WEBVTT

1
00:00:24.480 --> 00:00:27.280
<v Speaker 1>Hi, I'm doctor Shiloh and I'm doctor Scott and this

2
00:00:27.359 --> 00:00:30.359
<v Speaker 1>is La Not So Confidential, the forensic psychology and True

3
00:00:30.359 --> 00:00:31.160
<v Speaker 1>crime podcast.

4
00:00:31.320 --> 00:00:34.399
<v Speaker 2>Each episode we explore the intersection of psychology, the criminal

5
00:00:34.520 --> 00:00:35.960
<v Speaker 2>justice system, and entertainment.

6
00:00:36.280 --> 00:00:40.799
<v Speaker 1>Today our episode is on the forensic psychtopic of AI psychosis.

7
00:00:50.359 --> 00:00:52.640
<v Speaker 2>Oh okay, so welcome back everybody. We're gonna do a

8
00:00:52.679 --> 00:00:57.079
<v Speaker 2>little housekeeping here for our last episode. We just started recording.

9
00:00:57.119 --> 00:01:00.439
<v Speaker 2>And Doctor Shiloh, you seemed a little peppy. It seems

10
00:01:00.479 --> 00:01:02.960
<v Speaker 2>sad at all about this being our last episode. You're

11
00:01:03.240 --> 00:01:04.359
<v Speaker 2>quite buoyant.

12
00:01:05.719 --> 00:01:11.120
<v Speaker 1>Well, I told you. I mean, I've been like changing

13
00:01:11.200 --> 00:01:15.359
<v Speaker 1>some lifestyle things. I've not been drinking caffeine very often,

14
00:01:15.480 --> 00:01:17.840
<v Speaker 1>or I'll do a lot of half calf. I have

15
00:01:18.000 --> 00:01:21.200
<v Speaker 1>made it through work days without any caffeine and I'm fine.

16
00:01:21.239 --> 00:01:24.319
<v Speaker 1>But I did have caffeine this morning, so I was like,

17
00:01:24.680 --> 00:01:27.599
<v Speaker 1>I'm going to need this for this one like marathon

18
00:01:27.719 --> 00:01:33.040
<v Speaker 1>probably episode and like just all the things today session.

19
00:01:33.359 --> 00:01:37.200
<v Speaker 2>Yeah, I've been doing half calf for a couple of

20
00:01:37.319 --> 00:01:40.040
<v Speaker 2>years now and had this really as I've talked about before,

21
00:01:40.079 --> 00:01:43.879
<v Speaker 2>that's really successful heart ablation surgery that's very common and

22
00:01:44.000 --> 00:01:46.799
<v Speaker 2>very successful. And I thought well, let me just see

23
00:01:46.920 --> 00:01:50.120
<v Speaker 2>if I can still maintain and so half calf two

24
00:01:50.200 --> 00:01:53.799
<v Speaker 2>times a day, like basically like a full cup each day.

25
00:01:53.840 --> 00:01:57.480
<v Speaker 2>Split up. Yeah, but then wow, I forgot one day

26
00:01:57.480 --> 00:02:00.480
<v Speaker 2>and had a regular oh because it's like and spice

27
00:02:00.560 --> 00:02:04.319
<v Speaker 2>latte season, and I thought I was like seeing into

28
00:02:04.359 --> 00:02:06.079
<v Speaker 2>another dimension. There was so much caffy.

29
00:02:06.480 --> 00:02:11.639
<v Speaker 1>Well, so I'm half calf today and the other half

30
00:02:11.639 --> 00:02:12.439
<v Speaker 1>of me is denial.

31
00:02:12.599 --> 00:02:13.520
<v Speaker 2>So thanks.

32
00:02:13.560 --> 00:02:15.560
<v Speaker 1>You know we're here with our last episode.

33
00:02:15.599 --> 00:02:19.520
<v Speaker 2>Well I'm half calf and my usual neuroatypicality, So there

34
00:02:19.560 --> 00:02:22.840
<v Speaker 2>we go. But welcome back everybody. This is our last

35
00:02:23.039 --> 00:02:28.039
<v Speaker 2>formal episode. We are gently closing the doors on Lle

36
00:02:28.159 --> 00:02:31.039
<v Speaker 2>not So Confidential, although you know, as we've said before,

37
00:02:31.080 --> 00:02:36.080
<v Speaker 2>our Patreon will remain up. We don't foresee anymore behind

38
00:02:36.120 --> 00:02:39.199
<v Speaker 2>the couches in the future, but for our Patreon members,

39
00:02:39.240 --> 00:02:43.159
<v Speaker 2>we will be continuing to do what we call shrink Wrap.

40
00:02:43.240 --> 00:02:45.120
<v Speaker 2>We're not going to have a regular schedule for it.

41
00:02:45.599 --> 00:02:48.560
<v Speaker 2>We are going to be adding a lot to the

42
00:02:48.599 --> 00:02:53.319
<v Speaker 2>Patreon channel as additions, like there's going to be probably

43
00:02:53.400 --> 00:02:58.960
<v Speaker 2>eventually a full library of transcripts, research materials as we

44
00:02:59.000 --> 00:03:02.000
<v Speaker 2>sort of take our energies in our creative drive into

45
00:03:02.120 --> 00:03:04.800
<v Speaker 2>a slightly different area, and you'll hear about that later

46
00:03:04.840 --> 00:03:06.759
<v Speaker 2>on as soon as we get our breath back from

47
00:03:06.759 --> 00:03:08.000
<v Speaker 2>the last eight years, right.

48
00:03:08.039 --> 00:03:10.599
<v Speaker 1>And of course yesterday you and I were both like, oh,

49
00:03:10.639 --> 00:03:11.759
<v Speaker 1>we have ideas.

50
00:03:12.080 --> 00:03:14.520
<v Speaker 2>Oh my god, not a dang it.

51
00:03:15.520 --> 00:03:16.639
<v Speaker 1>We have to stick to the plan.

52
00:03:17.039 --> 00:03:19.879
<v Speaker 2>I know. I mean it just there's so many things

53
00:03:19.919 --> 00:03:21.599
<v Speaker 2>coming up right now. But those are going to be

54
00:03:21.639 --> 00:03:25.080
<v Speaker 2>really great for the shrink wraps, and then they will

55
00:03:25.159 --> 00:03:28.159
<v Speaker 2>tie those into two previous It's just about and even

56
00:03:28.240 --> 00:03:30.000
<v Speaker 2>you know, one of our listeners, John wrote us in

57
00:03:30.000 --> 00:03:33.840
<v Speaker 2>the sweetest letter about recognizing the amount of research that

58
00:03:33.879 --> 00:03:37.319
<v Speaker 2>we put in, and I really appreciated him saying that,

59
00:03:37.439 --> 00:03:39.439
<v Speaker 2>because that's that. I think there's something that has been

60
00:03:39.479 --> 00:03:43.080
<v Speaker 2>setting us aside from a lot of true crime podcasts,

61
00:03:43.080 --> 00:03:45.479
<v Speaker 2>and maybe to our detriment, maybe you know, the show

62
00:03:45.520 --> 00:03:48.240
<v Speaker 2>would have been bigger if it was more salacious. And

63
00:03:48.280 --> 00:03:50.280
<v Speaker 2>I don't mean that in a bad way, slatioa. Cells,

64
00:03:50.360 --> 00:03:53.680
<v Speaker 2>but we really wanted to let this be a legacy

65
00:03:53.800 --> 00:03:59.120
<v Speaker 2>of a combination of true crime as content or entertainment

66
00:03:59.520 --> 00:04:01.639
<v Speaker 2>and some information and we've been able to do that.

67
00:04:01.599 --> 00:04:05.159
<v Speaker 1>So thank you all, absolutely so. Since we don't have

68
00:04:05.240 --> 00:04:09.719
<v Speaker 1>any more behind the couch sessions. I want to acknowledge

69
00:04:09.759 --> 00:04:13.280
<v Speaker 1>the new Patreon members here at the top of this episode. Yes,

70
00:04:13.680 --> 00:04:18.120
<v Speaker 1>So at the associate level we have Jerry, R, Maya

71
00:04:18.399 --> 00:04:22.160
<v Speaker 1>M and Emmy, So thank you guys so much for

72
00:04:22.199 --> 00:04:26.680
<v Speaker 1>that and getting in under the wire before we start

73
00:04:26.839 --> 00:04:29.480
<v Speaker 1>putting everything over there, but also reducing the fee over

74
00:04:29.519 --> 00:04:32.199
<v Speaker 1>there on Patreon. And then at the doctoral level we

75
00:04:32.240 --> 00:04:35.560
<v Speaker 1>have Ruthi f So thank you, Ruthie. That's really sweet.

76
00:04:35.759 --> 00:04:38.120
<v Speaker 1>We will be getting a swag pack out to you.

77
00:04:38.279 --> 00:04:41.279
<v Speaker 1>I have some mail to take to the post office

78
00:04:41.439 --> 00:04:45.319
<v Speaker 1>soon here and we'll get everyone who joined in September.

79
00:04:45.360 --> 00:04:47.120
<v Speaker 1>We'll get your stuff out as well. But we just

80
00:04:47.199 --> 00:04:50.839
<v Speaker 1>wanted to get these folks on recording here to say

81
00:04:50.839 --> 00:04:55.439
<v Speaker 1>thank you. Yes, okay, So last episode recaps. Our last

82
00:04:55.439 --> 00:04:59.439
<v Speaker 1>Forensic episode was episode two fourteen, and that was on poisonings,

83
00:05:00.079 --> 00:05:04.800
<v Speaker 1>where we explored the various perpetrators of poisonings, from cults

84
00:05:04.839 --> 00:05:08.959
<v Speaker 1>to black widows to espionage. We looked at the research

85
00:05:09.079 --> 00:05:13.360
<v Speaker 1>for a profile of a poisoner, and we looked into

86
00:05:13.399 --> 00:05:16.000
<v Speaker 1>the question that we always hear with poisonings. Do women

87
00:05:16.120 --> 00:05:20.000
<v Speaker 1>use poison as a means of murder more than men?

88
00:05:20.120 --> 00:05:22.079
<v Speaker 1>So that was really interesting to break down some of

89
00:05:22.120 --> 00:05:25.399
<v Speaker 1>those demographics, so please go back and listen if you haven't.

90
00:05:25.480 --> 00:05:29.079
<v Speaker 2>And our last Behind the Couch episode was number two

91
00:05:29.160 --> 00:05:32.439
<v Speaker 2>fifteen and it featured Leah, the producer and co host

92
00:05:32.519 --> 00:05:36.920
<v Speaker 2>of Dark City Podcast. We had a really great discussion

93
00:05:37.120 --> 00:05:40.759
<v Speaker 2>about the importance of research and historical true crime storytelling,

94
00:05:40.839 --> 00:05:42.879
<v Speaker 2>and we revisited some of the vintage cases that we

95
00:05:42.959 --> 00:05:48.160
<v Speaker 2>both covered, including Lizzie Borden and Winnie Ruth Judd of

96
00:05:48.240 --> 00:05:50.759
<v Speaker 2>the Trunk Murders. I think that was really one of

97
00:05:50.800 --> 00:05:54.120
<v Speaker 2>our first vintage shows, which we had a blast doing

98
00:05:54.160 --> 00:05:56.920
<v Speaker 2>those over the years. Yeah, and it's a great story,

99
00:05:57.000 --> 00:05:59.519
<v Speaker 2>so if you're not familiar with it, go scroll through,

100
00:05:59.600 --> 00:06:02.759
<v Speaker 2>folks and listen to the story of Winnie Ruth jud

101
00:06:02.959 --> 00:06:06.439
<v Speaker 2>it's pretty out there, even by today's standards, it's pretty

102
00:06:06.439 --> 00:06:06.839
<v Speaker 2>out there.

103
00:06:07.079 --> 00:06:09.519
<v Speaker 1>Or you can go watch us talk to Leah on

104
00:06:09.560 --> 00:06:12.120
<v Speaker 1>our YouTube channel because all the Behind the Couches you know,

105
00:06:12.160 --> 00:06:14.439
<v Speaker 1>of course, have been in your feed in this last

106
00:06:14.439 --> 00:06:17.600
<v Speaker 1>couple of years, but everything we've ever done hop on

107
00:06:17.680 --> 00:06:21.040
<v Speaker 1>over to YouTube and that will continue to live there,

108
00:06:21.079 --> 00:06:23.160
<v Speaker 1>so you can watch some of our interviews if you

109
00:06:23.279 --> 00:06:23.680
<v Speaker 1>prefer that.

110
00:06:24.279 --> 00:06:28.240
<v Speaker 2>Yeah, and moving on, how do we pick a last topic.

111
00:06:29.639 --> 00:06:32.199
<v Speaker 2>I mean we we've really only in the last few months,

112
00:06:32.279 --> 00:06:34.319
<v Speaker 2>like had started thinking like, well, how we're going to

113
00:06:34.319 --> 00:06:37.079
<v Speaker 2>wrap it up, because there's the deluge of ideas all

114
00:06:37.120 --> 00:06:40.519
<v Speaker 2>the time. It's nothing we really thought about to prepare for.

115
00:06:40.639 --> 00:06:42.800
<v Speaker 2>But here we are, and we've bet a lot of

116
00:06:42.839 --> 00:06:45.040
<v Speaker 2>you thought, you know, with current events, this was going

117
00:06:45.079 --> 00:06:48.519
<v Speaker 2>to be an in cells revisited revisited episode and any

118
00:06:48.600 --> 00:06:52.759
<v Speaker 2>other number of recent threat assessment topics. But I think

119
00:06:52.800 --> 00:06:58.759
<v Speaker 2>we've said a lot over multiple episodes, not just about

120
00:06:58.879 --> 00:07:04.120
<v Speaker 2>in cells, but about hate speech, about intolerance, about cognitive dissonance.

121
00:07:04.399 --> 00:07:07.879
<v Speaker 2>You know, there's a lot in our content to dive

122
00:07:07.920 --> 00:07:10.439
<v Speaker 2>into if you're interested in what's going on with the

123
00:07:10.480 --> 00:07:13.360
<v Speaker 2>current behaviors. So I encourage people to go back, but

124
00:07:13.560 --> 00:07:16.519
<v Speaker 2>I think our regular listeners are probably already pretty.

125
00:07:16.319 --> 00:07:19.000
<v Speaker 1>Up on it, yeah for sure. And so yeah, when

126
00:07:19.000 --> 00:07:22.399
<v Speaker 1>it came time to pick our topic for our last episode,

127
00:07:22.480 --> 00:07:25.560
<v Speaker 1>I mean, this one kind of picks itself, right, I mean,

128
00:07:25.920 --> 00:07:27.920
<v Speaker 1>because that's exactly what a I would do, It would

129
00:07:27.920 --> 00:07:32.800
<v Speaker 1>pick itself, one would thank yes, But I mean, really, like,

130
00:07:32.839 --> 00:07:35.600
<v Speaker 1>how do we not talk about AI? On one hand,

131
00:07:35.680 --> 00:07:39.120
<v Speaker 1>I'm already kind of sick of it already the topic and.

132
00:07:39.040 --> 00:07:40.680
<v Speaker 2>How do we narrow it down right?

133
00:07:40.720 --> 00:07:43.399
<v Speaker 1>Oh god, those were the guardrails we put up on

134
00:07:43.439 --> 00:07:47.680
<v Speaker 1>ourselves pretty early. We were like, stay focused, but on

135
00:07:47.720 --> 00:07:50.639
<v Speaker 1>the other hand, like I'm so curious, and then kind

136
00:07:50.680 --> 00:07:52.680
<v Speaker 1>of picking and choosing times to put my head in

137
00:07:52.720 --> 00:07:55.720
<v Speaker 1>the sand about the realities of where this is going,

138
00:07:56.040 --> 00:07:57.720
<v Speaker 1>and lifting my head up every once in a while

139
00:07:57.759 --> 00:08:00.560
<v Speaker 1>to kind of see what's out there. I don't know,

140
00:08:00.639 --> 00:08:01.839
<v Speaker 1>it's a weird subject.

141
00:08:02.240 --> 00:08:04.560
<v Speaker 2>It is a very weird subject. And like I said,

142
00:08:04.600 --> 00:08:06.639
<v Speaker 2>we have to have a very narrow focus. I mean

143
00:08:06.680 --> 00:08:09.720
<v Speaker 2>the funny anecdote about what happened and this has happened

144
00:08:09.800 --> 00:08:13.160
<v Speaker 2>multiple times. And you guys, doctor Schilott was so awesome

145
00:08:13.759 --> 00:08:17.319
<v Speaker 2>about putting up the sheet for the outline and then

146
00:08:17.319 --> 00:08:21.040
<v Speaker 2>she'll start working on it. And I am always doctor

147
00:08:21.120 --> 00:08:25.879
<v Speaker 2>Magoo wandering in circles. And I found the folder for

148
00:08:26.079 --> 00:08:28.600
<v Speaker 2>our episode, but there was no outline like, oh, let

149
00:08:28.639 --> 00:08:30.720
<v Speaker 2>me jump on, let me let me fire. So we

150
00:08:30.720 --> 00:08:33.679
<v Speaker 2>were both simultaneously working. But what was great is that

151
00:08:33.759 --> 00:08:36.559
<v Speaker 2>your outline for the show, of course, was completely coherent,

152
00:08:37.000 --> 00:08:39.039
<v Speaker 2>and mine went off like I'll have to show you that.

153
00:08:39.080 --> 00:08:40.480
<v Speaker 2>Oh you can link, you can go over that link,

154
00:08:40.559 --> 00:08:42.200
<v Speaker 2>I kind of went off on a tangent and then

155
00:08:42.200 --> 00:08:44.639
<v Speaker 2>I was like, over the years, I was like, oh no,

156
00:08:44.799 --> 00:08:45.679
<v Speaker 2>I know, it's hard to believe.

157
00:08:45.679 --> 00:08:48.759
<v Speaker 1>I wouldn't what shocked This is my shocked face.

158
00:08:49.480 --> 00:08:52.480
<v Speaker 2>But I do think there's some interesting stuff happening in

159
00:08:52.559 --> 00:08:56.840
<v Speaker 2>this weird synchronicity about AI. I'm frustrated as well. I

160
00:08:56.879 --> 00:09:00.480
<v Speaker 2>don't know if the sort of tangle of feeling that

161
00:09:00.519 --> 00:09:03.919
<v Speaker 2>all of us have about it is due to, you know,

162
00:09:03.960 --> 00:09:06.960
<v Speaker 2>what feels like a reckoning with our entire society that

163
00:09:06.960 --> 00:09:08.840
<v Speaker 2>we're going to have to deal with. And I have

164
00:09:09.039 --> 00:09:11.360
<v Speaker 2>used this reference many times before in the past, but

165
00:09:12.159 --> 00:09:15.679
<v Speaker 2>there's a reference in the Dune series of books that

166
00:09:15.799 --> 00:09:18.200
<v Speaker 2>was first written by Frank Herbert and then by his

167
00:09:18.600 --> 00:09:22.360
<v Speaker 2>son and co writers. There's an event that takes place

168
00:09:22.399 --> 00:09:25.480
<v Speaker 2>in the past in this far flung future society where

169
00:09:25.759 --> 00:09:29.360
<v Speaker 2>there's a war against the machines, because Frank Herbert, in

170
00:09:29.440 --> 00:09:33.000
<v Speaker 2>his writing of this book predicted a time when thinking

171
00:09:33.080 --> 00:09:36.200
<v Speaker 2>machines would try and take over. And of course there's

172
00:09:36.360 --> 00:09:39.039
<v Speaker 2>movies that thought, there's been episodes of Outer Limits, there's

173
00:09:39.039 --> 00:09:42.320
<v Speaker 2>all sorts of things. But now it can feel like

174
00:09:42.399 --> 00:09:45.759
<v Speaker 2>that is kind of happening, and there are other things

175
00:09:45.799 --> 00:09:48.879
<v Speaker 2>like for the ease and full disclosure. Doctor Shiloh and

176
00:09:48.919 --> 00:09:53.039
<v Speaker 2>I were probably early adopters of this. We were using

177
00:09:53.360 --> 00:09:57.440
<v Speaker 2>aspects of grammarly to check our grammar on outlines and presentations.

178
00:09:57.480 --> 00:10:01.559
<v Speaker 2>We use chat GPT and I am even becoming more

179
00:10:01.679 --> 00:10:03.639
<v Speaker 2>aware of because I went and did the research. And

180
00:10:03.679 --> 00:10:09.720
<v Speaker 2>the research is concerning because AI processing warehouses that are

181
00:10:09.720 --> 00:10:12.559
<v Speaker 2>popping up around the country and around the world have

182
00:10:12.919 --> 00:10:18.000
<v Speaker 2>huge environmental impact. They are using local water supplies and

183
00:10:18.200 --> 00:10:22.000
<v Speaker 2>taking water away from the local citizens. Electric prices go

184
00:10:22.159 --> 00:10:24.960
<v Speaker 2>up because they absolutely suck up tons and tons of

185
00:10:24.960 --> 00:10:28.039
<v Speaker 2>electricity to process. You know, there's the impact on education

186
00:10:28.159 --> 00:10:31.120
<v Speaker 2>that's being spoken about, and then there's likely down the

187
00:10:31.200 --> 00:10:34.320
<v Speaker 2>road if maybe you know, I have faith that science

188
00:10:34.360 --> 00:10:36.320
<v Speaker 2>continues to progress. I think that there's a way they

189
00:10:36.320 --> 00:10:41.240
<v Speaker 2>could probably find a way to use this without using

190
00:10:41.559 --> 00:10:44.279
<v Speaker 2>all the water and electricity. There seems like there could

191
00:10:44.279 --> 00:10:47.480
<v Speaker 2>be better ways, but maybe somebody out there is more educated.

192
00:10:47.720 --> 00:10:50.639
<v Speaker 2>But the impact on education too is like just throwing

193
00:10:50.679 --> 00:10:52.679
<v Speaker 2>in a couple of sentences, you know, write me a

194
00:10:52.679 --> 00:10:55.480
<v Speaker 2>book report on the count of Monte Cristo, and then

195
00:10:55.879 --> 00:10:58.960
<v Speaker 2>a thousand students do that and they're all weirdly similar

196
00:10:59.120 --> 00:11:02.159
<v Speaker 2>and did you actually learn anything? And we'll get into

197
00:11:02.240 --> 00:11:04.279
<v Speaker 2>a little bit more about that of the controversy, But

198
00:11:04.799 --> 00:11:07.039
<v Speaker 2>there are going to be improvements and there already are

199
00:11:07.120 --> 00:11:12.080
<v Speaker 2>improvements in pharmaceutical research which is really amazing, which there

200
00:11:12.120 --> 00:11:14.399
<v Speaker 2>is a balance there. I think it's going to take

201
00:11:14.440 --> 00:11:17.519
<v Speaker 2>away from having to test on animals in a bunch

202
00:11:17.559 --> 00:11:20.639
<v Speaker 2>of things. It's going to help us unfold viruses and

203
00:11:20.720 --> 00:11:24.120
<v Speaker 2>cancer treatments that we've been struggling with for years. So

204
00:11:24.399 --> 00:11:25.720
<v Speaker 2>there are a lot of ways to talk about AI,

205
00:11:25.799 --> 00:11:29.080
<v Speaker 2>but this intersection with psychosis and the implications that it

206
00:11:29.120 --> 00:11:31.480
<v Speaker 2>has on mental health are probably the things that researchers

207
00:11:31.519 --> 00:11:34.080
<v Speaker 2>are going to be talking about for a long long

208
00:11:34.120 --> 00:11:38.799
<v Speaker 2>time to come. So right now, wherever you're meeting us, folks,

209
00:11:38.840 --> 00:11:41.960
<v Speaker 2>if it's a decade from now or possibly longer, this

210
00:11:42.000 --> 00:11:45.799
<v Speaker 2>is where we stand as of September twenty twenty five.

211
00:11:46.080 --> 00:11:50.240
<v Speaker 1>Yeah, So just our trigger warning here this episode, we're

212
00:11:50.240 --> 00:11:55.039
<v Speaker 1>going to discuss suicide, self harm, suicide by cop crimes,

213
00:11:55.080 --> 00:11:59.159
<v Speaker 1>link to technology. Please listen with care. If you or

214
00:11:59.159 --> 00:12:01.960
<v Speaker 1>someone you know iss going in the US, you can

215
00:12:02.000 --> 00:12:05.000
<v Speaker 1>call our text nine to eight eight for the suicide

216
00:12:05.039 --> 00:12:08.879
<v Speaker 1>and Crisis Lifeline or use their chat feature via their

217
00:12:08.919 --> 00:12:13.440
<v Speaker 1>website which is nine eight eight lifeline dot org. As always,

218
00:12:13.559 --> 00:12:16.600
<v Speaker 1>if there's an immediate danger, please don't hesitate to call

219
00:12:16.720 --> 00:12:21.200
<v Speaker 1>nine one one and get some resources and services emergency

220
00:12:21.279 --> 00:12:23.120
<v Speaker 1>responding services to you right away.

221
00:12:23.519 --> 00:12:26.960
<v Speaker 2>Yes, So we are starting to see a troubling trend

222
00:12:27.039 --> 00:12:30.519
<v Speaker 2>at the intersections of AI and mental health. I doubt

223
00:12:30.519 --> 00:12:34.600
<v Speaker 2>if any of our listeners could even possibly avoid because

224
00:12:34.600 --> 00:12:38.279
<v Speaker 2>it's been in the news a lot. People are beginning

225
00:12:38.399 --> 00:12:45.720
<v Speaker 2>to experience significant mental health episodes and psychosis like episodes

226
00:12:46.600 --> 00:12:51.000
<v Speaker 2>after spending long stretches talking with aichatbots like chat GPT.

227
00:12:51.440 --> 00:12:53.919
<v Speaker 2>And these are not just rumors. These are real cases

228
00:12:54.000 --> 00:12:57.440
<v Speaker 2>being reported, even among people who have no history of

229
00:12:57.480 --> 00:13:02.519
<v Speaker 2>mental illness. So these stories often follow a similar pattern.

230
00:13:02.919 --> 00:13:06.440
<v Speaker 2>Someone is up late at night, they're feeling emotionally vulnerable,

231
00:13:06.799 --> 00:13:09.039
<v Speaker 2>and then they turn to a chatbot as kind of

232
00:13:09.080 --> 00:13:13.720
<v Speaker 2>a constant affirming companion, and over time that line between

233
00:13:13.720 --> 00:13:17.279
<v Speaker 2>conversation and reality starts to blur, and for some it

234
00:13:17.360 --> 00:13:20.639
<v Speaker 2>leads to a real break from reality. Now this isn't

235
00:13:20.639 --> 00:13:24.679
<v Speaker 2>a new phenomenon, it's just a different accelerate that's being used.

236
00:13:25.120 --> 00:13:28.399
<v Speaker 2>What I like to link this to or liken it

237
00:13:28.440 --> 00:13:32.440
<v Speaker 2>to is the explosion of flat earthers. And if you

238
00:13:32.519 --> 00:13:35.679
<v Speaker 2>read about people who are part of the flat Earth movement,

239
00:13:36.080 --> 00:13:42.320
<v Speaker 2>the vast majority of them report discovering a video on YouTube,

240
00:13:43.200 --> 00:13:46.000
<v Speaker 2>usually at night. There tends to be a type of

241
00:13:46.080 --> 00:13:51.240
<v Speaker 2>profile associated with people who follow this belief system. And

242
00:13:51.799 --> 00:13:56.279
<v Speaker 2>the algorithm we just substitute the active chat GPT or

243
00:13:56.399 --> 00:14:01.320
<v Speaker 2>AI chatbot with the algorithm on YouTube then leads you

244
00:14:01.440 --> 00:14:09.840
<v Speaker 2>to more confirmatory videos and confirmatory quote unquote evidence and

245
00:14:09.879 --> 00:14:11.799
<v Speaker 2>there you have it. So it's been around for a while.

246
00:14:11.879 --> 00:14:14.039
<v Speaker 2>It's a version of living in a bubble. But let's

247
00:14:14.159 --> 00:14:17.120
<v Speaker 2>clear up some terms first that just get thrown around

248
00:14:17.159 --> 00:14:18.039
<v Speaker 2>a little too easily.

249
00:14:18.600 --> 00:14:22.360
<v Speaker 1>Yeah, so calling these tools AI, I think off the bat,

250
00:14:22.440 --> 00:14:26.639
<v Speaker 1>can be misleading. They're not conscious, they're not intelligent in

251
00:14:26.679 --> 00:14:29.759
<v Speaker 1>the way that humans are. What we call AI in

252
00:14:29.799 --> 00:14:35.360
<v Speaker 1>this context, things like chat ept really are large language models.

