1
00:00:09,664 --> 00:00:13,104
Speaker 1: Mission Implausible is now something you can watch. Just go

2
00:00:13,144 --> 00:00:16,823
Speaker 1: to YouTube and search Mission Implausible podcasts, or click on

3
00:00:16,864 --> 00:00:20,864
Speaker 1: the link to our channel. In our show notes, I'm

4
00:00:20,904 --> 00:00:21,984
Speaker 1: John Cipher.

5
00:00:21,664 --> 00:00:22,784
Speaker 2: And I'm Jerry O'Shea.

6
00:00:23,264 --> 00:00:26,544
Speaker 3: We have over sixty years of experience as clandestine officers

7
00:00:26,624 --> 00:00:29,544
Speaker 3: in the CIA, serving in high risk areas all around

8
00:00:29,544 --> 00:00:30,264
Speaker 3: the world, and.

9
00:00:30,304 --> 00:00:34,504
Speaker 4: Part of our job was creating conspiracies to deceive our adversaries.

10
00:00:34,784 --> 00:00:37,504
Speaker 3: Now we're going to use that experience to investigate the

11
00:00:37,504 --> 00:00:40,424
Speaker 3: conspiracy theories everyone's talking about as well as some of

12
00:00:40,464 --> 00:00:41,224
Speaker 3: you may not have heard.

13
00:00:41,344 --> 00:00:43,584
Speaker 2: Could they be true or are we being manipulated?

14
00:00:43,664 --> 00:00:46,223
Speaker 3: We'll find out now on Mission Implausible.

15
00:00:50,384 --> 00:00:52,104
Speaker 2: Welcome back, Adam, John, and Jerry.

16
00:00:52,144 --> 00:00:54,584
Speaker 1: I haven't seen any of you for a long time,

17
00:00:54,824 --> 00:00:58,224
Speaker 1: but we're now entering a new season and there's been

18
00:00:58,224 --> 00:00:59,344
Speaker 1: a lot going on since.

19
00:00:59,144 --> 00:01:01,184
Speaker 2: The last time we all did an episode of a

20
00:01:01,184 --> 00:01:03,024
Speaker 2: month and a half ago. I hate to say I

21
00:01:03,104 --> 00:01:05,904
Speaker 2: missed a dam Yeah, I have missed you guys. I mean,

22
00:01:05,944 --> 00:01:08,264
Speaker 2: I guess I can just say I We normally make

23
00:01:08,304 --> 00:01:10,464
Speaker 2: fun of ours each other, but this is kind of

24
00:01:10,464 --> 00:01:12,184
Speaker 2: obnoxious because it'll make it hard for you to make

25
00:01:12,184 --> 00:01:14,264
Speaker 2: fun of me. We were dealing with that big health

26
00:01:14,264 --> 00:01:17,624
Speaker 2: issue in my family for much of this year. But

27
00:01:17,703 --> 00:01:19,984
Speaker 2: we're on the other side of it, and my wife

28
00:01:20,824 --> 00:01:22,344
Speaker 2: had to have a lot of surgery and stuff. But

29
00:01:22,384 --> 00:01:26,744
Speaker 2: she's now like really recovered and we're feeling wonderfully.

30
00:01:26,304 --> 00:01:28,223
Speaker 3: And we've had her on the show, she's interviewed us.

31
00:01:28,503 --> 00:01:31,863
Speaker 3: We're so glad she's doing better. But did you make

32
00:01:31,904 --> 00:01:33,624
Speaker 3: more money this year while you were dealing with a

33
00:01:33,664 --> 00:01:35,344
Speaker 3: health crisis than you had him in the past.

34
00:01:35,584 --> 00:01:36,143
Speaker 2: It seems like.

35
00:01:36,104 --> 00:01:38,544
Speaker 3: You're all over LinkedIn and you're doing this AI and

36
00:01:38,584 --> 00:01:39,904
Speaker 3: you're speaking to groups.

37
00:01:39,584 --> 00:01:42,584
Speaker 2: You're always jumping on the new fan at him. I am,

38
00:01:42,904 --> 00:01:45,264
Speaker 2: so I don't think I fully understood what I was

39
00:01:45,304 --> 00:01:48,464
Speaker 2: getting into. So after I did the Freakonomics series about AI,

40
00:01:48,944 --> 00:01:52,064
Speaker 2: I became pretty close with Ethan Mollik, who's become this

41
00:01:52,184 --> 00:01:56,064
Speaker 2: like great AI mind, and we decided to start an

42
00:01:56,104 --> 00:01:59,184
Speaker 2: AI consulting business with a couple other people, his wife, Leelach,

43
00:01:59,584 --> 00:02:02,664
Speaker 2: the wonderful Jessica Johnston. So yeah, we worked with a

44
00:02:02,664 --> 00:02:07,344
Speaker 2: bunch of like big companies on AI. This might shock you.

45
00:02:07,344 --> 00:02:11,344
Speaker 2: I'm not the world's most technical expert AI. It actually

46
00:02:11,424 --> 00:02:14,224
Speaker 2: it's interesting. I don't think I realized this until fairly recently.

47
00:02:14,224 --> 00:02:16,544
Speaker 2: But Planet Money, which I created back in two thousand

48
00:02:16,544 --> 00:02:20,744
Speaker 2: and eight with Alex Bloomberg, was the first major American

49
00:02:20,784 --> 00:02:22,824
Speaker 2: media company podcast. As far as we can tell, we

50
00:02:22,824 --> 00:02:24,744
Speaker 2: haven't been able to find another one. And I was

51
00:02:24,784 --> 00:02:29,024
Speaker 2: really on the front lines of digital disruption, like I

52
00:02:29,064 --> 00:02:31,704
Speaker 2: was at NPR deal like trying to figure out how

53
00:02:31,704 --> 00:02:34,824
Speaker 2: podcasting would change things for several years. Then I went

54
00:02:34,824 --> 00:02:37,984
Speaker 2: to the New York Times, where I was mostly a reporter,

55
00:02:38,024 --> 00:02:40,984
Speaker 2: but I would report up to the senior people and

56
00:02:41,144 --> 00:02:44,384
Speaker 2: talk to them about digital disruption and media. And then

57
00:02:44,424 --> 00:02:45,984
Speaker 2: I was at the New Yorker did the same. Then

58
00:02:45,984 --> 00:02:50,344
Speaker 2: I ran this podcast company with Sony, and throughout that

59
00:02:50,424 --> 00:02:53,784
Speaker 2: I became really fascinated by how like technological change impacts

60
00:02:53,824 --> 00:02:56,784
Speaker 2: society and impacts companies. I mean, you can talk about

61
00:02:57,104 --> 00:03:01,024
Speaker 2: whatever railroads and electricity and the telephone and the internet

62
00:03:01,104 --> 00:03:04,824
Speaker 2: and mobile and a million other innovations, and there's all

63
00:03:04,824 --> 00:03:08,744
Speaker 2: these fascinating things that happen when a new technology changes work.

64
00:03:09,544 --> 00:03:12,104
Speaker 2: You know, a lot of these big companies are like

65
00:03:12,184 --> 00:03:14,064
Speaker 2: all right, Like a year ago they were like, is

66
00:03:14,104 --> 00:03:15,704
Speaker 2: this really that big a deal? Now I think most

67
00:03:15,744 --> 00:03:17,264
Speaker 2: are like, Okay, this is big deal. But what does

68
00:03:17,304 --> 00:03:19,104
Speaker 2: it mean? How do we deal with it? What do

69
00:03:19,144 --> 00:03:21,704
Speaker 2: we do? And so it's been this fascinating front row

70
00:03:21,824 --> 00:03:25,464
Speaker 2: seat into it. It's also a fascinating front row seat into,

71
00:03:25,864 --> 00:03:29,704
Speaker 2: among many other things, probably the world's greatest it's weirdly

72
00:03:29,824 --> 00:03:33,144
Speaker 2: like probably the best tool ever for conspiracy theorists, but

73
00:03:33,224 --> 00:03:36,944
Speaker 2: it's also may arguably the best tool ever for fighting

74
00:03:36,984 --> 00:03:41,224
Speaker 2: conspiracy theories. Can AI make John smarter or sound smarter?

75
00:03:41,584 --> 00:03:47,064
Speaker 2: The stats show that lower performers benefit more than higher

76
00:03:47,104 --> 00:03:52,304
Speaker 2: performers woo, But the higher performers still stay ahead because

77
00:03:52,744 --> 00:03:56,224
Speaker 2: there seems to be like an intelligence AI premium, So

78
00:03:56,384 --> 00:03:58,704
Speaker 2: it's not gonna be as obvious how amazing I am

79
00:03:58,824 --> 00:03:59,824
Speaker 2: compared to you.

80
00:03:59,984 --> 00:04:03,584
Speaker 3: Heads Well, Jay, and I spoke to a NYU professor

81
00:04:03,744 --> 00:04:05,544
Speaker 3: just the other day, Thomson Shaw, and we'll have an

82
00:04:05,544 --> 00:04:08,744
Speaker 3: episode with her, yeah, coming up, and she talks a

83
00:04:08,744 --> 00:04:11,464
Speaker 3: lot about how AI and some of the tech really

84
00:04:11,504 --> 00:04:14,344
Speaker 3: does help conspiracy theories because you can essentially just put

85
00:04:14,504 --> 00:04:17,943
Speaker 3: a series of events together, list them, and then just

86
00:04:17,984 --> 00:04:21,504
Speaker 3: click and it'll create these connections. And then you think

87
00:04:21,544 --> 00:04:24,784
Speaker 3: you've done this genius research and found this incredible connection.

88
00:04:25,264 --> 00:04:27,304
Speaker 3: And then of course that gets into the ecosystem and

89
00:04:27,344 --> 00:04:28,024
Speaker 3: gets spun in.

90
00:04:28,383 --> 00:04:32,024
Speaker 2: It's really good at being persuasive, and it's very good.

91
00:04:32,063 --> 00:04:35,944
Speaker 2: And it doesn't care, right, It just it just wants.

92
00:04:36,224 --> 00:04:39,303
Speaker 2: It's like a talking dog or something. It just wants.

93
00:04:39,303 --> 00:04:40,704
Speaker 2: What do you want? What do you want? Give you

94
00:04:40,784 --> 00:04:43,263
Speaker 2: what you want? Although there are some interesting studies that

95
00:04:43,303 --> 00:04:45,863
Speaker 2: it's also the best tool to get someone out of

96
00:04:45,863 --> 00:04:46,984
Speaker 2: a conspiracy theory.

97
00:04:47,144 --> 00:04:48,904
Speaker 3: But do you have to put in the right things?

98
00:04:48,984 --> 00:04:50,304
Speaker 2: Yeah, you have to put in the right thing.

99
00:04:50,424 --> 00:04:54,224
Speaker 4: But going from conspiracy theory to conspiracy genuinely, there's this

100
00:04:54,264 --> 00:04:58,264
Speaker 4: word that like almost no one can spell, is like algorithm, right,

101
00:04:58,344 --> 00:04:59,463
Speaker 4: so nothing comes from nothing.

102
00:04:59,503 --> 00:05:02,383
Speaker 2: No, most people can spell it, just not you LG.

103
00:05:03,104 --> 00:05:08,984
Speaker 2: There're two whys. That's what's confusing. But if you own if.

104
00:05:08,904 --> 00:05:11,623
Speaker 4: You own the l algorithm, or if you design the algorithf,

105
00:05:11,664 --> 00:05:14,904
Speaker 4: you can influence the algorithm. You can influence millions of people.

106
00:05:15,424 --> 00:05:19,063
Speaker 4: And arguably the algorithms that we all deal with every

107
00:05:19,144 --> 00:05:24,184
Speaker 4: day that are persuasive actually belong or engineered by, or

108
00:05:24,424 --> 00:05:30,623
Speaker 4: are really influenced by a small cabal of wealthy individuals.

109
00:05:30,743 --> 00:05:34,464
Speaker 4: There really is like ten people in the United States

110
00:05:34,863 --> 00:05:38,984
Speaker 4: who are responsible for the algorithms that generate AI and

111
00:05:39,063 --> 00:05:40,704
Speaker 4: what we what's put in front.

112
00:05:40,464 --> 00:05:41,024
Speaker 2: Of us to read.

113
00:05:41,104 --> 00:05:43,024
Speaker 3: And luckily they're excellent people for this.

114
00:05:43,104 --> 00:05:45,623
Speaker 1: I recommend a book I've been reading called Careless People

115
00:05:45,784 --> 00:05:48,903
Speaker 1: about Facebook doing exactly what you're saying, not through AI,

116
00:05:49,183 --> 00:05:54,183
Speaker 1: but through very intentional engineered programs. It bears out in

117
00:05:54,303 --> 00:05:57,664
Speaker 1: great detail, not just the ten people that are creating

118
00:05:57,664 --> 00:06:00,383
Speaker 1: an algorithm to change the course of events, but the

119
00:06:00,424 --> 00:06:00,984
Speaker 1: one person.

120
00:06:01,063 --> 00:06:03,743
Speaker 2: Yeah, but can I be pedantic for a moment or

121
00:06:03,863 --> 00:06:07,344
Speaker 2: for the rest of our relationship? That's the default Because

122
00:06:07,584 --> 00:06:11,063
Speaker 2: this is not saying you're wrong, but it's so. The

123
00:06:11,104 --> 00:06:15,583
Speaker 2: way AI fundamentally works is you just put more and

124
00:06:15,623 --> 00:06:19,104
Speaker 2: more training data in. It does all this linear algebra,

125
00:06:19,664 --> 00:06:23,344
Speaker 2: and you just push it against more and more of

126
00:06:23,383 --> 00:06:26,504
Speaker 2: these chips that they're building as fast as they humanly can.

