1
00:00:00,640 --> 00:00:04,160
Speaker 1: Hi, I'm Molly John Fast and this is Fast Politics,

2
00:00:04,360 --> 00:00:07,120
Speaker 1: where we discussed the top political headlines with some of

3
00:00:07,160 --> 00:00:12,840
Speaker 1: today's best minds, and a PRRI poll says Americans who

4
00:00:12,880 --> 00:00:15,560
Speaker 1: say they are proud or very proud of being an

5
00:00:15,560 --> 00:00:19,200
Speaker 1: American has dropped thirty one percent over the last decade.

6
00:00:19,600 --> 00:00:22,279
Speaker 1: I wonder who's been president. We have such a great

7
00:00:22,320 --> 00:00:25,840
Speaker 1: show for you today, talking fed's own Harry Littman stops

8
00:00:25,840 --> 00:00:29,440
Speaker 1: by to talk about Trump's latest illegal behavior. Then we'll

9
00:00:29,480 --> 00:00:35,279
Speaker 1: talk to better off Lines Ed Zitron about the latest

10
00:00:35,360 --> 00:00:39,000
Speaker 1: AI news. But first the news Somali.

11
00:00:39,280 --> 00:00:41,040
Speaker 2: We have some election news.

12
00:00:41,240 --> 00:00:45,160
Speaker 3: Trump got all his picks that he was excited about

13
00:00:45,200 --> 00:00:47,479
Speaker 3: in a couple of states, showing that he's strong on

14
00:00:47,520 --> 00:00:51,239
Speaker 3: the party, but on the left we have the opposite response,

15
00:00:51,280 --> 00:00:55,480
Speaker 3: where in DC, the socialist that he is threatening one

16
00:00:55,680 --> 00:00:56,520
Speaker 3: there is going to win.

17
00:00:56,920 --> 00:01:04,880
Speaker 1: Yeah, DC is like an eighty percent and democratic right.

18
00:01:04,880 --> 00:01:05,840
Speaker 2: He's been generous.

19
00:01:05,880 --> 00:01:10,000
Speaker 1: I think, yeah, like this is We've got a Los

20
00:01:10,000 --> 00:01:14,160
Speaker 1: Angeles situation here where Spencer Pratt looking for the votes.

21
00:01:14,560 --> 00:01:18,560
Speaker 1: Of course a democratic socialist is going to win. She

22
00:01:18,880 --> 00:01:24,080
Speaker 1: has a lead. She is a member of the city council. Look,

23
00:01:24,160 --> 00:01:27,400
Speaker 1: we have a democratic Socialist in New York, and he

24
00:01:27,680 --> 00:01:31,920
Speaker 1: has been pretty great. People are really happy with him.

25
00:01:32,200 --> 00:01:36,080
Speaker 1: He's done a lot of cool stuff. So maybe DC

26
00:01:36,280 --> 00:01:37,280
Speaker 1: will get some of that.

27
00:01:37,720 --> 00:01:39,240
Speaker 2: Yeah, I mean, what do you think though?

28
00:01:39,319 --> 00:01:42,880
Speaker 3: It really feels like Trump is having a great season

29
00:01:42,880 --> 00:01:44,680
Speaker 3: of electing very beatable candidates.

30
00:01:44,680 --> 00:01:45,240
Speaker 2: Am I right?

31
00:01:45,440 --> 00:01:48,160
Speaker 1: Yeah? I mean this has like always been the problem

32
00:01:48,160 --> 00:01:50,960
Speaker 1: with Trump is that he doesn't really care. There's been

33
00:01:50,960 --> 00:01:52,880
Speaker 1: the problem for Republicans. It's a good thing for the

34
00:01:52,880 --> 00:01:55,400
Speaker 1: rest of us. He doesn't really care about any one

35
00:01:55,480 --> 00:02:00,200
Speaker 1: selection but his own. So he has supported people know

36
00:02:00,320 --> 00:02:03,320
Speaker 1: that are just you know, the Roy Moore types. So

37
00:02:03,360 --> 00:02:07,080
Speaker 1: we saw that in we saw that in the in

38
00:02:07,160 --> 00:02:13,239
Speaker 1: the in Georgia, the Georgia's Senate primary that Trump came

39
00:02:13,280 --> 00:02:16,519
Speaker 1: in at the eleventh hour and endorsed a guy called

40
00:02:16,600 --> 00:02:20,680
Speaker 1: Mike Collins, who is very mega and very nutty and

41
00:02:20,760 --> 00:02:24,640
Speaker 1: makes Asoff's campaign for reelection a little easier.

42
00:02:24,960 --> 00:02:25,200
Speaker 2: Yeah.