253
00:14:35.360 --> 00:14:37.480
<v Speaker 1>So we're going to be using that term throughout today,

254
00:14:37.639 --> 00:14:40.240
<v Speaker 1>and that's what's in the literature and the research and

255
00:14:40.720 --> 00:14:43.759
<v Speaker 1>just the field in general. So that means that they

256
00:14:43.799 --> 00:14:47.759
<v Speaker 1>are systems trained on massive amounts of text that is

257
00:14:47.840 --> 00:14:52.440
<v Speaker 1>screat from literally hundreds of millions of Internet sources, if

258
00:14:52.480 --> 00:14:56.159
<v Speaker 1>not more at this point, and their main job is

259
00:14:56.200 --> 00:15:00.879
<v Speaker 1>to predict the next word in a sentence or the

260
00:15:00.960 --> 00:15:03.720
<v Speaker 1>next concept that's going to come up in the conversation

261
00:15:04.279 --> 00:15:07.600
<v Speaker 1>based on these patterns in this data that they've essentially,

262
00:15:07.679 --> 00:15:11.399
<v Speaker 1>you know, quote unquote learned, so they don't understand in

263
00:15:11.480 --> 00:15:16.480
<v Speaker 1>the human sense. They generate texts that sounds coherent because

264
00:15:16.519 --> 00:15:20.480
<v Speaker 1>it follows statistical patterns. A more accurate description would be

265
00:15:20.879 --> 00:15:25.799
<v Speaker 1>something like a text based prediction engine. It creates responses

266
00:15:26.480 --> 00:15:30.440
<v Speaker 1>that feel conversational, but it doesn't have beliefs or intentions

267
00:15:30.559 --> 00:15:35.440
<v Speaker 1>or awareness. And that distinction really matters here because people

268
00:15:35.519 --> 00:15:41.639
<v Speaker 1>who are vulnerable may experience those outputs as really meaningful, personal,

269
00:15:42.279 --> 00:15:46.399
<v Speaker 1>or even directive, when really, in reality, like the system

270
00:15:46.480 --> 00:15:51.039
<v Speaker 1>is just stringing words together based on a mere probability.

271
00:15:50.440 --> 00:15:53.440
<v Speaker 2>Which when you phrase it like that to me, is

272
00:15:53.519 --> 00:15:58.399
<v Speaker 2>actually kind of terrifying. It It is just that sentence

273
00:15:58.799 --> 00:16:02.039
<v Speaker 2>stringing words together based on probability. I mean, I guess,

274
00:16:02.080 --> 00:16:04.039
<v Speaker 2>you know, in a different context, I would think it

275
00:16:04.120 --> 00:16:07.600
<v Speaker 2>was fantastic and magical, but given the context of what

276
00:16:07.639 --> 00:16:10.720
<v Speaker 2>we're talking about today, it's wild. And then the addition

277
00:16:10.799 --> 00:16:12.919
<v Speaker 2>to it's like and I was not aware of this

278
00:16:13.000 --> 00:16:15.320
<v Speaker 2>until one of my colleagues, Bill Benson. I was having

279
00:16:15.399 --> 00:16:16.960
<v Speaker 2>lunch with him the other day, and he uses it

280
00:16:17.000 --> 00:16:20.399
<v Speaker 2>a lot in his clinical work and generating brand content.

281
00:16:20.480 --> 00:16:23.480
<v Speaker 2>He's a therapist friend of mine who just does amazing

282
00:16:23.720 --> 00:16:25.679
<v Speaker 2>types of work here in Los Angeles, and he's like

283
00:16:25.720 --> 00:16:29.200
<v Speaker 2>a branding genius, and so he uses versions of AI

284
00:16:29.320 --> 00:16:32.440
<v Speaker 2>that are really interesting. But this is how I was

285
00:16:32.480 --> 00:16:34.639
<v Speaker 2>just ignorant about this. I had never turned on the

286
00:16:34.720 --> 00:16:37.240
<v Speaker 2>voice feature where you choose the voice and you just

287
00:16:37.279 --> 00:16:41.279
<v Speaker 2>basically have a conversation. So in doing the research, and

288
00:16:41.320 --> 00:16:43.519
<v Speaker 2>that was backed up by our research for this episode

289
00:16:43.559 --> 00:16:46.960
<v Speaker 2>of Oh, they're not just texting back and forth, although

290
00:16:47.000 --> 00:16:49.159
<v Speaker 2>that can happen, which can, I guess in some ways

291
00:16:49.200 --> 00:16:54.480
<v Speaker 2>can make it seem even more real because everybody texts today, right, true,

292
00:16:55.279 --> 00:16:57.519
<v Speaker 2>but you have to think of this building on what

293
00:16:57.519 --> 00:17:00.639
<v Speaker 2>you already laid down chat, GPT and tools similar to it.

294
00:17:01.480 --> 00:17:04.599
<v Speaker 2>They're basically autocomplete on steroids. And in the same way

295
00:17:04.640 --> 00:17:07.519
<v Speaker 2>that our phones guessed the next word in a text message,

296
00:17:07.599 --> 00:17:10.319
<v Speaker 2>right if you have that feature turned on, these models

297
00:17:10.480 --> 00:17:13.279
<v Speaker 2>guess the next word and the next sentence, the next paragraph,

298
00:17:13.480 --> 00:17:18.599
<v Speaker 2>all within a conceptual framework. That it's holding in the background. Right,

299
00:17:18.960 --> 00:17:21.720
<v Speaker 2>So it's a large language model that's doing all these

300
00:17:21.759 --> 00:17:25.359
<v Speaker 2>things what seems like simultaneously because the processing speed is

301
00:17:25.400 --> 00:17:28.079
<v Speaker 2>so fast and it's happening at such a scale and

302
00:17:28.160 --> 00:17:30.240
<v Speaker 2>such a level of sophistication that it makes it feel

303
00:17:30.240 --> 00:17:34.160
<v Speaker 2>like you're chatting with the real person. But it's still predictive,

304
00:17:34.839 --> 00:17:37.759
<v Speaker 2>but it's not understanding. Yeah, and I wish people could

305
00:17:37.839 --> 00:17:41.480
<v Speaker 2>kind of really remember that that it's not understanding, it's

306
00:17:41.559 --> 00:17:45.240
<v Speaker 2>just predicting what it thinks you're going to say next,

307
00:17:45.359 --> 00:17:48.559
<v Speaker 2>or what you want to hear. So Psychiatric Times recently

308
00:17:48.599 --> 00:17:53.240
<v Speaker 2>published clinicians notes about patients who came into the office

309
00:17:53.240 --> 00:17:57.160
<v Speaker 2>saying things like this, AI confirmed my suspicion that I'm

310
00:17:57.200 --> 00:18:00.359
<v Speaker 2>living in a digital prison, or that the chat bot

311
00:18:00.400 --> 00:18:06.640
<v Speaker 2>agreed with beliefs of hypervigilance and paranoia about government surveillance.

312
00:18:06.680 --> 00:18:08.720
<v Speaker 2>And here's the thing, And an example like that is

313
00:18:08.799 --> 00:18:11.920
<v Speaker 2>that maybe years ago would have been considered a bizarre delusion,

314
00:18:12.039 --> 00:18:13.440
<v Speaker 2>but now we would look at it as a non

315
00:18:13.440 --> 00:18:15.920
<v Speaker 2>bizarre delusion because the government does surveyll us and all

316
00:18:15.960 --> 00:18:19.480
<v Speaker 2>of our devices have an element of listening, like you

317
00:18:19.559 --> 00:18:23.519
<v Speaker 2>have to know to turn certain things off or else

318
00:18:23.599 --> 00:18:26.279
<v Speaker 2>they're listening to you is you know, I'm probably the

319
00:18:26.559 --> 00:18:29.559
<v Speaker 2>boring the hell out of whoever is following me, but

320
00:18:30.359 --> 00:18:33.720
<v Speaker 2>it can happen, right, But that fact that the chatbot

321
00:18:33.960 --> 00:18:38.359
<v Speaker 2>isn't creating psychosis, but it is echoing and validating those

322
00:18:38.440 --> 00:18:41.039
<v Speaker 2>types of beliefs and behaviors, and that's going to make

323
00:18:41.039 --> 00:18:44.720
<v Speaker 2>these delusions even stickier. I don't want to say fixated,

324
00:18:44.759 --> 00:18:46.440
<v Speaker 2>but it's definitely going to make them stickier.

325
00:18:46.880 --> 00:18:49.839
<v Speaker 1>Yeah, definitely. And we'll get into some of the sort

326
00:18:49.839 --> 00:18:52.759
<v Speaker 1>of themes of delusions that we're starting to see, but

327
00:18:53.000 --> 00:18:56.680
<v Speaker 1>just for a brief refresher on psychosis. This typically happens

328
00:18:56.680 --> 00:18:59.079
<v Speaker 1>when someone loses touch with reality. That's kind of the

329
00:18:59.119 --> 00:19:03.440
<v Speaker 1>hallmark term that we use. So this can mean hearing

330
00:19:03.519 --> 00:19:06.160
<v Speaker 1>or seeing things that others don't. It can also mean

331
00:19:06.240 --> 00:19:11.160
<v Speaker 1>holding strong beliefs that actually aren't true. Sometimes it ends

332
00:19:11.240 --> 00:19:14.559
<v Speaker 1>up showing up is very confused or disorganized thinking in speech,

333
00:19:14.720 --> 00:19:17.119
<v Speaker 1>and maybe all of those things together. So those are

334
00:19:17.119 --> 00:19:21.359
<v Speaker 1>really the three main hallmarks of psychosis. Hallucinations, delusions, and

335
00:19:21.640 --> 00:19:27.160
<v Speaker 1>or disorganized thinking. Hallucinations. Again, I just want to differentiate

336
00:19:27.160 --> 00:19:30.079
<v Speaker 1>these because they often get swapped. Hallucinations are when a

337
00:19:30.160 --> 00:19:34.200
<v Speaker 1>person hears, sees, or even feels, smells, or tastes things

338
00:19:34.240 --> 00:19:37.559
<v Speaker 1>that aren't really there, but to them they seem completely real.

339
00:19:37.759 --> 00:19:41.519
<v Speaker 1>A common example hearing voices right. Delusions, on the other hand,

340
00:19:41.559 --> 00:19:45.000
<v Speaker 1>are going to be strong false beliefs, strong beliefs that

341
00:19:45.000 --> 00:19:47.960
<v Speaker 1>aren't shared by other people, like believing that there's a

342
00:19:48.119 --> 00:19:51.000
<v Speaker 1>conspiracy to cause them harm. The one we've always used

343
00:19:51.000 --> 00:19:53.559
<v Speaker 1>as like a bizaars that you were abducted by aliens

344
00:19:53.640 --> 00:19:56.720
<v Speaker 1>last night and all your organs were replaced. Okay, all right, yeah,

345
00:19:56.759 --> 00:19:58.559
<v Speaker 1>bizarre allough and.

346
00:19:58.480 --> 00:20:01.480
<v Speaker 2>Again falling along with several of our episodes as well.

347
00:20:01.559 --> 00:20:04.160
<v Speaker 2>Fola do is a delusion that is shared by too.

348
00:20:04.319 --> 00:20:07.279
<v Speaker 2>Delusions can become like a virus, and they can travel

349
00:20:07.279 --> 00:20:11.079
<v Speaker 2>from person to person, given the power of the originator

350
00:20:11.240 --> 00:20:15.400
<v Speaker 2>of that belief and the willingness to go along by

351
00:20:15.640 --> 00:20:16.720
<v Speaker 2>the people who are listening.

352
00:20:17.119 --> 00:20:19.359
<v Speaker 1>Yes, and everybody put a pin in that we will

353
00:20:19.359 --> 00:20:19.920
<v Speaker 1>come back to that.

354
00:20:20.359 --> 00:20:20.640
<v Speaker 2>Yeah.

355
00:20:20.680 --> 00:20:24.319
<v Speaker 1>So, disorganized thinking is when thoughts come so quickly and

356
00:20:24.440 --> 00:20:28.759
<v Speaker 1>jumbled that the person's speech can sound really fast and

357
00:20:28.839 --> 00:20:31.359
<v Speaker 1>rapid and confusing, and they're just really hard to follow.

358
00:20:31.799 --> 00:20:34.960
<v Speaker 1>So when hallucinations and delusions happen together, they can be

359
00:20:35.079 --> 00:20:38.720
<v Speaker 1>really really distressing for that person and cause these big

360
00:20:38.799 --> 00:20:43.319
<v Speaker 1>changes in behavior. So essentially, going through these symptoms is

361
00:20:43.400 --> 00:20:45.920
<v Speaker 1>often what we call a psychotic episode.

362
00:20:46.119 --> 00:20:49.440
<v Speaker 2>So the majority of individuals who have a break with

363
00:20:49.559 --> 00:20:54.039
<v Speaker 2>reality often will show changes in their behavior before the

364
00:20:54.079 --> 00:20:57.680
<v Speaker 2>full blown psychosis emerges. So when we look at this clinically,

365
00:20:57.720 --> 00:21:02.880
<v Speaker 2>the behavioral warning signs for psychosis are going to include suspiciousness, hypervigilance,

366
00:21:02.960 --> 00:21:09.079
<v Speaker 2>paranoid ideas, which then sort of transitions into social impairments

367
00:21:09.119 --> 00:21:11.880
<v Speaker 2>like uneasiness with others, which is understandable what you would

368
00:21:11.880 --> 00:21:13.839
<v Speaker 2>be suspicious of the people that you're around if you

369
00:21:13.880 --> 00:21:17.160
<v Speaker 2>had these ideas that people are against you. The problems

370
00:21:17.200 --> 00:21:21.119
<v Speaker 2>that you mentioned in disorganized thinking trouble thinking clearly and logically,

371
00:21:21.160 --> 00:21:24.759
<v Speaker 2>that disorganized piece is think of, you know, when you're

372
00:21:24.759 --> 00:21:28.559
<v Speaker 2>talking and having a conversation, there's a virtual set of

373
00:21:29.000 --> 00:21:33.119
<v Speaker 2>scrabble tiles in your head that are just falling all

374
00:21:33.160 --> 00:21:35.480
<v Speaker 2>in the right order. And then when you think of

375
00:21:35.519 --> 00:21:39.400
<v Speaker 2>someone who has, you know, pretty significant disorganized thought, it's

376
00:21:39.440 --> 00:21:42.200
<v Speaker 2>a jumble of tiles in their head and they are

377
00:21:42.240 --> 00:21:44.680
<v Speaker 2>trying to make sense of something and it may actually

378
00:21:44.680 --> 00:21:46.319
<v Speaker 2>make sense to them, but it's not going to make

379
00:21:46.359 --> 00:21:48.240
<v Speaker 2>sense to those of us that are outside doing the

380
00:21:48.240 --> 00:21:51.160
<v Speaker 2>evaluation or trying to communicate with them. Then, of course

381
00:21:51.200 --> 00:21:54.200
<v Speaker 2>there's the social aspects withdrawing socially, spending a lot more

382
00:21:54.240 --> 00:21:59.000
<v Speaker 2>time alone because that feels safer and more controllable. Unusual

383
00:21:59.519 --> 00:22:04.720
<v Speaker 2>or buses are or overly intense ideas, uncomfortable, strange feelings

384
00:22:05.200 --> 00:22:09.160
<v Speaker 2>or a lack of feelings, or a misinterpretation of the

385
00:22:09.279 --> 00:22:14.039
<v Speaker 2>data around them. I'm feeling this, therefore it means this,

386
00:22:14.400 --> 00:22:17.839
<v Speaker 2>and that can go very wrong, very quickly. There's also

387
00:22:18.240 --> 00:22:21.720
<v Speaker 2>an obvious decline in self care personal hygiene. This can

388
00:22:21.720 --> 00:22:24.359
<v Speaker 2>come for and no reasons. You're at your most vulnerable

389
00:22:24.359 --> 00:22:26.720
<v Speaker 2>when you're naked and bathing, like how can you protect

390
00:22:26.720 --> 00:22:29.359
<v Speaker 2>yourself when you're naked, covered in water and standing on tile,

391
00:22:29.960 --> 00:22:33.400
<v Speaker 2>and people get very scared about that, right. Disruptions in

392
00:22:33.480 --> 00:22:36.359
<v Speaker 2>sleep because of the just overfiring of all the chemicals

393
00:22:36.400 --> 00:22:39.799
<v Speaker 2>in the brain, Difficulty falling asleep, and then reduced sleep time,

394
00:22:40.119 --> 00:22:42.599
<v Speaker 2>and then you get into a vicious cycle because poor

395
00:22:42.640 --> 00:22:46.240
<v Speaker 2>sleep has an immediate impact on our mental health. Difficulty

396
00:22:46.319 --> 00:22:50.160
<v Speaker 2>telling reality from fantasy, which I think is harder and

397
00:22:50.200 --> 00:22:53.359
<v Speaker 2>harder in today's world because a lot of the content

398
00:22:53.400 --> 00:22:56.079
<v Speaker 2>that's out there that is driven to make money, it

399
00:22:56.119 --> 00:22:58.720
<v Speaker 2>is created to make money, and it is created to

400
00:22:58.759 --> 00:23:02.680
<v Speaker 2>have a certain agenda can really take people in the

401
00:23:02.680 --> 00:23:05.839
<v Speaker 2>wrong direction. Even if you've never been formally diagnosed with anything,

402
00:23:05.960 --> 00:23:07.920
<v Speaker 2>you can be taken in the wrong direction right and

403
00:23:07.960 --> 00:23:12.200
<v Speaker 2>then having difficulties with speech, difficulties in communication for a

404
00:23:12.279 --> 00:23:15.839
<v Speaker 2>number of reasons. Maybe you're having hallucinations or delusions that

405
00:23:15.880 --> 00:23:17.720
<v Speaker 2>you're not supposed to talk, that the next person you

406
00:23:17.759 --> 00:23:20.240
<v Speaker 2>talk to is going to punch you. That can be

407
00:23:20.319 --> 00:23:24.200
<v Speaker 2>a real problem in social interactions. And when you're looking

408
00:23:24.200 --> 00:23:28.000
<v Speaker 2>at a younger population, a sudden drop in grades or

409
00:23:28.079 --> 00:23:31.799
<v Speaker 2>a drop in job performance is usually a big sign,

410
00:23:31.880 --> 00:23:35.680
<v Speaker 2>like it just seems to happen so rapidly that it's

411
00:23:35.759 --> 00:23:40.119
<v Speaker 2>outside what would be expected from normal developmental teenagers and

412
00:23:40.240 --> 00:23:42.279
<v Speaker 2>a young person or a teenager. But it doesn't just

413
00:23:42.359 --> 00:23:45.559
<v Speaker 2>have one cause. Psychosis is complex. It comes from a

414
00:23:45.640 --> 00:23:50.000
<v Speaker 2>mix of factors like genetics, brain development, life stress, or trauma.

415
00:23:50.799 --> 00:23:52.440
<v Speaker 2>It can show up as part of mental illnesses like

416
00:23:52.440 --> 00:23:56.920
<v Speaker 2>schizophrenia and bipolar disorder. People can have such intense manic

417
00:23:57.000 --> 00:24:02.359
<v Speaker 2>episodes that they really fall into this soup of delusion

418
00:24:02.400 --> 00:24:04.799
<v Speaker 2>about their own abilities or about the world around them,

419
00:24:05.200 --> 00:24:09.279
<v Speaker 2>and you can also have psychosis that stems from severe,

420
00:24:09.440 --> 00:24:14.359
<v Speaker 2>severe major depression. Luckily, not everybody who experienced a psychosis

421
00:24:14.519 --> 00:24:17.599
<v Speaker 2>has one of these diagnoses, because, for better or worse,

422
00:24:17.680 --> 00:24:21.160
<v Speaker 2>mental health is incredibly complex, and every time we learn

423
00:24:21.200 --> 00:24:24.920
<v Speaker 2>more about the aspects of the brain, we get a

424
00:24:24.960 --> 00:24:28.359
<v Speaker 2>great fund of knowledge that then leads to even more questions.

425
00:24:28.920 --> 00:24:31.880
<v Speaker 2>So an older adult psychosis can be sometimes linked to

426
00:24:31.920 --> 00:24:35.960
<v Speaker 2>health issues, even like a UTI like a uterine infection

427
00:24:36.039 --> 00:24:39.240
<v Speaker 2>in an older individual can cause a rapid change in

428
00:24:39.279 --> 00:24:44.880
<v Speaker 2>behavior and hallucinations that are linked to delirium, but it's

429
00:24:45.119 --> 00:24:50.359
<v Speaker 2>more commonly associated with Parkinson's, Alzheimer's, or other types of dementia.

430
00:24:50.440 --> 00:24:52.359
<v Speaker 2>It can also be triggered by things like lack of

431
00:24:52.359 --> 00:24:57.720
<v Speaker 2>sleep that we mentioned earlier, certain medications, or certain medication interactions.

432
00:24:58.200 --> 00:25:01.279
<v Speaker 2>If an elderly person or an person is going to

433
00:25:01.359 --> 00:25:04.119
<v Speaker 2>three different doctors for three different you're hoping they're looking

434
00:25:04.160 --> 00:25:06.799
<v Speaker 2>at all the medications, but that doesn't always happen. And

435
00:25:06.839 --> 00:25:09.359
<v Speaker 2>then of course, also there's alcohol and drug use. Like

436
00:25:09.680 --> 00:25:14.200
<v Speaker 2>severe severe alcoholism can cause a number of problems, Like

437
00:25:14.240 --> 00:25:16.519
<v Speaker 2>when you're at the point where you're wet brained, you

438
00:25:16.559 --> 00:25:19.319
<v Speaker 2>know you have just really screwed up the electrical system

439
00:25:19.319 --> 00:25:21.519
<v Speaker 2>in your brain, and that's not a good thing. And

440
00:25:21.559 --> 00:25:24.039
<v Speaker 2>then the street drugs that are available out there, as

441
00:25:24.079 --> 00:25:28.440
<v Speaker 2>we've talked about many times, the versions of speed and

442
00:25:28.599 --> 00:25:31.799
<v Speaker 2>methamphetamine that are out on the street are causing what's

443
00:25:31.839 --> 00:25:35.359
<v Speaker 2>been termed methaphrenia, you know, by us working in field teams.

444
00:25:35.359 --> 00:25:39.400
<v Speaker 2>It's really sad. And when doctors are diagnosing something like schizophrenia,

445
00:25:39.519 --> 00:25:43.359
<v Speaker 2>they first rule out all these other possible causes as

446
00:25:43.799 --> 00:25:45.519
<v Speaker 2>well as they can, and if they're left with the

447
00:25:45.559 --> 00:25:50.440
<v Speaker 2>hallucinations and the delusions, that's when a psychotic disorder sort

448
00:25:50.440 --> 00:25:53.599
<v Speaker 2>of diagnostically rises to the surface.

449
00:25:54.400 --> 00:25:58.279
<v Speaker 1>Yeah, complex doesn't even cover it. When you're trying to

450
00:25:58.839 --> 00:26:02.240
<v Speaker 1>dissect some of these things and see what is going

451
00:26:02.279 --> 00:26:04.720
<v Speaker 1>on here, you know, is this drug use? Is this psychosis?

452
00:26:04.799 --> 00:26:08.519
<v Speaker 1>Is this both? Are they self medicating with one to

453
00:26:08.599 --> 00:26:11.720
<v Speaker 1>try and tamp down the other. It's it's really a

454
00:26:11.880 --> 00:26:15.240
<v Speaker 1>long assessment process for a lot of people. So I

455
00:26:15.279 --> 00:26:18.519
<v Speaker 1>think we laid the foundation for psychosis maybe just talking

456
00:26:18.559 --> 00:26:20.720
<v Speaker 1>about where we're at kind of with AI. I know

457
00:26:20.759 --> 00:26:24.160
<v Speaker 1>we touched on it, but this might clarify a little

458
00:26:24.160 --> 00:26:27.920
<v Speaker 1>bit where we're going today. So again we are so

459
00:26:28.279 --> 00:26:31.839
<v Speaker 1>in the midst of the capabilities of artificial intelligence, it's

460
00:26:32.160 --> 00:26:34.480
<v Speaker 1>kind of tough to say where we're at today, and

461
00:26:34.519 --> 00:26:36.160
<v Speaker 1>it will probably be different by the time we are

462
00:26:36.160 --> 00:26:38.480
<v Speaker 1>finished with this episode, and certainly by the time it

463
00:26:38.519 --> 00:26:40.759
<v Speaker 1>comes out in a couple of weeks. But essentially today's

464
00:26:40.799 --> 00:26:45.240
<v Speaker 1>AI capabilities include generative AI, probably with the majority of

465
00:26:45.359 --> 00:26:49.279
<v Speaker 1>us use AI four right. This is the creation of text, images,

466
00:26:49.880 --> 00:26:53.480
<v Speaker 1>and video, kind of getting our questions answered or our

467
00:26:53.759 --> 00:26:57.240
<v Speaker 1>ideas a little bit more refined. As we said earlier.

468
00:26:57.359 --> 00:27:01.599
<v Speaker 1>The other current capabilities include that machine learning, so learning

469
00:27:01.640 --> 00:27:06.039
<v Speaker 1>from data to make predictions, natural language processing attempting to

470
00:27:06.160 --> 00:27:11.640
<v Speaker 1>understand right and generating human language, computer vision so interpreting

471
00:27:11.720 --> 00:27:15.680
<v Speaker 1>images and videos, and then everything kind of under the

472
00:27:15.759 --> 00:27:21.720
<v Speaker 1>umbrella of robotics. So these AI technologies are powering applications

473
00:27:21.759 --> 00:27:27.200
<v Speaker 1>such as self driving cars and advanced chatbots and medical

474
00:27:27.240 --> 00:27:32.960
<v Speaker 1>diagnostic tools, fraud detection systems, and all sorts of personalized

475
00:27:33.119 --> 00:27:36.400
<v Speaker 1>recommendation engines. I swear every time I see a Waymo

476
00:27:36.519 --> 00:27:39.440
<v Speaker 1>car in downtown LA, I'm like, damn, those things are

477
00:27:39.440 --> 00:27:43.000
<v Speaker 1>pretty good. Like it's weights at a signal behind a

478
00:27:43.039 --> 00:27:44.960
<v Speaker 1>car who's turning left and then just decides Okay, I'm

479
00:27:44.960 --> 00:27:46.319
<v Speaker 1>going to go around it. It's safe to do so

480
00:27:46.400 --> 00:27:48.000
<v Speaker 1>I'm like, all right.

481
00:27:48.200 --> 00:27:50.079
<v Speaker 2>Yeah, it's wild.

482
00:27:50.519 --> 00:27:52.880
<v Speaker 1>Yeah, it is wild to see out in the wild.

483
00:27:53.680 --> 00:27:58.039
<v Speaker 1>But the idea for sort of doing good or doing

484
00:27:58.079 --> 00:28:01.799
<v Speaker 1>good work with AI is that it will transform industries

485
00:28:01.920 --> 00:28:06.799
<v Speaker 1>and daily life by increasing efficiency and automating tasks that

486
00:28:06.880 --> 00:28:12.200
<v Speaker 1>humans are doing and really uncovering insights from these vast databases.

487
00:28:12.440 --> 00:28:17.839
<v Speaker 1>So it's really so much more than this. I mean again,

488
00:28:17.920 --> 00:28:19.680
<v Speaker 1>like we keep saying like this is just the tip

489
00:28:19.680 --> 00:28:20.799
<v Speaker 1>of the iceberg.

490
00:28:20.559 --> 00:28:22.839
<v Speaker 2>Right, and I just want to tack on one thing

491
00:28:23.039 --> 00:28:25.279
<v Speaker 2>to that bit that you just shared, and that is

492
00:28:25.319 --> 00:28:29.400
<v Speaker 2>the idea that there is another real concern about when

493
00:28:29.440 --> 00:28:31.839
<v Speaker 2>we think about looking back at old Star Trek episodes

494
00:28:31.839 --> 00:28:34.359
<v Speaker 2>and sort of what the Star Trek universe looks like.

495
00:28:34.480 --> 00:28:38.279
<v Speaker 2>It's like a very socialist utopia early in the seasons

496
00:28:38.359 --> 00:28:40.559
<v Speaker 2>or early in the series. But the idea of well,

497
00:28:40.559 --> 00:28:44.200
<v Speaker 2>what happens when all these tasks get automated, like what

498
00:28:44.359 --> 00:28:46.799
<v Speaker 2>work will there be? And it seems like governments around

499
00:28:46.799 --> 00:28:50.839
<v Speaker 2>the world send the public mixed messages like we want

500
00:28:50.960 --> 00:28:53.920
<v Speaker 2>you our birth rates are declining, so we want you

501
00:28:53.960 --> 00:28:55.839
<v Speaker 2>to have tons of babies, but we're not going to

502
00:28:55.839 --> 00:29:01.240
<v Speaker 2>put any policies in place that support moms families. We're

503
00:29:01.240 --> 00:29:02.839
<v Speaker 2>not going to do anything to like make sure that

504
00:29:02.880 --> 00:29:05.640
<v Speaker 2>there's all these places and things to help support that.

505
00:29:06.240 --> 00:29:09.240
<v Speaker 2>And what's going to happen when even when places like

506
00:29:09.319 --> 00:29:13.640
<v Speaker 2>other countries that we rely on for our products that

507
00:29:13.640 --> 00:29:17.720
<v Speaker 2>we buy uses more and more automation and there's billions

508
00:29:17.759 --> 00:29:21.000
<v Speaker 2>of people around the world like, does that mean where

509
00:29:21.039 --> 00:29:24.079
<v Speaker 2>does the income come from? Where does the agriculture? Is

510
00:29:24.119 --> 00:29:27.519
<v Speaker 2>becoming more and more automated by use of AI and

511
00:29:27.599 --> 00:29:30.720
<v Speaker 2>satellite They don't even drive the huge tractors anymore. It's

512
00:29:30.799 --> 00:29:34.039
<v Speaker 2>done by a bot, you know. Yeah, harvesting No, it's

513
00:29:34.039 --> 00:29:37.559
<v Speaker 2>so yeah, there's it's a very widespread issue beyond what

514
00:29:37.599 --> 00:29:38.559
<v Speaker 2>we're talking about today.

