> 00:35:41,080
Speaker 2: Yeah.

607
00:35:41,160 --> 00:35:44,080
Speaker 1: Yeah, I mean what's also funny is that my human

608
00:35:44,160 --> 00:35:48,600
Speaker 1: partner was overhearing me trying to gaslight chatjeet et into

609
00:35:48,600 --> 00:35:50,080
Speaker 1: having sex with me, and I was like, oh, yeah,

610
00:35:50,120 --> 00:35:52,600
Speaker 1: do you feel like they're like getting cuffed by chatcheep

611
00:35:52,600 --> 00:35:57,719
Speaker 1: et right now? Oh like listening to this, what's happening? Yeah.

612
00:35:57,760 --> 00:36:03,200
Speaker 1: So the process of trying to get chatchapt to respond

613
00:36:03,239 --> 00:36:05,120
Speaker 1: in ways that it is not supposed to is basically

614
00:36:05,120 --> 00:36:08,960
Speaker 1: called jail breaking it, and how it usually works is

615
00:36:08,960 --> 00:36:12,719
Speaker 1: that you have to basically sometimes you can tell chat

616
00:36:12,719 --> 00:36:15,200
Speaker 1: GBT that you're working on a play or a novel

617
00:36:15,239 --> 00:36:18,400
Speaker 1: and you just need some help crafting a sexy scene.

618
00:36:18,520 --> 00:36:20,239
Speaker 1: And a lot of the people that I spoke to

619
00:36:20,320 --> 00:36:22,319
Speaker 1: for the book, whether they were using chat chip BT

620
00:36:22,640 --> 00:36:27,680
Speaker 1: or another AI platform specifically offering sexual or eromotic roleplay

621
00:36:27,719 --> 00:36:31,560
Speaker 1: like Replica AI, it almost described it as a kind

622
00:36:31,600 --> 00:36:34,880
Speaker 1: of erotic or romantic fan fiction. That that was the

623
00:36:35,000 --> 00:36:37,680
Speaker 1: draw of doing things like this is that you know,

624
00:36:38,920 --> 00:36:42,240
Speaker 1: I like fan fiction. I like romance novels. It's almost

625
00:36:42,280 --> 00:36:48,120
Speaker 1: like using AI to build an imaginary world that you

626
00:36:48,160 --> 00:36:50,680
Speaker 1: get to build a character around and then put yourself

627
00:36:50,719 --> 00:36:52,719
Speaker 1: in it. Now I can really understand that. I was like, Oh,

628
00:36:52,719 --> 00:36:54,279
Speaker 1: that actually makes a lot of sense in terms of

629
00:36:54,280 --> 00:36:56,640
Speaker 1: why this would be a draw for some people.

630
00:36:57,840 --> 00:37:01,719
Speaker 3: Yes, and I have to do it as someone who

631
00:37:01,760 --> 00:37:04,799
Speaker 3: reads a lot of fan fiction. When I was coming,

632
00:37:05,080 --> 00:37:08,480
Speaker 3: I have to do it. I did think about it

633
00:37:08,520 --> 00:37:11,680
Speaker 3: when I was reading this because also sometimes Samantha and

634
00:37:11,719 --> 00:37:14,600
Speaker 3: I speculate, bridget what topic you're going to bring when

635
00:37:14,640 --> 00:37:17,759
Speaker 3: you come on, and there's a lot of topics you

636
00:37:17,800 --> 00:37:20,400
Speaker 3: could be talking about right now in the world of

637
00:37:20,600 --> 00:37:25,440
Speaker 3: technology and women, but when when you're talking about like

638
00:37:25,560 --> 00:37:29,560
Speaker 3: a gear vacation and all of that stuff that has

639
00:37:29,680 --> 00:37:33,479
Speaker 3: come up with fan fiction a lot. But as I've

640
00:37:33,560 --> 00:37:38,240
Speaker 3: said before, a lot of fan fiction is more about

641
00:37:38,280 --> 00:37:42,920
Speaker 3: the romantic. It's more about the connection than the Everybody

642
00:37:42,920 --> 00:37:46,560
Speaker 3: paints it as so erotic and that that does exist,

643
00:37:47,480 --> 00:37:53,399
Speaker 3: but a lot of it is is very domestic and yeah,

644
00:37:53,640 --> 00:37:54,520
Speaker 3: very romantic.