43
00:02:25,600 --> 00:02:29,440
Speaker 3: So we're seeing the Senate delay j Clayton's nomination for

44
00:02:29,600 --> 00:02:30,440
Speaker 3: Intel director.

45
00:02:30,880 --> 00:02:34,680
Speaker 1: So it's not entirely clear what's happening here, whether Trump

46
00:02:35,000 --> 00:02:38,480
Speaker 1: is trying to get Trump had a sort of laundry

47
00:02:38,520 --> 00:02:42,239
Speaker 1: list of things he wanted with the JA Clayton stuff.

48
00:02:42,639 --> 00:02:45,880
Speaker 1: So basically, Bill Polty is going to be the acting

49
00:02:46,000 --> 00:02:51,919
Speaker 1: DNI until Jake Clayton is confirmed. Both Democrats and Republicans

50
00:02:52,080 --> 00:02:56,960
Speaker 1: decided to jump on the j Clayton confirmation hearings as

51
00:02:57,040 --> 00:03:00,840
Speaker 1: quickly as possible because they hoped that this would end

52
00:03:00,919 --> 00:03:04,760
Speaker 1: up creating a scenario where Poulty didn't get the job

53
00:03:04,800 --> 00:03:09,760
Speaker 1: at all. But Trump, in his infinite whatever weirdness, decided

54
00:03:10,200 --> 00:03:13,520
Speaker 1: that he just wanted to tie a bunch of stuff

55
00:03:13,680 --> 00:03:19,400
Speaker 1: to it, including Phiza and Save Act, which, by the way,

56
00:03:19,560 --> 00:03:23,200
Speaker 1: I am convinced that the reason why Trump is so

57
00:03:23,320 --> 00:03:26,880
Speaker 1: obsessed with the Save Act is because he just knows

58
00:03:26,880 --> 00:03:29,280
Speaker 1: that he can't win any other way. By the way,

59
00:03:29,440 --> 00:03:32,960
Speaker 1: there's no world in which the Save Act gets passed, right,

60
00:03:33,000 --> 00:03:35,400
Speaker 1: because they'd have to kill the filibuster in order to

61
00:03:35,440 --> 00:03:38,040
Speaker 1: do that, and everyone has said they have no The

62
00:03:38,080 --> 00:03:41,480
Speaker 1: Republicans have said there's no appetite to kill the filibuster.

63
00:03:41,760 --> 00:03:45,240
Speaker 1: So I don't know exactly what Trump is thinking here.

64
00:03:45,320 --> 00:03:47,120
Speaker 1: I mean, I think he thinks that if he gets

65
00:03:47,240 --> 00:03:50,360
Speaker 1: Bill Poulti in there, it's good for him. I still

66
00:03:50,400 --> 00:03:53,080
Speaker 1: am not convinced, but we'll definitely see more of this.

67
00:03:53,360 --> 00:03:56,480
Speaker 1: You know, this is this continual thing where you know,

68
00:03:56,520 --> 00:04:00,000
Speaker 1: Trump has put everybody in a terrible position as always.

69
00:04:00,560 --> 00:04:03,640
Speaker 3: So you know what I really hope is planning our

70
00:04:03,680 --> 00:04:07,720
Speaker 3: wars for us A racist chatbot designed by Elon Musk.

71
00:04:07,760 --> 00:04:11,360
Speaker 3: But unfortunately that is actually the case that the Pentagon

72
00:04:11,480 --> 00:04:13,280
Speaker 3: used that to target around.

73
00:04:13,640 --> 00:04:15,760
Speaker 1: There are a few people have gotten more from this

74
00:04:15,880 --> 00:04:18,960
Speaker 1: government than Elon Musk. I mean, maybe the Trump boy is,

75
00:04:19,400 --> 00:04:23,480
Speaker 1: but that's it. So top defense officials say data centers

76
00:04:23,760 --> 00:04:29,760
Speaker 1: powering trillionaire's chatbots are critical to national security, and they

77
00:04:30,000 --> 00:04:35,040
Speaker 1: used glack Ai to fire two thousand missiles at Iranians.

78
00:04:35,360 --> 00:04:37,680
Speaker 1: You know what we're going to see here. So this

79
00:04:37,800 --> 00:04:42,760
Speaker 1: is from this lawsuit about Elon's data centers illegally polluting

80
00:04:42,800 --> 00:04:46,480
Speaker 1: black communities. This is like so on the nose right,

81
00:04:46,560 --> 00:04:51,480
Speaker 1: Trump's DOJ defends Musks Colossus data centers despite the fact

82
00:04:51,520 --> 00:04:56,000
Speaker 1: that it threatens black communities. Elon Musk, the guy who's

83
00:04:56,200 --> 00:05:00,800
Speaker 1: anti anti racism and feels the real all victims of

84
00:05:00,880 --> 00:05:06,880
Speaker 1: racism are white people. That guy is illegally polluting black communities.