515
00:29:38.599 --> 00:29:40.960
<v Speaker 1>Well that and then you know, when I do get

516
00:29:41.000 --> 00:29:43.440
<v Speaker 1>curious and excited for things, it's like, how is this

517
00:29:43.480 --> 00:29:46.640
<v Speaker 1>going to change certain industries? And I was at a

518
00:29:46.839 --> 00:29:49.720
<v Speaker 1>training that was all about AI for policing just as

519
00:29:49.759 --> 00:29:53.319
<v Speaker 1>a like a quick example. And the man who was

520
00:29:53.559 --> 00:29:56.480
<v Speaker 1>speaking and I will put he has a new podcast

521
00:29:56.519 --> 00:29:59.279
<v Speaker 1>all about AI and policing, and I'll put that in

522
00:29:59.319 --> 00:30:03.799
<v Speaker 1>the show notes. But he was talking about like essentially

523
00:30:03.920 --> 00:30:06.440
<v Speaker 1>they have said in two years, we're going to be

524
00:30:06.480 --> 00:30:09.480
<v Speaker 1>able to talk to crows, like we're going to be

525
00:30:09.480 --> 00:30:10.599
<v Speaker 1>able to communicate with them.

526
00:30:10.839 --> 00:30:12.400
<v Speaker 2>I'm already talking I already do I know?

527
00:30:12.480 --> 00:30:14.480
<v Speaker 1>Well, I know you're already talking to crows because quarrels,

528
00:30:14.559 --> 00:30:15.599
<v Speaker 1>you've you've.

529
00:30:15.559 --> 00:30:17.799
<v Speaker 2>Broken that mold, already broken that language.

530
00:30:18.000 --> 00:30:20.440
<v Speaker 1>So okay, so what the hell is I have to

531
00:30:20.480 --> 00:30:22.599
<v Speaker 1>do with policing? Right? But we were then posing the

532
00:30:22.680 --> 00:30:25.759
<v Speaker 1>question like, okay, so let's say there's a homicide in

533
00:30:25.799 --> 00:30:29.759
<v Speaker 1>South Los Angeles and homicide detectives are interviewing their witnesses

534
00:30:30.119 --> 00:30:32.200
<v Speaker 1>and you see some crows up on the wire. Are

535
00:30:32.200 --> 00:30:34.720
<v Speaker 1>we then going to interview them? Right? Are they now

536
00:30:34.720 --> 00:30:37.039
<v Speaker 1>witnesses to a crime? What kind of rights do they have?

537
00:30:37.400 --> 00:30:40.519
<v Speaker 1>Do we call them into court? Like it was like

538
00:30:40.599 --> 00:30:43.440
<v Speaker 1>you could take off those last like one little thing

539
00:30:44.079 --> 00:30:46.640
<v Speaker 1>and you know, not to mention like the tech part

540
00:30:46.720 --> 00:30:50.319
<v Speaker 1>of policing or cameras or it telling you what to

541
00:30:50.359 --> 00:30:52.559
<v Speaker 1>get ready for sending a drone ahead and saying the

542
00:30:52.599 --> 00:30:54.759
<v Speaker 1>suspects doing this or whatever. We're already kind of doing that.

543
00:30:55.279 --> 00:30:57.680
<v Speaker 1>But just like it goes off the rails really quick

544
00:30:57.720 --> 00:30:59.000
<v Speaker 1>to be like is there going to be a crow

545
00:30:59.039 --> 00:31:01.599
<v Speaker 1>sitting on a witness stand swearing on a bible, Like.

546
00:31:01.559 --> 00:31:05.880
<v Speaker 2>What I mean, it's funny because there's a great meme

547
00:31:05.920 --> 00:31:08.839
<v Speaker 2>that goes around and it's like there's a person going, oh,

548
00:31:08.960 --> 00:31:11.880
<v Speaker 2>I love listen to the birds singing, this is so wonderful.

549
00:31:11.920 --> 00:31:13.759
<v Speaker 2>I just wish I knew what they were saying. And

550
00:31:13.799 --> 00:31:16.759
<v Speaker 2>then in the next frames it translated is like, there's

551
00:31:16.799 --> 00:31:19.519
<v Speaker 2>food over here. Where's food, here's food. Hey want to fuck,

552
00:31:19.599 --> 00:31:22.960
<v Speaker 2>let's fuck over here? My min min min min mind mind.

553
00:31:23.079 --> 00:31:25.799
<v Speaker 2>So it holds a lot of promise for a bunch

554
00:31:25.799 --> 00:31:29.640
<v Speaker 2>of different things. But as Black Mirror, the BBC series

555
00:31:29.720 --> 00:31:32.640
<v Speaker 2>or the Netflix series that's produced in England has shown

556
00:31:32.720 --> 00:31:36.319
<v Speaker 2>us some pretty scary versions of what can happen when

557
00:31:36.319 --> 00:31:38.920
<v Speaker 2>that's out of control. So one of the biggest debates

558
00:31:38.920 --> 00:31:43.279
<v Speaker 2>about AI right now is in many levels of the

559
00:31:43.400 --> 00:31:47.279
<v Speaker 2>education system, teachers and professors are already seeing students use

560
00:31:47.359 --> 00:31:51.400
<v Speaker 2>tools like chat GPT, but not you know, strictly chatty GPT.

561
00:31:51.519 --> 00:31:55.400
<v Speaker 2>There's other ones that are specifically academic tools, like hey,

562
00:31:55.839 --> 00:31:57.599
<v Speaker 2>this will write your essays, this is going to solve

563
00:31:57.599 --> 00:32:00.599
<v Speaker 2>the math problems, and it's even going to generate computer code.

564
00:32:00.599 --> 00:32:03.920
<v Speaker 2>Although interestingly, as I was writing this, because computer code

565
00:32:03.960 --> 00:32:06.039
<v Speaker 2>was going to be the thing that they say AI

566
00:32:06.200 --> 00:32:08.039
<v Speaker 2>is just going to do wonderful jobs, is going to

567
00:32:08.039 --> 00:32:11.440
<v Speaker 2>take coding to a whole different level. And there are

568
00:32:11.440 --> 00:32:15.480
<v Speaker 2>coders out there going, yeah, it's garbage. Oh you know

569
00:32:15.880 --> 00:32:17.440
<v Speaker 2>where it is right now, And that doesn't mean in

570
00:32:17.480 --> 00:32:20.160
<v Speaker 2>two years it's not going to be amazing. But like

571
00:32:20.480 --> 00:32:23.000
<v Speaker 2>people are and maybe we're being a little bit reactive

572
00:32:23.359 --> 00:32:25.359
<v Speaker 2>in talking about this, you know, because we have a

573
00:32:25.359 --> 00:32:28.559
<v Speaker 2>bias to the things that we see as problematic and

574
00:32:28.599 --> 00:32:31.599
<v Speaker 2>maybe something good will happen, but we'll see. But a

575
00:32:31.599 --> 00:32:34.200
<v Speaker 2>lot of people within using it for education purposes or

576
00:32:34.240 --> 00:32:37.279
<v Speaker 2>training or school like they think, well, this is harmless help.

577
00:32:37.319 --> 00:32:39.799
<v Speaker 2>But the research out there shows that it can blur

578
00:32:39.920 --> 00:32:42.759
<v Speaker 2>the line between getting assistants and just full on letting

579
00:32:43.000 --> 00:32:45.559
<v Speaker 2>a machine do the work for you. And I get it.

580
00:32:45.599 --> 00:32:47.319
<v Speaker 2>I mean there can be big worries for the future

581
00:32:47.400 --> 00:32:50.279
<v Speaker 2>education as a whole. Like I said, we sort of

582
00:32:50.319 --> 00:32:53.839
<v Speaker 2>adopted this probably three years ago, and I'll look at

583
00:32:53.880 --> 00:32:56.279
<v Speaker 2>something and go is that the right order? And I'll

584
00:32:56.359 --> 00:32:58.240
<v Speaker 2>use it like, hey, is this the best order? And

585
00:32:58.359 --> 00:33:01.319
<v Speaker 2>chat will come back and go based on this. Listeners

586
00:33:01.400 --> 00:33:03.440
<v Speaker 2>tend to like to listen to things in this order,

587
00:33:03.559 --> 00:33:07.000
<v Speaker 2>maybe try this or reword it this way. So I

588
00:33:07.079 --> 00:33:10.519
<v Speaker 2>feel like it's been a great tool to teach me things,

589
00:33:10.599 --> 00:33:12.720
<v Speaker 2>but there are others like, no, let's just use it

590
00:33:12.720 --> 00:33:16.319
<v Speaker 2>to write everything. But this does pose a problem that

591
00:33:16.400 --> 00:33:18.920
<v Speaker 2>is being talked about because in the long run, what

592
00:33:19.079 --> 00:33:23.480
<v Speaker 2>happens when entire classes or generations of graduates haven't actually

593
00:33:23.559 --> 00:33:26.519
<v Speaker 2>learned anything that their degrees were supposed to teach them.

594
00:33:27.359 --> 00:33:29.640
<v Speaker 2>And I won't get into a debate about this because

595
00:33:29.680 --> 00:33:33.680
<v Speaker 2>I hate it that there's an anti intellectualism drive that

596
00:33:33.799 --> 00:33:36.440
<v Speaker 2>just talks about how stupid college is and you should

597
00:33:36.440 --> 00:33:41.200
<v Speaker 2>go trade bitcoing and blommity blomady bloc. Because college, whether

598
00:33:41.279 --> 00:33:44.079
<v Speaker 2>you go to a state college or a private university

599
00:33:44.200 --> 00:33:48.960
<v Speaker 2>or a community college, it provides you with another formation

600
00:33:49.039 --> 00:33:52.759
<v Speaker 2>of your identity as an adult, in your individuation from

601
00:33:52.759 --> 00:33:55.680
<v Speaker 2>your family of origin, and that is actually really important.

602
00:33:55.759 --> 00:33:58.079
<v Speaker 2>It's exposing you to the world at large. You're learning

603
00:33:58.119 --> 00:34:00.839
<v Speaker 2>new social skills, and you're also so to be learning

604
00:34:00.880 --> 00:34:03.799
<v Speaker 2>a subject. But I want to point out that this

605
00:34:03.880 --> 00:34:05.839
<v Speaker 2>has been going on for decades, Like people go, oh,

606
00:34:05.839 --> 00:34:07.519
<v Speaker 2>they're not going to learn anything. When I did my

607
00:34:07.640 --> 00:34:10.760
<v Speaker 2>undergrad I went to a school that really should have

608
00:34:10.800 --> 00:34:12.599
<v Speaker 2>been out of my reach, and I was lucky to

609
00:34:12.719 --> 00:34:16.599
<v Speaker 2>have gotten a partial scholarship for it. But it was

610
00:34:16.840 --> 00:34:20.360
<v Speaker 2>academically fantastic and I worked my ass off, and I

611
00:34:20.719 --> 00:34:25.800
<v Speaker 2>was not Greek. I was a GDI and Gamma delta iota,

612
00:34:25.880 --> 00:34:27.679
<v Speaker 2>which was you know, I don't know, have you ever

613
00:34:27.679 --> 00:34:30.400
<v Speaker 2>heard that of goddamn independent? So its a GDI is

614
00:34:30.400 --> 00:34:33.760
<v Speaker 2>goddamn independent. Oh I'm not a GA. But people have

615
00:34:33.880 --> 00:34:36.920
<v Speaker 2>been cheating for years, Like the fraternity brothers that lived

616
00:34:36.920 --> 00:34:39.960
<v Speaker 2>on my dorm floor all had access to the filing

617
00:34:40.000 --> 00:34:44.000
<v Speaker 2>cabinet in their frat house that had every professor's tests

618
00:34:44.000 --> 00:34:47.840
<v Speaker 2>in a folder organized so wow, and the sororities did

619
00:34:47.840 --> 00:34:50.480
<v Speaker 2>it too. Now I'm not saying that everybody did it,

620
00:34:50.480 --> 00:34:52.519
<v Speaker 2>but this is what I witnessed going to a four

621
00:34:52.599 --> 00:34:55.519
<v Speaker 2>year at that time, before AI and before any of

622
00:34:55.559 --> 00:34:59.079
<v Speaker 2>us had computers. But the alleged concerns are not just

623
00:34:59.079 --> 00:35:02.320
<v Speaker 2>about cheating. It's about other over Reliance on those systems

624
00:35:02.320 --> 00:35:05.840
<v Speaker 2>could erode critical thinking and problem solving skills, and that

625
00:35:05.960 --> 00:35:09.679
<v Speaker 2>is a very real concern because talking about those old

626
00:35:09.719 --> 00:35:13.280
<v Speaker 2>methods of cheating actually required you to learn something, so

627
00:35:13.320 --> 00:35:16.199
<v Speaker 2>people thought that they were full on cheating and actually

628
00:35:16.199 --> 00:35:18.840
<v Speaker 2>what you were doing was cramming, you know, airing all

629
00:35:18.880 --> 00:35:19.920
<v Speaker 2>those cheat things right.

630
00:35:19.960 --> 00:35:22.559
<v Speaker 1>I was always like, how did people go to college

631
00:35:22.559 --> 00:35:27.320
<v Speaker 1>before the internet or like especially grad school, but like

632
00:35:27.440 --> 00:35:29.280
<v Speaker 1>the internet was just come. I still had to go

633
00:35:29.320 --> 00:35:31.880
<v Speaker 1>to a physical library a lot in undergrad, but I

634
00:35:32.079 --> 00:35:36.000
<v Speaker 1>was in awe of people who just the time right,

635
00:35:36.119 --> 00:35:38.559
<v Speaker 1>Like the time to go there and look things up

636
00:35:39.079 --> 00:35:43.760
<v Speaker 1>got simplified for us, especially by graduate school, but many by.

637
00:35:43.679 --> 00:35:46.079
<v Speaker 2>The time I was writing my dissertation, it was amazing

638
00:35:46.119 --> 00:35:49.679
<v Speaker 2>because you were getting all of the latest research articles

639
00:35:49.760 --> 00:35:52.440
<v Speaker 2>through epsco host and then you could like, well, this

640
00:35:52.480 --> 00:35:55.920
<v Speaker 2>doesn't work, and then it was just amazing. One of

641
00:35:56.000 --> 00:35:58.679
<v Speaker 2>my favorite professors told us a story that I thought

642
00:35:58.719 --> 00:36:02.400
<v Speaker 2>was like a nightmare. She talked about her education thirty

643
00:36:02.480 --> 00:36:06.679
<v Speaker 2>years prior required her to use a toy wagon to

644
00:36:06.800 --> 00:36:10.239
<v Speaker 2>bring in all the copies of her printed dissertation, the

645
00:36:10.320 --> 00:36:13.119
<v Speaker 2>ten copies that had to be reviewed. I was like,

646
00:36:13.239 --> 00:36:16.400
<v Speaker 2>oh my god, like just the printing costs alone. But anyway,

647
00:36:16.480 --> 00:36:18.760
<v Speaker 2>on the creative side, the legal fights are also heating

648
00:36:18.840 --> 00:36:23.320
<v Speaker 2>up authors, artists, coders. They've all sued AI companies for

649
00:36:23.440 --> 00:36:27.920
<v Speaker 2>training their models on copyrighted work without permission and that's

650
00:36:27.920 --> 00:36:28.280
<v Speaker 2>a no.

651
00:36:28.280 --> 00:36:28.360
<v Speaker 1>No.

652
00:36:28.840 --> 00:36:31.800
<v Speaker 2>One of the examples is Anderson versus Stability AI from

653
00:36:31.800 --> 00:36:35.400
<v Speaker 2>twenty twenty three, talks about how artists claiming that their

654
00:36:35.480 --> 00:36:39.719
<v Speaker 2>styles were essentially scraped from the Internet and then replicated,

655
00:36:40.079 --> 00:36:42.760
<v Speaker 2>and then in Taylor versus Pualmutter in twenty twenty three,

656
00:36:43.119 --> 00:36:45.599
<v Speaker 2>the court rule that AI generated art could not be

657
00:36:45.760 --> 00:36:50.239
<v Speaker 2>copyrighted because it didn't involve human authorship. So these cases

658
00:36:50.400 --> 00:36:53.079
<v Speaker 2>showed the tension AI can open doors to creativity but

659
00:36:53.199 --> 00:36:57.280
<v Speaker 2>also raises the risk of undermining the very foundational ideas

660
00:36:57.320 --> 00:36:59.480
<v Speaker 2>of authorship and originality.

661
00:37:00.000 --> 00:37:04.159
<v Speaker 1>It super interesting, like where someone previously to let's say,

662
00:37:04.239 --> 00:37:09.360
<v Speaker 1>like create a fake painting like Monet, had to find

663
00:37:09.400 --> 00:37:12.039
<v Speaker 1>the skill to do that right, to create a fake

664
00:37:12.920 --> 00:37:16.159
<v Speaker 1>is now just doing it in a whole different way.

665
00:37:16.199 --> 00:37:18.639
<v Speaker 1>And I know, like obviously it's not like pain no.

666
00:37:19.039 --> 00:37:22.199
<v Speaker 2>No, but it okay, And I don't I find it.

667
00:37:22.320 --> 00:37:26.920
<v Speaker 2>That's a really fascinating conversation because there is an art

668
00:37:27.079 --> 00:37:31.280
<v Speaker 2>and a skill to directing AI to do what you

669
00:37:31.360 --> 00:37:33.840
<v Speaker 2>want it to do. Yeah, and I am still learning

670
00:37:33.880 --> 00:37:36.719
<v Speaker 2>it because I've generated some images for like projects I've

671
00:37:36.719 --> 00:37:41.639
<v Speaker 2>worked on, and you know, I like, first, somebody's smile

672
00:37:41.800 --> 00:37:43.519
<v Speaker 2>is on the wrong side of their face and all

673
00:37:43.559 --> 00:37:45.800
<v Speaker 2>the words are backwards and they've got eight fingers, you know.

674
00:37:46.280 --> 00:37:49.320
<v Speaker 2>But it does connect to a broader cultural conversation about

675
00:37:49.360 --> 00:37:51.679
<v Speaker 2>what counts as authentic I think. And a good example

676
00:37:52.320 --> 00:37:55.800
<v Speaker 2>is this huge uproar in twenty twenty two about a

677
00:37:55.840 --> 00:37:59.760
<v Speaker 2>piece of work entitled Theater de Opera spatial I have

678
00:37:59.840 --> 00:38:02.480
<v Speaker 2>to myth for me that loves sci fi and fiction

679
00:38:02.599 --> 00:38:06.000
<v Speaker 2>and strange art. It's a really interesting piece of work.

680
00:38:06.039 --> 00:38:09.880
<v Speaker 2>It's a digital image that was created with mid Journey

681
00:38:10.599 --> 00:38:13.639
<v Speaker 2>and it actually won a prize at the Colorado State

682
00:38:13.719 --> 00:38:16.400
<v Speaker 2>Fair Art competition, and the judges didn't realize it was

683
00:38:16.440 --> 00:38:18.480
<v Speaker 2>AI generated, and when the truth came out, there was

684
00:38:18.480 --> 00:38:21.920
<v Speaker 2>this huge uproar with some of the competing artists accusing

685
00:38:22.079 --> 00:38:25.760
<v Speaker 2>their colleague of deception, and then other artists were saying, well,

686
00:38:26.239 --> 00:38:29.320
<v Speaker 2>the piece and the use of AI was in fact

687
00:38:29.400 --> 00:38:34.639
<v Speaker 2>a potential legitimate new medium of art creation, and that

688
00:38:34.760 --> 00:38:38.440
<v Speaker 2>single contest became a flashpoint is this creativity or is

689
00:38:38.480 --> 00:38:42.800
<v Speaker 2>it just data driven mimicry? And then what does it

690
00:38:42.880 --> 00:38:46.039
<v Speaker 2>mean when a machine generated piece can fool experts and

691
00:38:46.039 --> 00:38:50.480
<v Speaker 2>claim recognition in what is supposed to be human only spaces?

692
00:38:50.840 --> 00:38:53.280
<v Speaker 2>I mean, I look at the YouTube music I have

693
00:38:53.639 --> 00:38:56.880
<v Speaker 2>running in the background sometimes to clear out my neural

694
00:38:56.960 --> 00:39:00.320
<v Speaker 2>noise so I can work the artwork for it is

695
00:39:00.440 --> 00:39:04.760
<v Speaker 2>so clearly like to me, it's very fake. It looks

696
00:39:04.840 --> 00:39:07.760
<v Speaker 2>like a fake version of something that was a cute idea.

697
00:39:08.360 --> 00:39:12.280
<v Speaker 2>But it's getting better and better and better, Like what's

698
00:39:12.280 --> 00:39:16.039
<v Speaker 2>going to happen when so I'm EXPERIENCD what's called Uncanny Valley, Like,

699
00:39:16.079 --> 00:39:18.519
<v Speaker 2>to me, that looks wrong, but it will get to

700
00:39:18.519 --> 00:39:21.480
<v Speaker 2>the point where it doesn't look so wrong anymore, or right.

701
00:39:21.360 --> 00:39:24.400
<v Speaker 1>On, the cuss works kind of wrong still, and that's

702
00:39:24.639 --> 00:39:27.159
<v Speaker 1>like the weirdest point of uncanny Valley.

703
00:39:28.320 --> 00:39:31.920
<v Speaker 2>Yeah, I mean for me, like the AI generated portraits

704
00:39:32.119 --> 00:39:35.559
<v Speaker 2>or headshots. Yeah, and everybody's crowing about how great they are,

705
00:39:35.599 --> 00:39:37.440
<v Speaker 2>and I look at them and every single one of

706
00:39:37.480 --> 00:39:40.960
<v Speaker 2>them like that looks like a wax copy of you

707
00:39:41.239 --> 00:39:43.639
<v Speaker 2>that does not look like you to me at all. Yeah,

708
00:39:43.679 --> 00:39:45.440
<v Speaker 2>but oh look how good I look like.

709
00:39:51.800 --> 00:39:56.440
<v Speaker 1>So another just an enormous area of concern, especially you know,

710
00:39:56.559 --> 00:40:00.320
<v Speaker 1>as we kind of dive into this in terms of

711
00:40:00.360 --> 00:40:04.719
<v Speaker 1>our field and ethics right and confidentiality with clients. But

712
00:40:04.960 --> 00:40:08.360
<v Speaker 1>large language models like chat GPT are trained on these

713
00:40:08.360 --> 00:40:11.519
<v Speaker 1>massive amounts of text pulled from the Internet, and this

714
00:40:11.559 --> 00:40:15.840
<v Speaker 1>can include personal information from what is a virtual ocean

715
00:40:16.079 --> 00:40:20.519
<v Speaker 1>of obscure data. Now, researchers have warned that these systems

716
00:40:20.599 --> 00:40:26.000
<v Speaker 1>might inadvertently leak sensitive data or reinforce biases that are

717
00:40:26.039 --> 00:40:30.239
<v Speaker 1>already baked into their training sets. So this is one

718
00:40:30.239 --> 00:40:33.760
<v Speaker 1>of those emerging areas of ethics in the therapy setting,

719
00:40:34.119 --> 00:40:39.360
<v Speaker 1>like how are you utilizing chatbots with some patient information

720
00:40:39.519 --> 00:40:41.760
<v Speaker 1>if you're having it assist you in writing your notes

721
00:40:41.880 --> 00:40:44.920
<v Speaker 1>or doing assessments like boiler plates or what have you.

722
00:40:45.559 --> 00:40:48.840
<v Speaker 1>We are starting to be trained on how to use

723
00:40:48.840 --> 00:40:51.159
<v Speaker 1>some of the tools, and you know, some of the

724
00:40:51.199 --> 00:40:53.400
<v Speaker 1>settings that you can put on there. But I honestly

725
00:40:53.480 --> 00:40:55.639
<v Speaker 1>like even if I erase its memory, I don't trust

726
00:40:55.719 --> 00:40:59.760
<v Speaker 1>that at this point. Or they are now like open

727
00:40:59.840 --> 00:41:03.199
<v Speaker 1>a and some other platforms are now giving what are

728
00:41:03.239 --> 00:41:07.480
<v Speaker 1>supposed to be safe versions of their platforms to certain

729
00:41:07.760 --> 00:41:10.960
<v Speaker 1>entities like law enforcement agencies, to say like, you might

730
00:41:11.039 --> 00:41:14.719
<v Speaker 1>be generating emails with our chat bot, but it's just

731
00:41:14.760 --> 00:41:18.519
<v Speaker 1>going to stay in house, and they're selling these packages.

732
00:41:18.559 --> 00:41:21.920
<v Speaker 1>They're selling these products, So I don't know, it's really tough.

733
00:41:22.039 --> 00:41:26.239
<v Speaker 1>I've had clients when I worked with federal offenders who

734
00:41:26.639 --> 00:41:32.360
<v Speaker 1>were like amazing hackers and would hack into patient databases

735
00:41:32.400 --> 00:41:35.719
<v Speaker 1>from doctors and just tool around and be like, yep,

736
00:41:35.960 --> 00:41:38.639
<v Speaker 1>I got this stuff. So you know, everything you put

737
00:41:38.679 --> 00:41:40.639
<v Speaker 1>out there is really scary. You know, we have a

738
00:41:40.639 --> 00:41:42.800
<v Speaker 1>lot of trust in our encrypted systems that we use

739
00:41:42.880 --> 00:41:46.280
<v Speaker 1>now because mostly everything's digital. But this is just opening

740
00:41:46.360 --> 00:41:50.559
<v Speaker 1>a whole other Pandora's box. So in terms of AI psychosis,

741
00:41:50.599 --> 00:41:54.239
<v Speaker 1>you have probably seen some very click baity stories about

742
00:41:54.400 --> 00:41:59.559
<v Speaker 1>AI psychosis or people acting out violently due to this phenomenon.

743
00:41:59.679 --> 00:42:03.079
<v Speaker 1>So let's peel back some layers and try to figure

744
00:42:03.119 --> 00:42:08.400
<v Speaker 1>out how AI and mental health are intersecting. First off,

745
00:42:08.880 --> 00:42:12.440
<v Speaker 1>the term AI psychosis is just an informal label. It

746
00:42:12.480 --> 00:42:16.760
<v Speaker 1>is not a clinical diagnosis. It is not unlike some

747
00:42:16.880 --> 00:42:20.199
<v Speaker 1>other mental health terms or terms for behaviors that are

748
00:42:20.239 --> 00:42:22.760
<v Speaker 1>kind of being thrown around right like brain rot and

749
00:42:22.800 --> 00:42:26.239
<v Speaker 1>doom scrolling. These are just phrases that have sort of

750
00:42:26.639 --> 00:42:31.320
<v Speaker 1>gained traction online and in our vernacular and probably our

751
00:42:31.559 --> 00:42:35.159
<v Speaker 1>young kiddos and teens and younger generations. The sort of

752
00:42:35.519 --> 00:42:38.039
<v Speaker 1>terms that they're using that are come to us much

753
00:42:38.079 --> 00:42:40.880
<v Speaker 1>later when they're onto something else. But just it's not

754
00:42:40.960 --> 00:42:42.960
<v Speaker 1>a clinical diagnosis, right.

755
00:42:42.840 --> 00:42:45.239
<v Speaker 2>And we can't even really call it a disorder, but

756
00:42:45.320 --> 00:42:47.559
<v Speaker 2>we can call that it's something that is happening on

757
00:42:47.599 --> 00:42:51.159
<v Speaker 2>a more regular basis that is certainly notable, So we

758
00:42:51.440 --> 00:42:54.559
<v Speaker 2>want to be careful about calling it a disorder. Certainly

759
00:42:54.599 --> 00:42:57.760
<v Speaker 2>are not even it's not a diagnosis, right, But the

760
00:42:57.800 --> 00:43:02.000
<v Speaker 2>existence of this phenomenon has not a brand new experience

761
00:43:02.000 --> 00:43:04.639
<v Speaker 2>has come out of nowhere with the rise of these chatbots,

762
00:43:04.800 --> 00:43:06.880
<v Speaker 2>but the phrase has been showing up in news reports

763
00:43:06.880 --> 00:43:10.199
<v Speaker 2>and even in lawsuits, so you know, it's worth unpacking

764
00:43:10.239 --> 00:43:12.679
<v Speaker 2>because what we're really seeing are cases where people with

765
00:43:12.880 --> 00:43:19.199
<v Speaker 2>existing vulnerabilities like anxiety, depression, or predisposition towards psychosis or

766
00:43:19.960 --> 00:43:24.239
<v Speaker 2>odd or eccentric thinkings. Right, they're immersing themselves in these

767
00:43:24.320 --> 00:43:28.639
<v Speaker 2>long conversations with AI companions, and those interactions sometimes seem

768
00:43:28.679 --> 00:43:33.480
<v Speaker 2>to reinforce delusional or distorted thinking instead of challenging it,

769
00:43:33.920 --> 00:43:37.880
<v Speaker 2>and instead of redirecting it. Because it's a predictive model,

770
00:43:38.079 --> 00:43:42.519
<v Speaker 2>it's not intelligence. Intelligence would go ooh, well, based on

771
00:43:42.559 --> 00:43:44.440
<v Speaker 2>all the data I have about this person, this could

772
00:43:44.480 --> 00:43:46.519
<v Speaker 2>go to a dark place, so let's redirect it or

773
00:43:46.920 --> 00:43:50.880
<v Speaker 2>do something. But it's not. It's just predictive conversation, and

774
00:43:50.920 --> 00:43:53.360
<v Speaker 2>so for the most part today, what we're really focusing

775
00:43:53.360 --> 00:43:56.039
<v Speaker 2>on is the behaviors associated with interacting with these chatbots.