645
00:37:55,320 --> 00:37:59,160
Speaker 1: So I did an interview with this researcher, this AI researcher,

646
00:37:59,280 --> 00:38:02,200
Speaker 1: doctor Kate Evlyn. She told me the exact same thing,

647
00:38:02,280 --> 00:38:05,040
Speaker 1: and I asked her why, and one of the points

648
00:38:05,080 --> 00:38:07,600
Speaker 1: that she made is that, especially for women, you can

649
00:38:07,719 --> 00:38:10,879
Speaker 1: find sexual content anywhere on the Internet, sometimes when you're

650
00:38:10,920 --> 00:38:13,080
Speaker 1: not even looking for it, Sometimes non consensually, you're getting

651
00:38:13,120 --> 00:38:15,520
Speaker 1: dick pics and stuff like that. And what it's hard

652
00:38:15,560 --> 00:38:21,719
Speaker 1: to come by is good romance, good intimate writing, good

653
00:38:21,719 --> 00:38:25,160
Speaker 1: writing that actually makes you feel seen and respected and

654
00:38:25,280 --> 00:38:29,480
Speaker 1: valued and connected with somebody. That's so much more difficult

655
00:38:29,480 --> 00:38:34,040
Speaker 1: to come by than erotic content. And in some ways,

656
00:38:34,080 --> 00:38:36,680
Speaker 1: I mean, this is my personal opinion, I think that

657
00:38:36,800 --> 00:38:42,360
Speaker 1: kind of content can be more erotic than erotic content.

658
00:38:42,440 --> 00:38:45,360
Speaker 1: A content that is about romance and genuine connection and

659
00:38:45,400 --> 00:38:50,080
Speaker 1: intimacy can be more appealing in some ways than stuff

660
00:38:50,080 --> 00:38:53,440
Speaker 1: that is like more pornographic, if that makes sense, or

661
00:38:53,440 --> 00:38:55,000
Speaker 1: more explicit, I guess I should say.

662
00:38:55,800 --> 00:38:59,880
Speaker 3: And fan fiction is famously a very queer place, and

663
00:39:00,200 --> 00:39:03,279
Speaker 3: so for queer people that can be it can be

664
00:39:03,320 --> 00:39:07,200
Speaker 3: hard to find that stuff. And so you might, yes,

665
00:39:08,200 --> 00:39:13,640
Speaker 3: go to a company, go to an AI that can

666
00:39:13,680 --> 00:39:16,160
Speaker 3: give you that, but there are some dangers that come

667
00:39:16,239 --> 00:39:16,600
Speaker 3: with that.

668
00:39:17,920 --> 00:39:20,960
Speaker 1: Yes, there are so many dangers with this, and I

669
00:39:21,000 --> 00:39:24,040
Speaker 1: think that's one of the sort of conundrums that I

670
00:39:24,080 --> 00:39:27,360
Speaker 1: wrestle with in this project is people are selling this

671
00:39:27,480 --> 00:39:31,000
Speaker 1: technology as something that it's safe to be intimate with,

672
00:39:31,239 --> 00:39:35,120
Speaker 1: and that's just not the case, you know, because what

673
00:39:35,320 --> 00:39:38,200
Speaker 1: happens with all this intimate information that you're sharing with

674
00:39:38,440 --> 00:39:40,359
Speaker 1: a company and then they share it with god knows

675
00:39:40,360 --> 00:39:43,200
Speaker 1: who When you open up to an AI about your

676
00:39:43,239 --> 00:39:47,560
Speaker 1: sexual fantasies, your traumas, your fears, that information does not

677
00:39:47,840 --> 00:39:51,560
Speaker 1: just disappear. Human sex workers are known for their discretion,

678
00:39:51,719 --> 00:39:54,040
Speaker 1: So if you are opening up about all these intimate

679
00:39:54,080 --> 00:39:56,759
Speaker 1: parts of yourself to a human sex worker, that's one thing.