85
00:05:07,080 --> 00:05:10,479
Speaker 1: Who could have seen this coming? Everyone, everyone could have

86
00:05:10,520 --> 00:05:15,800
Speaker 1: seen this coming. Again, what's interesting about this is that

87
00:05:16,160 --> 00:05:19,919
Speaker 1: nobody knows who. I mean, certainly some people know, but

88
00:05:20,279 --> 00:05:23,479
Speaker 1: you know, none of us sort of casual observers to

89
00:05:24,080 --> 00:05:29,000
Speaker 1: the Defense Department know what was used to fire more

90
00:05:29,040 --> 00:05:33,599
Speaker 1: than two thousand munitions at two thousand distinct targets within

91
00:05:33,680 --> 00:05:36,839
Speaker 1: ninety six hours. And we do know that we did

92
00:05:37,120 --> 00:05:42,080
Speaker 1: hit a girls school, and US military investigators believe American

93
00:05:42,120 --> 00:05:45,880
Speaker 1: forces were likely responsible for these strike on an Iranian

94
00:05:45,920 --> 00:05:48,520
Speaker 1: girls school that killed at leads to one hundred and seventy

95
00:05:48,520 --> 00:05:56,680
Speaker 1: five people, mostly children, mostly girls, Iranian girls. And you'll

96
00:05:56,720 --> 00:05:59,240
Speaker 1: remember that when Trump was asked about it, he said

97
00:05:59,279 --> 00:06:02,600
Speaker 1: it wasn't him, wasn't us, It was definitely us. And

98
00:06:02,720 --> 00:06:08,159
Speaker 1: so there is this weird scenario where these guys are

99
00:06:08,440 --> 00:06:15,479
Speaker 1: actually admitting that rock powered AI took out a girls' school.

100
00:06:15,480 --> 00:06:17,600
Speaker 1: I'm sure that's not what's happening though.

101
00:06:17,960 --> 00:06:19,520
Speaker 2: No, it doesn't seem like rock at all.

102
00:06:19,560 --> 00:06:21,320
Speaker 3: Like, what does it take it every time, like twenty

103
00:06:21,320 --> 00:06:23,400
Speaker 3: four hours to become a Nazi when they let it

104
00:06:23,440 --> 00:06:24,240
Speaker 3: be unrestrained.

105
00:06:24,480 --> 00:06:28,760
Speaker 1: I am very suspicious of this whole story. But these

106
00:06:28,839 --> 00:06:33,840
Speaker 1: AI data centers, at least the ones that Elon is building,

107
00:06:34,040 --> 00:06:36,480
Speaker 1: are polluting and also racist.

108
00:06:36,839 --> 00:06:40,760
Speaker 3: So speaking of racism, you remember when the Trump campaign

109
00:06:40,880 --> 00:06:43,160
Speaker 3: was trying to convince us, Oh no, it's the Democrats

110
00:06:43,160 --> 00:06:43,760
Speaker 3: were racist.

111
00:06:43,880 --> 00:06:45,520
Speaker 2: Let's do you should have a bleg sit.

112
00:06:45,760 --> 00:06:50,480
Speaker 3: Candice Owens says that the Democratic Party is the real racist. Well,

113
00:06:50,600 --> 00:06:53,480
Speaker 3: as all we've seen did this administration is, of course

114
00:06:53,720 --> 00:06:57,600
Speaker 3: they're blatant racism, and we have another piece of evidence

115
00:06:57,600 --> 00:06:59,720
Speaker 3: of it, as the Trump administration is going to try

116
00:06:59,720 --> 00:07:04,000
Speaker 3: to kill the first reparations program for black Americans Evidston, Illinois.

117
00:07:04,320 --> 00:07:07,400
Speaker 1: Yeah, I mean, I also think a really important part

118
00:07:07,440 --> 00:07:11,960
Speaker 1: of this story is that they don't believe in states rights.

119
00:07:12,080 --> 00:07:15,840
Speaker 1: They always pretend that they care about states rights, but

120
00:07:15,880 --> 00:07:20,920
Speaker 1: they really really really don't. And here is the story

121
00:07:21,440 --> 00:07:27,280
Speaker 1: of the DOJ is going after a reparations program in

122
00:07:27,360 --> 00:07:32,120
Speaker 1: a blue state, Illinois, where in a city in a

123
00:07:32,120 --> 00:07:35,680
Speaker 1: blue state, this is the states rights. States Rights Texas

124
00:07:36,240 --> 00:07:40,120
Speaker 1: was allowed to overturn Roe V. Wade Evanston should be

125
00:07:40,160 --> 00:07:46,560
Speaker 1: allowed to pay reparations like period paragraph and of course

126
00:07:47,040 --> 00:07:51,160
Speaker 1: Harmeat Dylan, who you may remember from being a parking attendant.