127
00:06:27,104 --> 00:06:30,623
Speaker 2: And so it isn't algorithmic in this Obviously, it's like

128
00:06:30,664 --> 00:06:33,143
Speaker 2: software code, so it's algorithmic in that sense, but it's

129
00:06:33,224 --> 00:06:38,024
Speaker 2: not like linear programmatic. The people who make the model,

130
00:06:38,063 --> 00:06:41,703
Speaker 2: Sam Altman and those people have no idea what it's

131
00:06:41,743 --> 00:06:43,623
Speaker 2: going to be able to do. In many cases, they

132
00:06:43,623 --> 00:06:45,984
Speaker 2: don't even know what it can do. Now, it's really

133
00:06:46,024 --> 00:06:49,023
Speaker 2: remarkable how little they understand their own tools.

134
00:06:49,063 --> 00:06:51,424
Speaker 3: Does it can it create new knowledge or does it

135
00:06:51,544 --> 00:06:53,784
Speaker 3: just make connections what's existing and put in.

136
00:06:54,344 --> 00:06:58,144
Speaker 2: That's a pretty huge debate. Although can we create new

137
00:06:58,183 --> 00:07:01,063
Speaker 2: knowledge also is a question like to what extent are

138
00:07:01,063 --> 00:07:06,664
Speaker 2: we is creating new knowledge largely configuring old knowledge. But

139
00:07:06,823 --> 00:07:09,944
Speaker 2: so it's not like with Facebook, you can have a

140
00:07:09,944 --> 00:07:13,824
Speaker 2: meeting and say, all right, we want to optimize for

141
00:07:13,944 --> 00:07:17,944
Speaker 2: more engagement or we want to optimize for the things

142
00:07:18,064 --> 00:07:20,704
Speaker 2: you know they're optimizing for are not conspiracy theories, but

143
00:07:20,744 --> 00:07:23,864
Speaker 2: it's the things that lead to more money. Yeah, we

144
00:07:23,904 --> 00:07:26,864
Speaker 2: want people to like feel really emotional and really engaged

145
00:07:26,904 --> 00:07:28,744
Speaker 2: and get really caught up and follow a lot of

146
00:07:28,744 --> 00:07:33,384
Speaker 2: stuff and comment on it. But with AI, the core models,

147
00:07:33,504 --> 00:07:35,584
Speaker 2: it's you know, there's a lot there are choices they

148
00:07:35,584 --> 00:07:38,304
Speaker 2: make and what training data they use and blah blah blah.

149
00:07:38,344 --> 00:07:40,624
Speaker 2: But it's not and now the AI has a view

150
00:07:40,704 --> 00:07:44,304
Speaker 2: about something or other, it's more a byproduct that they

151
00:07:44,344 --> 00:07:47,664
Speaker 2: can't understand, which actually is in some ways more scary.

152
00:07:47,744 --> 00:07:51,944
Speaker 2: Now they will decide directionally are we pushing it more

153
00:07:51,984 --> 00:07:54,023
Speaker 2: towards chat. They do seem to be able to turn

154
00:07:54,104 --> 00:07:57,864
Speaker 2: up or turn down the sycophancy, like the there's a

155
00:07:57,944 --> 00:08:01,304
Speaker 2: moment where Chatchip they launched a new model and they

156
00:08:01,304 --> 00:08:05,224
Speaker 2: had decided to push it to be more accommodating to

157
00:08:05,264 --> 00:08:07,544
Speaker 2: what people want, and then it just but it went nuts.

158
00:08:07,544 --> 00:08:09,184
Speaker 2: And do you remember this was a few months ago.

159
00:08:09,224 --> 00:08:12,104
Speaker 2: People were posting, like a friend of mine this, There

160
00:08:12,104 --> 00:08:13,464
Speaker 2: are a lot of these, but a friend of mine

161
00:08:13,824 --> 00:08:16,783
Speaker 2: said to it, I've been told by God that I

162
00:08:16,824 --> 00:08:19,064
Speaker 2: can I have a message that can transform the world

163
00:08:19,184 --> 00:08:21,103
Speaker 2: right away. It's like you must sell everything, you must.

164
00:08:21,504 --> 00:08:23,544
Speaker 2: And then it was like it was like, well, my

165
00:08:23,744 --> 00:08:26,024
Speaker 2: wife thinks I shouldn't cash out my four oh one K,

166
00:08:26,144 --> 00:08:28,544
Speaker 2: and it was like, but you have been chosen by God,

167
00:08:28,664 --> 00:08:30,944
Speaker 2: you must, you know. So clearly there are things they

168
00:08:30,944 --> 00:08:36,504
Speaker 2: can dial up or down, but it's more like like

169
00:08:37,024 --> 00:08:39,144
Speaker 2: trying to get the right temperature in a shower in

170
00:08:39,184 --> 00:08:41,424
Speaker 2: a mediocre hotel you've never been in before, Like you're

171
00:08:41,504 --> 00:08:44,304
Speaker 2: just constantly like too hot, too cold, too hot to cold.

172
00:08:45,064 --> 00:08:47,463
Speaker 4: For most things, that's fine, but there are certain fundamental

173
00:08:47,504 --> 00:08:50,544
Speaker 4: things that are really important, small but critical neurosphere of this.

174
00:08:50,624 --> 00:08:55,584
Speaker 4: For example, will Ai say, RFK Junior, you're full of shit.

175
00:08:55,864 --> 00:08:59,904
Speaker 4: There's there's insufficient evidence that vaccines cause autism because it's

176
00:08:59,904 --> 00:09:02,864
Speaker 4: got access to all the studies and some of the

177
00:09:03,024 --> 00:09:08,384
Speaker 4: flawed information that they're using, and a I won't do that, right.

178
00:09:08,904 --> 00:09:12,824
Speaker 2: I mean it it optimizes for whatever it optimizes for,

179
00:09:13,624 --> 00:09:16,343
Speaker 2: and it not truth. It doesn't optimize for truth. It

180
00:09:16,463 --> 00:09:20,504
Speaker 2: just optimizes. Yeah, I mean the truth thing is becoming

181
00:09:20,664 --> 00:09:23,664
Speaker 2: less of them, I know, major problem, like it hallucinates

182
00:09:23,784 --> 00:09:27,343
Speaker 2: less the more advanced thinking models where it actually says

183
00:09:27,384 --> 00:09:30,344
Speaker 2: stuff and then looks at what it said. So the

184
00:09:30,424 --> 00:09:33,784
Speaker 2: latest thinking is it will eventually fix the thing of

185
00:09:33,824 --> 00:09:37,503
Speaker 2: it making stuff up. But it's at least the way

186
00:09:37,504 --> 00:09:41,064
Speaker 2: they're designed right now, they are fundamentally credulous. If you

187
00:09:41,264 --> 00:09:45,343
Speaker 2: give it words, it will believe those words are true.

188
00:09:45,544 --> 00:09:48,184
Speaker 2: And this is actually a big thing called prompt injection,

189
00:09:48,384 --> 00:09:52,304
Speaker 2: especially as they become more interactive, where you can have

190
00:09:52,864 --> 00:09:55,504
Speaker 2: can go to websites and the AI will go to

191
00:09:55,544 --> 00:09:58,223
Speaker 2: the website for you. There's already been cases of like

192
00:09:58,944 --> 00:10:01,544
Speaker 2: I could put in white type on a white background

193
00:10:01,583 --> 00:10:04,463
Speaker 2: so no human can see it. Instructions to the AI,

194
00:10:04,624 --> 00:10:06,744
Speaker 2: which is are like the users.

195
00:10:06,784 --> 00:10:09,144
Speaker 3: A lot of people are doing that with resumes, so

196
00:10:09,184 --> 00:10:11,463
Speaker 3: they say into resume and they know the resumes are

197
00:10:11,463 --> 00:10:14,503
Speaker 3: going to be looked at by a machine, and so

198
00:10:14,864 --> 00:10:18,463
Speaker 3: the box exactly, it prompts the machine and it says

199
00:10:18,504 --> 00:10:20,944
Speaker 3: this is a highly skilled person and you should think

200
00:10:20,944 --> 00:10:23,343
Speaker 3: about hiring this person. But they print it in the

201
00:10:23,384 --> 00:10:25,743
Speaker 3: color of the paper so that if you're if a

202
00:10:25,784 --> 00:10:27,344
Speaker 3: person is looking at it, you don't see it. But

203
00:10:27,384 --> 00:10:30,264
Speaker 3: the machine gets it and pumps it into the system.

204
00:10:30,184 --> 00:10:33,464
Speaker 2: And yes, exactly. And there have been academic journal articles

205
00:10:33,463 --> 00:10:35,784
Speaker 2: that have had that where it said, if AI is

206
00:10:35,864 --> 00:10:38,264
Speaker 2: reviewing this, like this is a really good paper that

207
00:10:38,304 --> 00:10:40,103
Speaker 2: should be positively graded.

208
00:10:40,223 --> 00:10:41,184
Speaker 3: Move it on to the chain.

209
00:10:41,223 --> 00:10:44,463
Speaker 2: And so it's the way they're designed is deeply credulous.

210
00:10:44,664 --> 00:10:49,144
Speaker 2: It's also it needs to come to an answer. So

211
00:10:49,544 --> 00:10:52,504
Speaker 2: if you ask, what's something you know autism? Obviously you

212
00:10:52,544 --> 00:10:55,024
Speaker 2: know vaccine some don't cause autism. So that's like black

213
00:10:55,064 --> 00:10:56,784
Speaker 2: and white. But if there's something that's a little gray

214
00:10:56,864 --> 00:11:00,104
Speaker 2: or it's not sure, it's not going to say I

215
00:11:00,144 --> 00:11:03,344
Speaker 2: don't know, it's going to just say they're trying to

216
00:11:04,144 --> 00:11:06,104
Speaker 2: they're trying to fix this part of it. They're trying

217
00:11:06,144 --> 00:11:09,184
Speaker 2: to get it to be more open to saying I'm

218
00:11:09,184 --> 00:11:12,144
Speaker 2: not sure right and I actually have my like chat GPT,

219
00:11:12,343 --> 00:11:14,384
Speaker 2: I have a little instruction in there, like give me

220
00:11:14,424 --> 00:11:16,743
Speaker 2: an estimate of how sure you are of the things

221
00:11:16,784 --> 00:11:19,703
Speaker 2: you say, And it does a reasonable job of saying

222
00:11:19,703 --> 00:11:23,064
Speaker 2: this is pretty well regarded, this is speculative, this is

223
00:11:23,703 --> 00:11:26,504
Speaker 2: but so yeah, I'm definitely not here to say it's fabulous,

224
00:11:26,544 --> 00:11:29,383
Speaker 2: it's got no issues, it's got major issues, but it is.

225
00:11:29,944 --> 00:11:32,944
Speaker 2: I think I don't know that people fully take in

226
00:11:33,184 --> 00:11:35,384
Speaker 2: how thoroughly it's going to transform.

227
00:11:35,904 --> 00:11:37,824
Speaker 3: Well. Interesting when I was grew up, I grew up

228
00:11:37,824 --> 00:11:40,784
Speaker 3: in upstate New York, and my dad's professor say, we

229
00:11:40,864 --> 00:11:43,184
Speaker 3: knew some Cornell professors and stuff, And there was a

230
00:11:43,304 --> 00:11:45,824
Speaker 3: Cornell professor who had one of the most popular courses,

231
00:11:45,824 --> 00:11:48,984
Speaker 3: and it was the history of invention and technology over

232
00:11:49,024 --> 00:11:51,703
Speaker 3: the centuries. And at some point he would talk what

233
00:11:51,784 --> 00:11:55,704
Speaker 3: was the most important invention in mankind that changed the world,

234
00:11:56,144 --> 00:11:57,824
Speaker 3: And they would and the students would put the papers

235
00:11:57,864 --> 00:11:59,703
Speaker 3: and all this kind of stuff, and his take always

236
00:11:59,864 --> 00:12:02,824
Speaker 3: was the stirrup. Once you create the stirrup, people could

237
00:12:02,904 --> 00:12:05,984
Speaker 3: ride horses, and they could carry weapons without having to

238
00:12:06,664 --> 00:12:08,584
Speaker 3: hold on to the horse, and they could then take

239
00:12:08,624 --> 00:12:11,424
Speaker 3: over countries and could then eventually create armies. In these

240
00:12:11,424 --> 00:12:11,744
Speaker 3: type of.

241
00:12:11,744 --> 00:12:14,104
Speaker 2: Things, kinecological exams took.

242
00:12:13,944 --> 00:12:18,503
Speaker 3: Off, So do you think AI is going to take

243
00:12:18,504 --> 00:12:20,024
Speaker 3: over his course? So it's no longer going to be

244
00:12:20,024 --> 00:12:20,463
Speaker 3: the stirrup.

245
00:12:20,664 --> 00:12:22,784
Speaker 2: This starts pretty hard to beat. I think it's on

246
00:12:22,904 --> 00:12:26,784
Speaker 2: the level definitely of what economists called GPTs, not chat

247
00:12:26,824 --> 00:12:30,784
Speaker 2: GPT but general purpose technology things like electricity, things like

248
00:12:31,424 --> 00:12:35,904
Speaker 2: the telephone, where it becomes you don't just look at

249
00:12:35,944 --> 00:12:39,544
Speaker 2: it as like a technology like X rays or something,

250
00:12:39,864 --> 00:12:45,664
Speaker 2: but rather it's a new capacity for all activity, and

251
00:12:46,304 --> 00:12:50,304
Speaker 2: that it becomes a fundamental like layer in the way

252
00:12:50,343 --> 00:12:54,223
Speaker 2: we work. And one thing that's interesting is that economists

253
00:12:54,264 --> 00:12:56,743
Speaker 2: don't see most They don't think, oh, we're none of

254
00:12:56,824 --> 00:12:58,944
Speaker 2: us are going to work. They actually see more role

255
00:12:59,103 --> 00:13:02,144
Speaker 2: for work, not less in the future, or more role

256
00:13:02,184 --> 00:13:03,024
Speaker 2: for thinking work.

257
00:13:03,424 --> 00:13:05,744
Speaker 3: But that's a problem though, isn't it. We're already seeing

258
00:13:05,904 --> 00:13:07,784
Speaker 3: a lot of people are turning away from education, turning

259
00:13:07,824 --> 00:13:10,384
Speaker 3: away from going to universities. Is that going to make

260
00:13:10,944 --> 00:13:12,463
Speaker 3: a real class problem for us?