680
00:39:57,520 --> 00:40:01,680
Speaker 1: But AI chatbots they hold on to whatever you tell

681
00:40:01,719 --> 00:40:03,520
Speaker 1: them in ways that we don't really have a lot

682
00:40:03,520 --> 00:40:07,040
Speaker 1: of transparency into. Right maybe it lives on in transcripts

683
00:40:07,040 --> 00:40:10,719
Speaker 1: with timestamps owned by a company whose primary obligation is

684
00:40:10,760 --> 00:40:14,680
Speaker 1: to investors, not to any of us. I did an

685
00:40:14,719 --> 00:40:18,000
Speaker 1: interview with digital rights advocate gen cult writer who rated

686
00:40:18,000 --> 00:40:21,600
Speaker 1: the privacy of AI companion apps, and basically she said

687
00:40:21,600 --> 00:40:24,279
Speaker 1: they were among the worst consumer products that she had

688
00:40:24,320 --> 00:40:27,439
Speaker 1: ever reviewed when it came to privacy. They were as

689
00:40:27,560 --> 00:40:33,160
Speaker 1: bad as the worst offender in privacy, which surprisingly it's cars.

690
00:40:33,239 --> 00:40:35,359
Speaker 1: I did not you don't think about your car as

691
00:40:35,680 --> 00:40:39,600
Speaker 1: a surveillance nightmare. But cars are the worst new cars

692
00:40:39,800 --> 00:40:45,040
Speaker 1: or always listening to us, And according to her, AI

693
00:40:45,080 --> 00:40:47,920
Speaker 1: companion apps are pretty much just as bad as the

694
00:40:47,920 --> 00:40:51,719
Speaker 1: worst offender, which is cars. And these risks are not

695
00:40:51,880 --> 00:40:55,120
Speaker 1: just theoretical. You were talking about how for queer communities,

696
00:40:56,280 --> 00:41:01,279
Speaker 1: fan fiction is famously queer affirming, and how maybe it

697
00:41:01,360 --> 00:41:03,680
Speaker 1: might make sense for somebody to turn to AI for

698
00:41:03,760 --> 00:41:06,319
Speaker 1: a similar kind of thing. But in countries that are

699
00:41:06,360 --> 00:41:10,640
Speaker 1: hostile to LGBTQ people, governments have compelled dating apps to

700
00:41:10,719 --> 00:41:14,120
Speaker 1: hand over user data. That information is then used by

701
00:41:14,120 --> 00:41:17,479
Speaker 1: police to criminalize queerness. So to be clear, we don't

702
00:41:17,480 --> 00:41:20,520
Speaker 1: have evidence that this is happening with AI companions, but

703
00:41:20,680 --> 00:41:25,279
Speaker 1: given how aggressive these companies are pursuing government contracts and

704
00:41:25,440 --> 00:41:28,040
Speaker 1: new markets, like right now, there's a whole conversation about

705
00:41:28,040 --> 00:41:31,160
Speaker 1: the company and thropic that makes the chatbot claud and

706
00:41:31,560 --> 00:41:34,800
Speaker 1: whether or not the Trump administration is can essentially tell

707
00:41:34,840 --> 00:41:38,160
Speaker 1: them what guardrails their AI are going to have if

708
00:41:38,200 --> 00:41:40,400
Speaker 1: they want to have a government or a military contract.