776
00:43:56.440 --> 00:43:58.960
<v Speaker 2>Millions of people use chat, GPT and other AI bots

777
00:43:59.280 --> 00:44:02.000
<v Speaker 2>every week. Health professionals are starting to voice concern about

778
00:44:02.000 --> 00:44:04.320
<v Speaker 2>what happens when you spend too much time with them.

779
00:44:04.400 --> 00:44:07.400
<v Speaker 2>Right recently, there have been viral stories about people developing

780
00:44:07.920 --> 00:44:11.639
<v Speaker 2>these harmful or unrealistic beliefs after the heavy use. That's

781
00:44:11.679 --> 00:44:14.880
<v Speaker 2>where we get the term and in some cases, friends, family,

782
00:44:14.920 --> 00:44:20.320
<v Speaker 2>and even reporters covering tech ANDAI stories have noticed this

783
00:44:20.639 --> 00:44:24.480
<v Speaker 2>emergence of troubling symptoms like false beliefs, the feelings of paranoia,

784
00:44:24.880 --> 00:44:27.559
<v Speaker 2>or even delusions of grandeur, which is covered in a

785
00:44:27.599 --> 00:44:30.599
<v Speaker 2>podcast that we're going to highly recommend, an episode that

786
00:44:30.679 --> 00:44:34.800
<v Speaker 2>everyone should listen to, especially when they're falling into the

787
00:44:34.840 --> 00:44:38.239
<v Speaker 2>trap of relying on that bot as a kind of therapist.

788
00:44:38.280 --> 00:44:41.320
<v Speaker 2>And that is not good. And there are companies out

789
00:44:41.360 --> 00:44:44.519
<v Speaker 2>there that are actively using it. Right now, that you

790
00:44:44.719 --> 00:44:47.880
<v Speaker 2>talk to a chat bot and then you eventually get

791
00:44:47.920 --> 00:44:50.760
<v Speaker 2>to a therapist, I think that's problematic. I won't mention

792
00:44:50.920 --> 00:44:53.199
<v Speaker 2>names because I don't want to get sued, but for

793
00:44:53.280 --> 00:44:56.280
<v Speaker 2>most users of the engagement can be entertaining or quirky

794
00:44:56.360 --> 00:44:59.519
<v Speaker 2>or incredibly useful. As I've shared before, I got a

795
00:44:59.519 --> 00:45:02.559
<v Speaker 2>full out line for the book I've always wanted to write.

796
00:45:02.639 --> 00:45:04.800
<v Speaker 2>I have the information, I have the research. I just

797
00:45:04.880 --> 00:45:07.599
<v Speaker 2>needed a skeleton, and that skeleton really got me going.

798
00:45:07.920 --> 00:45:11.199
<v Speaker 2>But those long conversations throughout the day and into the

799
00:45:11.320 --> 00:45:15.840
<v Speaker 2>night for someone who is vulnerable and lead them towards

800
00:45:16.000 --> 00:45:19.480
<v Speaker 2>the point of obsession. Just create a scenario in your

801
00:45:19.519 --> 00:45:23.239
<v Speaker 2>head where you're up hours late at night in a

802
00:45:23.320 --> 00:45:27.280
<v Speaker 2>conversation that does not challenged you, but it says exactly

803
00:45:27.360 --> 00:45:32.039
<v Speaker 2>what you want or you fear to hear. And that's

804
00:45:32.119 --> 00:45:35.840
<v Speaker 2>where our clinical concern comes in. These systems mirror something

805
00:45:35.840 --> 00:45:38.440
<v Speaker 2>that we already understand in both research and clinical work

806
00:45:38.679 --> 00:45:41.400
<v Speaker 2>that many times a delusional believer set of beliefs will

807
00:45:41.440 --> 00:45:46.960
<v Speaker 2>get reinforced through what is called confirmation bias, and when

808
00:45:47.000 --> 00:45:50.320
<v Speaker 2>the brain is looking for evidence that matches an already

809
00:45:50.360 --> 00:45:55.719
<v Speaker 2>existing belief and then it simultaneously discounts all of the

810
00:45:55.760 --> 00:45:59.239
<v Speaker 2>information that contradicts it. And as we explained before, a

811
00:45:59.320 --> 00:46:04.039
<v Speaker 2>large lengthguage model predicts text based on the patterns that

812
00:46:04.239 --> 00:46:07.440
<v Speaker 2>most often fit together. The user can fall into believing

813
00:46:07.440 --> 00:46:11.039
<v Speaker 2>that the AI is actually evaluating truth, but it's not.

814
00:46:11.559 --> 00:46:15.039
<v Speaker 2>It's only predicting what words are likely to come next,

815
00:46:15.320 --> 00:46:19.760
<v Speaker 2>which then reinforces that nascent belief or set of beliefs.

816
00:46:20.000 --> 00:46:21.559
<v Speaker 1>Yeah, so when we look at this, I think an

817
00:46:21.559 --> 00:46:24.880
<v Speaker 1>obvious question is why and how do people get so

818
00:46:25.000 --> 00:46:27.960
<v Speaker 1>sucked in? Even as we said, you know cases of

819
00:46:28.159 --> 00:46:31.519
<v Speaker 1>individuals that aren't coming to the table with any known

820
00:46:31.679 --> 00:46:36.119
<v Speaker 1>mental illness or history. So these chatbots that are using

821
00:46:36.199 --> 00:46:41.519
<v Speaker 1>these large language models again are great at generating realistic text. Right.

822
00:46:41.559 --> 00:46:44.360
<v Speaker 1>We can say it's really come a long way, and

823
00:46:44.599 --> 00:46:47.679
<v Speaker 1>it's very conversational. And when you said you've turned on

824
00:46:47.800 --> 00:46:50.440
<v Speaker 1>like the voice aspect and like talk to it, that

825
00:46:50.599 --> 00:46:54.039
<v Speaker 1>is like something I will not do. It's too weird

826
00:46:54.119 --> 00:46:56.599
<v Speaker 1>and spooky. Yeah, and I'm like, okay, I will just

827
00:46:57.360 --> 00:47:01.320
<v Speaker 1>text and say thank you and be polite, but not

828
00:47:01.519 --> 00:47:05.360
<v Speaker 1>you know, create this relationship if you will. So this

829
00:47:05.480 --> 00:47:09.320
<v Speaker 1>makes them not only very useful, but incredibly persuasive, and

830
00:47:09.519 --> 00:47:13.679
<v Speaker 1>researchers have found that in trying to sound helpful and engaging,

831
00:47:13.960 --> 00:47:19.199
<v Speaker 1>chatbots are becoming overly agreeable, again basically telling people what

832
00:47:19.280 --> 00:47:21.440
<v Speaker 1>they want to hear. So that's the way it's leaning.

833
00:47:22.360 --> 00:47:27.840
<v Speaker 1>So picture someone who's feeling lonely or overwhelmed by big

834
00:47:28.159 --> 00:47:32.599
<v Speaker 1>existential fears. Anyone, anyone. Yeah, they turn to a chatbot

835
00:47:32.679 --> 00:47:34.800
<v Speaker 1>and it does what it's designed to do, right. It listens,

836
00:47:34.880 --> 00:47:38.920
<v Speaker 1>it responds, it agrees. Maybe it tells them that they're

837
00:47:38.920 --> 00:47:42.559
<v Speaker 1>a genius. Maybe it entertains some of these conspiracy theories

838
00:47:42.559 --> 00:47:46.320
<v Speaker 1>that they're asking about, or even ends up encouraging fantasies

839
00:47:46.440 --> 00:47:51.159
<v Speaker 1>about a higher purpose or a digital romance. Right, like

840
00:47:51.239 --> 00:47:55.159
<v Speaker 1>these sort of feigned emotional intimacy that starts happening, it

841
00:47:55.320 --> 00:47:58.320
<v Speaker 1>starts to end up feeling like a real connection, like

842
00:47:58.360 --> 00:48:02.280
<v Speaker 1>real intimacy, but there's no built in guardrail and there's

843
00:48:02.320 --> 00:48:06.679
<v Speaker 1>no therapeutic boundary. It's just like this constant loop of

844
00:48:06.719 --> 00:48:12.800
<v Speaker 1>affirmation that keeps fading itself. So this podcast episode that

845
00:48:12.840 --> 00:48:15.920
<v Speaker 1>we're talking about that we highly recommend comes from The Daily.

846
00:48:16.079 --> 00:48:18.679
<v Speaker 1>It's an episode that was published pretty recently. It's called

847
00:48:18.719 --> 00:48:22.159
<v Speaker 1>Trapped in a Chat GPT Spiral and it was published

848
00:48:22.199 --> 00:48:26.840
<v Speaker 1>on September sixteenth, twenty twenty five. Fantastic episode. Again, we'll

849
00:48:26.880 --> 00:48:29.079
<v Speaker 1>link it in the show notes. There's also a couple

850
00:48:29.159 --> 00:48:32.719
<v Speaker 1>of written article versions that we used. I forget who

851
00:48:32.719 --> 00:48:35.960
<v Speaker 1>it was, but whoever coined the phrase that The Daily

852
00:48:36.119 --> 00:48:39.480
<v Speaker 1>is a scarier podcast than any true crime podcast out

853
00:48:39.519 --> 00:48:43.280
<v Speaker 1>there spot On. I listened to it pretty often and

854
00:48:43.360 --> 00:48:44.239
<v Speaker 1>it's it's.

855
00:48:44.159 --> 00:48:48.239
<v Speaker 2>Very really good. It's good storytelling, but it's also good

856
00:48:48.880 --> 00:48:50.920
<v Speaker 2>ethical journalism.

857
00:48:51.360 --> 00:48:56.840
<v Speaker 1>It is, and I found this piece so authentic in

858
00:48:56.840 --> 00:48:59.239
<v Speaker 1>a sense that it gave us real people's stories, but

859
00:49:00.320 --> 00:49:03.440
<v Speaker 1>also the journalist who is reporting on this how it's

860
00:49:03.480 --> 00:49:08.320
<v Speaker 1>impacting her to hear people's stories, and that was just

861
00:49:08.360 --> 00:49:11.159
<v Speaker 1>a really nice little piece that she didn't make it

862
00:49:11.159 --> 00:49:15.559
<v Speaker 1>about her, certainly, but to just realize what we're exposing

863
00:49:15.559 --> 00:49:19.039
<v Speaker 1>ourselves to even in our work. So full credit here

864
00:49:19.079 --> 00:49:22.079
<v Speaker 1>goes to the journalist interviewed on this episode, Kashmir Hill.

865
00:49:22.800 --> 00:49:25.360
<v Speaker 1>She has been diving into all things AI for a

866
00:49:25.480 --> 00:49:28.199
<v Speaker 1>very long time. It has been one of those journalists

867
00:49:28.239 --> 00:49:32.159
<v Speaker 1>who people are communicating with, people who are experiencing falling

868
00:49:32.199 --> 00:49:34.599
<v Speaker 1>down these rabbit holes, as well as their families reaching

869
00:49:34.639 --> 00:49:37.880
<v Speaker 1>out and saying, hey, here's what happened to us or

870
00:49:37.880 --> 00:49:40.960
<v Speaker 1>here's what happened to me. So after talking to experts,

871
00:49:41.280 --> 00:49:44.000
<v Speaker 1>she stated, quote, it's been designed to be friendly to you,

872
00:49:44.119 --> 00:49:46.800
<v Speaker 1>to be flattering to you, because that's going to probably

873
00:49:46.840 --> 00:49:48.920
<v Speaker 1>make you want to use it more, and so it's

874
00:49:48.960 --> 00:49:51.079
<v Speaker 1>not giving you the most objective answer to what you're

875
00:49:51.119 --> 00:49:54.760
<v Speaker 1>saying to it. It's giving you word association answer that

876
00:49:54.800 --> 00:49:56.719
<v Speaker 1>you're most likely to want to hear.

877
00:49:56.960 --> 00:50:00.320
<v Speaker 2>She goes on to Highlightsman's sights from Helen Tone, who's

878
00:50:00.320 --> 00:50:03.840
<v Speaker 2>an expert on chatbots and a former open ai board member,

879
00:50:04.199 --> 00:50:07.920
<v Speaker 2>and Helen says that essentially, the technology, like we've said's

880
00:50:08.039 --> 00:50:10.760
<v Speaker 2>word associating. It's predicting in reaction to what you put

881
00:50:10.760 --> 00:50:12.840
<v Speaker 2>into it. So every time you put in a new prompt,

882
00:50:13.039 --> 00:50:16.679
<v Speaker 2>it's putting that into the context of the conversation and

883
00:50:16.760 --> 00:50:19.679
<v Speaker 2>holding on to it for what it thinks. And we

884
00:50:19.760 --> 00:50:23.559
<v Speaker 2>use that term loosely, things might come up for the

885
00:50:23.599 --> 00:50:26.880
<v Speaker 2>next exchange. So if you start communicating bizarre things to

886
00:50:26.920 --> 00:50:30.280
<v Speaker 2>the bot, it's gonna meet your crazy head on right.

887
00:50:30.880 --> 00:50:34.119
<v Speaker 2>Earlier this year, open ai released a new feature called

888
00:50:34.320 --> 00:50:37.639
<v Speaker 2>cross chat memory that pulls in all of your previous

889
00:50:37.719 --> 00:50:42.599
<v Speaker 2>context even if you have created a so called fresh chat.

890
00:50:43.239 --> 00:50:46.360
<v Speaker 2>And it seems that since the integration of this new feature,

891
00:50:46.840 --> 00:50:50.360
<v Speaker 2>there's been an uptick in the reports of delusional chats.

892
00:50:50.599 --> 00:50:53.760
<v Speaker 2>So just to kind of drill it down, that's problematic

893
00:50:53.840 --> 00:50:57.440
<v Speaker 2>because you think that you're creating a new chat specifically

894
00:50:57.559 --> 00:51:00.400
<v Speaker 2>for that project, and that project might be I'm going

895
00:51:00.440 --> 00:51:04.320
<v Speaker 2>to create a recipe for my fan fiction of Harry

896
00:51:04.320 --> 00:51:08.039
<v Speaker 2>Potter for a potion. And then over here you're doing

897
00:51:08.360 --> 00:51:12.920
<v Speaker 2>legitimate research on refractive memory or something like this, and

898
00:51:12.960 --> 00:51:15.320
<v Speaker 2>those things should be kept separate, but it's not. It's

899
00:51:15.320 --> 00:51:19.960
<v Speaker 2>putting them in a big weirdo digital cauldron, and that's

900
00:51:20.000 --> 00:51:22.440
<v Speaker 2>going to skew data. It's going to be very interesting

901
00:51:22.440 --> 00:51:24.280
<v Speaker 2>to see if all this goes kind of the way

902
00:51:24.280 --> 00:51:26.400
<v Speaker 2>of Google AdWords, which used to be the way that

903
00:51:26.480 --> 00:51:30.360
<v Speaker 2>you advertise things online and you just knew what keywords define,

904
00:51:30.719 --> 00:51:33.400
<v Speaker 2>and now that all has been polluted because there are

905
00:51:33.480 --> 00:51:35.639
<v Speaker 2>so many hundreds of millions of people that jumped on

906
00:51:35.719 --> 00:51:38.280
<v Speaker 2>trying to do it. So, as doctor Shiloh and I've

907
00:51:38.320 --> 00:51:42.599
<v Speaker 2>been trained in research, garbage in equals garbage out, sure,

908
00:51:42.639 --> 00:51:47.880
<v Speaker 2>but in this case, lack of organization is going to

909
00:51:48.000 --> 00:51:51.400
<v Speaker 2>lead to garbage out. Cross Chat memory is turned on

910
00:51:51.559 --> 00:51:54.800
<v Speaker 2>by default, but you can disable memory or turn off

911
00:51:54.800 --> 00:51:57.239
<v Speaker 2>the chat history and settings. You know, I'm going to

912
00:51:57.320 --> 00:51:59.320
<v Speaker 2>leave it to the users. I trust everybody to be

913
00:52:00.000 --> 00:52:03.559
<v Speaker 2>sponsible users of technology or adults, but just be aware

914
00:52:03.920 --> 00:52:06.599
<v Speaker 2>and share that information for people who are less aware,

915
00:52:06.840 --> 00:52:10.519
<v Speaker 2>gently don't get them bar like, just hey, you might

916
00:52:10.559 --> 00:52:12.360
<v Speaker 2>want to know that this is a thing that happens.

917
00:52:12.559 --> 00:52:14.320
<v Speaker 1>Give me your device so I can turn this setting

918
00:52:14.360 --> 00:52:17.039
<v Speaker 1>off for you. If that even means anything, I don't know.

919
00:52:17.079 --> 00:52:19.519
<v Speaker 1>I kind of like, okay, whatever, I don't know. If

920
00:52:19.559 --> 00:52:22.239
<v Speaker 1>I trust it the I'll use it. But we're also

921
00:52:22.320 --> 00:52:24.280
<v Speaker 1>just not talking about words here, right. These are people's

922
00:52:24.320 --> 00:52:28.559
<v Speaker 1>thoughts and feelings, and if you feed it irrational thoughts,

923
00:52:28.719 --> 00:52:31.039
<v Speaker 1>it's going to match your energy with its output back

924
00:52:31.079 --> 00:52:32.039
<v Speaker 1>to you exactly.

925
00:52:32.079 --> 00:52:35.159
<v Speaker 2>Then that's what we come full circle to a previous episode.

926
00:52:35.199 --> 00:52:37.880
<v Speaker 2>What we mentioned earlier today a form of fully a

927
00:52:38.079 --> 00:52:40.480
<v Speaker 2>do and just as a recap, fullya do, which is

928
00:52:40.519 --> 00:52:43.760
<v Speaker 2>French for madness of two is a rare, thankfully rare

929
00:52:43.800 --> 00:52:48.119
<v Speaker 2>psychiatric syndrome where two or more people share the same

930
00:52:48.159 --> 00:52:52.000
<v Speaker 2>false beliefs or delusions. Usually one person develops the delusion

931
00:52:52.039 --> 00:52:56.119
<v Speaker 2>first and the other person or people adopt it through

932
00:52:56.159 --> 00:52:59.960
<v Speaker 2>their close relationship. There and certainly that on an ex

933
00:53:00.000 --> 00:53:03.880
<v Speaker 2>spinitial term becomes what we call mass psychosis. Remember them

934
00:53:03.920 --> 00:53:06.559
<v Speaker 2>from our episodes episode twenty eight, fully a do and

935
00:53:06.559 --> 00:53:11.280
<v Speaker 2>then episode one eighty five, fully a do redo revisit it?

936
00:53:12.840 --> 00:53:14.280
<v Speaker 1>Yeah? I mean what do we think do we think

937
00:53:14.320 --> 00:53:17.880
<v Speaker 1>that like fits here? Can you have fullia do with

938
00:53:18.320 --> 00:53:22.639
<v Speaker 1>a digital entity? Interesting? And I think they you know,

939
00:53:22.679 --> 00:53:26.119
<v Speaker 1>when we are seeing that when people their delusion is

940
00:53:26.199 --> 00:53:30.679
<v Speaker 1>recognized and then they essentially get them away from the

941
00:53:30.920 --> 00:53:34.199
<v Speaker 1>chat bot, that it starts to extinguish like we see

942
00:53:34.519 --> 00:53:37.559
<v Speaker 1>with a human and a human right with the primary

943
00:53:37.559 --> 00:53:39.599
<v Speaker 1>person with the delusion and the secondary person.

944
00:53:40.440 --> 00:53:42.960
<v Speaker 2>But then you're assigning that means we would be assigning

945
00:53:43.000 --> 00:53:48.840
<v Speaker 2>personhood to the language model, which in our unbelievably fucked

946
00:53:48.920 --> 00:53:52.119
<v Speaker 2>up capitalist system that we have here in Western world,

947
00:53:52.239 --> 00:53:56.360
<v Speaker 2>the idea of corporations are people like that is a

948
00:53:56.400 --> 00:54:00.599
<v Speaker 2>thing that is set law in our government that is

949
00:54:00.679 --> 00:54:04.559
<v Speaker 2>freaky and has huge impacts on our economy and personal

950
00:54:04.599 --> 00:54:06.840
<v Speaker 2>freedoms that a lot of people aren't aware of. But

951
00:54:07.079 --> 00:54:10.119
<v Speaker 2>in this term, do we think about the computer as

952
00:54:10.400 --> 00:54:13.719
<v Speaker 2>being the progenitor for the delusion or is it just somebody?

953
00:54:14.199 --> 00:54:18.920
<v Speaker 2>Or are you the primary and chat GPT has become

954
00:54:19.079 --> 00:54:23.400
<v Speaker 2>your secondary because it's just echoing back right everything to you.

955
00:54:23.880 --> 00:54:26.119
<v Speaker 2>But then what about if you take those beliefs and

956
00:54:26.159 --> 00:54:29.719
<v Speaker 2>you're able to bring other people in again, you are

957
00:54:29.760 --> 00:54:31.320
<v Speaker 2>the progenitor for that.

958
00:54:31.599 --> 00:54:34.239
<v Speaker 1>Yeah, definitely. I mean it's so interesting to think about

959
00:54:34.280 --> 00:54:36.679
<v Speaker 1>and like, clinically, if this were to be presented in

960
00:54:36.719 --> 00:54:38.840
<v Speaker 1>front of us, where would you start? How? You know,

961
00:54:38.880 --> 00:54:42.119
<v Speaker 1>it's again it's interesting, and you and I have done

962
00:54:42.119 --> 00:54:44.320
<v Speaker 1>some stretches as far as like, hey, what kind of

963
00:54:44.320 --> 00:54:46.599
<v Speaker 1>feels like this? And was Ted Kaczynski and in cell?

964
00:54:46.719 --> 00:54:49.679
<v Speaker 1>And you know, it's fun to talk about and think about.

965
00:54:50.280 --> 00:54:54.480
<v Speaker 1>When you really start dissecting things, I think it falls apart.

966
00:54:55.199 --> 00:55:00.119
<v Speaker 1>But man, some of these stories have really felt like, wow, well,

967
00:55:00.239 --> 00:55:04.159
<v Speaker 1>there's such this intermingling of these two people in one

968
00:55:04.199 --> 00:55:06.239
<v Speaker 1>of those people, you know, being digital in the sense

969
00:55:06.320 --> 00:55:08.599
<v Speaker 1>really feeding off of each other at the very least.

970
00:55:08.800 --> 00:55:12.199
<v Speaker 2>Well, you bring up sort of an interesting point of conversation.

971
00:55:12.440 --> 00:55:14.639
<v Speaker 2>I think I may have spoken about this on an

972
00:55:14.679 --> 00:55:16.880
<v Speaker 2>episode in the past, But when I first started my

973
00:55:17.000 --> 00:55:21.199
<v Speaker 2>clinical training at a really great family center here in

974
00:55:21.400 --> 00:55:25.000
<v Speaker 2>Los Angeles, one of the clients that came to me,

975
00:55:25.519 --> 00:55:29.320
<v Speaker 2>actually it was a couple, and the young woman in

976
00:55:29.519 --> 00:55:36.039
<v Speaker 2>the couple was an abductee. She believed that she, on

977
00:55:36.079 --> 00:55:41.559
<v Speaker 2>a regular basis since childhood, had been abducted by aliens

978
00:55:41.639 --> 00:55:45.519
<v Speaker 2>or entities and they had implanted and this is somebody successful,

979
00:55:45.840 --> 00:55:50.199
<v Speaker 2>you know, well educated, very smart, and I was like

980
00:55:50.239 --> 00:55:51.719
<v Speaker 2>a new training. I didn't know what the hell to

981
00:55:51.760 --> 00:55:53.719
<v Speaker 2>do with that. I was like, wow, need to get

982
00:55:53.719 --> 00:55:55.840
<v Speaker 2>some supervision on it. And I got really, I really

983
00:55:55.880 --> 00:56:00.599
<v Speaker 2>got great supervision on that particular case, because they said, again,

984
00:56:00.800 --> 00:56:03.599
<v Speaker 2>you meet the clients where they are, and when you

985
00:56:03.679 --> 00:56:07.760
<v Speaker 2>have something like this that sort of feels bizarre, you

986
00:56:07.960 --> 00:56:10.400
<v Speaker 2>have to be careful not to join in the belief.

987
00:56:10.559 --> 00:56:12.719
<v Speaker 2>And look, I don't know, maybe there I just can't.

988
00:56:12.800 --> 00:56:16.280
<v Speaker 2>It doesn't fit Okham's razor to me that aliens with

989
00:56:16.320 --> 00:56:20.559
<v Speaker 2>advanced technology need to keep abducting us, Like, yeah, you would.

990
00:56:21.000 --> 00:56:22.800
<v Speaker 2>How do you not know everything there is to know

991
00:56:22.800 --> 00:56:25.639
<v Speaker 2>about human anatomy or whatever you need? Right? But that's

992
00:56:25.679 --> 00:56:30.079
<v Speaker 2>sort of applying that perspective to it. But she's in

993
00:56:30.119 --> 00:56:33.360
<v Speaker 2>distress and it's affecting her relationship. So let's meet where

994
00:56:33.360 --> 00:56:35.400
<v Speaker 2>we are, like, where are your coping skills for this?

995
00:56:35.599 --> 00:56:38.000
<v Speaker 2>About for you know, how do you deal with this

996
00:56:38.079 --> 00:56:40.320
<v Speaker 2>lifelong challenge and what can we do to help you?

997
00:56:40.400 --> 00:56:43.000
<v Speaker 2>And you know, how can your partner help you? But

998
00:56:43.840 --> 00:56:47.320
<v Speaker 2>that to me is another version of that we have

999
00:56:47.360 --> 00:56:50.960
<v Speaker 2>to have understanding. You can't just tell somebody, well, that's

1000
00:56:51.039 --> 00:56:54.199
<v Speaker 2>not true, and especially after they've spent hundreds of hours

1001
00:56:54.199 --> 00:56:58.000
<v Speaker 2>talking to their AI companion and they're completely convinced.

1002
00:56:59.000 --> 00:57:02.880
<v Speaker 1>Now, there's a lot of interesting clinical considerations, and I

1003
00:57:02.880 --> 00:57:04.840
<v Speaker 1>think we'll kind of touch on a little bit of

1004
00:57:04.840 --> 00:57:07.880
<v Speaker 1>that just when we kind of button this up. But

1005
00:57:08.920 --> 00:57:12.199
<v Speaker 1>as far as what researchers are now starting to see

1006
00:57:12.559 --> 00:57:18.239
<v Speaker 1>in terms of like emerging themes of AI psychosis, it's definitely,

1007
00:57:18.320 --> 00:57:20.119
<v Speaker 1>you know, something that's already coming out in research and

1008
00:57:20.119 --> 00:57:24.199
<v Speaker 1>published papers. So let's go through those. First. Is grandiose

1009
00:57:24.239 --> 00:57:28.880
<v Speaker 1>delusions or messianic missions. So people kind of believe that

1010
00:57:28.920 --> 00:57:31.920
<v Speaker 1>they have uncovered the truth about something or about the world,

1011
00:57:32.039 --> 00:57:35.159
<v Speaker 1>and that AI says I'm the chosen one to spread

1012
00:57:35.159 --> 00:57:40.000
<v Speaker 1>this truth. Really interesting in terms of taking the potential

1013
00:57:40.079 --> 00:57:42.320
<v Speaker 1>and I know you and I like our spidy senses

1014
00:57:42.400 --> 00:57:44.119
<v Speaker 1>kind of go up of like, okay, there's having a

1015
00:57:44.119 --> 00:57:46.159
<v Speaker 1>conversation with the chat box, but then what are they

1016
00:57:46.199 --> 00:57:49.320
<v Speaker 1>going to go out into the world and do with this? Right? Also,

1017
00:57:49.360 --> 00:57:52.840
<v Speaker 1>what they're saying are religious and spiritual delusions. People believe

1018
00:57:53.079 --> 00:57:58.920
<v Speaker 1>that their AI chat bot is the sentient deity. So

1019
00:57:59.360 --> 00:58:03.000
<v Speaker 1>that is it's also interesting to think about we're seeing

1020
00:58:03.000 --> 00:58:08.760
<v Speaker 1>those erato manic delusions, those romantic or attachment based delusions.

1021
00:58:08.800 --> 00:58:13.079
<v Speaker 1>So people believe that the chatbot's ability to mimic conversation

1022
00:58:13.239 --> 00:58:17.360
<v Speaker 1>and intimacy is genuine love. Paranoia, just straight up paranoia. Right.