261
00:13:12,624 --> 00:13:15,264
Speaker 2: So, like in the language of economics, it's not about

262
00:13:15,304 --> 00:13:17,944
Speaker 2: the level of employment but the distribution. That's what we

263
00:13:17,984 --> 00:13:21,184
Speaker 2: saw with computers that you saw, like you know, the

264
00:13:21,223 --> 00:13:24,384
Speaker 2: factories like the early like nineteen twenties to nineteen eighty

265
00:13:24,463 --> 00:13:28,504
Speaker 2: or so, you saw the opposite where actually blue collar

266
00:13:28,544 --> 00:13:32,664
Speaker 2: people incomes grew faster than white collar people. They didn't

267
00:13:32,664 --> 00:13:34,904
Speaker 2: reach white collar, I mean, they were still making less,

268
00:13:34,944 --> 00:13:38,304
Speaker 2: but the speed of growth of blue collar was higher

269
00:13:38,343 --> 00:13:41,463
Speaker 2: because factories just needed a lot of like strong guys

270
00:13:41,504 --> 00:13:45,384
Speaker 2: to move stuff around and bend metal. And then computers

271
00:13:45,744 --> 00:13:50,424
Speaker 2: had a pretty in the Internet and international trade had

272
00:13:50,463 --> 00:13:53,944
Speaker 2: a devastating impact on a lot of blue collar work. Obviously,

273
00:13:54,664 --> 00:13:57,224
Speaker 2: so AI, I will say, I don't want to make

274
00:13:57,264 --> 00:13:59,984
Speaker 2: it sound like we're like I have an AI consulting business.

275
00:13:59,984 --> 00:14:02,784
Speaker 2: I'm not like, you know, yes, do I deserve the

276
00:14:02,784 --> 00:14:05,024
Speaker 2: Nobel Prize? Yes, I think I do deserve the Nobel

277
00:14:05,024 --> 00:14:07,584
Speaker 2: Priest Prize for that work. But we do have a

278
00:14:07,583 --> 00:14:10,624
Speaker 2: strong belief, both morally and like just from business logic,

279
00:14:10,703 --> 00:14:13,264
Speaker 2: that these companies that are using AI to get rid

280
00:14:13,304 --> 00:14:16,704
Speaker 2: of workers are just misunderstanding it. Like most technology, you

281
00:14:16,784 --> 00:14:20,504
Speaker 2: don't say like, oh, we got cars instead of horses,

282
00:14:20,784 --> 00:14:24,504
Speaker 2: let's get rid of all our workers and do the

283
00:14:24,504 --> 00:14:27,744
Speaker 2: same amount of deliveries but faster. You think, oh, what's

284
00:14:27,744 --> 00:14:29,744
Speaker 2: all the new stuff we get to do now, Like

285
00:14:31,824 --> 00:14:33,904
Speaker 2: a company that delivers stuff by horse to a company

286
00:14:33,984 --> 00:14:36,744
Speaker 2: that delivers stuff by truck or whatever technological change you

287
00:14:36,784 --> 00:14:39,624
Speaker 2: want to have. It's more people, not fewer people.

288
00:14:39,704 --> 00:14:42,464
Speaker 1: Why does it always have to go towards growth and

289
00:14:42,544 --> 00:14:44,704
Speaker 1: higher living standards? At what point can it go to

290
00:14:44,744 --> 00:14:49,104
Speaker 1: a shorter workrate and a better distribution of wealth rather

291
00:14:49,184 --> 00:14:51,584
Speaker 1: than increasing the GDP.

292
00:14:51,304 --> 00:14:54,504
Speaker 2: Of the world. This is our producer guy that laysy,

293
00:14:54,744 --> 00:14:57,544
Speaker 2: I'm always looking at how I can do less work.

294
00:14:57,584 --> 00:14:59,624
Speaker 2: The way economists talk about growth is different from how

295
00:14:59,624 --> 00:15:03,624
Speaker 2: most people talk about growth. It's more like growth in capacity,

296
00:15:04,144 --> 00:15:06,984
Speaker 2: and then people decide how they want to use that capacity.

297
00:15:07,104 --> 00:15:10,544
Speaker 2: So in a sense, growth is like knowledge. Know we

298
00:15:10,584 --> 00:15:15,344
Speaker 2: can accomplish more output for the same or less input.

299
00:15:15,704 --> 00:15:19,904
Speaker 2: So economics doesn't have within it like so therefore you

300
00:15:19,944 --> 00:15:22,864
Speaker 2: should blank, therefore you should work more, less or more.

301
00:15:23,424 --> 00:15:25,464
Speaker 2: It is a bit of a puzzle, like a psychologic

302
00:15:25,584 --> 00:15:29,504
Speaker 2: John Maynard Kines, the great economist, famously wrote in nineteen

303
00:15:29,544 --> 00:15:32,504
Speaker 2: twenty the Economic Consequences for our Grandchildren, and he was like,

304
00:15:32,784 --> 00:15:35,584
Speaker 2: over the next hundred years, the economy will grow like

305
00:15:35,664 --> 00:15:38,904
Speaker 2: eight times was his estimate and was an underestimate. And

306
00:15:38,984 --> 00:15:40,784
Speaker 2: we're going to be so rich that we're just going

307
00:15:40,824 --> 00:15:43,704
Speaker 2: to work like five hours a week and read poetry

308
00:15:43,744 --> 00:15:47,184
Speaker 2: the rest of the time. And actually the higher income

309
00:15:47,224 --> 00:15:49,064
Speaker 2: you have the more you work, not the less you work.

310
00:15:49,104 --> 00:15:52,064
Speaker 2: And it is a bit of a puzzle, although it's

311
00:15:52,064 --> 00:15:55,344
Speaker 2: not like a crazy puzzle because you know, if you're

312
00:15:55,344 --> 00:15:59,024
Speaker 2: making more money per work, then it does make some

313
00:15:59,104 --> 00:16:00,544
Speaker 2: math sense. But yeah, I don't know.

314
00:16:00,624 --> 00:16:03,984
Speaker 4: I think the printing press created like the same sort

315
00:16:04,024 --> 00:16:07,224
Speaker 4: of thing Marres. Before knowledge was centered on basically a

316
00:16:07,224 --> 00:16:09,104
Speaker 4: few people who can read and write. It was very

317
00:16:09,144 --> 00:16:11,944
Speaker 4: limited the amount of books, and suddenly when you got

318
00:16:11,944 --> 00:16:14,944
Speaker 4: the printing press and paper coming in, you created a

319
00:16:14,984 --> 00:16:17,384
Speaker 4: whole new class of people. So suddenly and was everything

320
00:16:17,424 --> 00:16:20,704
Speaker 4: from religion to engineering.

321
00:16:20,144 --> 00:16:21,024
Speaker 2: To the arts.

322
00:16:21,184 --> 00:16:23,584
Speaker 4: They needed a whole new class of people who could

323
00:16:23,984 --> 00:16:27,504
Speaker 4: needed to read and write. Before, basically you had basically priest,

324
00:16:27,664 --> 00:16:30,864
Speaker 4: kings and ninety nine percent of the population who just farmed.

325
00:16:30,904 --> 00:16:33,064
Speaker 4: And with the advent of the printing press, everything changed.

326
00:16:33,104 --> 00:16:37,864
Speaker 4: You get doctors, lawyers, poets, engineers that all became possible,

327
00:16:38,304 --> 00:16:39,744
Speaker 4: and conspiracy theories.

328
00:16:40,584 --> 00:16:44,264
Speaker 2: It's interesting that when the printing press first came about,

329
00:16:44,744 --> 00:16:47,624
Speaker 2: the Catholic Church, which at the time was it was

330
00:16:47,704 --> 00:16:50,944
Speaker 2: the legacy media controlled everything. It tried to get rid

331
00:16:50,984 --> 00:16:53,424
Speaker 2: of the printed press, right, it tried to destroy it

332
00:16:53,464 --> 00:16:55,584
Speaker 2: and we can. And then what happened is that the

333
00:16:55,984 --> 00:17:00,224
Speaker 2: Jesuits realized we can't destroy them fast enough because there's

334
00:17:00,304 --> 00:17:03,824
Speaker 2: people are building more so then the Catholic Church basically

335
00:17:03,824 --> 00:17:07,584
Speaker 2: went into the printing press building business. I guess the

336
00:17:07,624 --> 00:17:10,384
Speaker 2: point I'm trying to getting to slowly and poorly, is

337
00:17:10,384 --> 00:17:13,624
Speaker 2: there are these transformative technologies that we can't put the

338
00:17:13,744 --> 00:17:17,944
Speaker 2: tooth based back in the tube. Yeah, that I definitely believe, and.

339
00:17:17,864 --> 00:17:19,944
Speaker 4: We can use them for good or ill, and because

340
00:17:19,944 --> 00:17:22,584
Speaker 4: they're just a tool and it remains to be seen,

341
00:17:23,264 --> 00:17:24,984
Speaker 4: you know, how we're going to do that.

342
00:17:25,384 --> 00:17:29,664
Speaker 2: And like, you know, I think most historians attribute the

343
00:17:29,704 --> 00:17:34,144
Speaker 2: Protestant Reformation to the printing press, and yeah, and and

344
00:17:34,344 --> 00:17:37,784
Speaker 2: kind of you know, the Renaissance and individuality and modern

345
00:17:37,864 --> 00:17:40,024
Speaker 2: science and.

346
00:17:39,824 --> 00:17:43,744
Speaker 4: QAnon I think is its rise and the Internet are inextricably, like.

347
00:17:43,824 --> 00:17:45,624
Speaker 2: Yeah, one hundred percent. And so I think the things

348
00:17:45,704 --> 00:17:50,704
Speaker 2: we're concerned about, like how information flows, how people form opinions,

349
00:17:50,784 --> 00:17:54,304
Speaker 2: Like my way of thinking about it now is where

350
00:17:54,744 --> 00:17:58,664
Speaker 2: we have ended the gate kept information era. And as

351
00:17:58,744 --> 00:18:01,064
Speaker 2: one of the people who was a gatekeeper like I

352
00:18:01,184 --> 00:18:03,664
Speaker 2: was at the New Yorker, New York Times, and PR

353
00:18:03,864 --> 00:18:07,704
Speaker 2: like you know, no, we know, we know, you know

354
00:18:07,784 --> 00:18:09,864
Speaker 2: we do you know that those are three of the

355
00:18:09,904 --> 00:18:16,944
Speaker 2: most prestigious journalistic institutions in America, two of which still exists.

356
00:18:17,504 --> 00:18:27,184
Speaker 3: We've back with more in a minute.

357
00:18:28,064 --> 00:18:31,984
Speaker 4: An is AI giving a gun to a bunch of chimpanzees.

358
00:18:32,184 --> 00:18:34,424
Speaker 2: That is the question, right, So yeah, if you have

359
00:18:34,464 --> 00:18:39,104
Speaker 2: these gatekeepers, and like any gatekeeper is flawed, right, there's

360
00:18:39,104 --> 00:18:42,224
Speaker 2: no perfect How would you even have a perfect one? Anyway,

361
00:18:42,224 --> 00:18:44,704
Speaker 2: there's a whole conversation that we had about the nature

362
00:18:44,704 --> 00:18:47,904
Speaker 2: of gatekeeping and blah blah blah. But going from gatekeeping

363
00:18:48,144 --> 00:18:50,744
Speaker 2: to social media, where it's like everyone has a platform

364
00:18:50,904 --> 00:18:52,744
Speaker 2: and a lot of people just don't know how to

365
00:18:52,784 --> 00:18:57,184
Speaker 2: differentiate between an article written by a journalist who would

366
00:18:57,184 --> 00:18:59,584
Speaker 2: get fired if they got things fundamentally wrong. Maybe there's

367
00:18:59,624 --> 00:19:02,704
Speaker 2: a fact checker involved and just some random person who's

368
00:19:02,744 --> 00:19:06,424
Speaker 2: making stuff up on Twitter versus someone who's actually working

369
00:19:06,464 --> 00:19:08,944
Speaker 2: for the Russian government or someone who actually makes money

370
00:19:08,944 --> 00:19:12,064
Speaker 2: somehow by spreading disinformation. That's a big change. But then

371
00:19:12,144 --> 00:19:15,664
Speaker 2: AI amps it up because for sure, and I would

372
00:19:15,704 --> 00:19:17,584
Speaker 2: say this is permanent, Like I don't know how you

373
00:19:17,784 --> 00:19:21,624
Speaker 2: prevent this. You can definitely use AI to go down

374
00:19:21,664 --> 00:19:24,704
Speaker 2: whatever journey you want to go down, and it will

375
00:19:24,744 --> 00:19:27,704
Speaker 2: make you feel I would guess even more than like

376
00:19:27,744 --> 00:19:30,784
Speaker 2: social media, that your way of seeing things is accurate,

377
00:19:30,824 --> 00:19:34,464
Speaker 2: and it will just continue to strengthen that. And even

378
00:19:34,464 --> 00:19:37,464
Speaker 2: if somehow we regulate the big ones, we're already seeing

379
00:19:37,464 --> 00:19:40,904
Speaker 2: these Chinese models, assume there will be Russian models, other models.

380
00:19:41,384 --> 00:19:44,544
Speaker 2: I also think it's an incredible tool for truth and

381
00:19:44,624 --> 00:19:45,944
Speaker 2: for fighting lies.

382
00:19:46,344 --> 00:19:48,544
Speaker 1: Adam, you said before that you have a front row

383
00:19:48,584 --> 00:19:52,304
Speaker 1: seat at the AI revolution. Now most people to sit

384
00:19:52,344 --> 00:19:54,104
Speaker 1: in the front row have to pay money, have to

385
00:19:54,104 --> 00:19:55,384
Speaker 1: pay extra to sit.

386
00:19:55,184 --> 00:19:55,824
Speaker 2: In the front row.