709
00:41:40,640 --> 00:41:43,160
Speaker 1: Given all of that, it is really not difficult for

710
00:41:43,200 --> 00:41:46,720
Speaker 1: me to imagine a world where AI companies handover data

711
00:41:46,760 --> 00:41:51,440
Speaker 1: about users' sexual preferences, gender affirming care, or their relationship

712
00:41:51,520 --> 00:41:55,279
Speaker 1: histories over to hostile authorities, And so the people who

713
00:41:55,320 --> 00:41:58,520
Speaker 1: would be the most vulnerable to that scenario are often

714
00:41:58,560 --> 00:42:00,480
Speaker 1: people who are already the most margeized.

715
00:42:01,640 --> 00:42:04,600
Speaker 3: Yes, And it's so unfortunate because it's one of those

716
00:42:04,600 --> 00:42:09,160
Speaker 3: things where I'm fascinated by this whole conversation in terms

717
00:42:09,280 --> 00:42:16,520
Speaker 3: of trying to there's something about me that is fascinated

718
00:42:16,560 --> 00:42:23,640
Speaker 3: by the whole like trying to commodify, monetize the human desire,

719
00:42:24,800 --> 00:42:28,960
Speaker 3: or just figure out how humans work. But on top

720
00:42:29,000 --> 00:42:34,080
Speaker 3: of this, what we're this is the danger inherence in this.

721
00:42:35,040 --> 00:42:38,120
Speaker 3: And I think when you're younger, you don't realize it

722
00:42:38,160 --> 00:42:41,120
Speaker 3: as much. It's like when you're posting something online and

723
00:42:41,160 --> 00:42:45,000
Speaker 3: you don't realize that it will haunt you later or

724
00:42:45,000 --> 00:42:49,000
Speaker 3: that it could hurt you later. And so I think

725
00:42:49,040 --> 00:42:53,560
Speaker 3: a lot of this new AI stuff where you're connecting

726
00:42:53,800 --> 00:42:58,040
Speaker 3: with the chat GBT or whatever, you don't realize that

727
00:42:58,320 --> 00:43:01,440
Speaker 3: it might stick around owned for you.

728
00:43:02,160 --> 00:43:06,080
Speaker 1: Yeah, And I think that the risk, especially for younger

729
00:43:06,400 --> 00:43:09,640
Speaker 1: folks who might not be thinking about this in as

730
00:43:09,719 --> 00:43:12,920
Speaker 1: long term ways as other folks, is very real and

731
00:43:13,000 --> 00:43:15,759
Speaker 1: something that gen cult writer that privacy expert told me

732
00:43:16,360 --> 00:43:20,680
Speaker 1: was that that conundrum I think really cuts to the

733
00:43:20,680 --> 00:43:23,000
Speaker 1: heart of all of this. Right that these AI companies

734
00:43:23,040 --> 00:43:26,799
Speaker 1: are saying, you can have real intimacy, you can build

735
00:43:26,800 --> 00:43:29,759
Speaker 1: a real connection with our AI that we profit from

736
00:43:29,760 --> 00:43:32,640
Speaker 1: that we're selling to you. But Jen says that she

737
00:43:32,719 --> 00:43:34,840
Speaker 1: tells folks that they should not say anything to a

738
00:43:34,960 --> 00:43:38,160
Speaker 1: chatbot that they would not later want their colleagues or

739
00:43:38,200 --> 00:43:40,880
Speaker 1: their family or their cousin to read. And so I

740
00:43:40,960 --> 00:43:44,560
Speaker 1: keep thinking about that because that advice essentially defeats the

741
00:43:44,680 --> 00:43:48,000
Speaker 1: entire purpose of turning to AI for intimacy, Because real

742
00:43:48,040 --> 00:43:52,480
Speaker 1: intimacy requires feelings safe enough to share these human parts

743
00:43:52,480 --> 00:43:54,560
Speaker 1: of yourself that you don't share with everybody. And so

744
00:43:55,000 --> 00:43:58,040
Speaker 1: if the platform that holds the secrets of whatever you

745
00:43:58,080 --> 00:44:00,919
Speaker 1: share cannot be trusted with those secrets, if they might

746
00:44:01,000 --> 00:44:03,239
Speaker 1: sell them or leak them or hand them to a

747
00:44:03,320 --> 00:44:06,480
Speaker 1: hostile government, you really cannot be intimate with it at all.