127
00:07:51,520 --> 00:07:55,720
Speaker 1: She said, there are sound ways for a city to

128
00:07:55,880 --> 00:08:00,480
Speaker 1: remedy past discrimination or direct resources to its most vulnerableth

129
00:08:00,520 --> 00:08:03,160
Speaker 1: citizens and neighborhoods. I don't think she gives a fuck

130
00:08:03,200 --> 00:08:06,880
Speaker 1: about that. Simply handing out money based on race, however,

131
00:08:07,040 --> 00:08:09,560
Speaker 1: is not the answer. That is not what's happening here.

132
00:08:09,760 --> 00:08:14,120
Speaker 1: Evanston has a lot of twenty million dollars the program

133
00:08:14,200 --> 00:08:18,320
Speaker 1: raised from attacks on legal marijuana, and seven million has

134
00:08:18,440 --> 00:08:21,680
Speaker 1: been distributed for use on down payments, house repairs, and

135
00:08:21,720 --> 00:08:25,240
Speaker 1: paying interest in late penalties on property in the city.

136
00:08:25,440 --> 00:08:28,560
Speaker 1: And again this is I think what's so important about

137
00:08:28,600 --> 00:08:32,320
Speaker 1: this is the Trump administration really does not want to

138
00:08:32,400 --> 00:08:37,240
Speaker 1: set a precedent of any kind of reparations because they

139
00:08:37,400 --> 00:08:42,719
Speaker 1: do not want, in any which way to admit that

140
00:08:42,760 --> 00:08:52,479
Speaker 1: there was racism or tried to make reparations for that racism.

141
00:08:52,559 --> 00:08:55,199
Speaker 1: Harry Littman is a former US attorney in the host

142
00:08:55,240 --> 00:08:57,000
Speaker 1: of the podcast Talking Feds.

143
00:08:57,200 --> 00:08:59,320
Speaker 4: Welcome, Welcome, ed, Thank you for having me.

144
00:08:59,559 --> 00:09:02,200
Speaker 1: You've been an AI skeptic for a while, which as

145
00:09:02,240 --> 00:09:04,440
Speaker 1: have I been, and I'd love you to talk about

146
00:09:04,480 --> 00:09:07,280
Speaker 1: sort of where we are at this moment with AI.

147
00:09:07,720 --> 00:09:10,959
Speaker 4: So are a very hysterical moment for the industry. They

148
00:09:11,040 --> 00:09:14,160
Speaker 4: have spent years promising that this stuff would become this

149
00:09:14,559 --> 00:09:18,520
Speaker 4: autonomous panacea for every capitalist element. It would automate every

150
00:09:18,640 --> 00:09:20,880
Speaker 4: job it would do all these things. In reality, what

151
00:09:20,920 --> 00:09:24,760
Speaker 4: we've got is something that can generate some code sometimes

152
00:09:24,800 --> 00:09:29,720
Speaker 4: that can do a poor man's Google Search, and costs

153
00:09:29,760 --> 00:09:31,800
Speaker 4: more than anything has ever cost in the world.

154
00:09:32,080 --> 00:09:33,720
Speaker 5: Ever, it is larger.

155
00:09:34,000 --> 00:09:37,800
Speaker 4: It is ten fifteen times the size of Amazon Web

156
00:09:37,840 --> 00:09:42,520
Speaker 4: Services as far as costs go, trillion dollars plus already invested,

157
00:09:42,520 --> 00:09:46,199
Speaker 4: they're planning to invest trillions of dollars more data centers

158
00:09:46,240 --> 00:09:49,840
Speaker 4: being built for demand for AI compute that doesn't exist,

159
00:09:50,360 --> 00:09:53,440
Speaker 4: and the two largest companies, Open Ai and Anthropic, lose

160
00:09:53,480 --> 00:09:57,320
Speaker 4: billions of dollars. Every single AI startup is unprofitable and

161
00:09:57,400 --> 00:10:01,520
Speaker 4: none of it resembles any previous era. Will say, oh, well,

162
00:10:01,559 --> 00:10:04,000
Speaker 4: Amazon Web Services lost a lot of money fifty three

163
00:10:04,040 --> 00:10:07,600
Speaker 4: billion dollars over ten years, actually eleven years. Uber lost

164
00:10:07,600 --> 00:10:10,200
Speaker 4: a lot of money thirty two billion dollars over five

165
00:10:10,240 --> 00:10:13,559
Speaker 4: to six years, then they became profitable. Open Ai they

166
00:10:13,600 --> 00:10:16,600
Speaker 4: lost twenty one billion dollars in twenty twenty five. These

167
00:10:16,640 --> 00:10:19,720
Speaker 4: companies burn money, They have no path to profitability, and

168
00:10:19,760 --> 00:10:24,200
Speaker 4: the actual underlying technology large language models. No one can

169
00:10:24,280 --> 00:10:27,000
Speaker 4: prove the actual return on investment. No one can do it.