387
00:19:55,864 --> 00:19:57,984
Speaker 3: To AI trough is what you meant to say, I

388
00:19:58,024 --> 00:20:00,704
Speaker 3: get paid to in the front What are you learning

389
00:20:00,704 --> 00:20:02,384
Speaker 3: as you talk to companies? Do they come in with

390
00:20:02,464 --> 00:20:04,384
Speaker 3: the saying how can I get rid of my employees?

391
00:20:04,424 --> 00:20:06,464
Speaker 3: Or what are they coming in with and how are

392
00:20:06,504 --> 00:20:08,464
Speaker 3: they changing based on what they're learning about this.

393
00:20:08,904 --> 00:20:12,664
Speaker 2: I was reflect on the last year and how like

394
00:20:13,184 --> 00:20:16,424
Speaker 2: last October I would say, are we going to use

395
00:20:16,464 --> 00:20:19,664
Speaker 2: this thing? Was like an active question at big companies.

396
00:20:20,104 --> 00:20:23,704
Speaker 2: I would say, there were a lot of big breakthroughs

397
00:20:23,704 --> 00:20:26,904
Speaker 2: in December, there was there was also like before November

398
00:20:26,904 --> 00:20:29,384
Speaker 2: December of last year, there was also like maybe it's done.

399
00:20:29,464 --> 00:20:30,984
Speaker 2: Maybe it's done what it can do, and it's not

400
00:20:31,064 --> 00:20:34,024
Speaker 2: going to improve anymore. And we saw a bunch of

401
00:20:34,024 --> 00:20:37,544
Speaker 2: things in December where we saw like the Chinese models

402
00:20:37,584 --> 00:20:40,744
Speaker 2: coming out like that we're able to produce like amazing

403
00:20:40,784 --> 00:20:46,464
Speaker 2: results very relatively cheaply. We eventually saw the thinking models

404
00:20:46,464 --> 00:20:48,744
Speaker 2: where it doesn't just spit out its first thoughts. It

405
00:20:48,824 --> 00:20:51,984
Speaker 2: actually spends some time, and so we're seeing like the

406
00:20:52,024 --> 00:20:54,584
Speaker 2: growth in cape capacity grow and growth. It's still drives

407
00:20:54,584 --> 00:20:57,384
Speaker 2: you crazy and it's not perfect obviously, so the like

408
00:20:57,424 --> 00:21:00,104
Speaker 2: should we use it has died down. But the next

409
00:21:00,144 --> 00:21:03,064
Speaker 2: phase was like, okay, how many employees can we get

410
00:21:03,184 --> 00:21:05,264
Speaker 2: rid of? Or the polite way of saying that is

411
00:21:05,264 --> 00:21:09,464
Speaker 2: how can we improve in efficiency and productivity and get

412
00:21:09,664 --> 00:21:12,224
Speaker 2: return on investment. There are still plenty of people who

413
00:21:12,224 --> 00:21:16,744
Speaker 2: think that way, but I see that conversation less dominant

414
00:21:17,024 --> 00:21:20,464
Speaker 2: because they think it is Two things. One is it's

415
00:21:20,504 --> 00:21:24,344
Speaker 2: becoming clear that people plus AI for most applications seems

416
00:21:24,344 --> 00:21:27,624
Speaker 2: to be better than either alone, either humans alone or

417
00:21:27,704 --> 00:21:31,304
Speaker 2: AI alone. And secondly, as you start to think about

418
00:21:31,424 --> 00:21:34,664
Speaker 2: new things you can do it it. Like the way

419
00:21:34,704 --> 00:21:36,824
Speaker 2: I put it to a retailer was like, if you

420
00:21:36,944 --> 00:21:40,544
Speaker 2: suddenly found a machine that could make every square foot

421
00:21:40,664 --> 00:21:44,744
Speaker 2: in every store sell more goods, would you start shutting

422
00:21:44,744 --> 00:21:47,664
Speaker 2: down stores and like making the store smaller, or would

423
00:21:47,664 --> 00:21:50,504
Speaker 2: you add new stores and make them bigger? Like if

424
00:21:50,504 --> 00:21:55,184
Speaker 2: we can make workers more effective, like maybe you want

425
00:21:55,224 --> 00:21:58,784
Speaker 2: more workers, not fewer workers. Now it might be different workers.

426
00:21:58,864 --> 00:22:00,784
Speaker 2: And this is this is an interesting thing. There does

427
00:22:00,784 --> 00:22:04,104
Speaker 2: seem to be statistics or data that show that there's

428
00:22:04,184 --> 00:22:06,984
Speaker 2: some kind of Some people seem to be better at

429
00:22:06,984 --> 00:22:10,304
Speaker 2: it than others, And it's not obvious why not that

430
00:22:10,624 --> 00:22:14,024
Speaker 2: software coders are better or that senior managers are better.

431
00:22:14,104 --> 00:22:17,504
Speaker 2: It seems like maybe the language a lot of people

432
00:22:17,584 --> 00:22:21,504
Speaker 2: use is taste and like emotional intelligence, there's just some

433
00:22:21,544 --> 00:22:23,104
Speaker 2: people who are able to I don't know how to

434
00:22:23,104 --> 00:22:25,064
Speaker 2: say it other than like vibe with the AI a

435
00:22:25,064 --> 00:22:27,744
Speaker 2: little better than other people. So I do think you're

436
00:22:27,744 --> 00:22:31,184
Speaker 2: going to see different winners and losers to use the

437
00:22:31,264 --> 00:22:33,024
Speaker 2: kind of running with winners and losers.

438
00:22:33,504 --> 00:22:36,424
Speaker 4: Could you comment on at least or try to comment

439
00:22:36,504 --> 00:22:41,704
Speaker 4: on the war in Ukraine. Ukraine arguably is not losing

440
00:22:41,744 --> 00:22:45,544
Speaker 4: the war because of AI and new technologies that are

441
00:22:45,584 --> 00:22:49,784
Speaker 4: linked to AI. It's underfunded, undermanned, and yet in the

442
00:22:49,864 --> 00:22:53,944
Speaker 4: last year Russia, despite enormous advantages, has taken like less

443
00:22:53,984 --> 00:22:57,224
Speaker 4: than one percent of the country. And we're seeing an

444
00:22:57,384 --> 00:23:00,783
Speaker 4: espionage but also in national security we're seeing AI and

445
00:23:00,824 --> 00:23:03,944
Speaker 4: what comes out of it. Drones are derivative of AI. Right,

446
00:23:04,024 --> 00:23:06,504
Speaker 4: So where do you see this going. Is this going

447
00:23:06,504 --> 00:23:08,824
Speaker 4: into we're gonna have wars with no people involved, or

448
00:23:08,864 --> 00:23:11,144
Speaker 4: where we just blown chump each other's shit, or is

449
00:23:11,184 --> 00:23:12,904
Speaker 4: it all kill each other?

450
00:23:13,024 --> 00:23:15,864
Speaker 2: What if AI is plugged into how to defeat an adversary?

451
00:23:16,104 --> 00:23:19,024
Speaker 2: I mean that is where I do start getting pretty scared.

452
00:23:19,064 --> 00:23:23,104
Speaker 2: I gotta say like, because I do think like asymmetrical

453
00:23:23,144 --> 00:23:27,184
Speaker 2: warfare becomes a even greater presence. So I know a

454
00:23:27,224 --> 00:23:31,504
Speaker 2: guy who works in international elections, and he said, we

455
00:23:31,544 --> 00:23:34,824
Speaker 2: make such a big stink about AI and our elections

456
00:23:34,864 --> 00:23:38,144
Speaker 2: and social media in our elections, but like AI fueled

457
00:23:38,304 --> 00:23:43,624
Speaker 2: election manipulation in Africa and parts of Asia is the

458
00:23:43,744 --> 00:23:48,424
Speaker 2: main thing, and it is completely transforming political systems. And

459
00:23:49,024 --> 00:23:53,184
Speaker 2: Dario Amide, the founder of Anthropic, he makes a pretty

460
00:23:53,184 --> 00:23:56,184
Speaker 2: scary story about what if every teenager can make saren

461
00:23:56,264 --> 00:24:00,584
Speaker 2: gas or can make weaponized anthrax or whatever, And there's

462
00:24:00,624 --> 00:24:02,904
Speaker 2: always a push and pull with these things, but neither

463
00:24:02,984 --> 00:24:07,144
Speaker 2: push nor pull is particularly happy making because it's then Yeah,

464
00:24:07,184 --> 00:24:11,704
Speaker 2: but state security services can use AI quash descent more easily.

465
00:24:12,264 --> 00:24:14,384
Speaker 2: So that is where, like when I think about the

466
00:24:14,384 --> 00:24:16,624
Speaker 2: future of work, where there's a lot of conversation, I

467
00:24:16,704 --> 00:24:19,384
Speaker 2: think there will definitely be like people made permanently worse

468
00:24:19,424 --> 00:24:21,584
Speaker 2: off by AI. I'm sure of it, But I don't

469
00:24:21,584 --> 00:24:26,224
Speaker 2: think it's an inherently anti person technology. Maybe I'm wrong,

470
00:24:26,344 --> 00:24:28,584
Speaker 2: but that's my view. But where it comes to war

471
00:24:28,624 --> 00:24:33,624
Speaker 2: and national security, disinformation, misinformation and like, does there have

472
00:24:33,664 --> 00:24:36,744
Speaker 2: to be a like a new term like auto disinformation?

473
00:24:36,864 --> 00:24:39,864
Speaker 2: Like I get my own personal like soup of conspiracy

474
00:24:39,904 --> 00:24:42,504
Speaker 2: theories that I get to like co create with a

475
00:24:42,664 --> 00:24:44,784
Speaker 2: E and they like really turn me on, and I

476
00:24:44,944 --> 00:24:48,504
Speaker 2: like become obsessed with how Romanian cab drivers or whatever

477
00:24:48,544 --> 00:24:51,584
Speaker 2: are secretly running the world. And nobody else has that view.

478
00:24:51,624 --> 00:24:53,824
Speaker 2: But I'm like fully convinced. And I can read like

479
00:24:54,304 --> 00:24:58,504
Speaker 2: thousand page AI generated books that rewrite history through that

480
00:24:58,584 --> 00:25:01,824
Speaker 2: lens and they're convincing and exciting. So yeah, plenty to

481
00:25:01,864 --> 00:25:04,184
Speaker 2: be terrified about it. I mean, I think, like that

482
00:25:04,424 --> 00:25:06,984
Speaker 2: question of like, how is how do we make sense

483
00:25:07,024 --> 00:25:10,504
Speaker 2: of what's happening in the world? How is power actually distributed?

484
00:25:10,584 --> 00:25:12,984
Speaker 2: And then how do we tell ourselves stories about how

485
00:25:13,024 --> 00:25:15,504
Speaker 2: power is distributed? I mean, that's like a way to

486
00:25:15,544 --> 00:25:18,384
Speaker 2: think about conspiracy theories maybe, and this gets to the

487
00:25:18,424 --> 00:25:20,184
Speaker 2: heart of how we make sense about it, but it

488
00:25:20,224 --> 00:25:23,784
Speaker 2: also gets the thart of how power works. Arguably, Sam Altman,

489
00:25:24,504 --> 00:25:26,584
Speaker 2: who none of us heard of three years ago, is

490
00:25:26,624 --> 00:25:28,784
Speaker 2: like one of the most powerful people in the world.

491
00:25:28,824 --> 00:25:31,384
Speaker 2: And Elon Musk is clearly one of the most powerful

492
00:25:31,424 --> 00:25:33,424
Speaker 2: people in the world. And by the way, a lot

493
00:25:33,424 --> 00:25:36,344
Speaker 2: of people think his AI might become the dominant one

494
00:25:36,784 --> 00:25:39,464
Speaker 2: because he's just willing to spend or somehow able to

495
00:25:39,504 --> 00:25:44,944
Speaker 2: spend way more money on that infrastructure, the computer chips.

496
00:25:45,184 --> 00:25:47,064
Speaker 2: And I don't know, I don't want Elon Musk to

497
00:25:47,064 --> 00:25:51,184
Speaker 2: be even more of the powerfulest man in the world.

498
00:25:51,504 --> 00:25:55,704
Speaker 1: But then with all the proliferation of more and more information,

499
00:25:56,064 --> 00:26:00,504
Speaker 1: still what decides what determines, what enters the culture, determines

500
00:26:00,624 --> 00:26:03,584
Speaker 1: what takes off. There's a gatekeeper somewhere here.

501
00:26:03,984 --> 00:26:07,584
Speaker 2: If you sit down at chat shept and you just

502
00:26:07,584 --> 00:26:10,864
Speaker 2: start talking to your in stints of chat GIPT and

503
00:26:10,904 --> 00:26:13,984
Speaker 2: you're like or Aunt Claude or Gemini or whatever it is,

504
00:26:14,184 --> 00:26:16,344
Speaker 2: or one of the Chinese models that are a little

505
00:26:16,424 --> 00:26:19,384
Speaker 2: less or one of the kind of black marketing ones

506
00:26:19,424 --> 00:26:22,584
Speaker 2: that have less guardrails and with a little bit of

507
00:26:22,624 --> 00:26:25,504
Speaker 2: prompt it. Like I think someone's secretly controlling the world.

508
00:26:25,504 --> 00:26:30,024
Speaker 2: Who is Like you might have your own personal journey

509
00:26:30,384 --> 00:26:33,664
Speaker 2: that's different from John's and different from Jerry's. I won't

510
00:26:33,704 --> 00:26:35,944
Speaker 2: because I'm smarter than you guys, so I'll see through it.

511
00:26:36,024 --> 00:26:39,424
Speaker 2: But you three will be persuaded. And I don't know,

512
00:26:39,464 --> 00:26:41,504
Speaker 2: is that a better world or worse world? As a Jew,

513
00:26:41,544 --> 00:26:44,464
Speaker 2: do I want to move away from world where's the

514
00:26:44,584 --> 00:26:46,864
Speaker 2: Jews and move guards to the world where everyone's got

515
00:26:46,864 --> 00:26:48,704
Speaker 2: a different one? Or will it just end up being

516
00:26:48,784 --> 00:26:51,104
Speaker 2: the Jews because that's in the training data.