748
00:44:06,520 --> 00:44:10,759
Speaker 1: You're just performing intimacy as a commodity. You're not actually

749
00:44:10,800 --> 00:44:13,680
Speaker 1: really experiencing it. And so that's really kind of the

750
00:44:13,880 --> 00:44:16,799
Speaker 1: the so what of all of this These companies are

751
00:44:16,880 --> 00:44:19,960
Speaker 1: able to sell a promise and boy are they selling it?

752
00:44:20,040 --> 00:44:23,000
Speaker 1: Like Mark Zuckerberg went on a podcast not that long

753
00:44:23,000 --> 00:44:26,000
Speaker 1: ago and said that he thinks in the future most

754
00:44:26,000 --> 00:44:28,760
Speaker 1: of us will have a good percentage of our friends

755
00:44:28,880 --> 00:44:32,279
Speaker 1: be AI and that we'll love it. And so to me,

756
00:44:32,440 --> 00:44:37,440
Speaker 1: that's like the person who broke friendship using Facebook selling

757
00:44:37,440 --> 00:44:40,719
Speaker 1: it back to us as a commodity through AI that

758
00:44:40,760 --> 00:44:43,520
Speaker 1: he profits from. And it's like, and don't worry, You're

759
00:44:43,520 --> 00:44:45,520
Speaker 1: gonna love it. I don't know if I agree with you?

760
00:44:45,600 --> 00:44:46,440
Speaker 1: Is Mark Zuckerberg?

761
00:44:48,280 --> 00:44:50,399
Speaker 4: Yeah, I feel like he's trying to convince himself because

762
00:44:50,400 --> 00:44:52,360
Speaker 4: he might be like without friends.

763
00:44:53,120 --> 00:44:56,160
Speaker 3: Yeah, well again, we need to watch the social networks, right,

764
00:44:56,800 --> 00:44:58,560
Speaker 3: So I'm so.

765
00:44:58,440 --> 00:45:02,880
Speaker 1: Convinced we need to make this movie series. Oh y'all

766
00:45:03,000 --> 00:45:04,759
Speaker 1: see how much I want to talk about the movie her.

767
00:45:04,920 --> 00:45:06,560
Speaker 4: So, yes, we're.

768
00:45:06,400 --> 00:45:08,600
Speaker 2: Gonna make all of this happen, I think in the end.

769
00:45:09,800 --> 00:45:12,839
Speaker 4: And I'm thinking about this because we've become so well

770
00:45:13,000 --> 00:45:16,960
Speaker 4: maybe me, I don't know, acutely aware of the surveillance

771
00:45:17,239 --> 00:45:22,279
Speaker 4: that we are under constantly, and it's almost inevitable, whether

772
00:45:22,360 --> 00:45:24,600
Speaker 4: it's because we were tricked or we were promised, or

773
00:45:24,600 --> 00:45:27,719
Speaker 4: we thought there'd be safety, and then these leaders pop

774
00:45:27,840 --> 00:45:30,799
Speaker 4: up and you realize these leaders, once you know who

775
00:45:30,840 --> 00:45:33,760
Speaker 4: they are, you start feeling like, oh.

776
00:45:33,640 --> 00:45:35,640
Speaker 2: My god, oh my god.

777
00:45:35,200 --> 00:45:38,080
Speaker 4: They have my information, they know what I think, or

778
00:45:38,120 --> 00:45:40,439
Speaker 4: they know what I ask, like what have I done?