170
00:10:27,000 --> 00:10:30,880
Speaker 4: It's not measurable, and so everyone's very hysterical right now.

171
00:10:30,920 --> 00:10:34,560
Speaker 4: They're making bigger claims. They're claiming it's too scary to

172
00:10:34,640 --> 00:10:37,719
Speaker 4: release mythos, which has now been banned kind of as

173
00:10:37,720 --> 00:10:39,400
Speaker 4: a result of that, which I'm sure we'll get into.

174
00:10:39,920 --> 00:10:42,200
Speaker 4: Long and short of it is is we're in the

175
00:10:42,240 --> 00:10:45,840
Speaker 4: crazy time. We're in theory the irrational exuberance. But there's

176
00:10:45,880 --> 00:10:48,720
Speaker 4: a much more desperation than I think than during the

177
00:10:48,760 --> 00:10:50,240
Speaker 4: Great Financial Crisis.

178
00:10:50,800 --> 00:10:53,200
Speaker 1: Well there, we haven't gotten to the crisis yet, so

179
00:10:53,920 --> 00:10:57,400
Speaker 1: or we've got to the spending but not the crisis.

180
00:10:57,400 --> 00:11:00,520
Speaker 1: So when I talk to people about AI, they say

181
00:11:00,559 --> 00:11:04,160
Speaker 1: to me, is they say it's getting better and better

182
00:11:04,480 --> 00:11:08,120
Speaker 1: and better, and so it's in a process of getting better,

183
00:11:08,480 --> 00:11:11,840
Speaker 1: and it will get so much better that eventually its

184
00:11:11,920 --> 00:11:14,840
Speaker 1: betterness will get better. Right that it'll be a sort

185
00:11:14,880 --> 00:11:20,560
Speaker 1: of infinite loop of smartness that will get itself smarter

186
00:11:20,679 --> 00:11:23,839
Speaker 1: that will continue. Is there any truth to that?

187
00:11:24,320 --> 00:11:24,960
Speaker 5: To be clear?

188
00:11:25,240 --> 00:11:28,839
Speaker 4: When they say better AI models are measured on benchmarks

189
00:11:29,040 --> 00:11:32,600
Speaker 4: specifically made for large language models because you can't really

190
00:11:32,640 --> 00:11:35,320
Speaker 4: rely on them to do real tasks. You can't just

191
00:11:35,360 --> 00:11:37,600
Speaker 4: be like, go and do my job, because they don't

192
00:11:37,640 --> 00:11:40,880
Speaker 4: do that. They are generative models. That create text or

193
00:11:41,000 --> 00:11:44,320
Speaker 4: images in some cases, and they can sometimes do Excel spreadsheets.

194
00:11:44,880 --> 00:11:47,000
Speaker 4: But the thing is when they say sometimes it's because

195
00:11:47,000 --> 00:11:50,720
Speaker 4: of this inherent think of hallucinations, because these models are probabilistic.

196
00:11:51,000 --> 00:11:53,840
Speaker 4: Even at their most complex level. What they're doing is

197
00:11:53,880 --> 00:11:55,920
Speaker 4: working out what the most likely thing they're meant to

198
00:11:55,920 --> 00:11:59,120
Speaker 4: do is. They don't have consciousness, they don't think, So

199
00:11:59,160 --> 00:12:02,560
Speaker 4: those hallucinations are when they authoritatively state something or do

200
00:12:02,640 --> 00:12:05,680
Speaker 4: something that isn't true or the ideal action long story

201
00:12:05,720 --> 00:12:09,320
Speaker 4: short is getting better is not something that anyone's able

202
00:12:09,360 --> 00:12:12,319
Speaker 4: to measure in an RO license in the return on

203
00:12:12,440 --> 00:12:15,760
Speaker 4: investment sense. They can write code. The code might work,

204
00:12:15,840 --> 00:12:18,400
Speaker 4: the code might not work. But code is a much

205
00:12:18,440 --> 00:12:22,160
Speaker 4: more complex thing than just thing that does thing and

206
00:12:22,200 --> 00:12:26,520
Speaker 4: then program happens. It is especially for larger companies. A

207
00:12:27,160 --> 00:12:30,240
Speaker 4: long term project which needs to be read by other people,

208
00:12:30,440 --> 00:12:32,840
Speaker 4: needs to be understood by other people. It's meant to

209
00:12:32,880 --> 00:12:35,240
Speaker 4: have some degree of commentary that tells you why it

210
00:12:35,320 --> 00:12:38,240
Speaker 4: was built. Except you're writing this code using an LLM