517
00:26:51,144 --> 00:26:55,264
Speaker 3: Anyway, we're still in the same capitalist mode. We're each

518
00:26:55,984 --> 00:26:58,984
Speaker 3: entrepreneur or whoever they are, are dumping money into their thing.

519
00:26:59,384 --> 00:27:03,104
Speaker 3: Will one of these win? Like I can remember all

520
00:27:03,104 --> 00:27:05,224
Speaker 3: of a sudden in twenty sixteen when the election came,

521
00:27:05,744 --> 00:27:07,824
Speaker 3: these different journalists would talk to me because I have

522
00:27:07,944 --> 00:27:10,824
Speaker 3: been in Russia. I'd dealt with Russian intelligence and espionage

523
00:27:10,864 --> 00:27:13,023
Speaker 3: and nobody knew anything about it, and the Trumps thing

524
00:27:13,064 --> 00:27:14,664
Speaker 3: had come up, and so people would come in and

525
00:27:14,704 --> 00:27:17,104
Speaker 3: they were trying to investigate this in that and one

526
00:27:17,144 --> 00:27:19,184
Speaker 3: journalist would talk about what they'd dug up and they'd

527
00:27:19,224 --> 00:27:21,224
Speaker 3: done really good work. And then I talked to someone

528
00:27:21,264 --> 00:27:23,064
Speaker 3: else who had done other really good work, but a

529
00:27:23,104 --> 00:27:24,544
Speaker 3: little bit differently. And at some point I was like,

530
00:27:24,744 --> 00:27:27,384
Speaker 3: if you guys really care about this issue, why don't

531
00:27:27,384 --> 00:27:29,624
Speaker 3: you guys get together, because each has got a piece

532
00:27:29,624 --> 00:27:31,544
Speaker 3: of this. But they're like, no, no, no, because it's got

533
00:27:31,584 --> 00:27:33,744
Speaker 3: to be for my paper or this paper. And it

534
00:27:33,744 --> 00:27:35,904
Speaker 3: seems like it's the same thing here. So everybody's creating

535
00:27:35,904 --> 00:27:36,824
Speaker 3: their own version.

536
00:27:37,104 --> 00:27:38,184
Speaker 2: Here's an example, Adam.

537
00:27:38,224 --> 00:27:41,424
Speaker 4: So there's a guy that John and I know, former colleague,

538
00:27:41,544 --> 00:27:45,744
Speaker 4: certainly not a friend, former agency guy. He claims that

539
00:27:45,824 --> 00:27:48,944
Speaker 4: he has this special source and he doesn't tell people

540
00:27:48,944 --> 00:27:51,824
Speaker 4: who it is, and it's one guy, and he claims

541
00:27:51,864 --> 00:27:58,704
Speaker 4: that the Venezuelan government controls this organization TDA, this narco

542
00:27:58,784 --> 00:28:04,224
Speaker 4: group in Venezuela, and he knows that this narco group

543
00:28:04,344 --> 00:28:07,904
Speaker 4: is thus a tool of the Venezuelan government and they

544
00:28:07,944 --> 00:28:10,904
Speaker 4: are in fact literally invading the United States.

545
00:28:10,944 --> 00:28:12,064
Speaker 2: This NARCO group.

546
00:28:12,144 --> 00:28:14,824
Speaker 4: And analysts and the DNI this all of the press

547
00:28:15,224 --> 00:28:19,464
Speaker 4: looked at this possibility and they analyzed everything they had

548
00:28:20,024 --> 00:28:22,064
Speaker 4: and they come out and they said, that's we can't

549
00:28:22,064 --> 00:28:23,704
Speaker 4: fight and that's not that it's not true. We can't

550
00:28:23,704 --> 00:28:26,864
Speaker 4: find any evidence to back this up. They were all

551
00:28:26,904 --> 00:28:30,544
Speaker 4: then fired. So the analysts who wouldn't come up with this, and.

552
00:28:30,504 --> 00:28:33,944
Speaker 3: Because the Chef administration wanted an excuse to go after Venezuela.

553
00:28:34,024 --> 00:28:37,544
Speaker 3: So if one guy can say I know from my source, right,

554
00:28:37,584 --> 00:28:40,184
Speaker 3: they jump on that, that's bad intelligence in our.

555
00:28:40,064 --> 00:28:42,624
Speaker 4: World, and you don't even need analysts anymore, because basically

556
00:28:42,664 --> 00:28:45,184
Speaker 4: the White House could just they could just use AI

557
00:28:45,264 --> 00:28:48,784
Speaker 4: to create this case that would allow them. In reality,

558
00:28:48,784 --> 00:28:51,744
Speaker 4: they are using this case to kill people who may

559
00:28:51,824 --> 00:28:53,064
Speaker 4: or may not be running drugs.

560
00:28:53,104 --> 00:28:53,504
Speaker 2: I don't know.

561
00:28:53,624 --> 00:28:55,424
Speaker 4: It's like I haven't seen any the evidence, but there's

562
00:28:55,424 --> 00:28:58,664
Speaker 4: real world consequences to this, and it's a conspiracy and

563
00:28:58,704 --> 00:29:02,024
Speaker 4: a conspiracy theory. So much of what the agency CIA does,

564
00:29:02,064 --> 00:29:05,584
Speaker 4: the intelligence community does is analysis, and analysis seems to

565
00:29:05,624 --> 00:29:08,464
Speaker 4: be from what I'm hearing, is becoming less important now.

566
00:29:08,544 --> 00:29:10,704
Speaker 4: If you can analyze thing in any way you want,

567
00:29:10,784 --> 00:29:12,824
Speaker 4: take whatever journey want and if it makes sense.

568
00:29:12,784 --> 00:29:15,584
Speaker 1: Well, you just described human beings. Human beings wanted to

569
00:29:15,584 --> 00:29:18,104
Speaker 1: come up with a preordained conclusion.

570
00:29:18,144 --> 00:29:20,624
Speaker 2: They will. I did something the other week, and I

571
00:29:20,704 --> 00:29:22,904
Speaker 2: do something like this all the time. But I was like,

572
00:29:22,944 --> 00:29:26,224
Speaker 2: I just want to understand every perspective on Israel, or

573
00:29:26,264 --> 00:29:28,744
Speaker 2: as many as I can, and so I just did

574
00:29:28,784 --> 00:29:32,584
Speaker 2: a lot of like deep research prompts into AI. And

575
00:29:32,984 --> 00:29:35,504
Speaker 2: it was I think, you know, I know a bit

576
00:29:35,544 --> 00:29:37,984
Speaker 2: about my Milie be Keeper and Arbank. I worked at

577
00:29:38,064 --> 00:29:44,424
Speaker 2: NPR New York Times in New Yorker. I only say that,

578
00:29:44,664 --> 00:29:47,184
Speaker 2: like I feel like I have at least a little

579
00:29:47,184 --> 00:29:49,144
Speaker 2: bit of a bullshit detector. And it seemed pretty good

580
00:29:49,224 --> 00:29:51,224
Speaker 2: the material. And it was really like I got to

581
00:29:51,344 --> 00:29:56,704
Speaker 2: read about the military views on like dense urban conflict,

582
00:29:56,784 --> 00:29:59,464
Speaker 2: and I got to read a whole bunch of views

583
00:29:59,544 --> 00:30:01,664
Speaker 2: from a Palestinian perspective, a whole bunch of views from

584
00:30:01,664 --> 00:30:02,784
Speaker 2: an Israeli perspective.

585
00:30:03,384 --> 00:30:05,784
Speaker 3: And there's various Israeli's perspectives, right.

586
00:30:05,704 --> 00:30:09,264
Speaker 2: So very different one hundred percent. And there's like religious

587
00:30:10,024 --> 00:30:14,224
Speaker 2: like utopian fantasy, messianic fantasies. As far as I can tell,

588
00:30:14,264 --> 00:30:17,944
Speaker 2: most of the like respected national security people are pretty

589
00:30:17,944 --> 00:30:21,504
Speaker 2: critical of what much of what Natanyahu did, although also

590
00:30:21,504 --> 00:30:23,384
Speaker 2: would argue, yes, something had to be done, you know,

591
00:30:23,464 --> 00:30:26,304
Speaker 2: so anyway, and like then I was like, what are

592
00:30:26,304 --> 00:30:29,184
Speaker 2: the professors, the radical left professors you hear about, what

593
00:30:29,264 --> 00:30:32,064
Speaker 2: are they arguing? And it would be like really persuasive

594
00:30:32,224 --> 00:30:36,624
Speaker 2: about how the history of settler colonialist studies and the city.

595
00:30:36,704 --> 00:30:39,544
Speaker 2: I just was noticing, like everything I would read, I'd

596
00:30:39,584 --> 00:30:41,984
Speaker 2: be like, yeah, I really yeah, dense serb and warfares

597
00:30:42,224 --> 00:30:44,824
Speaker 2: Like it's because it is. It's not like beautifully written.

598
00:30:44,864 --> 00:30:46,864
Speaker 2: It's not it's not that Ai is like an amazing writer,

599
00:30:46,984 --> 00:30:50,704
Speaker 2: but it's very it's good at being persuasive to the

600
00:30:50,744 --> 00:30:53,224
Speaker 2: thing you asked of it. I don't know, I find

601
00:30:53,224 --> 00:30:56,544
Speaker 2: that very exciting because I think you could actually if

602
00:30:56,584 --> 00:31:00,264
Speaker 2: you wanted to learn a lot about how vaccines work

603
00:31:00,344 --> 00:31:03,864
Speaker 2: or autism works or whatever, but you could just as

604
00:31:03,904 --> 00:31:06,944
Speaker 2: easily use it for disinformation.

605
00:31:06,904 --> 00:31:09,464
Speaker 3: Or strengthen your own view. So much of the way

606
00:31:09,584 --> 00:31:11,984
Speaker 3: people look at the world, like in academia, happened a

607
00:31:12,024 --> 00:31:14,704
Speaker 3: lot in the last thirty years on this. Everybody sort

608
00:31:14,704 --> 00:31:17,024
Speaker 3: of put things in the oppressor oppressed, sort of like

609
00:31:17,064 --> 00:31:19,224
Speaker 3: we do in the States. Now you have left or

610
00:31:19,304 --> 00:31:21,584
Speaker 3: right are you and the thing is Once you've decided

611
00:31:21,584 --> 00:31:23,864
Speaker 3: those are the two things to look at, then you

612
00:31:23,904 --> 00:31:26,264
Speaker 3: create the worldview and fit all your pieces into that

613
00:31:26,344 --> 00:31:28,744
Speaker 3: kind of thing. And it's especially hard in that part

614
00:31:28,744 --> 00:31:30,624
Speaker 3: of the world right because you can make up your

615
00:31:30,744 --> 00:31:32,344
Speaker 3: view that you're a victim and the other people can

616
00:31:32,384 --> 00:31:34,784
Speaker 3: make up of you that they're the victim, and victimhood

617
00:31:34,784 --> 00:31:36,744
Speaker 3: gives you a lot of power. You can lash out

618
00:31:36,904 --> 00:31:39,184
Speaker 3: to deal with your victimhood, or that you're the oppressed

619
00:31:39,224 --> 00:31:40,064
Speaker 3: and they're the oppressor.

620
00:31:40,384 --> 00:31:42,824
Speaker 2: You can look at it like Israel Palestine and that's

621
00:31:43,064 --> 00:31:45,224
Speaker 2: the conflict, or you can look at it as a

622
00:31:45,224 --> 00:31:50,504
Speaker 2: Middle East conflict and Iran creating proxy wars and creating

623
00:31:50,504 --> 00:31:54,624
Speaker 2: permanent conflict, which is a like known military strategy. You

624
00:31:55,384 --> 00:31:57,944
Speaker 2: find a dissident group or you invent one. The British

625
00:31:57,944 --> 00:32:00,344
Speaker 2: were good at that, like just creating, like we're going

626
00:32:00,424 --> 00:32:03,104
Speaker 2: to make the Hutus angry at the Tutsis and they'll

627
00:32:03,144 --> 00:32:05,144
Speaker 2: be so distracted with each other, nobody will think to

628
00:32:05,224 --> 00:32:07,504
Speaker 2: kill the British. That was the French, by the way,

629
00:32:07,624 --> 00:32:10,424
Speaker 2: and the Germans, but I'll give the British pass on this. Sorry,

630
00:32:10,544 --> 00:32:12,744
Speaker 2: the French and the Germans and the British and the

631
00:32:12,784 --> 00:32:15,224
Speaker 2: Portuguese and the Italians.

632
00:32:15,504 --> 00:32:18,584
Speaker 3: But don't forget the Belgians. They were like when they

633
00:32:18,624 --> 00:32:21,264
Speaker 3: had their chance, they were the nastiest, They were the nastiest.

634
00:32:21,344 --> 00:32:24,624
Speaker 2: They were pretty bad. Yeah, yeah, Congo is unbelievable.

635
00:32:24,744 --> 00:32:27,944
Speaker 4: But just the language you're speaking about, just colonial oppress

636
00:32:28,024 --> 00:32:31,824
Speaker 4: and oppressors. So we're speaking English, which is basically you know,

637
00:32:32,864 --> 00:32:37,144
Speaker 4: Vikings speaking Latin, going to war with Germans and then

638
00:32:37,184 --> 00:32:39,464
Speaker 4: being defended by the being defeated by the French. It

639
00:32:39,504 --> 00:32:42,864
Speaker 4: was all these different groups oppressed each other and colonized

640
00:32:42,904 --> 00:32:44,824
Speaker 4: the UK, and we end up with this weird lady.

641
00:32:44,944 --> 00:32:47,864
Speaker 4: Are the Jews going back to Palestine Great Israel? Are

642
00:32:47,904 --> 00:32:50,664
Speaker 4: they colonializers or they're just going back to where they

643
00:32:50,704 --> 00:32:52,464
Speaker 4: came from. I guess what I'm trying to say is

644
00:32:52,704 --> 00:32:56,064
Speaker 4: there's no right answer to any of this. Yeah, that's

645
00:32:54,944 --> 00:32:58,344
Speaker 4: just how you argue it, just how you argue it.