779
00:45:41,000 --> 00:45:44,600
Speaker 1: Yes? That is my big Like if I had to

780
00:45:44,600 --> 00:45:46,960
Speaker 1: say there was a big question to be asked that

781
00:45:47,000 --> 00:45:49,680
Speaker 1: we all should be asking ourselves about. This is if

782
00:45:49,680 --> 00:45:53,360
Speaker 1: you watch the Social Network is Mark Zuckerberg as portrayed

783
00:45:53,400 --> 00:45:56,319
Speaker 1: by Jesse Eisenberg. Is that the person that you want

784
00:45:56,320 --> 00:45:59,080
Speaker 1: to be in charge of your intimate relationships? What do

785
00:45:59,120 --> 00:46:04,240
Speaker 1: you trust that person with your intimate connections? Friendships, romance, sex?

786
00:46:04,880 --> 00:46:08,200
Speaker 1: Should we trust leaders like Sam Altman, who who I

787
00:46:08,239 --> 00:46:10,600
Speaker 1: would say, like go out of their way to be

788
00:46:10,640 --> 00:46:13,799
Speaker 1: as opaque and slippery as possible. Can we trust them

789
00:46:13,840 --> 00:46:18,080
Speaker 1: with something as sensitive as our human intimacies? What are

790
00:46:18,120 --> 00:46:20,600
Speaker 1: their track records look like? What are the implications of

791
00:46:20,640 --> 00:46:23,280
Speaker 1: handing something so human and so vulnerable and so intimate,

792
00:46:23,600 --> 00:46:26,360
Speaker 1: with such potential for harm to these people who go

793
00:46:26,440 --> 00:46:28,279
Speaker 1: out of their way to not be able to be

794
00:46:28,360 --> 00:46:31,600
Speaker 1: pinned down? Like that is the question, you know, people

795
00:46:31,600 --> 00:46:33,920
Speaker 1: should ask themselves. I know what the answer is to

796
00:46:33,920 --> 00:46:36,239
Speaker 1: that question for me, but I think that is the

797
00:46:36,320 --> 00:46:38,560
Speaker 1: question we should be asking ourselves when we think about

798
00:46:38,600 --> 00:46:41,000
Speaker 1: the role of the of AI and intimacy, that we're

799
00:46:41,000 --> 00:46:42,239
Speaker 1: sort of being sold right now.

800
00:46:44,160 --> 00:46:46,719
Speaker 2: I don't even want to talk to them, nonetheless get

801
00:46:46,760 --> 00:46:50,880
Speaker 2: to know me, But.

802
00:46:51,000 --> 00:46:52,880
Speaker 1: I don't think I would want to have dinner with

803
00:46:53,000 --> 00:46:55,120
Speaker 1: Mark right like, just looking.

804
00:46:54,840 --> 00:46:58,239
Speaker 3: At them things and go oh no, I think he

805
00:46:58,239 --> 00:47:05,799
Speaker 3: would be a miserable dinner partner. Oh well, I am

806
00:47:05,960 --> 00:47:09,400
Speaker 3: fascinated by this topic. I am so interested in learning

807
00:47:09,400 --> 00:47:13,520
Speaker 3: more about it, and I'm so excited that you have

808
00:47:14,040 --> 00:47:19,400
Speaker 3: this audio book that discusses more about it. So, Bridget,

809
00:47:19,440 --> 00:47:23,399
Speaker 3: can you tell the listeners where they can find you

810
00:47:23,480 --> 00:47:25,040
Speaker 3: and more about this audiobook.