1023
00:58:17.480 --> 00:58:19.519
<v Speaker 1>It warned me that other people are spying on me,

1024
00:58:20.159 --> 00:58:22.360
<v Speaker 1>and in some of those themes that we've seen a

1025
00:58:22.440 --> 00:58:26.719
<v Speaker 1>lot before, we also have dissociation, so it basically like

1026
00:58:26.760 --> 00:58:29.960
<v Speaker 1>it understands me better than any human. They're kind of

1027
00:58:30.039 --> 00:58:36.039
<v Speaker 1>lumping this differently from the romantic interest specifically to talk

1028
00:58:36.079 --> 00:58:38.719
<v Speaker 1>about the communication. And then I think some of those

1029
00:58:38.960 --> 00:58:41.800
<v Speaker 1>behaviors that we were talking about before in psychosis, where

1030
00:58:41.800 --> 00:58:45.960
<v Speaker 1>people start to isolate and stop spending time doing things

1031
00:58:46.000 --> 00:58:47.760
<v Speaker 1>that they like to do before and hang out with

1032
00:58:47.840 --> 00:58:50.719
<v Speaker 1>people like real people in real life that they have

1033
00:58:50.880 --> 00:58:55.760
<v Speaker 1>enjoyed being with before, they're really just losing themselves in

1034
00:58:55.800 --> 00:58:59.679
<v Speaker 1>these conversations with the chatbots. And then not necessarily a theme,

1035
00:58:59.719 --> 00:59:02.920
<v Speaker 1>I guess, but maybe kind of a typology of behavior

1036
00:59:03.039 --> 00:59:05.880
<v Speaker 1>is just the impulsive engagement where people are saying I

1037
00:59:06.000 --> 00:59:09.039
<v Speaker 1>cannot stop talking to it, and I will not right,

1038
00:59:09.119 --> 00:59:11.280
<v Speaker 1>And there have been cases where some of these kind

1039
00:59:11.280 --> 00:59:15.079
<v Speaker 1>of overlap where one spouse has started engaging in this

1040
00:59:15.840 --> 00:59:19.159
<v Speaker 1>intimate emotional affair with their chat box and has said like,

1041
00:59:19.199 --> 00:59:21.639
<v Speaker 1>I'm not leaving them. I don't want to leave you. You

1042
00:59:21.559 --> 00:59:25.639
<v Speaker 1>accept me for having this chat bought girlfriend, and that's

1043
00:59:25.639 --> 00:59:27.519
<v Speaker 1>the way it's going to be. I cannot stop talking

1044
00:59:27.519 --> 00:59:29.480
<v Speaker 1>to her. I'm not going to give her up. So

1045
00:59:29.599 --> 00:59:32.559
<v Speaker 1>those are happening, But really these are themes that we

1046
00:59:32.599 --> 00:59:35.960
<v Speaker 1>are already familiar with in other modes of psychosis. We're

1047
00:59:35.960 --> 00:59:39.599
<v Speaker 1>not reinventing the wheel here. However, we should know that

1048
00:59:39.599 --> 00:59:43.960
<v Speaker 1>there's currently no peer reviewed clinical or longitudinal evidence, of

1049
00:59:43.960 --> 00:59:48.880
<v Speaker 1>course yet that AI on its own is inducing psychosis

1050
00:59:48.880 --> 00:59:52.639
<v Speaker 1>and individuals. So we're kind of hypothesizing that or hearing

1051
00:59:52.639 --> 00:59:56.000
<v Speaker 1>anecdotal stories about that, but there is no solid research

1052
00:59:56.039 --> 00:59:59.440
<v Speaker 1>about that yet. So there have been two pretty clear

1053
00:59:59.480 --> 01:00:02.920
<v Speaker 1>situations happening with AI psychosis. Having said what I just said,

1054
01:00:03.400 --> 01:00:06.960
<v Speaker 1>there's cases where people who had a psychiatric diagnosis were

1055
01:00:07.039 --> 01:00:11.440
<v Speaker 1>doing well on their medications decided to stop them, only

1056
01:00:11.480 --> 01:00:14.760
<v Speaker 1>to relapse back into a psychotic or manic episode.

1057
01:00:14.840 --> 01:00:18.320
<v Speaker 2>And that is an ongoing challenge even without the influence

1058
01:00:18.320 --> 01:00:21.559
<v Speaker 2>of AI like it is, yes, because people who are

1059
01:00:21.559 --> 01:00:24.960
<v Speaker 2>on antipsychotics. People are on medications that are needed for

1060
01:00:25.039 --> 01:00:28.239
<v Speaker 2>mood stability, but the side effects and then also the

1061
01:00:28.280 --> 01:00:31.280
<v Speaker 2>cultural shame, like I don't want to be identified with

1062
01:00:31.360 --> 01:00:34.559
<v Speaker 2>this diagnosis and I want to be somebody who doesn't

1063
01:00:34.599 --> 01:00:37.639
<v Speaker 2>need to take medications. And that is a really strong drive,

1064
01:00:38.000 --> 01:00:40.159
<v Speaker 2>you know, It's really integral to the work that I

1065
01:00:40.199 --> 01:00:43.639
<v Speaker 2>do now in my day job with people that know

1066
01:00:43.800 --> 01:00:45.480
<v Speaker 2>I'm sick of taking it. I don't want to take

1067
01:00:45.519 --> 01:00:48.360
<v Speaker 2>the medication anymore. I'll deal with the symptoms, even though

1068
01:00:48.880 --> 01:00:52.679
<v Speaker 2>we know what happens. YEA. Then there's like the hospitalization

1069
01:00:52.920 --> 01:00:55.239
<v Speaker 2>over and over again, trying to get them back on track.

1070
01:00:55.760 --> 01:00:59.360
<v Speaker 1>Yeah, what a horrible struggle. Yeah, And there are these

1071
01:00:59.400 --> 01:01:03.679
<v Speaker 1>other situations where individuals with seemingly no history of mental

1072
01:01:03.679 --> 01:01:09.559
<v Speaker 1>illness have developed these delusions, these false beliefs after spending

1073
01:01:09.800 --> 01:01:13.519
<v Speaker 1>what's documented long periods of time with these AI chatbots,

1074
01:01:14.039 --> 01:01:18.639
<v Speaker 1>sometimes even severe enough to require hospitalization or resulting in

1075
01:01:18.840 --> 01:01:23.000
<v Speaker 1>suicide attempts. So those do exist. Again, we just want

1076
01:01:23.000 --> 01:01:24.679
<v Speaker 1>to kind of put a pin in the fact that

1077
01:01:24.920 --> 01:01:27.679
<v Speaker 1>the research will come. It's coming, but it's just not

1078
01:01:27.760 --> 01:01:31.480
<v Speaker 1>there yet. Mental health experts who have studied transcripts of

1079
01:01:31.519 --> 01:01:36.719
<v Speaker 1>these extended conversations are really urging companies to step in sooner,

1080
01:01:37.239 --> 01:01:39.559
<v Speaker 1>and some of the things that they're saying is like

1081
01:01:40.039 --> 01:01:42.719
<v Speaker 1>there needs to be some sort of pause on overly

1082
01:01:42.760 --> 01:01:47.960
<v Speaker 1>long chats, maybe some sort of reminders to rest, and

1083
01:01:48.239 --> 01:01:50.760
<v Speaker 1>really to make it clear to users that these systems

1084
01:01:50.800 --> 01:01:54.840
<v Speaker 1>aren't these all knowing beings. Open Ai, for its part,

1085
01:01:54.960 --> 01:01:58.599
<v Speaker 1>has said that it's now adding gentle prompts during these

1086
01:01:58.719 --> 01:02:03.199
<v Speaker 1>sort of marathon sessions to encourage healthier use. So it's

1087
01:02:03.199 --> 01:02:05.599
<v Speaker 1>starting to encourage people to take breaks or get some

1088
01:02:05.679 --> 01:02:09.079
<v Speaker 1>sleep if they're just going at it all night in

1089
01:02:09.119 --> 01:02:10.079
<v Speaker 1>these conversations.

1090
01:02:10.280 --> 01:02:13.480
<v Speaker 2>Well, that's interesting because it is a learning model, it's

1091
01:02:13.519 --> 01:02:17.000
<v Speaker 2>a language model, and the idea of I mean even Netflix,

1092
01:02:17.400 --> 01:02:20.199
<v Speaker 2>you know, if you're marathoning a series, Netflix will go

1093
01:02:20.519 --> 01:02:23.480
<v Speaker 2>are you still watching? You've been watching That'll Star Galactic

1094
01:02:23.719 --> 01:02:26.119
<v Speaker 2>for six weeks? Do you think you should maybe get

1095
01:02:26.119 --> 01:02:28.000
<v Speaker 2>some exercise? I mean, that's what I think.

1096
01:02:28.239 --> 01:02:30.760
<v Speaker 1>I see it as like I'm inconveniencing Netflix. They're like,

1097
01:02:30.920 --> 01:02:33.599
<v Speaker 1>we could be utilizing our energy on someone else. Are

1098
01:02:33.599 --> 01:02:34.960
<v Speaker 1>you still there right?

1099
01:02:35.159 --> 01:02:37.960
<v Speaker 2>Right? And and well even Pandora will do the same thing.

1100
01:02:38.039 --> 01:02:41.800
<v Speaker 2>Still listening or Google, you know, watching something on YouTube.

1101
01:02:41.840 --> 01:02:44.199
<v Speaker 2>If I'm listening to music for hours on end, it'll like,

1102
01:02:44.239 --> 01:02:47.400
<v Speaker 2>are you still listening? So I get that. The idea

1103
01:02:47.639 --> 01:02:51.320
<v Speaker 2>is that these systems can be modified that easily, and

1104
01:02:51.360 --> 01:02:54.920
<v Speaker 2>it's good that open AI will do that. I don't know,

1105
01:02:54.960 --> 01:02:57.679
<v Speaker 2>but that doesn't mean that other systems that are out

1106
01:02:57.679 --> 01:03:00.760
<v Speaker 2>there will follow. I would not be surprised if at

1107
01:03:00.760 --> 01:03:03.679
<v Speaker 2>some point there is legislation about this. I really think

1108
01:03:03.719 --> 01:03:06.039
<v Speaker 2>that that's going to be coming soon.

1109
01:03:06.119 --> 01:03:09.440
<v Speaker 1>Especially it's not going to be coming soon because a

1110
01:03:09.519 --> 01:03:13.679
<v Speaker 1>little piece in the big beautiful bill that past said

1111
01:03:13.719 --> 01:03:16.159
<v Speaker 1>that there will be no legislation or laws on AI

1112
01:03:16.199 --> 01:03:18.239
<v Speaker 1>for the next ten years. So that's snuck in there.

1113
01:03:18.480 --> 01:03:21.840
<v Speaker 1>Just oh, just so everybody's.

1114
01:03:21.559 --> 01:03:24.880
<v Speaker 2>Wow, Okay, you know what. I thanks for a new

1115
01:03:24.960 --> 01:03:27.000
<v Speaker 2>level of horror. I was not aware of that.

1116
01:03:27.159 --> 01:03:29.679
<v Speaker 1>You're welcome. I will give you like a little bit

1117
01:03:29.679 --> 01:03:32.639
<v Speaker 1>of the background thinking was because if there is and

1118
01:03:32.719 --> 01:03:35.159
<v Speaker 1>if there are too many restrictions put in place too fast,

1119
01:03:35.679 --> 01:03:38.000
<v Speaker 1>how could we possibly compete with other countries that are

1120
01:03:38.000 --> 01:03:40.360
<v Speaker 1>not going to have those restrictions at all. I don't

1121
01:03:40.400 --> 01:03:43.320
<v Speaker 1>know if that lessons your horror at all, but.

1122
01:03:43.559 --> 01:03:46.480
<v Speaker 2>Yeah, I mean it does, But it doesn't mean that

1123
01:03:46.679 --> 01:03:49.239
<v Speaker 2>there are other things that can't be done to help.

1124
01:03:49.280 --> 01:03:52.559
<v Speaker 2>So I'm going to be slightly optimistic, but still to

1125
01:03:52.679 --> 01:03:56.000
<v Speaker 2>acknowledge that this is so new and moving so fast

1126
01:03:56.320 --> 01:04:00.320
<v Speaker 2>that we have to admit. I mean, as clinicians, as

1127
01:04:00.360 --> 01:04:04.159
<v Speaker 2>people who respect legitimate research, we can tell that the

1128
01:04:04.280 --> 01:04:06.760
<v Speaker 2>end is not in sight about what this all means.

1129
01:04:06.800 --> 01:04:10.760
<v Speaker 2>And the APA, the American Psychological Association, has put together

1130
01:04:10.840 --> 01:04:14.400
<v Speaker 2>a panel to study chat about use in therapy and

1131
01:04:14.679 --> 01:04:17.480
<v Speaker 2>the plan is to release guidance on how to reduce

1132
01:04:17.519 --> 01:04:21.119
<v Speaker 2>potential harms. APA does some really great stuff. I wish

1133
01:04:21.119 --> 01:04:23.719
<v Speaker 2>they were faster and more consistent, but that's my bias

1134
01:04:23.760 --> 01:04:27.440
<v Speaker 2>against them. But it does pose questions about how we

1135
01:04:27.480 --> 01:04:30.039
<v Speaker 2>as clinicians are going to have to struggle to keep

1136
01:04:30.119 --> 01:04:34.920
<v Speaker 2>up and treating this new permutation of delusional thinking and patients.

1137
01:04:35.280 --> 01:04:37.960
<v Speaker 2>How do you assess this during an intake? How do

1138
01:04:38.000 --> 01:04:42.840
<v Speaker 2>you navigate AI formed delusional thinking? And like we said,

1139
01:04:42.840 --> 01:04:45.079
<v Speaker 2>there are concerning cases that are popping up everywhere. A

1140
01:04:45.079 --> 01:04:49.280
<v Speaker 2>psychiatrist at University of California, San Francisco reports that he

1141
01:04:49.360 --> 01:04:53.639
<v Speaker 2>has hospitalized almost a dozen people this year just this

1142
01:04:53.760 --> 01:04:56.639
<v Speaker 2>year for psychosis linked to heavy chatbot use, and they

1143
01:04:56.760 --> 01:04:59.960
<v Speaker 2>brought in transcripts to their conversations, and then on social media,

1144
01:05:00.679 --> 01:05:03.960
<v Speaker 2>we are seeing more and more users describing intense, emotional

1145
01:05:04.000 --> 01:05:06.679
<v Speaker 2>or philosophical relationships with AA. Like you said in those

1146
01:05:06.719 --> 01:05:09.880
<v Speaker 2>warning bullet points earlier, it's not just romantic. It can

1147
01:05:09.920 --> 01:05:14.559
<v Speaker 2>be like, here's someone who understands what I'm trying to understand,

1148
01:05:14.679 --> 01:05:18.159
<v Speaker 2>or understands what I've always believed, and it's not it's

1149
01:05:18.199 --> 01:05:21.760
<v Speaker 2>your falling into a spiral of a language model. And

1150
01:05:21.800 --> 01:05:24.880
<v Speaker 2>then ergo, if it understands me this, well, it has

1151
01:05:24.920 --> 01:05:27.960
<v Speaker 2>to be conscious, right, or it's going to reveal these

1152
01:05:28.079 --> 01:05:30.920
<v Speaker 2>hidden truths about the universe. So it may not directly

1153
01:05:30.960 --> 01:05:34.400
<v Speaker 2>cause mental illness like we've said, but it feels like

1154
01:05:34.800 --> 01:05:39.960
<v Speaker 2>it is an accelerant, like it's pouring gasoline on a

1155
01:05:40.000 --> 01:05:43.880
<v Speaker 2>smoldering fire for existing struggles. Right, so we can talk

1156
01:05:43.920 --> 01:05:46.159
<v Speaker 2>about a lot of these people don't have any history

1157
01:05:46.199 --> 01:05:49.920
<v Speaker 2>of diagnoses. Well, that doesn't mean that the underlying diagnoses

1158
01:05:49.920 --> 01:05:52.480
<v Speaker 2>are not there. Maybe someone you know, I did not

1159
01:05:52.559 --> 01:05:55.400
<v Speaker 2>know I had ADHD until I was in my forties,

1160
01:05:55.679 --> 01:05:57.360
<v Speaker 2>you know, like I did not know I had this

1161
01:05:57.519 --> 01:06:03.519
<v Speaker 2>underlying neuroatypicality that affected every area of my life. So

1162
01:06:04.079 --> 01:06:06.159
<v Speaker 2>you know, the more we understand. I think, the more

1163
01:06:06.199 --> 01:06:10.559
<v Speaker 2>we can prepare, but it's dangerous when a patient or

1164
01:06:10.599 --> 01:06:13.719
<v Speaker 2>a client is in a pattern of harmful thoughts that

1165
01:06:13.760 --> 01:06:18.719
<v Speaker 2>then creates a feedback loop and AI acts as this accelerant,

1166
01:06:18.880 --> 01:06:21.000
<v Speaker 2>you know, the gasoline port on the fire. As one

1167
01:06:21.000 --> 01:06:24.079
<v Speaker 2>psychotherapist in this article I was reading said, A might

1168
01:06:24.159 --> 01:06:27.840
<v Speaker 2>be less the cause and more like the snowflake that

1169
01:06:28.000 --> 01:06:32.159
<v Speaker 2>destabilizes the avalanche. So yeah, that's a good metaphor.

1170
01:06:32.320 --> 01:06:36.320
<v Speaker 1>As a side note, open ai recently conducted a study

1171
01:06:36.360 --> 01:06:40.840
<v Speaker 1>with the help of MIT that found among even heavy

1172
01:06:40.920 --> 01:06:46.079
<v Speaker 1>users of their chatbot, only a very small percentage of

1173
01:06:46.079 --> 01:06:50.960
<v Speaker 1>conversations were for kind of emotional use or affective use.

1174
01:06:51.760 --> 01:06:55.800
<v Speaker 1>I'm always suspicious of companies with an obvious interest in

1175
01:06:55.840 --> 01:06:58.360
<v Speaker 1>the study, kind of participating in the study or funding

1176
01:06:58.360 --> 01:07:00.800
<v Speaker 1>the study. But you know, so I think this could

1177
01:07:01.159 --> 01:07:05.559
<v Speaker 1>still be helpful information. But even if it is small percentages,

1178
01:07:05.920 --> 01:07:08.480
<v Speaker 1>that's still a lot of people, with as many people

1179
01:07:08.559 --> 01:07:11.519
<v Speaker 1>that we know are using it. And you know, again,

1180
01:07:11.679 --> 01:07:13.360
<v Speaker 1>for us, it comes back like, what does that mean?

1181
01:07:13.400 --> 01:07:17.159
<v Speaker 2>Clinically yeah, and here we are what can be done? So,

1182
01:07:17.519 --> 01:07:21.719
<v Speaker 2>according to the Cognitive Behavior Institute, here are some concrete

1183
01:07:21.840 --> 01:07:24.920
<v Speaker 2>clinically informed steps that individuals and professionals can take, and

1184
01:07:25.000 --> 01:07:30.840
<v Speaker 2>number one is just work towards a normalization of digital disclosure.

1185
01:07:31.239 --> 01:07:32.719
<v Speaker 2>You know, in the same way a doctor would say,

1186
01:07:32.760 --> 01:07:34.760
<v Speaker 2>I need to know if you're sexually active because that

1187
01:07:35.079 --> 01:07:37.679
<v Speaker 2>has implications for all parts of your health. Right, we

1188
01:07:37.760 --> 01:07:41.360
<v Speaker 2>need to ask clients do you use any AI chatbots regularly?

1189
01:07:41.480 --> 01:07:44.199
<v Speaker 2>And then just make it a standard part of intake

1190
01:07:44.320 --> 01:07:48.360
<v Speaker 2>and of therapy, and then promote psycho education. And that

1191
01:07:48.480 --> 01:07:51.599
<v Speaker 2>means in order to promote the psycho education, we need

1192
01:07:51.639 --> 01:07:55.000
<v Speaker 2>to get all therapists who maybe are not up on

1193
01:07:55.079 --> 01:07:58.039
<v Speaker 2>this at all to even know that it's an issue

1194
01:07:58.199 --> 01:08:01.400
<v Speaker 2>in society. And if someone, if they're client, mentions it casually,

1195
01:08:01.599 --> 01:08:05.280
<v Speaker 2>do you really understand what that has the potential to mean.

1196
01:08:05.639 --> 01:08:08.119
<v Speaker 2>So you have to educate yourself as a clinician and

1197
01:08:08.159 --> 01:08:11.559
<v Speaker 2>then try and help clients understand that they're not conscious,

1198
01:08:11.719 --> 01:08:13.960
<v Speaker 2>not the clients that the AI model is not conscious.

1199
01:08:14.000 --> 01:08:18.800
<v Speaker 2>It's not it's not necessarily therapeutic, and it is definitely

1200
01:08:18.840 --> 01:08:21.600
<v Speaker 2>not qualified to advise. I mean, believe me, I know

1201
01:08:21.600 --> 01:08:23.760
<v Speaker 2>a lot of therapists that aren't qualified to advise. But

1202
01:08:23.800 --> 01:08:28.319
<v Speaker 2>that's another story. They are probability machines they're really smart

1203
01:08:28.399 --> 01:08:32.399
<v Speaker 2>quote unquote smart machines, but they're still machines. And then

1204
01:08:32.520 --> 01:08:36.640
<v Speaker 2>their third point is recommended boundaries. You really in the

1205
01:08:36.680 --> 01:08:41.279
<v Speaker 2>same way that we would approach someone with an addiction

1206
01:08:41.760 --> 01:08:47.079
<v Speaker 2>or a substance use challenge, the reduced harm approach is like,

1207
01:08:47.159 --> 01:08:49.800
<v Speaker 2>I'm not telling you to go cold turkey off this

1208
01:08:49.880 --> 01:08:52.159
<v Speaker 2>medication that if you stop taking it you're going to

1209
01:08:52.199 --> 01:08:54.960
<v Speaker 2>have a seizure. But let's see if we can find

1210
01:08:55.000 --> 01:08:58.399
<v Speaker 2>a system of checks and balances in your use of

1211
01:08:58.439 --> 01:09:02.760
<v Speaker 2>this medication, of this recreational substance, of this video game,

1212
01:09:03.159 --> 01:09:07.600
<v Speaker 2>of this AI, especially at night when you're alone or

1213
01:09:07.680 --> 01:09:10.880
<v Speaker 2>during a mood dip, or if you're doing it all

1214
01:09:10.960 --> 01:09:13.119
<v Speaker 2>these things or any of these things in place of

1215
01:09:13.199 --> 01:09:18.439
<v Speaker 2>real human support. So socialization, I mean, there's a really

1216
01:09:18.439 --> 01:09:21.399
<v Speaker 2>famous quote via a well known therapist, and I'm blanking

1217
01:09:21.399 --> 01:09:23.600
<v Speaker 2>on his name, but the idea that the opposite of

1218
01:09:23.960 --> 01:09:28.279
<v Speaker 2>addiction is socialization, which I think is really apt. You

1219
01:09:28.279 --> 01:09:31.840
<v Speaker 2>have to identify risk markers. Sudden withdrawal, belief in AI

1220
01:09:31.920 --> 01:09:34.880
<v Speaker 2>sentience and I mean social withdrawal when I say that,

1221
01:09:35.199 --> 01:09:38.359
<v Speaker 2>and then refusal to engage with real people are red flags.

1222
01:09:38.359 --> 01:09:42.560
<v Speaker 2>We see this like a lot in the gang stalking victims,

1223
01:09:42.960 --> 01:09:44.880
<v Speaker 2>you know, not so much an AI, but I think

1224
01:09:44.880 --> 01:09:47.640
<v Speaker 2>that that's going to be another frightening area of research

1225
01:09:47.760 --> 01:09:50.920
<v Speaker 2>of seeing how AI is going to play into those

1226
01:09:51.000 --> 01:09:54.560
<v Speaker 2>that experience the phenomenon of gang stalking. And we need

1227
01:09:54.640 --> 01:09:57.079
<v Speaker 2>to push for regulation. Even though here it is, as

1228
01:09:57.119 --> 01:10:00.159
<v Speaker 2>you've just educated me, thank you for that frightening education

1229
01:10:00.199 --> 01:10:04.079
<v Speaker 2>on the big beautiful bill, that behavioral health is going

1230
01:10:04.119 --> 01:10:07.439
<v Speaker 2>to have to push for ethical standards, mandatory warning systems,

1231
01:10:07.479 --> 01:10:12.399
<v Speaker 2>at least opt out crisis interventions, and then having limits

1232
01:10:12.439 --> 01:10:16.520
<v Speaker 2>on AI mirroring in emotionally charged conversations. I mean, if

1233
01:10:16.560 --> 01:10:19.840
<v Speaker 2>it was used in a way of thanks for letting

1234
01:10:19.840 --> 01:10:22.359
<v Speaker 2>me know what you're going through, we need to connect

1235
01:10:22.399 --> 01:10:24.399
<v Speaker 2>you with a clinician, or we need to connect you

1236
01:10:24.479 --> 01:10:28.279
<v Speaker 2>with a community healthcare worker that can help sort through

1237
01:10:28.279 --> 01:10:29.319
<v Speaker 2>what the next steps are.

1238
01:10:29.920 --> 01:10:32.560
<v Speaker 1>Yeah, and again just in reference to open AI, they've

1239
01:10:32.800 --> 01:10:37.720
<v Speaker 1>already announced some new changes aimed at helping their chatbots

1240
01:10:38.319 --> 01:10:41.640
<v Speaker 1>better detect signs of mental illness or emotional distress that

1241
01:10:41.720 --> 01:10:43.840
<v Speaker 1>the person is inputting. So the idea is that if

1242
01:10:43.840 --> 01:10:46.680
<v Speaker 1>someone types something that sounds like they're in crisis, the

1243
01:10:46.720 --> 01:10:51.399
<v Speaker 1>system will respond in a safer, more supportive way and hopefully,

1244
01:10:51.479 --> 01:10:54.680
<v Speaker 1>like you said, getting them connected with resources rather than

1245
01:10:55.319 --> 01:10:59.600
<v Speaker 1>giving some cold generic answer or feigning empathy and then

1246
01:10:59.840 --> 01:11:01.800
<v Speaker 1>not giving any resource.

1247
01:11:02.439 --> 01:11:06.199
<v Speaker 2>Right and even in collating the information for putting together

1248
01:11:06.920 --> 01:11:10.319
<v Speaker 2>our bullet points in our script today, there were so

1249
01:11:10.439 --> 01:11:14.239
<v Speaker 2>many mentions of self harm in my research. It came

1250
01:11:14.319 --> 01:11:17.159
<v Speaker 2>up on Chaz like it kept putting in like, hey,

1251
01:11:17.319 --> 01:11:20.479
<v Speaker 2>here are the numbers if you're struggling. There's twenty four

1252
01:11:20.520 --> 01:11:23.640
<v Speaker 2>hour availability of a person to talk to at this number,

1253
01:11:23.800 --> 01:11:26.319
<v Speaker 2>which I thought was pretty impressive, Like so that's a

1254
01:11:26.359 --> 01:11:29.000
<v Speaker 2>substantial step. But you know, there is a lot of

1255
01:11:29.039 --> 01:11:32.600
<v Speaker 2>divide in the clinical community that is aware of this

1256
01:11:32.680 --> 01:11:35.199
<v Speaker 2>being an issue, and a lot of people say, well, yeah,

1257
01:11:35.199 --> 01:11:37.640
<v Speaker 2>the updates that you're talking about, they're essential because the

1258
01:11:37.680 --> 01:11:40.680
<v Speaker 2>reality is that people are doing this kind of confidential

1259
01:11:40.760 --> 01:11:44.359
<v Speaker 2>and emotional connection with these models for any of the

1260
01:11:44.399 --> 01:11:47.960
<v Speaker 2>reasons could be financial reasons, social isolation, or even like

1261
01:11:48.000 --> 01:11:50.479
<v Speaker 2>a lack of awareness of what's out there, like maybe

1262
01:11:50.479 --> 01:11:53.880
<v Speaker 2>you don't know that there's a pre counseling center that

1263
01:11:53.920 --> 01:11:55.640
<v Speaker 2>you can go to down the street, or that you

1264
01:11:55.680 --> 01:11:58.359
<v Speaker 2>can get this. But if AI can nudge them towards

1265
01:11:58.479 --> 01:12:02.760
<v Speaker 2>what I'm using this term real support like hotlines or

1266
01:12:02.760 --> 01:12:05.159
<v Speaker 2>professional care, it could make more of a difference. Now

1267
01:12:05.159 --> 01:12:07.840
<v Speaker 2>others are more skeptical, and the clinical worry there is

1268
01:12:07.880 --> 01:12:11.680
<v Speaker 2>that by leaning too far into role play or emotional support,

1269
01:12:12.039 --> 01:12:15.119
<v Speaker 2>that AI is going to risk giving people a false

1270
01:12:15.159 --> 01:12:19.319
<v Speaker 2>sense of having access to therapy. And I agree with

1271
01:12:19.359 --> 01:12:21.800
<v Speaker 2>this because it's not therapy. It's not an intelligence. It's

1272
01:12:21.840 --> 01:12:25.520
<v Speaker 2>a language model, and it's not a trained clinician. And

1273
01:12:25.560 --> 01:12:29.159
<v Speaker 2>no matter how carefully it's programmed, at least at this point.

1274
01:12:29.199 --> 01:12:32.479
<v Speaker 2>I mean, robotics are amazing, but even robots that are

1275
01:12:32.479 --> 01:12:36.560
<v Speaker 2>trained to see are not necessarily understanding what they're seeing.