211
00:12:38,280 --> 00:12:42,000
Speaker 4: with no intention but no thoughts. This eventually becomes cumbersome,

212
00:12:42,120 --> 00:12:44,240
Speaker 4: It becomes over verbose, just meaning there's a lot of

213
00:12:44,240 --> 00:12:47,520
Speaker 4: code and doesn't need to be. And also the more

214
00:12:47,600 --> 00:12:49,800
Speaker 4: you use a large language model as a software engineer,

215
00:12:49,840 --> 00:12:52,160
Speaker 4: the less you know about writing software, the more dependent

216
00:12:52,160 --> 00:12:54,560
Speaker 4: you become on it. This doesn't mean that it's writing

217
00:12:54,600 --> 00:12:57,439
Speaker 4: good code. It just means you, as a person, are

218
00:12:57,840 --> 00:13:01,280
Speaker 4: less able to write it yourself. So you'll notice that

219
00:13:01,280 --> 00:13:05,320
Speaker 4: I'm only mentioning code, and that's because it's okay with code.

220
00:13:05,360 --> 00:13:07,880
Speaker 4: There are a guy called Carle Brown from the Internet

221
00:13:07,960 --> 00:13:10,000
Speaker 4: Bugs describes it well as it makes the easy things

222
00:13:10,040 --> 00:13:12,480
Speaker 4: easier and the harder things harder. Very simple things it

223
00:13:12,520 --> 00:13:15,400
Speaker 4: can do. The problem is with large language models is

224
00:13:15,440 --> 00:13:16,720
Speaker 4: they're not very good at anything.

225
00:13:17,040 --> 00:13:17,679
Speaker 5: You can't there.

226
00:13:17,880 --> 00:13:20,200
Speaker 4: There are not many jobs that are just texts. They're

227
00:13:20,240 --> 00:13:23,080
Speaker 4: not good at writing. You've got a lot of slop

228
00:13:23,120 --> 00:13:26,000
Speaker 4: on substack as a result, they've really eroded the quality

229
00:13:26,040 --> 00:13:29,480
Speaker 4: of substack. But large language models are not good at jobs.

230
00:13:29,840 --> 00:13:33,080
Speaker 4: They're not good at deterministic actions or a replicable one.

231
00:13:33,480 --> 00:13:38,200
Speaker 4: They're not really good at anything, which is why no

232
00:13:38,240 --> 00:13:40,320
Speaker 4: one seems to be able to actually show any proof

233
00:13:40,320 --> 00:13:42,680
Speaker 4: that they're replacing jobs or that they're good at something.

234
00:13:42,960 --> 00:13:45,240
Speaker 4: You hear a lot of a theemo of being like, oh,

235
00:13:45,320 --> 00:13:48,559
Speaker 4: it's right, Spotify CEO said, oh, our best code is

236
00:13:48,559 --> 00:13:51,080
Speaker 4: are not writing any tode anymore. You'll notice that doesn't

237
00:13:51,120 --> 00:13:55,360
Speaker 4: result in better software, more stable software across the board

238
00:13:55,400 --> 00:13:59,880
Speaker 4: use any app. Everything feels less stable, everything feels more brittle.

239
00:14:00,040 --> 00:14:02,560
Speaker 4: And that's partly because of AI and partly because the

240
00:14:02,600 --> 00:14:06,960
Speaker 4: companies themselves don't really care anymore. So you can't really

241
00:14:07,520 --> 00:14:10,040
Speaker 4: do things with AI, despite all the promises, despite all

242
00:14:10,080 --> 00:14:10,840
Speaker 4: the data centers.

243
00:14:11,040 --> 00:14:14,480
Speaker 1: Okay, so this story is cybersecurity executives urged the Trump

244
00:14:14,559 --> 00:14:20,360
Speaker 1: administration to ease restrictions on Entropic. Can you explain what

245
00:14:20,720 --> 00:14:25,920
Speaker 1: is happening with Anthropic and why that has I mean,

246
00:14:26,080 --> 00:14:29,200
Speaker 1: is that just partisan or is that that they're actually

247
00:14:29,240 --> 00:14:30,080
Speaker 1: cooking something?

248
00:14:30,800 --> 00:14:30,880
Speaker 2: So?

249
00:14:31,400 --> 00:14:34,400
Speaker 4: Okay, So to explain rolling back a few years twenty

250
00:14:34,480 --> 00:14:37,200
Speaker 4: twenty three, both Altman and Sam Ltman of Open Ai

251
00:14:37,320 --> 00:14:40,480
Speaker 4: Dario Ami Day of Anthropic both were saying, Oh, our

252
00:14:40,520 --> 00:14:42,960
Speaker 4: models are too scary, They're too powerful, They're too scary.