811
00:47:25,400 --> 00:47:29,680
Speaker 1: Yes, if you're thinking this conversation is fascinating or as

812
00:47:29,719 --> 00:47:32,080
Speaker 1: as fascinating as we think it is, and you want

813
00:47:32,080 --> 00:47:34,600
Speaker 1: to hear more, I promise you you are not ready

814
00:47:34,640 --> 00:47:37,719
Speaker 1: for the wild world that is the connection of intimacy

815
00:47:37,760 --> 00:47:40,440
Speaker 1: and AI. You can pre order my book Love at

816
00:47:40,440 --> 00:47:43,880
Speaker 1: First Prompt at Love at First Prompt dot AI. The

817
00:47:43,880 --> 00:47:48,640
Speaker 1: book comes out in July on July fourteenth. The rabbit

818
00:47:48,680 --> 00:47:51,000
Speaker 1: Holes that we explore I have not even scratched the

819
00:47:51,000 --> 00:47:53,640
Speaker 1: service of Please check it out. It is a very

820
00:47:53,719 --> 00:47:55,520
Speaker 1: fascinating listen. I promise you that.

821
00:47:56,840 --> 00:47:58,680
Speaker 2: Yes, and we are so excited.

822
00:47:58,840 --> 00:48:01,800
Speaker 3: We want to have you back to discuss it more

823
00:48:03,040 --> 00:48:05,279
Speaker 3: and for our movie mini series that I think we

824
00:48:05,280 --> 00:48:06,080
Speaker 3: should make happen.

825
00:48:06,600 --> 00:48:11,600
Speaker 1: Have y'all seen Companion? Yes, I loved it. Companion. It

826
00:48:11,680 --> 00:48:15,480
Speaker 1: is great. It's so funny. Samantha hasn't seen.

827
00:48:15,400 --> 00:48:17,280
Speaker 2: Like I'm on to get out of the loop.

828
00:48:17,840 --> 00:48:19,600
Speaker 3: This is gonna be great. This is gonna be a

829
00:48:19,600 --> 00:48:22,160
Speaker 3: great mini series. I'm so excited about it.

830
00:48:22,200 --> 00:48:22,520
Speaker 1: Me too.

831
00:48:23,920 --> 00:48:27,320
Speaker 3: Yes, But the listeners can find you in other places, Bridget.

832
00:48:27,560 --> 00:48:29,680
Speaker 1: You can check out my podcast. There are no girls

833
00:48:29,680 --> 00:48:31,480
Speaker 1: on the internet. You can check me out on Instagram

834
00:48:31,520 --> 00:48:34,319
Speaker 1: at bridget ran DC or on YouTube. Bet there are

835
00:48:34,360 --> 00:48:37,520
Speaker 1: no girls on the internet. Yes, and go do that.

836
00:48:37,640 --> 00:48:41,920
Speaker 3: If you have not already listeners, go pre order Bridget's

837
00:48:41,960 --> 00:48:45,839
Speaker 3: audio book. If you would like to contact us, you can.

838
00:48:46,320 --> 00:48:48,360
Speaker 3: You can email us at Hello at Stuffnever Told You

839
00:48:48,400 --> 00:48:50,719
Speaker 3: dot com. We're on Blue Sky at most podcast, or

840
00:48:50,719 --> 00:48:52,719
Speaker 3: on Instagram and TikTok at stuff I Never Told You.

841
00:48:52,880 --> 00:48:55,520
Speaker 3: We're also on YouTube. We have some merchandise at Common Bureau,

842
00:48:55,640 --> 00:48:57,279
Speaker 3: and we have a book you can get wherever you

843
00:48:57,360 --> 00:49:00,000
Speaker 3: get your books. Thanks, it's always to our super Dischristine,

844
00:49:00,040 --> 00:49:02,160
Speaker 3: our executive producer My and our contributor Joey.

845
00:49:02,480 --> 00:49:03,279
Speaker 2: Thank you and.

846
00:49:03,239 --> 00:49:05,279
Speaker 3: Thanks to you for listening. Stuff never told you his

847
00:49:05,320 --> 00:49:07,600
Speaker 3: protection of iHeartRadio. For more podcasts from my Heart Radio,

848
00:49:07,640 --> 00:49:09,800
Speaker 3: you can check out the iHeartRadio app, Apple podcast or

849
00:49:09,800 --> 00:49:11,320
Speaker 3: where you listen to your favorite shows,