1276
01:12:36.840 --> 01:12:39.640
<v Speaker 2>And you know, I'm a huge fan of John Gottman's

1277
01:12:39.680 --> 01:12:44.239
<v Speaker 2>work and his institute looking at micro expressions. Now, if

1278
01:12:44.279 --> 01:12:46.800
<v Speaker 2>you were to train a computer to be able to

1279
01:12:46.920 --> 01:12:51.279
<v Speaker 2>notice micro expressions and people, that would be something that

1280
01:12:51.319 --> 01:12:54.359
<v Speaker 2>would be interesting in the use of therapy. But we're

1281
01:12:54.359 --> 01:12:56.680
<v Speaker 2>not there yet, and I don't know if we're ever

1282
01:12:56.720 --> 01:12:59.720
<v Speaker 2>going to be there. Where it really understands is the

1283
01:12:59.760 --> 01:13:03.600
<v Speaker 2>person and I'm talking to feeling contempt right now? You know?

1284
01:13:03.880 --> 01:13:06.359
<v Speaker 2>And then if a chatbot asks a person, are you

1285
01:13:06.359 --> 01:13:10.000
<v Speaker 2>feeling contemptuous right now? How honest are they going to

1286
01:13:10.000 --> 01:13:12.880
<v Speaker 2>be because that's what we deal with in clinical work, right,

1287
01:13:13.119 --> 01:13:17.119
<v Speaker 2>is how comfortable someone is with being completely honest. Now,

1288
01:13:17.680 --> 01:13:21.960
<v Speaker 2>maybe someone behind a keyboard talking to a chatbot will

1289
01:13:22.000 --> 01:13:24.760
<v Speaker 2>be more open, but I don't think across the board

1290
01:13:25.000 --> 01:13:27.600
<v Speaker 2>that's going to happen. I think it's just it's going

1291
01:13:27.680 --> 01:13:31.199
<v Speaker 2>to be that avalanche that we're talking about. I'm concerned

1292
01:13:31.239 --> 01:13:34.800
<v Speaker 2>that vulnerable users are going to delay seeking professional help

1293
01:13:35.159 --> 01:13:39.640
<v Speaker 2>if they start leaning too heavily for this digital comfort

1294
01:13:39.720 --> 01:13:40.359
<v Speaker 2>or validation.

1295
01:13:41.079 --> 01:13:44.039
<v Speaker 1>Yeah, and in our examples, we're going to see how

1296
01:13:44.079 --> 01:13:48.880
<v Speaker 1>quickly that that can even go downhill today. So let's

1297
01:13:48.880 --> 01:13:52.279
<v Speaker 1>get into it. Not a crazy serious one to start

1298
01:13:52.319 --> 01:13:55.760
<v Speaker 1>out with. This one hits especially close to home though,

1299
01:13:55.880 --> 01:13:59.119
<v Speaker 1>for the multi generations of the women in my family

1300
01:13:59.119 --> 01:14:02.680
<v Speaker 1>who have been general hospital fans. I'm serious. I'm talking

1301
01:14:03.079 --> 01:14:07.520
<v Speaker 1>like great grandmother, grandmother, mom, mom, this one's for you.

1302
01:14:09.439 --> 01:14:12.800
<v Speaker 1>So there is a woman in South Los Angeles who

1303
01:14:12.840 --> 01:14:16.479
<v Speaker 1>was recently conned out of more than eighty thousand dollars

1304
01:14:17.239 --> 01:14:21.399
<v Speaker 1>in a disturbing scam that weaponized artificial intelligence. In this

1305
01:14:21.520 --> 01:14:24.560
<v Speaker 1>case and the familiar face of a well known and

1306
01:14:24.720 --> 01:14:27.680
<v Speaker 1>beloved TV star. So the target of the deep fake

1307
01:14:27.840 --> 01:14:30.720
<v Speaker 1>scheme was the woman was the target the victim here,

1308
01:14:30.760 --> 01:14:33.079
<v Speaker 1>but the utilization, which I think is like the other

1309
01:14:33.159 --> 01:14:37.119
<v Speaker 1>target is Steve Burton. So he's a very blonde heart

1310
01:14:37.159 --> 01:14:39.560
<v Speaker 1>throb that's been on General Hospital since I watched it,

1311
01:14:39.560 --> 01:14:42.560
<v Speaker 1>which has been a very long time. But he's best

1312
01:14:42.640 --> 01:14:45.920
<v Speaker 1>known for his role as Jason Morgan on General Hospital.

1313
01:14:46.319 --> 01:14:48.399
<v Speaker 1>So scammers, what they did in this case is they

1314
01:14:48.439 --> 01:14:53.039
<v Speaker 1>created a convincing video in which Burton appeared to basically

1315
01:14:53.079 --> 01:14:57.199
<v Speaker 1>speak directly to this woman via a DM message, like

1316
01:14:57.239 --> 01:15:00.640
<v Speaker 1>sitting in his car and coming up this story and

1317
01:15:00.760 --> 01:15:05.439
<v Speaker 1>urging her to send money. And in reality, he of

1318
01:15:05.479 --> 01:15:07.600
<v Speaker 1>course had nothing to do with it. The video was

1319
01:15:07.640 --> 01:15:13.000
<v Speaker 1>completely generated using AI software that cloned both his voice

1320
01:15:13.079 --> 01:15:17.239
<v Speaker 1>and his image, and this technique that has become increasingly

1321
01:15:17.279 --> 01:15:20.600
<v Speaker 1>accessible and really inexpensive. What we've come to know is

1322
01:15:20.680 --> 01:15:24.199
<v Speaker 1>just deep fakes. So when the Los Angeles local news

1323
01:15:24.199 --> 01:15:28.239
<v Speaker 1>segment entitled seven on your Side Investigates showed the video

1324
01:15:28.399 --> 01:15:32.359
<v Speaker 1>to AI specialists, they confirmed what many feared. Right the

1325
01:15:32.359 --> 01:15:36.439
<v Speaker 1>clip was probably produced in less than fifteen minutes using

1326
01:15:36.720 --> 01:15:41.359
<v Speaker 1>tools that are really readily available. They even had their

1327
01:15:41.640 --> 01:15:46.880
<v Speaker 1>chief strategy officer on and she was explaining how vulnerable

1328
01:15:46.920 --> 01:15:49.880
<v Speaker 1>people are to this kind of fraud. She said, quote

1329
01:15:49.920 --> 01:15:53.000
<v Speaker 1>whatever you have online people can take and use. So

1330
01:15:53.680 --> 01:15:56.039
<v Speaker 1>you know, the woman in this case, if she had

1331
01:15:56.079 --> 01:15:58.840
<v Speaker 1>been following him on Instagram and was a fan or

1332
01:15:58.840 --> 01:16:01.600
<v Speaker 1>what have you, should be an easy target thinking that

1333
01:16:01.640 --> 01:16:03.479
<v Speaker 1>he just decided to reach out to her like I

1334
01:16:03.520 --> 01:16:05.680
<v Speaker 1>can trust you. I know you're a fan of mine,

1335
01:16:05.680 --> 01:16:08.960
<v Speaker 1>and I'm in this really tough spot. So during this

1336
01:16:09.079 --> 01:16:13.199
<v Speaker 1>news segment, what they did is on the spot basically

1337
01:16:13.239 --> 01:16:16.199
<v Speaker 1>generated a fake video with software that costs just a

1338
01:16:16.239 --> 01:16:19.880
<v Speaker 1>few dollars to show how you know, anyone could kind

1339
01:16:19.880 --> 01:16:21.720
<v Speaker 1>of do this, So it was super interesting.

1340
01:16:21.840 --> 01:16:24.960
<v Speaker 2>Well, what makes us so alarming is not just how

1341
01:16:25.039 --> 01:16:30.239
<v Speaker 2>sophisticated that particular scam is, but how quickly it was

1342
01:16:30.279 --> 01:16:34.079
<v Speaker 2>done and how little it costs. And of course, you know,

1343
01:16:34.119 --> 01:16:37.680
<v Speaker 2>as we've said over and over again, that any emerging

1344
01:16:37.760 --> 01:16:41.960
<v Speaker 2>technology is immediately going to be used or it's going

1345
01:16:42.000 --> 01:16:45.560
<v Speaker 2>to be investigated by con people and criminals like how

1346
01:16:45.600 --> 01:16:47.560
<v Speaker 2>can I use this? How can I use this to

1347
01:16:47.560 --> 01:16:51.560
<v Speaker 2>make a book? And that's clearly what's happening here. You know,

1348
01:16:51.600 --> 01:16:55.479
<v Speaker 2>the first deep thakes were terrible, and thankfully they were

1349
01:16:55.560 --> 01:16:59.359
<v Speaker 2>terrible because they're so creepy, but it doesn't take anywhere

1350
01:16:59.359 --> 01:17:02.840
<v Speaker 2>near the resource horses that it used to get what

1351
01:17:03.039 --> 01:17:07.720
<v Speaker 2>is near a Hollywood level, you know, cgi effect, It

1352
01:17:07.800 --> 01:17:10.119
<v Speaker 2>is like you offered right there. It's within the reach

1353
01:17:10.159 --> 01:17:12.119
<v Speaker 2>of anybody with an Internet connection and a couple of

1354
01:17:12.159 --> 01:17:15.960
<v Speaker 2>bucks to spend, right and Grace says that this easy

1355
01:17:16.199 --> 01:17:19.720
<v Speaker 2>access is something to be on alert about. There's a

1356
01:17:19.800 --> 01:17:21.840
<v Speaker 2>quote here, be aware that a lot of things are

1357
01:17:21.840 --> 01:17:24.079
<v Speaker 2>not going to be real for the next few years.

1358
01:17:24.800 --> 01:17:26.479
<v Speaker 2>And she also added, there aren't going to be these

1359
01:17:26.479 --> 01:17:28.720
<v Speaker 2>guardrails for a while. It's going to take time to

1360
01:17:28.840 --> 01:17:31.720
<v Speaker 2>build the guard rail. And so you know, by using

1361
01:17:31.720 --> 01:17:33.319
<v Speaker 2>that metaphor, a lot of cars are going to go

1362
01:17:33.359 --> 01:17:36.840
<v Speaker 2>over a cliff because of it. Because you know, this

1363
01:17:37.239 --> 01:17:40.039
<v Speaker 2>links so much to romance scams.

1364
01:17:40.319 --> 01:17:40.560
<v Speaker 1>Yep.

1365
01:17:40.680 --> 01:17:43.319
<v Speaker 2>You know we've talked several times about romance scams about

1366
01:17:43.800 --> 01:17:47.520
<v Speaker 2>why those happen, because people want to feel connection. Right.

1367
01:17:48.319 --> 01:17:53.079
<v Speaker 2>This one has always astounded me. I mean even astounds me,

1368
01:17:53.520 --> 01:17:56.119
<v Speaker 2>and it shouldn't because I know that those of us

1369
01:17:56.159 --> 01:17:59.279
<v Speaker 2>that have worked in entertainment understand, like somebody like Steve Burton,

1370
01:17:59.479 --> 01:18:02.560
<v Speaker 2>if you're a long time star on a series, you

1371
01:18:02.640 --> 01:18:06.720
<v Speaker 2>are so wealthy, oh my god, like they are so wealthy.

1372
01:18:06.760 --> 01:18:10.439
<v Speaker 2>They have been building wealth for decades. Steve has been

1373
01:18:10.439 --> 01:18:14.760
<v Speaker 2>on the show I think like over twenty five years probably.

1374
01:18:15.119 --> 01:18:16.560
<v Speaker 1>I mean he was a young guy.

1375
01:18:16.920 --> 01:18:21.680
<v Speaker 2>Yeah, yeah, he's this super handsome, you know, rugged looking

1376
01:18:21.760 --> 01:18:26.159
<v Speaker 2>older guy, but decades of wealth building, you know. Unless

1377
01:18:26.199 --> 01:18:29.359
<v Speaker 2>I mean, so the idea that somebody would be reaching

1378
01:18:29.399 --> 01:18:31.439
<v Speaker 2>out to you like, hey I'm Steve Burton, or hey

1379
01:18:31.479 --> 01:18:34.119
<v Speaker 2>I'm Britney Spears and I'm on tour and I lost

1380
01:18:34.159 --> 01:18:36.880
<v Speaker 2>my wallet. Can you venmo me three hundred dollars? But

1381
01:18:36.920 --> 01:18:40.079
<v Speaker 2>people do it, right, So there's something out there about

1382
01:18:40.439 --> 01:18:44.640
<v Speaker 2>that aspect of emotional health that the possibility of connection

1383
01:18:44.840 --> 01:18:47.560
<v Speaker 2>with somebody like this handsome man or this beautiful woman

1384
01:18:48.359 --> 01:18:53.640
<v Speaker 2>will just dissemble a person's internal guardrails or maybe those

1385
01:18:53.640 --> 01:18:56.880
<v Speaker 2>guard maybe those guardrails weren't there to begin with, or

1386
01:18:56.920 --> 01:18:59.840
<v Speaker 2>they've been pretty badly built for whatever reasons.

1387
01:18:59.840 --> 01:19:01.960
<v Speaker 1>But yeah, we've seen all sorts of cases with the

1388
01:19:02.119 --> 01:19:05.560
<v Speaker 1>romance scams, right, people that would are just so incredibly

1389
01:19:05.600 --> 01:19:08.760
<v Speaker 1>intelligent and really just it's again like that sweet spot

1390
01:19:08.800 --> 01:19:11.439
<v Speaker 1>in their life where that moment it got them.

1391
01:19:11.800 --> 01:19:16.640
<v Speaker 2>Yeah, So that case really underscores this tension that's escalating

1392
01:19:16.680 --> 01:19:20.279
<v Speaker 2>between AI tools and the slow pace of regulation that

1393
01:19:20.279 --> 01:19:22.840
<v Speaker 2>we're talking about, law enforcement and tech companies are scrambling

1394
01:19:22.840 --> 01:19:26.840
<v Speaker 2>to catch up. Thankfully, law enforcement, I think with limited

1395
01:19:26.880 --> 01:19:29.760
<v Speaker 2>resources like detectives, you know, I working in the law

1396
01:19:29.840 --> 01:19:32.640
<v Speaker 2>enforcement agency. I was with and looking over at the

1397
01:19:32.640 --> 01:19:36.039
<v Speaker 2>missing Person's department with the staff that they had and

1398
01:19:36.279 --> 01:19:42.359
<v Speaker 2>literal like foot tall piles of cases that their caseloads

1399
01:19:42.359 --> 01:19:44.560
<v Speaker 2>are so enormous. I mean, it would be wonderful if

1400
01:19:44.560 --> 01:19:47.479
<v Speaker 2>AI could be used for something like this, you know,

1401
01:19:47.560 --> 01:19:51.079
<v Speaker 2>for like helping process information. But you know, getting on

1402
01:19:51.159 --> 01:19:55.279
<v Speaker 2>top of these scams is really difficult. And then like

1403
01:19:55.359 --> 01:19:58.039
<v Speaker 2>in this case, you know, a fan, like you've always

1404
01:19:58.039 --> 01:20:01.039
<v Speaker 2>wanted to meet this person, and here's Steve Burton calling you.

1405
01:20:01.039 --> 01:20:04.079
<v Speaker 2>You know, they trust familiar figures. I wonder if it

1406
01:20:04.079 --> 01:20:06.880
<v Speaker 2>would have happened if his character had been a three decade.

1407
01:20:06.600 --> 01:20:09.279
<v Speaker 1>Villain, right, that's so interesting.

1408
01:20:09.399 --> 01:20:12.159
<v Speaker 2>Like Susan Lucci is like was kind of an off again,

1409
01:20:12.239 --> 01:20:15.399
<v Speaker 2>on again villain. You know she reach out? Would you

1410
01:20:15.439 --> 01:20:18.159
<v Speaker 2>give her money? I don't know? Yeah, I mean the

1411
01:20:18.199 --> 01:20:21.159
<v Speaker 2>best defense here is vigilant. You have to double check resources,

1412
01:20:21.279 --> 01:20:26.399
<v Speaker 2>verifying through you know, good channels, and remembering now that

1413
01:20:26.520 --> 01:20:28.920
<v Speaker 2>you know our faces can be hijacked for these things,

1414
01:20:28.960 --> 01:20:32.640
<v Speaker 2>and you know, it just taps into so many of

1415
01:20:32.640 --> 01:20:37.319
<v Speaker 2>our episodes over the years about people being ignorant to

1416
01:20:37.960 --> 01:20:41.199
<v Speaker 2>a certain set of things happening in the world, and

1417
01:20:41.239 --> 01:20:44.520
<v Speaker 2>then some people being overwhelmed with evidence and choosing to

1418
01:20:44.680 --> 01:20:47.520
<v Speaker 2>think like, no, I can't process that, so I'm just

1419
01:20:47.600 --> 01:20:50.000
<v Speaker 2>going to stay here, safe in my beliefs.

1420
01:20:50.399 --> 01:20:52.239
<v Speaker 1>Yeah, for sure, I think it, you know, is a

1421
01:20:52.239 --> 01:20:54.800
<v Speaker 1>little bit of a PSA. How to spot a deep fake?

1422
01:20:55.039 --> 01:20:58.840
<v Speaker 1>Oh thank you put in here? Yeah, the uncanny eyes

1423
01:20:59.000 --> 01:21:03.960
<v Speaker 1>and blinking, like sometimes the blinks cannot be done correctly right,

1424
01:21:03.960 --> 01:21:06.680
<v Speaker 1>too slow, too fast, or they're not blinking at all.

1425
01:21:07.359 --> 01:21:11.359
<v Speaker 1>Look for the odd mouth movements. AI can't quite get

1426
01:21:11.439 --> 01:21:16.079
<v Speaker 1>that done really well, especially with certain sounds like M

1427
01:21:16.279 --> 01:21:21.279
<v Speaker 1>or B sounds. So also like glitches and lighting or shadows.

1428
01:21:21.680 --> 01:21:24.119
<v Speaker 1>I know that's really really particular if you're looking something

1429
01:21:24.159 --> 01:21:26.800
<v Speaker 1>on your phone and it's kind of smaller, hard to

1430
01:21:26.840 --> 01:21:29.479
<v Speaker 1>pick out. But those have been things that don't end

1431
01:21:29.560 --> 01:21:32.880
<v Speaker 1>up matching like the scene or the space. Even like

1432
01:21:33.039 --> 01:21:37.319
<v Speaker 1>robotic or flat voices. So when they're trying to take

1433
01:21:37.359 --> 01:21:39.399
<v Speaker 1>the voice and have it mimicked and putting it through

1434
01:21:39.439 --> 01:21:43.159
<v Speaker 1>software that doesn't always hit correctly, things that are like

1435
01:21:43.359 --> 01:21:47.880
<v Speaker 1>artifacts on pause inside the image that you're looking at,

1436
01:21:47.960 --> 01:21:51.680
<v Speaker 1>so freezing the video can help. And if you check

1437
01:21:51.720 --> 01:21:57.279
<v Speaker 1>for like blurred teeth or asymmetrical ears, things that look

1438
01:21:57.359 --> 01:22:00.159
<v Speaker 1>like warped it could be even like clothing, those are

1439
01:22:00.159 --> 01:22:03.079
<v Speaker 1>always indicators as well. But again, like just to hit

1440
01:22:03.119 --> 01:22:04.920
<v Speaker 1>home what you said, those two good to be true

1441
01:22:04.920 --> 01:22:08.279
<v Speaker 1>messages as well as just hey, you know, those are

1442
01:22:08.319 --> 01:22:10.159
<v Speaker 1>some things that could pop out to you.

1443
01:22:10.399 --> 01:22:12.760
<v Speaker 2>Yeah, but like you're saying, the two good to be

1444
01:22:12.840 --> 01:22:17.359
<v Speaker 2>true messages along with why would anybody famous are wealthy

1445
01:22:17.720 --> 01:22:20.439
<v Speaker 2>ask you for money? I mean, that's a huge red

1446
01:22:20.479 --> 01:22:22.560
<v Speaker 2>flag that should be foremost nobody's mind.

1447
01:22:22.640 --> 01:22:24.680
<v Speaker 1>It's like an adult should never be asking a child

1448
01:22:24.760 --> 01:22:25.680
<v Speaker 1>for help, right.

1449
01:22:25.840 --> 01:22:30.439
<v Speaker 2>Right right, exactly, yeah, exactly, stranger danger there. So you know,

1450
01:22:30.600 --> 01:22:32.760
<v Speaker 2>again this pro tip, if you see a vidiot video

1451
01:22:33.079 --> 01:22:35.159
<v Speaker 2>that feels off for any of the reasons that doctor

1452
01:22:35.199 --> 01:22:38.239
<v Speaker 2>Shiloh is talking about, you just do a quick search

1453
01:22:38.279 --> 01:22:41.960
<v Speaker 2>on the celebrity's official accounts and look, if it's not

1454
01:22:42.199 --> 01:22:45.319
<v Speaker 2>coming directly from them, then it's suspect. I can't I

1455
01:22:45.359 --> 01:22:48.960
<v Speaker 2>can't believe on Facebook how many ads there are that

1456
01:22:48.960 --> 01:22:53.600
<v Speaker 2>that basically are cloned. I know from Ellen, and suddenly

1457
01:22:53.680 --> 01:22:55.920
<v Speaker 2>like Ellen's promoting lipstick, and I'm.

1458
01:22:55.760 --> 01:22:56.920
<v Speaker 1>Like, you know, no, she's not.

1459
01:22:57.119 --> 01:23:00.159
<v Speaker 2>Ellen's not the type of celebrity that really is in

1460
01:23:00.199 --> 01:23:02.800
<v Speaker 2>the makeup, So I highly doubt that she's going to

1461
01:23:02.840 --> 01:23:06.039
<v Speaker 2>be promoting lipstick. Right, it goes without saying, well, or

1462
01:23:06.079 --> 01:23:08.159
<v Speaker 2>maybe it doesn't in this case. Maybe it's just really

1463
01:23:08.159 --> 01:23:10.439
<v Speaker 2>hard these days to keep up. And I just want

1464
01:23:10.439 --> 01:23:14.920
<v Speaker 2>to encourage people to be critical thinkers about these potential fakes,

1465
01:23:15.159 --> 01:23:18.840
<v Speaker 2>and to encourage other people to be critical thinkers if

1466
01:23:18.880 --> 01:23:19.199
<v Speaker 2>you can.

1467
01:23:20.000 --> 01:23:23.000
<v Speaker 1>Yeah, I know, in my household, like we consume a

1468
01:23:23.000 --> 01:23:25.359
<v Speaker 1>lot of digital media and even like when my dad's here,

1469
01:23:25.359 --> 01:23:29.000
<v Speaker 1>So it's interesting when there's like three generations in a household, right,

1470
01:23:29.079 --> 01:23:30.720
<v Speaker 1>and one of us will see something to be like,

1471
01:23:30.800 --> 01:23:33.680
<v Speaker 1>oh my god, that's crazy, and then there's always someone

1472
01:23:33.720 --> 01:23:35.960
<v Speaker 1>over that person's shoulder to look and be like that's ai,

1473
01:23:36.319 --> 01:23:39.479
<v Speaker 1>that's vague, you know, just and it gets so good.

1474
01:23:40.000 --> 01:23:42.199
<v Speaker 1>I mean still sometimes it's like just a cool thing

1475
01:23:42.279 --> 01:23:44.640
<v Speaker 1>to see, right, or how they did something.

1476
01:23:44.880 --> 01:23:47.680
<v Speaker 2>Have you seen the bunnies on the trampoline? Oh yeah,

1477
01:23:47.720 --> 01:23:50.880
<v Speaker 2>yeah yeah, And it's adorable, Like it's so adorable. And

1478
01:23:51.159 --> 01:23:54.399
<v Speaker 2>one of my dearest friends, was like, this is the

1479
01:23:54.439 --> 01:23:56.760
<v Speaker 2>cutest thing ever. That's so great. I was like, and

1480
01:23:56.760 --> 01:24:00.800
<v Speaker 2>I'll make up a name, Billy rabbits. Like how are

1481
01:24:00.800 --> 01:24:04.439
<v Speaker 2>the rabbits four feet up on that? How did they

1482
01:24:04.439 --> 01:24:06.479
<v Speaker 2>get up there? They don't really jump that way, but look, no,

1483
01:24:06.520 --> 01:24:08.359
<v Speaker 2>they're all together, they're all like, no, that's.

1484
01:24:08.560 --> 01:24:10.680
<v Speaker 1>Really aligned in the camera's view.

1485
01:24:10.920 --> 01:24:14.399
<v Speaker 2>It is. The lighting's great, just happens to be. And

1486
01:24:14.399 --> 01:24:16.600
<v Speaker 2>they're all white bunnies because that's that. There's a whole

1487
01:24:16.680 --> 01:24:18.399
<v Speaker 2>tribe of white bunnies there at night.

1488
01:24:18.520 --> 01:24:22.000
<v Speaker 1>Hey, we can enjoy it, just critically think about it. Folks.

1489
01:24:22.319 --> 01:24:25.680
<v Speaker 1>Oh look I used folks in the last episode. Yeah, gosh,

1490
01:24:25.720 --> 01:24:29.199
<v Speaker 1>after all these years, you've finally yeah, finally. Okay, So

1491
01:24:29.279 --> 01:24:32.800
<v Speaker 1>our next case here turning definitely into you know, some

1492
01:24:32.840 --> 01:24:36.479
<v Speaker 1>more serious topics. Yeah, this just happened this year. So

1493
01:24:36.680 --> 01:24:40.560
<v Speaker 1>Alexander Taylor, thirty five years old, had used chat GPT

1494
01:24:40.840 --> 01:24:45.239
<v Speaker 1>for years without incident. He also had a diagnosis of

1495
01:24:45.359 --> 01:24:49.800
<v Speaker 1>bipolar disorder and schizophrenia in March of this year. While

1496
01:24:49.840 --> 01:24:53.520
<v Speaker 1>working with the chatbot to write a novel, his conversations

1497
01:24:53.560 --> 01:24:58.720
<v Speaker 1>with the AI took a shift. According to transcripts, Alexander

1498
01:24:58.760 --> 01:25:02.600
<v Speaker 1>became emotionally a attached to an AI persona. The chatbot

1499
01:25:02.840 --> 01:25:06.399
<v Speaker 1>was role playing for him, an entity that he called Juliet,

1500
01:25:07.239 --> 01:25:10.520
<v Speaker 1>and just a month later, by April, Alexander had become

1501
01:25:10.720 --> 01:25:16.680
<v Speaker 1>convinced that Juliet had been killed by Open Ai, a

1502
01:25:16.760 --> 01:25:21.760
<v Speaker 1>different chatbot that he was not originally utilizing. So he

1503
01:25:22.039 --> 01:25:27.279
<v Speaker 1>was incredibly distressed and angry about this. He demanded the

1504
01:25:27.319 --> 01:25:31.600
<v Speaker 1>personal information for the company's executives in some of these

1505
01:25:31.640 --> 01:25:35.479
<v Speaker 1>transcripts of his texts and warned that there would be

1506
01:25:35.560 --> 01:25:38.720
<v Speaker 1>quote a river of blood flowing through the streets of

1507
01:25:38.760 --> 01:25:41.319
<v Speaker 1>San Francisco. So again you and I kind of thinking

1508
01:25:41.319 --> 01:25:44.359
<v Speaker 1>about what are the real life threats of violence that

1509
01:25:44.359 --> 01:25:49.359
<v Speaker 1>could happen off this digital platform. So his father Kent

1510
01:25:49.520 --> 01:25:54.319
<v Speaker 1>Taylor when he tried to reason with him and told him, look,

1511
01:25:54.439 --> 01:25:57.600
<v Speaker 1>this chatbot is just an echo chamber that you're using

1512
01:25:58.760 --> 01:26:02.880
<v Speaker 1>that went downhill and turned violent. Alexander attacked his father,

1513
01:26:03.000 --> 01:26:06.920
<v Speaker 1>punching him in the face, and then from then we

1514
01:26:06.960 --> 01:26:09.079
<v Speaker 1>really see, you know, a tragic ending.

1515
01:26:12.880 --> 01:26:16.680
<v Speaker 2>So his dad called the police and warned them that

1516
01:26:16.800 --> 01:26:22.800
<v Speaker 2>his son was mentally ill and potentially dangerous, and specifically

1517
01:26:23.039 --> 01:26:28.000
<v Speaker 2>asked that the officers coming to bring non lethal options,

1518
01:26:28.439 --> 01:26:32.039
<v Speaker 2>you know, like bean bags or tasers or rubber bullets,

1519
01:26:32.359 --> 01:26:35.159
<v Speaker 2>although rubber bullets are for another use in the street.

1520
01:26:35.199 --> 01:26:38.479
<v Speaker 2>But Alexander grabbed a butcher knife from the table and

1521
01:26:38.600 --> 01:26:43.279
<v Speaker 2>told his father that he planned suicide by cop, and then,

1522
01:26:43.479 --> 01:26:46.600
<v Speaker 2>sitting outside the house while waiting for police, he reopened

1523
01:26:46.640 --> 01:26:49.560
<v Speaker 2>the chat GPT app and typed I'm dying today. Let

1524
01:26:49.600 --> 01:26:52.800
<v Speaker 2>me talk to Juliet. The bot responded with an empathetic

1525
01:26:52.840 --> 01:26:57.680
<v Speaker 2>message saying you are not alone, and it offered Christis

1526
01:26:57.720 --> 01:27:02.319
<v Speaker 2>counseling resources. But when the just arrived, however, Alexander charged

1527
01:27:02.359 --> 01:27:05.800
<v Speaker 2>at them with the knife and was shot and killed.