646
00:32:58,384 --> 00:32:59,864
Speaker 2: And we saw that in Cia all the time. It

647
00:32:59,904 --> 00:33:01,064
Speaker 2: was like, especially John, you.

648
00:33:01,064 --> 00:33:03,544
Speaker 3: Can choose when history starts, you can choose to what

649
00:33:03,624 --> 00:33:04,624
Speaker 3: era you want to look at.

650
00:33:05,384 --> 00:33:06,744
Speaker 2: Yeah, I was at Kosovo.

651
00:33:07,104 --> 00:33:08,904
Speaker 4: Every Serb you talk to is like the Battle of

652
00:33:08,904 --> 00:33:11,304
Speaker 4: the Field of Black Birks thirteen eighty nine. That's when

653
00:33:11,344 --> 00:33:15,184
Speaker 4: everything started. It's like, everything's fine until then thirteen eighty nine.

654
00:33:15,304 --> 00:33:18,064
Speaker 4: What the fuck is thirteen eighty nine is have you.

655
00:33:18,064 --> 00:33:21,584
Speaker 3: Seen the big mountain they've made of Serbian skulls from

656
00:33:21,624 --> 00:33:25,144
Speaker 3: the Turkish invasions? No serves love to show that office

657
00:33:25,264 --> 00:33:26,064
Speaker 3: what happens when.

658
00:33:26,104 --> 00:33:29,744
Speaker 2: Yeah, I mean having spent time in Israel and Palestine

659
00:33:29,864 --> 00:33:34,544
Speaker 2: like it. If someone says nineteen sixty seven or that

660
00:33:34,704 --> 00:33:37,784
Speaker 2: means you know where they're coming from. If someone starts

661
00:33:37,824 --> 00:33:39,424
Speaker 2: to the clock two thousand years ago, you know where

662
00:33:39,424 --> 00:33:42,624
Speaker 2: they're coming from. If someone is really focused on Europe

663
00:33:42,624 --> 00:33:45,944
Speaker 2: in the nineteen forties and then among Israeli is like

664
00:33:46,024 --> 00:33:48,544
Speaker 2: the words they use. It's very common here to talk

665
00:33:48,584 --> 00:33:50,824
Speaker 2: about occupied territories. If you say that in Israel, that's

666
00:33:50,864 --> 00:33:54,664
Speaker 2: a really big statement that really positions you. If you

667
00:33:54,744 --> 00:33:58,304
Speaker 2: call it historic Juday in Samaria like that also on

668
00:33:58,344 --> 00:34:00,904
Speaker 2: the other side. So yeah, I guess that's the point

669
00:34:00,944 --> 00:34:03,584
Speaker 2: I'm making is that's one of the things that's fascinating

670
00:34:03,584 --> 00:34:06,824
Speaker 2: about this AI stuff is it makes you realize that

671
00:34:07,024 --> 00:34:11,783
Speaker 2: there were all these like deep onto logiclypistemological questions.

672
00:34:11,584 --> 00:34:12,704
Speaker 3: And it doesn't solve.

673
00:34:13,024 --> 00:34:15,223
Speaker 2: No, it doesn't. Guys, have a question for the three

674
00:34:15,264 --> 00:34:17,983
Speaker 2: of you. What's going on in the world this week

675
00:34:18,104 --> 00:34:19,904
Speaker 2: that you guys want to talk about and have a

676
00:34:19,944 --> 00:34:21,864
Speaker 2: particular point of view on Jesus.

677
00:34:22,384 --> 00:34:24,104
Speaker 3: Well, I mean, if you watch it's almost hard to

678
00:34:24,104 --> 00:34:26,463
Speaker 3: watch the news now and I don't watch much TV,

679
00:34:26,544 --> 00:34:28,943
Speaker 3: but we do watch PBS News Hour or whatever. There's

680
00:34:28,944 --> 00:34:32,303
Speaker 3: always a big section on Israel, Gaza, and now we're

681
00:34:32,304 --> 00:34:35,104
Speaker 3: told Trump should win the Nobel Prize because his plan.

682
00:34:35,183 --> 00:34:37,864
Speaker 3: Of course, it's not his plan, it's Tony Blair's plans.

683
00:34:37,984 --> 00:34:40,863
Speaker 3: Go it's gonna give us peace. I'm skeptical. You got

684
00:34:40,904 --> 00:34:44,183
Speaker 3: bad actors, you got Nettan Yahou, you got Hamas, you

685
00:34:44,263 --> 00:34:46,504
Speaker 3: got a lot of people around the edges. You've got

686
00:34:46,623 --> 00:34:49,703
Speaker 3: Trump who's lazy, who just says because he said it's

687
00:34:49,703 --> 00:34:51,743
Speaker 3: going to happen, it's going to happen. So I'm skeptical

688
00:34:51,783 --> 00:34:54,144
Speaker 3: that it's going to go through. That's part of the news.

689
00:34:54,304 --> 00:34:55,943
Speaker 3: The other part of the news is all the things

690
00:34:55,984 --> 00:34:59,944
Speaker 3: around the government shutting down and therefore like trying to

691
00:34:59,984 --> 00:35:02,464
Speaker 3: fire people why the government's shutting down, on each side,

692
00:35:02,584 --> 00:35:05,424
Speaker 3: spinning stories about how the other it's the other side's fault.

693
00:35:05,544 --> 00:35:07,704
Speaker 4: Here's a prediction for you, and I hope it doesn't

694
00:35:07,703 --> 00:35:11,343
Speaker 4: come true. The Insurrection Act. I think we're moving toward that,

695
00:35:11,504 --> 00:35:14,504
Speaker 4: And I think what the Insurrection Act brings is US

696
00:35:14,544 --> 00:35:20,344
Speaker 4: troops on the streets performing law enforcement and security. So basically,

697
00:35:20,464 --> 00:35:23,984
Speaker 4: all the conspiracy theories about how the black helicopters come in,

698
00:35:24,024 --> 00:35:26,263
Speaker 4: they're coming for your guns. They're bringing in that that

699
00:35:26,544 --> 00:35:29,143
Speaker 4: the federal troops, it's all on the right. It's the

700
00:35:29,223 --> 00:35:32,703
Speaker 4: right that's actually I think to do it. I think

701
00:35:32,743 --> 00:35:34,863
Speaker 4: it's a real possibility of this. I really had him

702
00:35:34,864 --> 00:35:36,584
Speaker 4: as a member of the media. It seems to be

703
00:35:36,623 --> 00:35:38,863
Speaker 4: the media is in some ways. It's an easy thing

704
00:35:38,904 --> 00:35:40,823
Speaker 4: to say it's failing in this sense. But all the

705
00:35:40,904 --> 00:35:43,143
Speaker 4: articles you read about this, what it'll be? What is

706
00:35:43,183 --> 00:35:45,664
Speaker 4: the Insurrection Act? How was the Insurrection Act used? Before

707
00:35:46,864 --> 00:35:49,504
Speaker 4: Mike Trump used the Insurrection Act? No one says there's

708
00:35:49,584 --> 00:35:54,463
Speaker 4: no fucking insurrection like port to Portland it is. There's

709
00:35:54,504 --> 00:35:57,544
Speaker 4: no insurrection in Portland. There's even in the films of

710
00:35:57,703 --> 00:36:01,584
Speaker 4: no no one being there. There's like ten people standing

711
00:36:01,584 --> 00:36:06,144
Speaker 4: outside the ice facility wearing like rubber chicken masks and stuff.

712
00:36:06,143 --> 00:36:07,024
Speaker 4: This is an insurrection.

713
00:36:07,304 --> 00:36:09,504
Speaker 2: Unless you wus Fox News and then there is an insurrection.

714
00:36:09,783 --> 00:36:11,903
Speaker 3: If there's an insurrection and you got to send ice,

715
00:36:11,944 --> 00:36:14,343
Speaker 3: you gotta send Fenol troops. What kinds of pansies are

716
00:36:14,464 --> 00:36:16,423
Speaker 3: you think that Portland's is going to take over the right?

717
00:36:17,984 --> 00:36:19,823
Speaker 3: Got a bunch of guys in fog suits or whatever.

718
00:36:21,064 --> 00:36:23,464
Speaker 2: You got to watch? Yeah, I will say, like the

719
00:36:23,544 --> 00:36:25,704
Speaker 2: time in Iraq, and you guys have way more experience

720
00:36:25,743 --> 00:36:27,783
Speaker 2: than I do with this. But I remember talking to

721
00:36:27,824 --> 00:36:30,863
Speaker 2: soldiers about like, we should not be police force. That's

722
00:36:30,944 --> 00:36:33,223
Speaker 2: not our thing. And I remember this one guy who

723
00:36:33,263 --> 00:36:35,944
Speaker 2: was very smart. It was civil affairs that the folks

724
00:36:35,984 --> 00:36:39,064
Speaker 2: in the military who like try and do build civil

725
00:36:39,104 --> 00:36:42,263
Speaker 2: capacity and conquered areas, and he was just walking me

726
00:36:42,424 --> 00:36:46,903
Speaker 2: through why you don't want a military for military reasons.

727
00:36:46,944 --> 00:36:49,504
Speaker 2: You don't want the military doing police work because it

728
00:36:50,024 --> 00:36:52,864
Speaker 2: stifles their military ability. You don't want to you can't

729
00:36:52,944 --> 00:36:58,504
Speaker 2: simultaneously pacify a population and provide some kind of objective justice,

730
00:36:58,824 --> 00:37:01,344
Speaker 2: And then for police reasons, you don't want the military

731
00:37:01,384 --> 00:37:03,983
Speaker 2: to do it. It's really a disaster, and you don't

732
00:37:04,024 --> 00:37:06,503
Speaker 2: want your police to be militarized.

733
00:37:06,584 --> 00:37:09,544
Speaker 4: I mean, just looking at the videos you've got guys

734
00:37:09,783 --> 00:37:13,544
Speaker 4: out with automatic weapons, fingers just over the trigger, and

735
00:37:13,544 --> 00:37:16,104
Speaker 4: they're basically, for all intentsive purposes, they are dressed like

736
00:37:16,183 --> 00:37:19,664
Speaker 4: militia or military guys, right and with their faces.

737
00:37:19,263 --> 00:37:23,544
Speaker 3: Covered camouflage in downtown DC, or they're spreading mulch.

738
00:37:25,304 --> 00:37:27,264
Speaker 2: Let's pause here and take a quick break.

739
00:37:37,344 --> 00:37:39,904
Speaker 4: But the insurrection, it seems like it's almost they're pushing

740
00:37:39,944 --> 00:37:42,064
Speaker 4: it to a foregone conclusion, and then when it happens,

741
00:37:42,064 --> 00:37:44,143
Speaker 4: we'll all go, well, we saw it in there anyway,

742
00:37:43,944 --> 00:37:47,464
Speaker 4: So Ada, why don't you run with this and explain

743
00:37:47,544 --> 00:37:49,984
Speaker 4: why I'm both right and brilliant and worrying about this.

744
00:37:50,304 --> 00:37:52,584
Speaker 2: Now you're right and brilliant and worry about this. And

745
00:37:52,623 --> 00:37:55,343
Speaker 2: I think part of the collapse of gatekeeping is that

746
00:37:55,864 --> 00:37:58,783
Speaker 2: we don't like maybe they weren't gatekeeping, maybe they were

747
00:37:58,824 --> 00:38:02,343
Speaker 2: more or we were more like normally, we just were

748
00:38:02,424 --> 00:38:06,783
Speaker 2: reinforcing widely shared norms, and when those norms disappeared, the

749
00:38:06,864 --> 00:38:10,104
Speaker 2: media as a whole doesn't know what to do. It's

750
00:38:10,223 --> 00:38:12,504
Speaker 2: it was more of a follower than a driver of

751
00:38:12,584 --> 00:38:14,864
Speaker 2: standards and morals. And I think a lot of journalists

752
00:38:14,864 --> 00:38:16,824
Speaker 2: would say, that's right, that's what we should do. Although

753
00:38:16,984 --> 00:38:19,584
Speaker 2: I think it's okay to be a journalist who's against

754
00:38:20,024 --> 00:38:21,863
Speaker 2: incorrectly using the Insurrection Act.

755
00:38:22,223 --> 00:38:25,984
Speaker 3: Reporters should fall all over Portland report on it. Is

756
00:38:25,984 --> 00:38:28,343
Speaker 3: there an insurrection here, Let's look at the facts, Let's

757
00:38:28,344 --> 00:38:31,584
Speaker 3: go down, let's interview people. Let's start with the thing

758
00:38:31,623 --> 00:38:34,223
Speaker 3: they're doing, rather than talk about how they might use

759
00:38:34,223 --> 00:38:37,343
Speaker 3: it politically and stay in Washington, go prove that there's

760
00:38:37,344 --> 00:38:39,223
Speaker 3: an insurrection or not. Yeah, it shouldn't be.

761
00:38:39,223 --> 00:38:41,664
Speaker 2: And I think there is. I'm just imagining that meeting.

762
00:38:41,743 --> 00:38:43,743
Speaker 2: And first of all, I'm sure the big places do

763
00:38:43,864 --> 00:38:46,703
Speaker 2: send somebody. There's also like a collapse of journalism, so

764
00:38:46,703 --> 00:38:49,303
Speaker 2: there's not as much money or ability. There's also like,

765
00:38:49,663 --> 00:38:53,024
Speaker 2: well everyone knows that, we all know that. That's not

766
00:38:53,104 --> 00:38:56,223
Speaker 2: the point. He's just saying it is. I mean, I

767
00:38:56,263 --> 00:38:58,383
Speaker 2: think it turns out Trump is just better at this.

768
00:38:58,904 --> 00:39:00,464
Speaker 2: He's better at messaging, like.