253
00:14:43,000 --> 00:14:47,000
Speaker 4: They spent years building this up. Anthropic particularly nasty about this,

254
00:14:47,080 --> 00:14:50,560
Speaker 4: warryo Ama Day, of course, saying the fifty percent of

255
00:14:50,560 --> 00:14:51,200
Speaker 4: white collar.

256
00:14:51,080 --> 00:14:52,840
Speaker 5: Jobs be eliminated in a few years.

257
00:14:53,120 --> 00:14:56,320
Speaker 4: The reason I mentioned that is back in April, they

258
00:14:56,360 --> 00:14:59,040
Speaker 4: announced something called Mythos, a new model that they claimed

259
00:14:59,200 --> 00:15:03,200
Speaker 4: was too powerful to launch because of its cybersecurity capabilities.

260
00:15:03,280 --> 00:15:07,960
Speaker 4: Now independent researchers have now said, no, it's not good.

261
00:15:08,200 --> 00:15:10,160
Speaker 4: It's good at finding some things. It has a lot

262
00:15:10,200 --> 00:15:12,200
Speaker 4: of false positives, meaning it claims there's a bug when

263
00:15:12,200 --> 00:15:15,120
Speaker 4: there isn't. But putting all that aside, anthropics said it's

264
00:15:15,120 --> 00:15:17,480
Speaker 4: too powerful, we have to meter it out only a

265
00:15:17,520 --> 00:15:20,880
Speaker 4: few trusted organizations. A month later they're like, okay, we're

266
00:15:20,880 --> 00:15:24,680
Speaker 4: giving to one hundred and fifty more organizations. Then in

267
00:15:24,760 --> 00:15:28,200
Speaker 4: early June I believe, they announced something called Fable. Now

268
00:15:28,680 --> 00:15:32,440
Speaker 4: Claude Fable was Mythos, but you couldn't do any cybersecurity

269
00:15:32,480 --> 00:15:34,520
Speaker 4: on it. If you asked, it refused, it would drop

270
00:15:34,560 --> 00:15:37,680
Speaker 4: down to another model. A few days ago, a week

271
00:15:37,720 --> 00:15:40,080
Speaker 4: I think before this podcast will go out, there was

272
00:15:40,160 --> 00:15:44,080
Speaker 4: an announcement that by Anthropics saying we've had to cut

273
00:15:44,120 --> 00:15:47,640
Speaker 4: access to Fable and cut access to Mythos due to

274
00:15:47,680 --> 00:15:50,760
Speaker 4: the government putting export controls on them because the government

275
00:15:50,880 --> 00:15:54,119
Speaker 4: was told by Amazon CEO Andy Jesse.

276
00:15:54,760 --> 00:15:57,120
Speaker 5: That apparently was possible to jail break Fable.

277
00:15:57,200 --> 00:16:00,240
Speaker 4: So by which I mean, tell Fable something to give

278
00:16:00,280 --> 00:16:05,160
Speaker 4: you the power of mythos. Semophore reported that maybe China

279
00:16:05,200 --> 00:16:08,200
Speaker 4: got access. It's not clear, so what happened maybe partisan.

280
00:16:09,160 --> 00:16:12,200
Speaker 4: It may indeed, because there was a huge battle between

281
00:16:12,240 --> 00:16:14,240
Speaker 4: them a few months ago with the Department of Defense.

282
00:16:14,320 --> 00:16:18,800
Speaker 4: But partisan or not, they're blocking it because Anthropics spent

283
00:16:19,120 --> 00:16:21,920
Speaker 4: years saying our models are big and scary. Be scared

284
00:16:21,960 --> 00:16:24,080
Speaker 4: of the be scared of the scary models. Be scared,

285
00:16:24,240 --> 00:16:27,440
Speaker 4: and someone took them seriously. Now there probably is a

286
00:16:27,440 --> 00:16:30,680
Speaker 4: bit of hex Hegseth did a victory lap on Twitter

287
00:16:31,320 --> 00:16:34,320
Speaker 4: about it, So of course there is a degree of partisanship,

288
00:16:34,640 --> 00:16:36,880
Speaker 4: but the same and actually some of the reporting have

289
00:16:36,920 --> 00:16:39,320
Speaker 4: suggested it might be as well. But at the same time,

290
00:16:39,920 --> 00:16:41,800
Speaker 4: it does come back to this thing of the AI

291
00:16:41,880 --> 00:16:44,320
Speaker 4: industry has sold itself on fear. I think they thought

292
00:16:44,320 --> 00:16:46,560
Speaker 4: that they would just scare everyone into handing over their

293
00:16:46,560 --> 00:16:50,560
Speaker 4: credit cards, when actually what happened was they ended up

294
00:16:50,560 --> 00:16:53,920
Speaker 4: getting banned. And this is actually interesting because they could

295
00:16:54,000 --> 00:16:58,880
Speaker 4: theoretically get unbanned, but apparently Anthropic is refusing to fix

296
00:16:58,920 --> 00:17:02,360
Speaker 4: the jail break. Again it's coming from the government, who knows.