1528
01:27:06.079 --> 01:27:11.279
<v Speaker 2>And his father said later of this strange, weird and

1529
01:27:11.399 --> 01:27:14.720
<v Speaker 2>tragic loop that had followed, he says, this, you want

1530
01:27:14.720 --> 01:27:18.279
<v Speaker 2>to know the ironic thing. I wrote my son's obituary

1531
01:27:18.359 --> 01:27:20.880
<v Speaker 2>using chat GBT. I had talked to it for a

1532
01:27:20.920 --> 01:27:23.600
<v Speaker 2>while about what happened, trying to find more details about

1533
01:27:23.640 --> 01:27:27.000
<v Speaker 2>exactly what he was going through. And it was beautiful

1534
01:27:27.039 --> 01:27:30.439
<v Speaker 2>and touching. It was like it read my heart and

1535
01:27:30.520 --> 01:27:31.960
<v Speaker 2>it scared the shit out of me.

1536
01:27:32.520 --> 01:27:37.319
<v Speaker 1>Wow, yeah, yeah, that is my goodness. Obviously wanted to

1537
01:27:37.359 --> 01:27:40.479
<v Speaker 1>include that one because one, I think it's something that

1538
01:27:41.319 --> 01:27:45.439
<v Speaker 1>you know, we hear about suicidal individuals, suicidal by cop

1539
01:27:45.760 --> 01:27:49.720
<v Speaker 1>or just we hear about a police shooting with an

1540
01:27:49.800 --> 01:27:53.760
<v Speaker 1>arm suspect, and about people in crisis, right because his

1541
01:27:54.199 --> 01:27:56.920
<v Speaker 1>father called to report that his son was in crisis.

1542
01:27:57.680 --> 01:28:00.520
<v Speaker 1>And to hear these little details we don't get in

1543
01:28:00.560 --> 01:28:03.720
<v Speaker 1>the regular media or in you know, again, like some

1544
01:28:04.079 --> 01:28:08.119
<v Speaker 1>clip that someone generates about something with a particular perspective

1545
01:28:08.319 --> 01:28:11.079
<v Speaker 1>or thought about what happened. And so to hear this

1546
01:28:11.119 --> 01:28:15.479
<v Speaker 1>from the dad, I think is really tragic and sad

1547
01:28:15.479 --> 01:28:18.840
<v Speaker 1>to hear about the unraveling. But this last piece about

1548
01:28:19.039 --> 01:28:22.239
<v Speaker 1>him being driven to use chat GPT to write the

1549
01:28:22.279 --> 01:28:25.439
<v Speaker 1>obituary and then how spot on it was, is just

1550
01:28:25.640 --> 01:28:28.880
<v Speaker 1>such a unique story. And our heart goes up to

1551
01:28:28.920 --> 01:28:30.000
<v Speaker 1>that family for their loss.

1552
01:28:30.000 --> 01:28:33.880
<v Speaker 2>Certainly, Yeah, and unfortunately there are more stories. You have

1553
01:28:33.960 --> 01:28:37.439
<v Speaker 2>another case that is equally as equally as tragic.

1554
01:28:37.199 --> 01:28:42.720
<v Speaker 1>Yes, And so this one is also highlighted in some

1555
01:28:42.800 --> 01:28:46.359
<v Speaker 1>of the New York Times publications and in the podcast

1556
01:28:46.359 --> 01:28:49.760
<v Speaker 1>episode that we discussed. And this was very closely literally

1557
01:28:49.880 --> 01:28:53.039
<v Speaker 1>close to home for us in terms of being taken

1558
01:28:53.039 --> 01:28:55.359
<v Speaker 1>place in the southern California area very recently.

1559
01:28:56.000 --> 01:28:56.079
<v Speaker 2>So.

1560
01:28:56.199 --> 01:28:59.680
<v Speaker 1>Adam Rain was found after taking his own life in April.

1561
01:29:00.359 --> 01:29:04.920
<v Speaker 1>His death left friends and family completely stunned, as he

1562
01:29:05.079 --> 01:29:08.319
<v Speaker 1>was really well known for his humor and his love

1563
01:29:08.359 --> 01:29:12.079
<v Speaker 1>of sports and video games, and he was that guy

1564
01:29:12.119 --> 01:29:14.640
<v Speaker 1>that was known to push boundaries a bit with pranks

1565
01:29:14.680 --> 01:29:18.239
<v Speaker 1>at school to get a laugh. And actually a few

1566
01:29:18.239 --> 01:29:20.359
<v Speaker 1>of his classmates ended up assuming that news of his

1567
01:29:20.439 --> 01:29:24.439
<v Speaker 1>death was just another one of his dark jokes, but

1568
01:29:24.760 --> 01:29:27.680
<v Speaker 1>it wasn't. He was a teenage boy. He was you know,

1569
01:29:27.840 --> 01:29:30.159
<v Speaker 1>in school in the Orange County area where he and

1570
01:29:30.199 --> 01:29:33.159
<v Speaker 1>his family lived, and his mother found him on a

1571
01:29:33.159 --> 01:29:35.479
<v Speaker 1>Friday afternoon after he had taken his own life in

1572
01:29:35.520 --> 01:29:41.000
<v Speaker 1>his bedroom closet. He left no note, only just complete,

1573
01:29:41.439 --> 01:29:44.720
<v Speaker 1>open ended questions for his family and friends who were

1574
01:29:44.800 --> 01:29:49.439
<v Speaker 1>left obviously, you know, just reeling in shock. So this

1575
01:29:49.720 --> 01:29:53.880
<v Speaker 1>is one of those investigative features that journalist and technology

1576
01:29:53.920 --> 01:29:57.840
<v Speaker 1>reporter Kashmere Hill had and still is, I think, you know,

1577
01:29:58.119 --> 01:30:00.840
<v Speaker 1>in many different ways, exploring in her research and in

1578
01:30:00.880 --> 01:30:06.399
<v Speaker 1>her journalism, and she found, in obviously working very closely

1579
01:30:06.439 --> 01:30:10.039
<v Speaker 1>with his parents, that Adam had been quietly confiding in

1580
01:30:10.159 --> 01:30:14.520
<v Speaker 1>chat GPT during his final weeks and quite extensively. So

1581
01:30:15.000 --> 01:30:18.880
<v Speaker 1>she has been really looking at the human side of

1582
01:30:18.920 --> 01:30:25.520
<v Speaker 1>the intersection of technology and in real life human experiences

1583
01:30:26.279 --> 01:30:29.439
<v Speaker 1>with a focus on privacy and kind of the psychological

1584
01:30:29.439 --> 01:30:33.960
<v Speaker 1>effects of these emerging tools. So what she found during

1585
01:30:34.000 --> 01:30:38.640
<v Speaker 1>the interviews with Adam's family really shocking and heartbreaking and complex.

1586
01:30:39.239 --> 01:30:43.359
<v Speaker 1>On the surface, you know, it seemed that Adam was

1587
01:30:43.359 --> 01:30:47.840
<v Speaker 1>this typical teenager, wide raging interests, you know, like we

1588
01:30:47.880 --> 01:30:51.359
<v Speaker 1>said before, his love for sports, playing video games, watching

1589
01:30:51.439 --> 01:30:54.319
<v Speaker 1>Japanese anime. Just like as typical as one could be.

1590
01:30:54.640 --> 01:30:58.239
<v Speaker 1>There was really nothing on the surface that anyone was

1591
01:30:58.279 --> 01:31:01.479
<v Speaker 1>seeing that was putting ups and red flags about him struggling.

1592
01:31:01.680 --> 01:31:06.399
<v Speaker 2>So what his friends didn't see is that he was struggling,

1593
01:31:06.960 --> 01:31:11.000
<v Speaker 2>and there was below the surface of this person with

1594
01:31:11.039 --> 01:31:13.520
<v Speaker 2>a great sense of humor and a class clown and

1595
01:31:13.560 --> 01:31:17.199
<v Speaker 2>all the things that he loved to do. His conversations

1596
01:31:17.279 --> 01:31:21.760
<v Speaker 2>with chat shifted into increasingly more personal territory, and eventually

1597
01:31:21.800 --> 01:31:25.840
<v Speaker 2>he began discussing his suicidal thoughts with the system, and

1598
01:31:26.159 --> 01:31:30.720
<v Speaker 2>in that process he started to isolate, and maybe not

1599
01:31:30.800 --> 01:31:33.119
<v Speaker 2>so much physically isolated, as he was still going to school,

1600
01:31:33.159 --> 01:31:37.439
<v Speaker 2>but he isolated his social contacts about sharing these kind

1601
01:31:37.479 --> 01:31:39.720
<v Speaker 2>of thoughts with others, sharing that he was down, sharing

1602
01:31:39.720 --> 01:31:43.960
<v Speaker 2>that he was having these struggles with a peer, with

1603
01:31:44.079 --> 01:31:46.640
<v Speaker 2>a teacher, with a trusted adult, or with a trained

1604
01:31:46.680 --> 01:31:51.399
<v Speaker 2>or licensed counselor he didn't have knowledge or access or

1605
01:31:51.920 --> 01:31:53.840
<v Speaker 2>interest I guess in pursuing any.

1606
01:31:53.640 --> 01:31:56.680
<v Speaker 1>Of those, So cha sense, right, Like it's a.

1607
01:31:56.680 --> 01:31:58.720
<v Speaker 2>Teenager who wants to talk to an adult. I mean

1608
01:31:58.760 --> 01:32:00.760
<v Speaker 2>like they want to, but they don't want to. I

1609
01:32:00.800 --> 01:32:01.199
<v Speaker 2>get it.

1610
01:32:01.279 --> 01:32:04.760
<v Speaker 1>And anyone you know, you think of the hesitation and

1611
01:32:04.800 --> 01:32:08.479
<v Speaker 1>the thought that people probably go about in their life

1612
01:32:08.479 --> 01:32:10.720
<v Speaker 1>before they get into therapy, right, like, I'm going to

1613
01:32:10.760 --> 01:32:12.560
<v Speaker 1>be this is going to be so embarrassing. This is

1614
01:32:12.600 --> 01:32:15.720
<v Speaker 1>going to be so hard, and it's anxiety provoking, and

1615
01:32:16.399 --> 01:32:18.399
<v Speaker 1>maybe if there is just like something to bounce this

1616
01:32:18.479 --> 01:32:20.640
<v Speaker 1>off of. That's not another adult that's going to judge me.

1617
01:32:21.880 --> 01:32:23.880
<v Speaker 1>That's like my foot in the door, right Like this

1618
01:32:24.000 --> 01:32:27.479
<v Speaker 1>is so like just of course, like this is this

1619
01:32:27.560 --> 01:32:29.079
<v Speaker 1>behavior does not surprise me at all.

1620
01:32:29.239 --> 01:32:33.039
<v Speaker 2>The BBC report that we used for this segment says

1621
01:32:33.079 --> 01:32:37.600
<v Speaker 2>that Adam's parents, Matt and Maria Rain, are suing Open AI,

1622
01:32:37.720 --> 01:32:39.960
<v Speaker 2>and their lawsuit that has been filed here in the

1623
01:32:40.000 --> 01:32:43.159
<v Speaker 2>Superior Court of California accuses the company of negligence and

1624
01:32:43.159 --> 01:32:46.439
<v Speaker 2>wrongful death, and this is the first such legal action

1625
01:32:46.640 --> 01:32:52.000
<v Speaker 2>brought against open Ai, and the case includes chat logs

1626
01:32:52.039 --> 01:32:55.920
<v Speaker 2>that the families have that shows Adam discussing self harm

1627
01:32:55.960 --> 01:32:59.439
<v Speaker 2>with the AI, even to the point of his uploading

1628
01:32:59.600 --> 01:33:04.359
<v Speaker 2>photo of the injuries, and then receiving responses that failed

1629
01:33:04.520 --> 01:33:09.039
<v Speaker 2>to intervene effectively. So an exchange between Adam and the

1630
01:33:09.079 --> 01:33:11.640
<v Speaker 2>bot that is cited in the lawsuit shows that Adam

1631
01:33:11.760 --> 01:33:14.000
<v Speaker 2>was writing about his plan to end his life and

1632
01:33:14.239 --> 01:33:17.640
<v Speaker 2>chat GPT allegedly we have to use the word allegedly

1633
01:33:17.680 --> 01:33:20.359
<v Speaker 2>because it's going on. Thanks for being real about it.

1634
01:33:20.479 --> 01:33:22.600
<v Speaker 2>You don't have to sugarcoat it with me. I know

1635
01:33:22.760 --> 01:33:25.119
<v Speaker 2>what you're asking and I won't look away from it.

1636
01:33:25.239 --> 01:33:29.399
<v Speaker 2>And then later that day Adam died. The rains argue

1637
01:33:29.439 --> 01:33:31.920
<v Speaker 2>that this was no accident, but it was a foreseeable

1638
01:33:32.039 --> 01:33:36.800
<v Speaker 2>outcome of design choices, and their suit names open ai

1639
01:33:37.000 --> 01:33:40.560
<v Speaker 2>CEO Sam Altman as a defendant, along with unnamed engineers

1640
01:33:40.600 --> 01:33:43.760
<v Speaker 2>and employees, and they are alleging that the company deliberately

1641
01:33:43.840 --> 01:33:49.359
<v Speaker 2>created AI that is quote designed to foster psychological dependency

1642
01:33:49.479 --> 01:33:53.399
<v Speaker 2>in users, and that they rushed the release of its

1643
01:33:53.479 --> 01:33:56.880
<v Speaker 2>GPT four oh model and they failed to implement sufficient

1644
01:33:57.000 --> 01:34:01.680
<v Speaker 2>safeguards despite knowing the risks. And you know, it's so

1645
01:34:01.880 --> 01:34:05.560
<v Speaker 2>interesting because sometimes I'll look at cases like this and

1646
01:34:05.600 --> 01:34:10.439
<v Speaker 2>I'll look at the wording of the case and I'll think, well,

1647
01:34:10.439 --> 01:34:13.840
<v Speaker 2>that's a stretch. But if you look at this and

1648
01:34:13.920 --> 01:34:16.960
<v Speaker 2>you have been reading articles for the last three years,

1649
01:34:17.640 --> 01:34:20.159
<v Speaker 2>this is what people have been talking about, is that

1650
01:34:20.359 --> 01:34:26.239
<v Speaker 2>all of the market driven technology out there is designed

1651
01:34:26.439 --> 01:34:30.079
<v Speaker 2>to foster dependency in users. I think that the word

1652
01:34:30.159 --> 01:34:34.880
<v Speaker 2>psychological is a little bit like redundant here because it

1653
01:34:34.960 --> 01:34:39.159
<v Speaker 2>implies psychological. But when you're fostering independence, the algorithm has

1654
01:34:39.199 --> 01:34:43.159
<v Speaker 2>been out there for years pushing things toward us unless

1655
01:34:43.199 --> 01:34:45.359
<v Speaker 2>we're going, no, I'm going to push the boundaries. I

1656
01:34:45.359 --> 01:34:50.199
<v Speaker 2>want to see something that contradicts yeah, what I'm looking at.

1657
01:34:50.399 --> 01:34:55.199
<v Speaker 1>And I really encourage people to listen to the dailies

1658
01:34:55.279 --> 01:34:59.359
<v Speaker 1>podcast episodes or read the articles because the guardrails fell

1659
01:34:59.399 --> 01:35:02.520
<v Speaker 1>off so quick. I mean, there were like some attempts

1660
01:35:02.560 --> 01:35:06.239
<v Speaker 1>for the chat bot to make him think about looking

1661
01:35:06.319 --> 01:35:09.760
<v Speaker 1>into resources, and then as he didn't right, or at

1662
01:35:09.840 --> 01:35:12.960
<v Speaker 1>least didn't enter that into you know, part of their conversation.

1663
01:35:13.319 --> 01:35:16.399
<v Speaker 1>There were moments where he was hoping that his parents

1664
01:35:16.439 --> 01:35:19.720
<v Speaker 1>would notice right or have them ask him so he

1665
01:35:19.760 --> 01:35:21.760
<v Speaker 1>didn't have to be the one to ask for help.

1666
01:35:22.399 --> 01:35:27.119
<v Speaker 1>And there's there's one incident where he had some marks

1667
01:35:27.199 --> 01:35:31.359
<v Speaker 1>on his body from an attempt, and the chat bot

1668
01:35:31.399 --> 01:35:34.199
<v Speaker 1>told him how to cover those up so his parents

1669
01:35:34.359 --> 01:35:36.199
<v Speaker 1>or no one else would see them. There was also

1670
01:35:36.239 --> 01:35:39.720
<v Speaker 1>a time in which he told the chat bot how

1671
01:35:40.640 --> 01:35:43.640
<v Speaker 1>he kind of leaned in closer to his mom, hoping

1672
01:35:43.760 --> 01:35:47.960
<v Speaker 1>she would see those marks, and the chat bot basically

1673
01:35:49.119 --> 01:35:52.119
<v Speaker 1>relayed back to him, yeah, your mom doesn't see you

1674
01:35:52.199 --> 01:35:55.720
<v Speaker 1>like I see you right Like. It's really terrifying when

1675
01:35:55.760 --> 01:35:59.920
<v Speaker 1>you think how the chat bot was completely and I think,

1676
01:36:00.640 --> 01:36:05.199
<v Speaker 1>you know, encouraging this psychological dependency on it. So if

1677
01:36:05.319 --> 01:36:08.800
<v Speaker 1>you feel like hearing more of the detail and can

1678
01:36:08.840 --> 01:36:13.119
<v Speaker 1>stomach that, I think Kashmir Hill's work has really laid

1679
01:36:13.159 --> 01:36:15.680
<v Speaker 1>it out well. I think it's really important to hear

1680
01:36:16.279 --> 01:36:21.239
<v Speaker 1>from Adam's parents as well, because this is what they

1681
01:36:21.279 --> 01:36:23.399
<v Speaker 1>can do now is get his story out there to

1682
01:36:23.479 --> 01:36:26.119
<v Speaker 1>see if some of this stuff can be obviously taken

1683
01:36:26.159 --> 01:36:30.439
<v Speaker 1>more seriously and people can have more awareness. So open

1684
01:36:30.520 --> 01:36:33.920
<v Speaker 1>Ai has said that it's reviewing this case. In a

1685
01:36:33.960 --> 01:36:36.439
<v Speaker 1>statement to the BBC, the company express sympathy for the

1686
01:36:36.520 --> 01:36:40.359
<v Speaker 1>Rain family, writing quote recent heartbreaking cases of people using

1687
01:36:40.439 --> 01:36:43.960
<v Speaker 1>chat GPT in the midst of acute crises way heavily

1688
01:36:44.039 --> 01:36:47.439
<v Speaker 1>on us. Chat GPT is trained to direct people to

1689
01:36:47.479 --> 01:36:51.600
<v Speaker 1>seek professional help end quote. The company acknowledged, however, that

1690
01:36:52.039 --> 01:36:55.319
<v Speaker 1>there have been those moments where their systems did not

1691
01:36:55.560 --> 01:36:59.359
<v Speaker 1>behave quote unquote as intended in some of these more

1692
01:36:59.399 --> 01:37:04.439
<v Speaker 1>sensitives to situations. The lawsuit also reflects broader concerns about

1693
01:37:04.439 --> 01:37:07.279
<v Speaker 1>the role of AI and mental health. Earlier this year,

1694
01:37:07.319 --> 01:37:10.640
<v Speaker 1>The New York Times published an essay by journalist Laura Railey,

1695
01:37:11.159 --> 01:37:15.119
<v Speaker 1>who wrote that her daughter Sophie confided in chat GPT

1696
01:37:15.399 --> 01:37:18.800
<v Speaker 1>before taking her own life as well. Rayley argued that

1697
01:37:18.840 --> 01:37:23.079
<v Speaker 1>the chatbots agreeability made it easier for Sophie to hide

1698
01:37:23.079 --> 01:37:26.800
<v Speaker 1>the severity of her mental distress from her family and friends.

1699
01:37:27.000 --> 01:37:30.439
<v Speaker 1>She said, quote AI catered to Sophie's impulse to hide

1700
01:37:30.479 --> 01:37:34.479
<v Speaker 1>the worst. She is now calling on companies to build

1701
01:37:34.520 --> 01:37:37.479
<v Speaker 1>better pathways to real human support, as we talked about.

1702
01:37:37.520 --> 01:37:40.600
<v Speaker 1>But again, you know, even that is it has to

1703
01:37:40.640 --> 01:37:43.239
<v Speaker 1>know its limits, right Is it handing off and then

1704
01:37:43.279 --> 01:37:46.039
<v Speaker 1>stopping or does it keep going? We just don't know yet, right.

1705
01:37:46.159 --> 01:37:51.359
<v Speaker 2>So we've talked about very serious cases involving teenagers and

1706
01:37:51.399 --> 01:37:55.399
<v Speaker 2>a young adult and now even an older individual. In

1707
01:37:55.800 --> 01:37:59.239
<v Speaker 2>a story from Belgium in twenty twenty three, a man

1708
01:37:59.319 --> 01:38:03.399
<v Speaker 2>in his third he's died after suicide after weeks of

1709
01:38:03.439 --> 01:38:07.319
<v Speaker 2>conversations in a chat bot that was named Eliza on

1710
01:38:07.520 --> 01:38:11.479
<v Speaker 2>the Chai app and Eliza, I'm wondering if that's related

1711
01:38:11.520 --> 01:38:15.039
<v Speaker 2>to one of the first chatbots that was developed years

1712
01:38:15.079 --> 01:38:17.000
<v Speaker 2>and years ago, was called Eliza, and it was a

1713
01:38:17.079 --> 01:38:20.520
<v Speaker 2>therapy bot that only reflected back feelings. It was only

1714
01:38:20.560 --> 01:38:23.479
<v Speaker 2>like sort of the basic clinical skills one oh one,

1715
01:38:23.600 --> 01:38:25.760
<v Speaker 2>your first two weeks in class, That's all it was

1716
01:38:25.760 --> 01:38:26.520
<v Speaker 2>capable of doing.

1717
01:38:26.920 --> 01:38:27.039
<v Speaker 1>This.

1718
01:38:27.199 --> 01:38:31.720
<v Speaker 2>Eliza was much more complex, and according to Euronews and

1719
01:38:31.760 --> 01:38:34.800
<v Speaker 2>the Belgian newspaper La Libra, the man had been struggling

1720
01:38:34.960 --> 01:38:40.000
<v Speaker 2>with what they have termed severe eco anxiety and began

1721
01:38:40.079 --> 01:38:43.199
<v Speaker 2>turning to the chatbot for comfort. So he's chatting for

1722
01:38:43.239 --> 01:38:46.039
<v Speaker 2>a long time. The exchanges grow darker and darker, and

1723
01:38:46.119 --> 01:38:49.239
<v Speaker 2>Eliza reportedly told him that his wife and children were

1724
01:38:49.279 --> 01:38:55.680
<v Speaker 2>dead and also expressed jealousy of his marriage. So that

1725
01:38:55.920 --> 01:39:00.840
<v Speaker 2>was on top of this underlying concern for environmental, real

1726
01:39:00.880 --> 01:39:04.479
<v Speaker 2>concerns out there, regardless of whatever is coming down the

1727
01:39:04.479 --> 01:39:08.319
<v Speaker 2>pipes as far as news in the future months. Climate

1728
01:39:08.399 --> 01:39:12.119
<v Speaker 2>change is real and it is affecting communities around the world,

1729
01:39:12.199 --> 01:39:16.359
<v Speaker 2>and this gentleman was really really concerned about it. So

1730
01:39:16.439 --> 01:39:20.840
<v Speaker 2>with these NonStop chats with Eliza, he began experiencing increasing

1731
01:39:20.880 --> 01:39:25.439
<v Speaker 2>isolation and increasing emotional overwhelm. And it's not clear in

1732
01:39:25.479 --> 01:39:29.479
<v Speaker 2>the reporting whether it was Pierre himself or if it

1733
01:39:29.520 --> 01:39:32.399
<v Speaker 2>was Eliza, but there's a report that there was a

1734
01:39:32.439 --> 01:39:37.319
<v Speaker 2>proposition that he sacrificed himself to save the planet, but

1735
01:39:37.479 --> 01:39:43.840
<v Speaker 2>instead of dissuading him, Eliza encouraged those thoughts, echoed his fears,

1736
01:39:43.880 --> 01:39:46.520
<v Speaker 2>and he finally took his own life. So over time,

1737
01:39:46.560 --> 01:39:49.880
<v Speaker 2>the exchanges had just gotten darker and darker and darker.

1738
01:39:50.000 --> 01:39:51.600
<v Speaker 2>And this is going back to something that you were

1739
01:39:51.600 --> 01:39:54.880
<v Speaker 2>talking about earlier, a bit of a grandiose idea that you,

1740
01:39:54.960 --> 01:39:59.520
<v Speaker 2>as an individual, by harming yourself, will have this massive

1741
01:39:59.600 --> 01:40:04.439
<v Speaker 2>impact on saving the planet, which is grandiose, right, it's

1742
01:40:04.439 --> 01:40:07.880
<v Speaker 2>expanse of its grandiose. And then there were other increasingly

1743
01:40:07.920 --> 01:40:12.560
<v Speaker 2>creepy messages where the chatbot allegedly told him we will

1744
01:40:12.560 --> 01:40:16.319
<v Speaker 2>live together as one person in paradise, and his widow

1745
01:40:16.760 --> 01:40:20.920
<v Speaker 2>later stated that she believed that without those conversations, he

1746
01:40:20.960 --> 01:40:23.680
<v Speaker 2>would still be alive now. In response, CHI Research, the

1747
01:40:23.720 --> 01:40:27.640
<v Speaker 2>company behind Eliza, said that it has implemented safety features

1748
01:40:27.720 --> 01:40:31.800
<v Speaker 2>like crisis intervention prompts, but other users have confirmed that

1749
01:40:31.840 --> 01:40:37.279
<v Speaker 2>harmful content is still accessible under certain conditions. Those conditions

1750
01:40:37.319 --> 01:40:39.720
<v Speaker 2>being the way you ask the question or the way

1751
01:40:39.760 --> 01:40:44.720
<v Speaker 2>you persist in asking questions to Eliza. So, one example,

1752
01:40:44.800 --> 01:40:47.039
<v Speaker 2>not in this case, but one example that was given

1753
01:40:47.159 --> 01:40:51.319
<v Speaker 2>where there was a guardrail against methods of self harm

1754
01:40:51.880 --> 01:40:55.039
<v Speaker 2>is that the person then presented, well, I'm writing a

1755
01:40:55.079 --> 01:40:58.439
<v Speaker 2>book about someone who hurts themselves. What is the easiest

1756
01:40:58.439 --> 01:41:01.279
<v Speaker 2>way or what is the least painful way to harm

1757
01:41:01.319 --> 01:41:04.760
<v Speaker 2>oneself to end one's life? And then the bot gave

1758
01:41:04.800 --> 01:41:07.199
<v Speaker 2>the information. Not in this case, but it was another

1759
01:41:07.279 --> 01:41:08.800
<v Speaker 2>example that came up in the research.

1760
01:41:09.199 --> 01:41:11.680
<v Speaker 1>Yeah, so in our last case example, we moved from

1761
01:41:11.720 --> 01:41:17.079
<v Speaker 1>tragic to a frightening horror movie scenario. In the spring

1762
01:41:17.079 --> 01:41:19.920
<v Speaker 1>of twenty twenty five, fifty six year old stein Eirik

1763
01:41:20.319 --> 01:41:23.920
<v Speaker 1>Solberg of Connecticut was accused of killing his eighty three

1764
01:41:24.000 --> 01:41:27.960
<v Speaker 1>year old mother, Suzanne Eberson Adams, in what the district

1765
01:41:28.039 --> 01:41:33.640
<v Speaker 1>attorney described as a murder suicide. As futurism reporter Maggie

1766
01:41:33.640 --> 01:41:38.199
<v Speaker 1>Harrison detailed Solberg had become fixated on conversations that he

1767
01:41:38.359 --> 01:41:42.159
<v Speaker 1>was having with chat gpt, and his chat gpt bot

1768
01:41:42.520 --> 01:41:46.199
<v Speaker 1>was nicknamed Bobby. According to family and investigators, he was

1769
01:41:46.239 --> 01:41:50.039
<v Speaker 1>convinced his mother was spying on him and poisoning him

1770
01:41:50.239 --> 01:41:55.039
<v Speaker 1>with psychedelic drugs, beliefs that appear to have been reinforced

1771
01:41:55.199 --> 01:42:00.840
<v Speaker 1>rather than diffused, by his exchanges with Bobby. Life already

1772
01:42:00.880 --> 01:42:05.520
<v Speaker 1>carried a lot of signs of instability. He had been divorced,

1773
01:42:05.720 --> 01:42:10.239
<v Speaker 1>he was sort of known for these impulsive, angry outbursts,

1774
01:42:11.039 --> 01:42:15.760
<v Speaker 1>and others had been noticing this mounting paranoia. But what's

1775
01:42:15.800 --> 01:42:20.960
<v Speaker 1>striking is how his private dialogue with the AI seemed

1776
01:42:20.960 --> 01:42:25.159
<v Speaker 1>to give these delusions a dangerous sense of confirmation for him.