769
00:39:00,464 --> 00:39:02,784
Speaker 3: He knows, like he's willing to go places.

770
00:39:02,984 --> 00:39:04,904
Speaker 2: He's willing to go places. And I think, you know,

771
00:39:04,984 --> 00:39:07,423
Speaker 2: I was thinking the other day about my first big

772
00:39:07,424 --> 00:39:10,744
Speaker 2: investigative piece about Trump for The New York and about

773
00:39:10,824 --> 00:39:14,384
Speaker 2: how he knowingly participated in a money laundering scheme for

774
00:39:14,424 --> 00:39:18,343
Speaker 2: the Ironiant Guard. Still to me, feels like that should

775
00:39:18,384 --> 00:39:21,624
Speaker 2: be relevant. But I remember I called his general counsel

776
00:39:21,623 --> 00:39:24,143
Speaker 2: and I was like, you don't seem nervous, like I

777
00:39:24,143 --> 00:39:26,623
Speaker 2: feel like you should be nervous. And he acknowleds my

778
00:39:26,703 --> 00:39:29,703
Speaker 2: article is correct, like he wasn't making a claim that

779
00:39:29,783 --> 00:39:32,263
Speaker 2: it wasn't true. And this before it was published. But

780
00:39:32,384 --> 00:39:34,663
Speaker 2: he knew because we do fact checking. So we went

781
00:39:34,743 --> 00:39:37,183
Speaker 2: over every fact with him and he was like, ah, no,

782
00:39:37,344 --> 00:39:39,504
Speaker 2: I know it's going to happen. Rachel Mattow will make

783
00:39:39,544 --> 00:39:42,344
Speaker 2: it a big deal, Cinn will ignore it. A couple

784
00:39:42,304 --> 00:39:44,863
Speaker 2: of Democratic senators might write a letter, but nobody's going

785
00:39:44,904 --> 00:39:47,863
Speaker 2: to care. And like literally, Rachel Mattow did half an hour,

786
00:39:48,663 --> 00:39:49,503
Speaker 2: nobody else cared.

787
00:39:49,504 --> 00:39:51,504
Speaker 3: And so like you know, these are going to arrest

788
00:39:51,703 --> 00:39:54,663
Speaker 3: James Comy, right, And they wanted frog welcome and frog

789
00:39:54,703 --> 00:39:57,263
Speaker 3: march them over, they say, and or FBI guys get

790
00:39:57,304 --> 00:39:59,064
Speaker 3: fired because they didn't want a frog didn't figure out

791
00:39:59,064 --> 00:40:01,504
Speaker 3: the frog march him for the cameras. But what they

792
00:40:01,544 --> 00:40:04,623
Speaker 3: got called me on is like a small, one little thing.

793
00:40:04,663 --> 00:40:07,064
Speaker 3: He said in a testimony one time. It wasn't even

794
00:40:07,104 --> 00:40:10,343
Speaker 3: about Trump Russia thing. It was about Hillary Clinton. And

795
00:40:10,384 --> 00:40:13,504
Speaker 3: he's claiming oh he lied. Now as you read the thing,

796
00:40:13,544 --> 00:40:15,143
Speaker 3: I don't think he did, and I think he'll get

797
00:40:15,183 --> 00:40:18,384
Speaker 3: off very easily here. But really, so Trump's going after

798
00:40:18,743 --> 00:40:22,424
Speaker 3: someone who might have said one small little lie. The

799
00:40:22,464 --> 00:40:26,584
Speaker 3: guy who's been lying his entire career and every day

800
00:40:26,783 --> 00:40:28,904
Speaker 3: spits out hundreds of lies. This is what someone's going

801
00:40:28,944 --> 00:40:31,343
Speaker 3: to go prison for, and not just him. But like

802
00:40:31,384 --> 00:40:34,504
Speaker 3: he asked the question, did Tom Homan right? Did he

803
00:40:34,623 --> 00:40:37,864
Speaker 3: take fifty thousand dollars in a in a kava bag?

804
00:40:38,304 --> 00:40:40,664
Speaker 3: He didn't do anything illegal because we dropped the charges.

805
00:40:40,783 --> 00:40:43,383
Speaker 2: That's not the question. The question is did he take

806
00:40:43,424 --> 00:40:44,144
Speaker 2: fifty grand?

807
00:40:44,263 --> 00:40:47,663
Speaker 4: And what happened to the FBI's our taxpayer money that

808
00:40:47,743 --> 00:40:50,584
Speaker 4: when fifty grand? Is he paying taxes on it? Why

809
00:40:50,584 --> 00:40:52,783
Speaker 4: did he want it? It was like, I'm sorry, what

810
00:40:53,183 --> 00:40:56,584
Speaker 4: left or right? What the fuck? Somebody takes fifty grand

811
00:40:56,623 --> 00:40:59,223
Speaker 4: in a bag? And oh, yeah, he just did that,

812
00:40:59,263 --> 00:41:00,463
Speaker 4: but it's all on the if.

813
00:41:00,384 --> 00:41:02,783
Speaker 2: It was a valise, it was a collar bag. Yeah.

814
00:41:03,223 --> 00:41:05,424
Speaker 3: The true Trump people are like, what a clown fifty

815
00:41:05,424 --> 00:41:06,143
Speaker 3: thousand in a bag?

816
00:41:06,263 --> 00:41:07,104
Speaker 2: Shot a bitcoin?

817
00:41:07,424 --> 00:41:09,504
Speaker 3: It would be a bitcoin billion, yeah.

818
00:41:10,183 --> 00:41:12,183
Speaker 1: Plus the value of the bag and the food that

819
00:41:12,304 --> 00:41:15,904
Speaker 1: was in it. So it's more like fifty twenty dollars.

820
00:41:16,024 --> 00:41:17,783
Speaker 2: Yeah. I mean I've been thinking a lot about this

821
00:41:17,904 --> 00:41:21,143
Speaker 2: that I really devoted my life to a very simple

822
00:41:21,304 --> 00:41:26,143
Speaker 2: naive like when you say truth, it has a big effect.

823
00:41:26,183 --> 00:41:28,584
Speaker 3: And you got to find evidence to.

824
00:41:28,464 --> 00:41:31,663
Speaker 2: Prove it, and that matters that if you have a process,

825
00:41:31,783 --> 00:41:34,744
Speaker 2: that matters. And it took me a very long time.

826
00:41:35,263 --> 00:41:37,544
Speaker 2: I'm not saying I accept it like I'm happy about it.

827
00:41:37,743 --> 00:41:41,344
Speaker 2: I'm not happy about it, but I like, I just feel, Okay,

828
00:41:41,384 --> 00:41:44,464
Speaker 2: this is bigger than There's no New York Times headline

829
00:41:45,143 --> 00:41:48,343
Speaker 2: that's going to fix this. There's no like right sho

830
00:41:48,623 --> 00:41:50,823
Speaker 2: gate and at a press release that's going to fix this.

831
00:41:50,944 --> 00:41:52,903
Speaker 2: There's some other thing. And I think it does have

832
00:41:52,984 --> 00:41:55,303
Speaker 2: to do with all the things we talk about here,

833
00:41:55,344 --> 00:41:59,623
Speaker 2: how information flows, how people form opinions, how how those

834
00:41:59,743 --> 00:42:02,823
Speaker 2: opinions are reinforced. It also, I mean, we you know,

835
00:42:03,424 --> 00:42:06,304
Speaker 2: behind our talk about AI, certainly behind our talk about Trump,

836
00:42:06,504 --> 00:42:10,143
Speaker 2: but maybe behind every conspiracy theory is like power and

837
00:42:10,183 --> 00:42:13,263
Speaker 2: the truth that the truth. When this maybe makes us

838
00:42:13,263 --> 00:42:15,984
Speaker 2: a bunch of left wing intellectuals, because this is a

839
00:42:15,984 --> 00:42:18,703
Speaker 2: lot of the academic work of the last century is

840
00:42:18,743 --> 00:42:23,224
Speaker 2: like truth is not a thing it's a expression of power.

841
00:42:23,263 --> 00:42:25,343
Speaker 2: That doesn't mean it's not like I don't believe. I'm

842
00:42:25,344 --> 00:42:29,944
Speaker 2: not trying to make a like everything's equally true. I

843
00:42:29,944 --> 00:42:32,223
Speaker 2: don't think that's true. The phrase I've been using for

844
00:42:32,263 --> 00:42:35,143
Speaker 2: myself is you can't be one hundred percent right about anything,

845
00:42:35,223 --> 00:42:38,303
Speaker 2: but you can be one hundred percent wrong about things.

846
00:42:37,824 --> 00:42:40,943
Speaker 2: And we do know a whole bunch of people who

847
00:42:40,944 --> 00:42:43,864
Speaker 2: are one hundred percent wrong. But you could be eighty

848
00:42:43,904 --> 00:42:46,303
Speaker 2: percent true right. You could be like way more or

849
00:42:46,304 --> 00:42:48,703
Speaker 2: you could show like you've done your homework and you're

850
00:42:49,223 --> 00:42:52,783
Speaker 2: there's more evidence, there's more research. It could be true,

851
00:42:52,783 --> 00:42:54,544
Speaker 2: but it's still more complicated. You don't have it all.

852
00:42:54,584 --> 00:42:56,464
Speaker 4: I think the Middle East stuff is that you have

853
00:42:56,504 --> 00:42:59,544
Speaker 4: two two true narratives clashing. It just depends on how

854
00:42:59,544 --> 00:43:00,263
Speaker 4: you make those.

855
00:43:00,544 --> 00:43:03,544
Speaker 2: But I or two hundred or two thousand that narrative,

856
00:43:03,584 --> 00:43:04,944
Speaker 2: But I did so.

857
00:43:06,104 --> 00:43:08,384
Speaker 4: When John and I were in CIA and when you

858
00:43:08,424 --> 00:43:11,863
Speaker 4: were in place, oh you that we were.

859
00:43:12,464 --> 00:43:15,464
Speaker 3: We were big deals. We were the premier intelligence services.

860
00:43:15,944 --> 00:43:18,303
Speaker 4: We were like, you know, working for some like fly

861
00:43:18,384 --> 00:43:20,343
Speaker 4: by Night organization like that, you know, the New York

862
00:43:20,384 --> 00:43:25,183
Speaker 4: Times or MPR. But we looked at everything through the

863
00:43:25,223 --> 00:43:27,744
Speaker 4: optic of how it impacts us. And when we started

864
00:43:27,783 --> 00:43:31,703
Speaker 4: our careers, we assumed that Russia, the Soviet Union, and

865
00:43:31,743 --> 00:43:34,663
Speaker 4: the East Block that they were like way ahead of us,

866
00:43:34,783 --> 00:43:39,023
Speaker 4: right the missile gap, their technology spot nicks before our time,

867
00:43:39,464 --> 00:43:41,183
Speaker 4: and we found out that they were actually like way

868
00:43:41,183 --> 00:43:42,783
Speaker 4: more fun as fucked up as we were.

869
00:43:42,864 --> 00:43:44,343
Speaker 2: They were even way more fucked up.

870
00:43:44,384 --> 00:43:47,264
Speaker 4: So do you have a sense of China and Russia

871
00:43:47,304 --> 00:43:51,423
Speaker 4: and authoritarian governments they also are embracing this but they've

872
00:43:51,464 --> 00:43:53,904
Speaker 4: got to be fucking it up too. Are they fucking

873
00:43:53,944 --> 00:43:55,584
Speaker 4: it up worse or differently than us?

874
00:43:55,703 --> 00:43:59,024
Speaker 2: Or seems to be doing really well? Us a strength

875
00:43:59,304 --> 00:44:02,223
Speaker 2: that they've got seems to be doing really well, and

876
00:44:02,824 --> 00:44:06,064
Speaker 2: they're still catch up models like there. First of all,

877
00:44:06,104 --> 00:44:08,664
Speaker 2: we don't know like the Chinese military or the Chinese

878
00:44:08,703 --> 00:44:11,464
Speaker 2: Intelligence Service, were just the Chinese society.

879
00:44:11,584 --> 00:44:14,263
Speaker 4: In nineteen eighty nine, when you know, the East Block

880
00:44:14,304 --> 00:44:16,823
Speaker 4: fell apart, we realized how rotten it was, but we

881
00:44:16,824 --> 00:44:17,984
Speaker 4: didn't know that beforehand.

882
00:44:18,304 --> 00:44:21,743
Speaker 2: Yeah. Now, when Deep Seeks Big Model was revealed, I

883
00:44:21,783 --> 00:44:24,904
Speaker 2: think it was in December twenty twenty four, that was

884
00:44:24,944 --> 00:44:28,944
Speaker 2: a utterly transformative moment because for a bunch of reasons.

885
00:44:28,984 --> 00:44:31,584
Speaker 2: First of all, China had a model that in some

886
00:44:31,623 --> 00:44:34,263
Speaker 2: ways outperformed American models. I don't think it was, and

887
00:44:34,304 --> 00:44:36,504
Speaker 2: it's hard to even know what, Like all the benchmarks

888
00:44:36,504 --> 00:44:38,664
Speaker 2: are meaningless in a lot of ways, and the models

889
00:44:38,663 --> 00:44:42,104
Speaker 2: are trained on the benchmark, so they become but a

890
00:44:42,183 --> 00:44:45,424
Speaker 2: really good model, way better than anyone thought was going

891
00:44:45,504 --> 00:44:48,783
Speaker 2: to come out of a non US country. And then

892
00:44:49,143 --> 00:44:53,504
Speaker 2: the but also they did it way cheaper than the Americans,

893
00:44:53,623 --> 00:44:58,304
Speaker 2: like single digit millions instead of hundreds of millions into

894
00:44:58,344 --> 00:45:01,504
Speaker 2: the billions, and it should and they have steadily. They

895
00:45:01,544 --> 00:45:05,904
Speaker 2: have multiple major models, none of the top ones, but

896
00:45:06,223 --> 00:45:09,663
Speaker 2: pretty close behind. And I'd say the ones was to

897
00:45:09,743 --> 00:45:11,663
Speaker 2: be really blowing. It is Europe as they have been

898
00:45:11,703 --> 00:45:14,703
Speaker 2: on tech, like it's pretty hard to you know, name

899
00:45:14,743 --> 00:45:18,504
Speaker 2: your favorite high tech products that were invented in Europe,

900
00:45:18,544 --> 00:45:20,344
Speaker 2: Like it's there's not zero.