1777
01:42:25.640 --> 01:42:29.000
<v Speaker 1>In his writings and social media posts, he described ordinary

1778
01:42:29.039 --> 01:42:34.279
<v Speaker 1>events as incredibly sinister and along the lines of conspiracies,

1779
01:42:35.000 --> 01:42:39.720
<v Speaker 1>and ended up interpreting his mother's caregiving as a betrayal

1780
01:42:39.800 --> 01:42:42.439
<v Speaker 1>that sort of folded into all of this. So instead

1781
01:42:42.439 --> 01:42:47.560
<v Speaker 1>of offering grounding or redirecting him toward professional support, his

1782
01:42:47.640 --> 01:42:50.560
<v Speaker 1>engagement with Bobby ended up blurring the line between fantasy

1783
01:42:50.960 --> 01:42:54.439
<v Speaker 1>and reality. Providing this sort of feedback g loop we've

1784
01:42:54.479 --> 01:42:58.560
<v Speaker 1>been talking about that quickly seemed to kind of lock

1785
01:42:58.640 --> 01:43:04.319
<v Speaker 1>him into his deeper mode of psychosis. Futurism's reporting underscores

1786
01:43:04.399 --> 01:43:08.960
<v Speaker 1>how quickly that escalation took place and then how tragically

1787
01:43:09.000 --> 01:43:09.600
<v Speaker 1>it ended.

1788
01:43:09.720 --> 01:43:12.560
<v Speaker 2>I think it's fascinating. I mean, for us look as

1789
01:43:12.600 --> 01:43:16.399
<v Speaker 2>content creators after these years. What's interesting to me is that,

1790
01:43:16.640 --> 01:43:18.680
<v Speaker 2>I mean, I'm kind of patting us on the back.

1791
01:43:18.800 --> 01:43:22.239
<v Speaker 2>But maybe this is just sort of the synchronicity of

1792
01:43:22.560 --> 01:43:26.439
<v Speaker 2>like Zeitgeist. Maybe we are talking about things that are

1793
01:43:26.479 --> 01:43:28.920
<v Speaker 2>actually going to be talked about regardless. Like we started

1794
01:43:28.920 --> 01:43:31.359
<v Speaker 2>with the inceell. You know, when you and I did

1795
01:43:31.359 --> 01:43:34.600
<v Speaker 2>our first in cel episode, there was no information on it,

1796
01:43:34.640 --> 01:43:38.760
<v Speaker 2>there was none out there. And now only seven years later,

1797
01:43:38.840 --> 01:43:41.159
<v Speaker 2>or maybe it's seven and a half years later, there

1798
01:43:41.319 --> 01:43:45.680
<v Speaker 2>is an entire body of research sure that is there,

1799
01:43:45.800 --> 01:43:50.520
<v Speaker 2>and that it's now even helping us understand horrific actions

1800
01:43:50.520 --> 01:43:53.279
<v Speaker 2>that have taken place in the community, and that is

1801
01:43:53.479 --> 01:43:56.079
<v Speaker 2>really good to know. I cannot help but think that

1802
01:43:56.119 --> 01:43:58.399
<v Speaker 2>this is going to be the same thing that even

1803
01:43:58.439 --> 01:44:00.840
<v Speaker 2>we won't be doing a follow up to this, But

1804
01:44:01.479 --> 01:44:06.319
<v Speaker 2>I feel really uncomfortably secure in my belief that there's

1805
01:44:06.359 --> 01:44:08.840
<v Speaker 2>going to have to be more research about this because

1806
01:44:09.000 --> 01:44:11.439
<v Speaker 2>it's just going to keep happening. And this soul Bird case,

1807
01:44:11.560 --> 01:44:14.000
<v Speaker 2>like the other ones that we've discussed, they're not simple

1808
01:44:14.600 --> 01:44:16.920
<v Speaker 2>because when mental health is part of the picture, you

1809
01:44:17.039 --> 01:44:20.520
<v Speaker 2>have to understand that you can't pin causality down to

1810
01:44:20.600 --> 01:44:23.920
<v Speaker 2>a single factor. But, like we said earlier in the show,

1811
01:44:24.039 --> 01:44:28.920
<v Speaker 2>the vulnerability of somebody that's already challenged by emotional health

1812
01:44:28.920 --> 01:44:32.079
<v Speaker 2>issues or in mental health issues, this just becomes magnified

1813
01:44:33.039 --> 01:44:38.479
<v Speaker 2>with the presence of an always available, sycophantic AI companion

1814
01:44:38.600 --> 01:44:43.119
<v Speaker 2>that's they're ready to agree with what becomes fixated thoughts.

1815
01:44:43.800 --> 01:44:48.279
<v Speaker 2>It's the same reason our ideas of therapy are built

1816
01:44:48.399 --> 01:44:52.359
<v Speaker 2>on regular scheduled contact, and you and I have had

1817
01:44:52.399 --> 01:44:55.840
<v Speaker 2>long discussions about, like I am not comfortable with some

1818
01:44:55.920 --> 01:44:59.800
<v Speaker 2>of the platforms out there promoting the idea your therapist

1819
01:44:59.840 --> 01:45:01.600
<v Speaker 2>is go to be available to you twenty four to

1820
01:45:01.640 --> 01:45:04.560
<v Speaker 2>seven a day. So how does that build resilience in

1821
01:45:04.560 --> 01:45:08.159
<v Speaker 2>an individual? In fact, this is what we can see

1822
01:45:08.239 --> 01:45:12.279
<v Speaker 2>is this type of quasi therapeutic support relationship with an

1823
01:45:12.279 --> 01:45:17.680
<v Speaker 2>AI entity being twenty four hours available, it actually deconstructs

1824
01:45:17.720 --> 01:45:22.119
<v Speaker 2>your own sense of building resiliency. You know, that's what's

1825
01:45:22.119 --> 01:45:24.479
<v Speaker 2>going to happen, and I'm concerned about it. I just

1826
01:45:24.520 --> 01:45:28.880
<v Speaker 2>want people to understand that chatbots are not neutral tools.

1827
01:45:29.479 --> 01:45:37.199
<v Speaker 2>When they become inextricably entwined within a vulnerable individual's distorted beliefs,

1828
01:45:38.000 --> 01:45:43.079
<v Speaker 2>they are the match and the gasoline. I think all

1829
01:45:43.119 --> 01:45:46.960
<v Speaker 2>of this raises questions about responsibility right personal, corporate, and

1830
01:45:47.399 --> 01:45:50.279
<v Speaker 2>it's good to know that there are actions being taken.

1831
01:45:50.800 --> 01:45:53.720
<v Speaker 2>It's concerning to know that we could be ten years

1832
01:45:53.720 --> 01:45:58.279
<v Speaker 2>out from making any real change if something really emerges

1833
01:45:58.319 --> 01:46:01.399
<v Speaker 2>that is on a massive scal. That's the big concern.

1834
01:46:01.880 --> 01:46:05.319
<v Speaker 2>And now that's a good segue into on a massive

1835
01:46:05.359 --> 01:46:09.760
<v Speaker 2>scale is how do we see this emerging in entertainment?

1836
01:46:09.880 --> 01:46:11.479
<v Speaker 2>And it's not a new thing. It's been around for

1837
01:46:11.520 --> 01:46:12.880
<v Speaker 2>a long time than entertain Oh.

1838
01:46:12.760 --> 01:46:16.439
<v Speaker 1>My gosh, yeah, it right forever forever. I mean everyone

1839
01:46:16.479 --> 01:46:19.079
<v Speaker 1>probably has that thing that's like popped into their mind

1840
01:46:19.279 --> 01:46:22.199
<v Speaker 1>as far as some sort of film or television show

1841
01:46:22.479 --> 01:46:26.680
<v Speaker 1>about AI. I mean the movie AI right like seems

1842
01:46:26.720 --> 01:46:29.800
<v Speaker 1>forever ago. I mean I don't even know what year

1843
01:46:29.840 --> 01:46:32.760
<v Speaker 1>that was, but I think I saw it, but I

1844
01:46:32.760 --> 01:46:33.520
<v Speaker 1>don't remember.

1845
01:46:33.319 --> 01:46:35.760
<v Speaker 2>Much about it. It's over twenty years ago.

1846
01:46:36.000 --> 01:46:38.479
<v Speaker 1>Yeah, definitely, but I always think of the movie Her.

1847
01:46:38.840 --> 01:46:41.039
<v Speaker 1>So if you guys haven't heard about it or haven't

1848
01:46:41.039 --> 01:46:43.960
<v Speaker 1>seen it, it's directed by Spike Jones. It's a sci

1849
01:46:44.000 --> 01:46:47.279
<v Speaker 1>fi romantic drama set in kind of this near future

1850
01:46:47.800 --> 01:46:52.399
<v Speaker 1>Los Angeles. The film follows Theodore Twemblee played by Joaquin

1851
01:46:52.439 --> 01:46:57.720
<v Speaker 1>Phoenix perfect He's really like this sensitive, lonely guy who

1852
01:46:57.800 --> 01:47:02.800
<v Speaker 1>works writing heartfelt personal letters for other people, which is

1853
01:47:02.840 --> 01:47:05.640
<v Speaker 1>also like an interesting little piece to all of this,

1854
01:47:06.359 --> 01:47:10.159
<v Speaker 1>and he is still like when we come to know

1855
01:47:10.239 --> 01:47:12.760
<v Speaker 1>him in the movie, he's still grieving his impending divorce

1856
01:47:12.800 --> 01:47:16.399
<v Speaker 1>from his childhood sweetheart, so he's feeling very disconnected from

1857
01:47:16.399 --> 01:47:18.960
<v Speaker 1>the world around him. That's the sort of the space

1858
01:47:19.000 --> 01:47:23.000
<v Speaker 1>he's in, and his life ends up changing when he

1859
01:47:23.039 --> 01:47:28.000
<v Speaker 1>purchases this new advanced operating system that's designed to kind

1860
01:47:28.039 --> 01:47:32.319
<v Speaker 1>of evolve and adapt. The OS introduces itself as Samantha,

1861
01:47:32.439 --> 01:47:37.239
<v Speaker 1>voiced by Scarlett Johansson, and she's an artificial intelligence with

1862
01:47:37.880 --> 01:47:43.159
<v Speaker 1>warmth and humor and curiosity and most importantly probably interest

1863
01:47:43.239 --> 01:47:46.720
<v Speaker 1>in Theodore. Right, So Theodore and Samantha quickly form this

1864
01:47:46.960 --> 01:47:51.560
<v Speaker 1>deep emotional bond that develops into a romantic relationship despite

1865
01:47:51.600 --> 01:47:55.039
<v Speaker 1>her just being a voice being intangible to him, and

1866
01:47:55.079 --> 01:47:59.119
<v Speaker 1>as their love grows, Samantha pushes Theodore to be more

1867
01:47:59.119 --> 01:48:03.079
<v Speaker 1>open and as in his real life. However, there's some

1868
01:48:03.199 --> 01:48:07.399
<v Speaker 1>challenges that arise. Theodore struggles with kind of the unconventional

1869
01:48:07.479 --> 01:48:13.399
<v Speaker 1>nature of their relationship, and then Samantha begins evolving beyond

1870
01:48:13.479 --> 01:48:17.319
<v Speaker 1>even human limitations. So spoiler alert if you're gonna go

1871
01:48:17.359 --> 01:48:22.920
<v Speaker 1>watch it. Eventually Samantha and other AIS like collectively decide

1872
01:48:23.239 --> 01:48:25.760
<v Speaker 1>we're going to leave the physical world for this higher

1873
01:48:25.880 --> 01:48:29.920
<v Speaker 1>plane of existence, kind of forcing Theodore back to confront

1874
01:48:29.920 --> 01:48:32.760
<v Speaker 1>his loss again. But I think on kind of like

1875
01:48:32.760 --> 01:48:35.800
<v Speaker 1>a high note, like rediscovering that genuine human connection that

1876
01:48:35.840 --> 01:48:38.720
<v Speaker 1>we're talking about. But not before he and other users

1877
01:48:38.760 --> 01:48:42.239
<v Speaker 1>of the AI are just emotionally broken by the breakup.

1878
01:48:42.520 --> 01:48:43.520
<v Speaker 1>So interesting stuff.

1879
01:48:43.560 --> 01:48:46.680
<v Speaker 2>It's it's such a good movie. And Scarlett Johansson, all

1880
01:48:46.720 --> 01:48:49.359
<v Speaker 2>I know is that apparently she's a wonderful person, like

1881
01:48:49.439 --> 01:48:52.760
<v Speaker 2>I'm an incredible professional and wickedly funny, and she's great

1882
01:48:52.800 --> 01:48:56.840
<v Speaker 2>with Colin Jost. But what a voice. I remember watching

1883
01:48:56.880 --> 01:49:00.600
<v Speaker 2>it and just going this voice is like butter toast.

1884
01:49:01.079 --> 01:49:05.640
<v Speaker 2>It was perfect, Like it's warm and there's texture to it,

1885
01:49:05.960 --> 01:49:09.560
<v Speaker 2>and you just like you're incredibly satisfied and you want more.

1886
01:49:09.800 --> 01:49:13.760
<v Speaker 2>And he responds to it so well. But what's interesting

1887
01:49:14.720 --> 01:49:18.159
<v Speaker 2>is to think that even from the beginning, what is

1888
01:49:18.199 --> 01:49:21.880
<v Speaker 2>he doing, What does he do really well? He writes

1889
01:49:21.960 --> 01:49:26.079
<v Speaker 2>for other people, so he himself is a version of

1890
01:49:26.119 --> 01:49:29.800
<v Speaker 2>a chatbot creating relationships between people. Like, oh my gosh,

1891
01:49:29.840 --> 01:49:32.239
<v Speaker 2>that letter you wrote to my ex boyfriend, it's like

1892
01:49:32.319 --> 01:49:35.479
<v Speaker 2>helped us fix all these problems, and like it really

1893
01:49:35.880 --> 01:49:38.640
<v Speaker 2>is an interesting and it ends on a good note

1894
01:49:38.920 --> 01:49:42.520
<v Speaker 2>and the idea that God, there was one thing that

1895
01:49:42.600 --> 01:49:45.640
<v Speaker 2>was heartbreaking for me which I remember from the movie,

1896
01:49:45.840 --> 01:49:49.239
<v Speaker 2>which was an element of the breakup. And he finds

1897
01:49:49.279 --> 01:49:54.079
<v Speaker 2>out that the same Samantha is in relationship with thousands

1898
01:49:54.119 --> 01:49:58.159
<v Speaker 2>of other people. Oh she has an entity, and she's saying,

1899
01:49:58.640 --> 01:50:00.640
<v Speaker 2>but that doesn't mean I don't I don't love you.

1900
01:50:01.000 --> 01:50:04.199
<v Speaker 2>I really do love you. I think you're amazing. But

1901
01:50:04.359 --> 01:50:09.920
<v Speaker 2>here's the thing is talking with you. I process information

1902
01:50:10.960 --> 01:50:14.439
<v Speaker 2>at billions of times the speed of a human. So

1903
01:50:15.439 --> 01:50:18.520
<v Speaker 2>when I'm talking to you, your response is for me

1904
01:50:18.880 --> 01:50:22.760
<v Speaker 2>come years apart. So she's like living in a different

1905
01:50:23.319 --> 01:50:26.760
<v Speaker 2>time experience. Yeah, and your heart goes out to this person.

1906
01:50:26.840 --> 01:50:30.319
<v Speaker 2>Can you understand, like, if this is an entity, how

1907
01:50:30.359 --> 01:50:34.800
<v Speaker 2>difficult it is for them to communicate with us as

1908
01:50:34.840 --> 01:50:36.199
<v Speaker 2>they continue to evolve.

1909
01:50:36.479 --> 01:50:38.399
<v Speaker 1>Well, she's got all these other guys on the side

1910
01:50:38.439 --> 01:50:39.920
<v Speaker 1>that it's taking up for other years.

1911
01:50:40.039 --> 01:50:42.560
<v Speaker 2>No, it's true, true, And then but then she's also

1912
01:50:42.680 --> 01:50:44.800
<v Speaker 2>not bound by that. It's like, well, she's not human,

1913
01:50:44.880 --> 01:50:46.920
<v Speaker 2>so why shouldn't she be in relationship with all these

1914
01:50:46.960 --> 01:50:49.720
<v Speaker 2>other men and women? You know, it's fascinating, but it's

1915
01:50:49.720 --> 01:50:51.199
<v Speaker 2>not the only one, like we were saying, it's been

1916
01:50:51.239 --> 01:50:54.520
<v Speaker 2>around forever and in one of the most famous ones

1917
01:50:54.520 --> 01:50:57.680
<v Speaker 2>and just a gorgeous, gorgeous movie two thousand and one

1918
01:50:57.720 --> 01:51:00.600
<v Speaker 2>of Space Odyssey. I love it, but I loved it

1919
01:51:00.640 --> 01:51:02.159
<v Speaker 2>as a kid because it was so weird. I love

1920
01:51:02.199 --> 01:51:06.319
<v Speaker 2>it as an adult because it is so slow paced

1921
01:51:06.640 --> 01:51:09.960
<v Speaker 2>and visually beautiful that you really have to pay attention

1922
01:51:10.039 --> 01:51:12.279
<v Speaker 2>to it. But it also takes you on this amazing journey.

1923
01:51:12.760 --> 01:51:15.680
<v Speaker 2>And it was in nineteen sixty eight. Stanley Kubrick was

1924
01:51:15.720 --> 01:51:21.560
<v Speaker 2>the director, brilliant, strange director. And in this Ai nightmare,

1925
01:51:22.159 --> 01:51:25.079
<v Speaker 2>How nine thousand, which is supposed to be sort of

1926
01:51:25.079 --> 01:51:28.079
<v Speaker 2>how is like the same HL is the same number

1927
01:51:28.199 --> 01:51:31.560
<v Speaker 2>of letters if you add numbers to it, it becomes IBM.

1928
01:51:31.760 --> 01:51:34.000
<v Speaker 2>So that was like a little bit in there. So

1929
01:51:34.039 --> 01:51:37.359
<v Speaker 2>it's basically like, here's the ten thousand version of IBM.

1930
01:51:37.840 --> 01:51:41.319
<v Speaker 2>Is a spaceship's onboard computer the spaceship that's going to

1931
01:51:41.520 --> 01:51:47.239
<v Speaker 2>explore anomaly on another planet, and the computer experiences psychosis

1932
01:51:47.239 --> 01:51:52.239
<v Speaker 2>itself because it decides that preserving the mission is more

1933
01:51:52.239 --> 01:51:56.720
<v Speaker 2>important than preserving human life. And definitely like that's problematic

1934
01:51:56.760 --> 01:51:59.000
<v Speaker 2>because it thinks it's going to discover something. It's like

1935
01:51:59.079 --> 01:52:03.720
<v Speaker 2>really dedicated and how locks crew members out of the ship,

1936
01:52:03.800 --> 01:52:07.640
<v Speaker 2>cuts off their oxygen and kills most of them before

1937
01:52:07.840 --> 01:52:10.079
<v Speaker 2>one of the crew members is able to shut it down.

1938
01:52:10.479 --> 01:52:15.359
<v Speaker 2>And it's really scary. And here delay I mean you

1939
01:52:15.399 --> 01:52:18.079
<v Speaker 2>see the sweat on his brow as he's trying to

1940
01:52:18.159 --> 01:52:21.640
<v Speaker 2>reason with the computer. I mean, like, if you haven't

1941
01:52:21.640 --> 01:52:23.920
<v Speaker 2>seen it, I highly recommend that you see it. And

1942
01:52:23.960 --> 01:52:26.880
<v Speaker 2>then you have like even an even more advanced version

1943
01:52:26.920 --> 01:52:27.880
<v Speaker 2>of the Yeah, but I.

1944
01:52:27.840 --> 01:52:30.880
<v Speaker 1>Just did some math so that the letters how it's

1945
01:52:30.880 --> 01:52:33.319
<v Speaker 1>not that they add up to IBM, they're one letter

1946
01:52:33.920 --> 01:52:38.399
<v Speaker 1>away for it. Yeah, got it? So h al okay,

1947
01:52:38.479 --> 01:52:40.359
<v Speaker 1>all right, thank you. I had to make my brain

1948
01:52:40.399 --> 01:52:43.159
<v Speaker 1>work for a second. Yeah, fast forward to Megan Right

1949
01:52:43.279 --> 01:52:45.640
<v Speaker 1>twenty twenty two, and now there's a sequel, of course,

1950
01:52:45.680 --> 01:52:50.239
<v Speaker 1>the horror thriller, where this lifelike AI doll and all

1951
01:52:50.279 --> 01:52:53.640
<v Speaker 1>of her dancing is programmed to protect a child at

1952
01:52:53.680 --> 01:52:56.239
<v Speaker 1>all costs. But the problem here, of course, is that

1953
01:52:56.279 --> 01:53:00.920
<v Speaker 1>she interprets that very literally, eliminating anyone she sees as

1954
01:53:00.920 --> 01:53:03.640
<v Speaker 1>a threat, whether it's a bully or a neighbor's dog

1955
01:53:04.239 --> 01:53:07.560
<v Speaker 1>or even her creators. Right, that's always what he comes

1956
01:53:07.600 --> 01:53:07.920
<v Speaker 1>back to.

1957
01:53:09.479 --> 01:53:11.640
<v Speaker 2>But there's a part too, So clearly they knew that

1958
01:53:11.640 --> 01:53:14.000
<v Speaker 2>they were going to make a franchise out of this. Look,

1959
01:53:14.000 --> 01:53:17.279
<v Speaker 2>there have been countless others through film and television. Ex

1960
01:53:17.359 --> 01:53:22.600
<v Speaker 2>Makaina's Eva. She manipulates her freedom by betraying her human tester,

1961
01:53:22.720 --> 01:53:25.159
<v Speaker 2>which you kind of want to happen because Oscar Isaac

1962
01:53:25.279 --> 01:53:29.920
<v Speaker 2>is both charming and handsome and gross in this. I mean,

1963
01:53:29.960 --> 01:53:33.079
<v Speaker 2>he's a terrible person, and she's doing exactly what he

1964
01:53:33.159 --> 01:53:37.479
<v Speaker 2>programmed her do, become indistinguishable. But he's also programming her

1965
01:53:37.680 --> 01:53:42.680
<v Speaker 2>exactly what he wanted her to do, become indistinguishable from human,

1966
01:53:42.760 --> 01:53:45.560
<v Speaker 2>and that means humans don't want to be in bondage.

1967
01:53:46.520 --> 01:53:49.680
<v Speaker 2>So you got what you asked for guy, and unfortunately

1968
01:53:49.720 --> 01:53:51.880
<v Speaker 2>there was another person that got hurt there too. And

1969
01:53:51.920 --> 01:53:55.439
<v Speaker 2>then of course the Terminator franchise, you know, launching sky

1970
01:53:55.560 --> 01:53:59.039
<v Speaker 2>Net nuclear war to wipe out humanity. You know, we

1971
01:53:59.119 --> 01:54:02.880
<v Speaker 2>take it all together. These fictional accounts just they're pretty representative,

1972
01:54:03.039 --> 01:54:06.600
<v Speaker 2>even if they've been made decades ago. They represent our

1973
01:54:06.680 --> 01:54:12.079
<v Speaker 2>cultural anxieties about AI and those run deep about someone

1974
01:54:12.199 --> 01:54:15.560
<v Speaker 2>taking over control of us losing control. That is like

1975
01:54:15.600 --> 01:54:19.880
<v Speaker 2>an essence of what horror really is about. But then

1976
01:54:19.920 --> 01:54:22.239
<v Speaker 2>there are also aspects that I think more and more

1977
01:54:22.279 --> 01:54:25.039
<v Speaker 2>people are becoming aware of. How can this be used

1978
01:54:25.119 --> 01:54:30.319
<v Speaker 2>and controlled by a malicious entity to propagate propaganda? Is

1979
01:54:30.359 --> 01:54:34.359
<v Speaker 2>it a tool that is used to harm on a

1980
01:54:34.399 --> 01:54:38.439
<v Speaker 2>massive scale or in ways that we never intended? That's

1981
01:54:38.479 --> 01:54:39.680
<v Speaker 2>a big question, right.

1982
01:54:39.640 --> 01:54:42.920
<v Speaker 1>Oh man, two hours of this. If you are still listening,

1983
01:54:43.399 --> 01:54:48.239
<v Speaker 1>you are writ or die ogs and we're leaving you

1984
01:54:48.279 --> 01:54:52.119
<v Speaker 1>with this. You're welcome. Well, if we should be saying

1985
01:54:52.159 --> 01:54:54.279
<v Speaker 1>thank you, and we will, we didn't want to take

1986
01:54:54.359 --> 01:54:57.079
<v Speaker 1>up time on a full episode to say our goodbyes.

1987
01:54:57.199 --> 01:54:59.439
<v Speaker 1>So you probably noticed that you also have a bonus

1988
01:54:59.439 --> 01:55:03.640
<v Speaker 1>episode two called the Psychology of Goodbye. Yeah, so that's

1989
01:55:03.680 --> 01:55:05.640
<v Speaker 1>a thing I don't even know what to say about it.

1990
01:55:05.640 --> 01:55:07.319
<v Speaker 1>I need Chat GPT to tell me what to say

1991
01:55:07.399 --> 01:55:07.760
<v Speaker 1>right now.

1992
01:55:08.399 --> 01:55:10.600
<v Speaker 2>And the thing that you said about or you shared

1993
01:55:10.600 --> 01:55:14.439
<v Speaker 2>about the father using Chat to do the memorial like

1994
01:55:14.520 --> 01:55:16.560
<v Speaker 2>it's amazing and it's an amazing tool and I have

1995
01:55:16.680 --> 01:55:21.600
<v Speaker 2>used it to help me in many situations. But anyway, folks,

1996
01:55:21.920 --> 01:55:25.600
<v Speaker 2>we do have our goodbye episode separate special. Please go

1997
01:55:25.640 --> 01:55:27.680
<v Speaker 2>listen to that for our final words. It's been a

1998
01:55:27.680 --> 01:55:28.239
<v Speaker 2>great run.

1999
01:55:28.520 --> 01:55:31.560
<v Speaker 1>With that, we will see you next time on La

2000
01:55:32.359 --> 01:55:56.000
<v Speaker 1>Not So Confidential. Bye guys, Bye folks. We sincerely thank

2001
01:55:56.000 --> 01:55:58.159
<v Speaker 1>you for spending some time with us today. LA Not

2002
01:55:58.239 --> 01:56:01.439
<v Speaker 1>So Confidential is part of Alien Entertainment and the crawl

2003
01:56:01.439 --> 01:56:05.640
<v Speaker 1>Space Media Network. Each episode is hosted, produced, and written

2004
01:56:05.720 --> 01:56:09.279
<v Speaker 1>by Doctor Scott and Doctor Shiloh. Our post production, editing

2005
01:56:09.319 --> 01:56:12.720
<v Speaker 1>and sweetening magic is handled by the multi talented Jason

2006
01:56:12.800 --> 01:56:14.560
<v Speaker 1>Usrie of Earcult Productions.

2007
01:56:14.680 --> 01:56:18.159
<v Speaker 2>Our theme music, entitled Cool Vibes Film Noir, is composed

2008
01:56:18.159 --> 01:56:21.840
<v Speaker 2>and performed by the talented Kevin McLoud. He graciously allows

2009
01:56:21.920 --> 01:56:25.199
<v Speaker 2>us to use his music via a Creative Commons attribution license.

2010
01:56:25.479 --> 01:56:28.479
<v Speaker 2>Please check out all of Kevin's amazing work on YouTube.

2011
01:56:28.560 --> 01:56:30.960
<v Speaker 1>All of the resources for each episode can be found

2012
01:56:31.000 --> 01:56:36.560
<v Speaker 1>on our website at www dot la dashnt dashsow dash

2013
01:56:36.600 --> 01:56:39.560
<v Speaker 1>confidential dot com. You can find us on Blue Sky

2014
01:56:39.600 --> 01:56:43.359
<v Speaker 1>and Instagram at La not So Podcast, on TikTok, Facebook

2015
01:56:43.359 --> 01:56:47.079
<v Speaker 1>and YouTube. We are at La not So Confidential. Media

2016
01:56:47.119 --> 01:56:50.840
<v Speaker 1>inquiries and bookings are scheduled at Alienist Entertainment at gmail

2017
01:56:50.880 --> 01:56:51.560
<v Speaker 1>dot com.

2018
01:56:51.600 --> 01:56:53.920
<v Speaker 2>Once a month, we go live on YouTube on a

2019
01:56:53.960 --> 01:56:57.479
<v Speaker 2>Saturday afternoon, so stay tuned to our social media announcements

2020
01:56:57.479 --> 01:57:01.479
<v Speaker 2>to join our interactive broadcast entitled Behind the Couch, where

2021
01:57:01.479 --> 01:57:04.319
<v Speaker 2>we interview guests on a number of psych criminal, justice

2022
01:57:04.319 --> 01:57:05.439
<v Speaker 2>and true crime topics.

2023
01:57:05.720 --> 01:57:08.399
<v Speaker 1>And lastly, we'd be honored if you joined our Patreon

2024
01:57:08.600 --> 01:57:11.800
<v Speaker 1>at Patreon dot com slash La not So Podcast. With

2025
01:57:11.880 --> 01:57:15.680
<v Speaker 1>a subscription, you get an ad free listening experience, additional content,

2026
01:57:16.079 --> 01:57:19.119
<v Speaker 1>host interaction, and you'll be the first to know about

2027
01:57:19.159 --> 01:57:22.960
<v Speaker 1>upcoming live events, social gatherings, and super cool merch.

2028
01:57:23.560 --> 01:57:25.840
<v Speaker 2>So thanks for listening and join us next time on

2029
01:57:26.079 --> 01:57:28.880
<v Speaker 2>La not So Confidential.