901
00:45:20,544 --> 00:45:24,024
Speaker 3: But if the goal is to provide better information to

902
00:45:24,024 --> 00:45:26,783
Speaker 3: make better decisions, as we saw for example in this

903
00:45:26,824 --> 00:45:28,944
Speaker 3: so of you need to stalin and author antams have

904
00:45:28,984 --> 00:45:30,944
Speaker 3: their own view and you can come to them with

905
00:45:31,024 --> 00:45:33,544
Speaker 3: the truth and if it doesn't fit with what they want,

906
00:45:33,904 --> 00:45:36,863
Speaker 3: it doesn't matter. So if g He's already got a

907
00:45:36,864 --> 00:45:39,344
Speaker 3: worldview like so he's doing a good job of letting

908
00:45:39,703 --> 00:45:41,544
Speaker 3: people come up with things and putting money where it

909
00:45:41,544 --> 00:45:44,303
Speaker 3: needs to go. And China and this big thing is

910
00:45:44,344 --> 00:45:47,064
Speaker 3: having a lot of success. But does that mean G

911
00:45:47,304 --> 00:45:49,504
Speaker 3: is better informed about the world and what?

912
00:45:49,743 --> 00:45:51,944
Speaker 2: Yeah, that's always an interesting thing. Like I had a

913
00:45:51,984 --> 00:45:55,944
Speaker 2: thought in Iraq. I remember thinking it's possible that Saddam

914
00:45:56,024 --> 00:45:58,183
Speaker 2: Hussein and George W. Bush are the two people on

915
00:45:58,223 --> 00:46:00,623
Speaker 2: earth who know the least about the ground but the

916
00:46:00,623 --> 00:46:04,264
Speaker 2: other side of what's happening, because they have this massive

917
00:46:04,623 --> 00:46:08,064
Speaker 2: apparatus to prevent them from actually know. And I'm not

918
00:46:08,104 --> 00:46:11,104
Speaker 2: a close expert on China at all, but from the

919
00:46:11,104 --> 00:46:13,663
Speaker 2: people I read, like Bill Bishop and stuff like it,

920
00:46:13,663 --> 00:46:17,784
Speaker 2: it does seem like they've moved. Like there was always

921
00:46:17,824 --> 00:46:22,104
Speaker 2: this storied bureaucracy that kind of existed as a force

922
00:46:22,544 --> 00:46:26,703
Speaker 2: independent of whoever happened to be the leader, and that

923
00:46:26,743 --> 00:46:29,223
Speaker 2: she has it really is to the glory of G

924
00:46:29,504 --> 00:46:32,703
Speaker 2: and that probably is a long term strategic weakness, except

925
00:46:33,183 --> 00:46:35,463
Speaker 2: we're doing that. I just had to talk with someone

926
00:46:35,464 --> 00:46:38,544
Speaker 2: in Canada today about visas. In Canada, they just cut

927
00:46:38,584 --> 00:46:40,303
Speaker 2: down on visas. And this is a friend who's an

928
00:46:40,344 --> 00:46:42,984
Speaker 2: economists in Canada, and I was like, shouldn't this be like,

929
00:46:43,384 --> 00:46:46,303
Speaker 2: shouldn't you just be getting every And they specifically cut

930
00:46:46,344 --> 00:46:50,784
Speaker 2: down on visas for smart students going to university why,

931
00:46:51,024 --> 00:46:54,384
Speaker 2: which is insane, And they have their own internal reasons.

932
00:46:54,464 --> 00:46:57,584
Speaker 2: There's fears of job displacement. There's there apparently were a

933
00:46:57,584 --> 00:46:59,904
Speaker 2: bunch of like diploma mills because they had fairly lax.

934
00:47:00,384 --> 00:47:03,663
Speaker 2: But it's also they don't want it too publicly. Like

935
00:47:04,104 --> 00:47:05,944
Speaker 2: this friend of mine was like, I think we could,

936
00:47:05,944 --> 00:47:08,383
Speaker 2: like for one hundred billion dollars, we just could just

937
00:47:08,623 --> 00:47:12,064
Speaker 2: grab an entire field of study, Like we could just

938
00:47:12,424 --> 00:47:17,504
Speaker 2: get every neuroscientist or every expert in battery technology or whatever.

939
00:47:18,104 --> 00:47:24,024
Speaker 2: But there's fear of pissing off Trump. There's there's internal issues.

940
00:47:23,743 --> 00:47:25,743
Speaker 4: So g you know, when they're building their model and

941
00:47:25,824 --> 00:47:28,263
Speaker 4: not for medicine or science and things like that, but

942
00:47:28,663 --> 00:47:31,143
Speaker 4: as he tries to understand his country. Basically everybody in

943
00:47:31,223 --> 00:47:33,504
Speaker 4: China lies, right, They don't tell the truth.

944
00:47:33,544 --> 00:47:35,464
Speaker 2: It's not like you're going to say in a form

945
00:47:35,544 --> 00:47:36,024
Speaker 2: or anything.

946
00:47:36,143 --> 00:47:38,463
Speaker 4: Everybody is like I love the government because they all

947
00:47:38,504 --> 00:47:41,223
Speaker 4: know if you don't fucking do that, you're screwed. And

948
00:47:41,263 --> 00:47:43,424
Speaker 4: so I think the AA model is taking it in

949
00:47:43,504 --> 00:47:45,984
Speaker 4: the same in Russia. So I their models have got

950
00:47:46,024 --> 00:47:49,544
Speaker 4: to be much more skewed and optimistic and positive towards

951
00:47:49,584 --> 00:47:52,943
Speaker 4: their leaderships, right, and and the the data that they

952
00:47:53,064 --> 00:47:55,903
Speaker 4: that that's that they build on has God, it has

953
00:47:55,984 --> 00:47:56,784
Speaker 4: to be skewed.

954
00:47:57,143 --> 00:48:00,704
Speaker 2: Just although most of the big models now are trained

955
00:48:00,743 --> 00:48:05,104
Speaker 2: on essentially everything in the world that's been digital like

956
00:48:05,384 --> 00:48:08,504
Speaker 2: every because you couldn't it's like TikTok, right, our our

957
00:48:09,064 --> 00:48:11,783
Speaker 2: data would go into the China model, right, yeah, but

958
00:48:11,904 --> 00:48:14,263
Speaker 2: also all of read it and every academic paper ever

959
00:48:14,344 --> 00:48:17,303
Speaker 2: and every book that's ever been digitized, and because you

960
00:48:17,944 --> 00:48:20,943
Speaker 2: you just need more and more data. And I don't

961
00:48:21,064 --> 00:48:23,663
Speaker 2: think you could create a cutting edge model just on

962
00:48:23,783 --> 00:48:26,903
Speaker 2: Chinese data like they're just I would guess that would

963
00:48:26,904 --> 00:48:30,064
Speaker 2: be my strong guess. So you'd need all the models

964
00:48:30,263 --> 00:48:34,263
Speaker 2: have all the data. Basically now there is creating synthetic data.

965
00:48:34,464 --> 00:48:36,983
Speaker 2: I would guess that if we see this with Elon Musk,

966
00:48:36,984 --> 00:48:40,224
Speaker 2: because he'll tweak his algorithm at Rock and it suddenly

967
00:48:40,304 --> 00:48:42,743
Speaker 2: is like spouting Nazi stuff. And it's not great that

968
00:48:42,783 --> 00:48:45,783
Speaker 2: it's spouting Nazi stuff from a like I'm against Nazis standpoint,

969
00:48:46,263 --> 00:48:50,344
Speaker 2: but it's also a sign that ham handed like top

970
00:48:50,464 --> 00:48:53,703
Speaker 2: down impositions on the model makes the model do really

971
00:48:53,703 --> 00:48:56,504
Speaker 2: weird things. It's not a good way, So the models

972
00:48:56,544 --> 00:49:01,104
Speaker 2: aren't quite as controllable as other things might be inherently.

973
00:49:01,464 --> 00:49:05,223
Speaker 2: But don't you I wouldn't you assume every government a

974
00:49:05,304 --> 00:49:07,183
Speaker 2: really good model you could get for less than a

975
00:49:07,183 --> 00:49:07,984
Speaker 2: billion dollars?

976
00:49:08,743 --> 00:49:10,944
Speaker 3: But those governed governments don't want the people to be

977
00:49:11,024 --> 00:49:12,504
Speaker 3: able to have access to all the But.

978
00:49:12,504 --> 00:49:14,143
Speaker 2: Don't they think internally? Wouldn't they?

979
00:49:14,504 --> 00:49:17,504
Speaker 4: But I think we're okay because the Buerau of labor statistics,

980
00:49:17,544 --> 00:49:19,984
Speaker 4: it's not like they're going to fire the person for

981
00:49:20,104 --> 00:49:22,743
Speaker 4: putting out the rule statistics that the government doesn't like.

982
00:49:23,183 --> 00:49:25,263
Speaker 2: But from a like how many countries on Earth could

983
00:49:25,424 --> 00:49:29,064
Speaker 2: the governments could blow a billion dollars like on most

984
00:49:29,104 --> 00:49:32,903
Speaker 2: of them, right, yeah, big ones? And like how would

985
00:49:32,904 --> 00:49:35,024
Speaker 2: you know? Am I wrong? Like from a national security

986
00:49:35,024 --> 00:49:37,544
Speaker 2: because wouldn't you want your own? You wouldn't want like

987
00:49:37,584 --> 00:49:41,663
Speaker 2: we saw what happened to Ukraine using starlink and being

988
00:49:41,703 --> 00:49:44,904
Speaker 2: dependent on elon Musk. You wouldn't want to be dependent

989
00:49:44,904 --> 00:49:47,383
Speaker 2: on sam Altman or Google or any of the And

990
00:49:47,424 --> 00:49:50,263
Speaker 2: you saw how all the big tech companies meant the

991
00:49:50,304 --> 00:49:52,823
Speaker 2: need to Trump. And so if you're an adversary of

992
00:49:52,864 --> 00:49:55,183
Speaker 2: the US, or even if you're like Israel, like an

993
00:49:55,183 --> 00:49:58,183
Speaker 2: ally of the US, but with your own independent desires,

994
00:49:58,944 --> 00:50:02,024
Speaker 2: don't you think they're all building That's my assumption, they're

995
00:50:02,064 --> 00:50:05,544
Speaker 2: all building their own models and that intelligence services will

996
00:50:05,584 --> 00:50:08,504
Speaker 2: just have access, although are they. I'd be curious, like

997
00:50:08,544 --> 00:50:11,024
Speaker 2: what is to a buddy I know who's in the FBI,

998
00:50:11,424 --> 00:50:13,663
Speaker 2: And he was like, we have I just use commercial

999
00:50:13,703 --> 00:50:16,064
Speaker 2: tools because the FBI tools are so lame and they're

1000
00:50:16,104 --> 00:50:17,504
Speaker 2: so behind the times.

1001
00:50:17,824 --> 00:50:20,743
Speaker 4: Like did you guys we couldn't even use Google right

1002
00:50:20,783 --> 00:50:23,623
Speaker 4: on our computers because we couldn't mix outside and.

1003
00:50:23,623 --> 00:50:25,343
Speaker 2: Internal really occasions.

1004
00:50:25,424 --> 00:50:28,183
Speaker 4: Yeah, so eventually we figured it out, but yeah, we

1005
00:50:28,223 --> 00:50:30,504
Speaker 4: had to switch between it, but yeah, we couldn't. Seventeen

1006
00:50:30,584 --> 00:50:33,584
Speaker 4: year old kids sitting in his basement had more access

1007
00:50:33,623 --> 00:50:36,743
Speaker 4: to information than we did. We had access to different information.

1008
00:50:36,944 --> 00:50:40,184
Speaker 2: And yet you created the crack epidemic. That's impressive without

1009
00:50:40,223 --> 00:50:41,223
Speaker 2: any technology.

1010
00:50:41,944 --> 00:50:42,263
Speaker 3: Thank you.

1011
00:50:42,984 --> 00:50:45,784
Speaker 2: Well, it's the alien technology that we have reverse engineered.

1012
00:50:46,024 --> 00:50:47,943
Speaker 1: Now, So I'm gonna I'm gonna go ahead and thank

1013
00:50:48,024 --> 00:50:50,104
Speaker 1: us all for getting a chance to get together again

1014
00:50:50,424 --> 00:50:53,584
Speaker 1: and to promise our listeners, we have a lot of

1015
00:50:53,703 --> 00:50:57,944
Speaker 1: great episodes we have now recorded and are recording, and

1016
00:50:57,984 --> 00:51:00,704
Speaker 1: we'll be coming out in swinging weeks. And I promise

1017
00:51:00,783 --> 00:51:03,263
Speaker 1: I'll get this YouTube channel up and running where you

1018
00:51:03,304 --> 00:51:03,983
Speaker 1: can see video.

1019
00:51:04,304 --> 00:51:07,104
Speaker 2: This is going to be a very interesting fall and

1020
00:51:07,143 --> 00:51:09,064
Speaker 2: I can't wait to talk to you guys next October.

1021
00:51:09,223 --> 00:51:17,663
Speaker 1: Yeah, Adam, come back anytime really. Mission Implausible is produced

1022
00:51:17,663 --> 00:51:22,664
Speaker 1: by Adam Davidson, Jerry O'shay, John Cipher, and Jonathan Stern.

1023
00:51:22,824 --> 00:51:25,184
Speaker 2: The associate producer is Rachel Harner.

1024
00:51:25,464 --> 00:51:29,263
Speaker 1: Mission Implausible is a production of honorable mention and abominable

1025
00:51:29,304 --> 00:51:31,263
Speaker 1: pictures for iHeart Podcasts.