297
00:17:02,720 --> 00:17:06,360
Speaker 1: Your theory of the case is that this is all

298
00:17:06,560 --> 00:17:11,159
Speaker 1: just bullshit like that. It's not so powerful that it

299
00:17:11,200 --> 00:17:16,720
Speaker 1: needs regulation. It's just so scammy that it needs the

300
00:17:16,760 --> 00:17:22,240
Speaker 1: government to believe that it needs regulation because investors will

301
00:17:22,280 --> 00:17:23,680
Speaker 1: pour more money into it that.

302
00:17:23,680 --> 00:17:26,959
Speaker 4: Way, Because if they had to talk about what the

303
00:17:27,000 --> 00:17:29,480
Speaker 4: thing he did today, they would have to be like, yeah,

304
00:17:29,520 --> 00:17:31,760
Speaker 4: it can do some code generation, and it can look

305
00:17:31,800 --> 00:17:34,320
Speaker 4: like it can summarize documents, and it can write emails.

306
00:17:34,440 --> 00:17:36,919
Speaker 4: It's quite boring when you talk about what it can do,

307
00:17:37,200 --> 00:17:39,960
Speaker 4: so they only I wish they'd regulate it so they

308
00:17:39,960 --> 00:17:41,040
Speaker 4: can only talk about.

309
00:17:40,800 --> 00:17:42,480
Speaker 5: What it can do. I'm tired of hearing about what

310
00:17:42,520 --> 00:17:42,960
Speaker 5: it will do.

311
00:17:43,440 --> 00:17:46,440
Speaker 4: But this is the thing, if they had to describe

312
00:17:46,480 --> 00:17:49,359
Speaker 4: it in today's terms, how would they possibly justify the

313
00:17:49,440 --> 00:17:53,040
Speaker 4: data centers? How would they possibly justify the valuations. The

314
00:17:53,080 --> 00:17:56,000
Speaker 4: whole economy is run on k fabe right now, the

315
00:17:56,000 --> 00:17:59,080
Speaker 4: whole restling thing of pretending this is big and serious

316
00:17:59,119 --> 00:18:02,639
Speaker 4: and also scared when we're dealing with very boring normal

317
00:18:02,680 --> 00:18:05,800
Speaker 4: cloud software. Except because and I have this theory called

318
00:18:05,800 --> 00:18:08,679
Speaker 4: the rock com bubble, where the tech industry is run.

319
00:18:08,560 --> 00:18:09,320
Speaker 5: Out of big ideas.

320
00:18:09,359 --> 00:18:12,040
Speaker 4: They don't have a smartphone, they don't have a new

321
00:18:12,080 --> 00:18:14,640
Speaker 4: cloud computing, they don't have a new thing to sell us,

322
00:18:14,920 --> 00:18:17,520
Speaker 4: a new thing to grow forever. So they have to

323
00:18:17,520 --> 00:18:19,640
Speaker 4: pretend that this is it because they've got nothing else.

324
00:18:19,800 --> 00:18:23,080
Speaker 1: And because the Magnificent seven drives so much of our economy,

325
00:18:23,520 --> 00:18:27,520
Speaker 1: if they can't grow, that means that the whole economy

326
00:18:27,560 --> 00:18:29,200
Speaker 1: doesn't grow exactly.

327
00:18:29,400 --> 00:18:31,639
Speaker 4: And in Video being the largest company on the stock

328
00:18:31,680 --> 00:18:35,960
Speaker 4: market is extremely scary because it's not just that they

329
00:18:36,000 --> 00:18:37,600
Speaker 4: have to sell a lot of GPUs.

330
00:18:38,400 --> 00:18:41,440
Speaker 1: It's explained to our listeners because remember this is we're

331
00:18:41,720 --> 00:18:44,440
Speaker 1: politics podcast, so you need to explain a little bit

332
00:18:44,440 --> 00:18:46,000
Speaker 1: more about NA Video what it does.

333
00:18:46,680 --> 00:18:50,240
Speaker 4: So in Video is a company that makes various things,

334
00:18:50,280 --> 00:18:52,480
Speaker 4: but the big thing that they're making money on is

335
00:18:52,520 --> 00:18:54,800
Speaker 4: something called a GPU. Now you don't need to know

336
00:18:54,880 --> 00:18:57,080
Speaker 4: much about it other than the fact that these things

337
00:18:57,119 --> 00:19:00,439
Speaker 4: are really expensive fifty to seventy grand apiece and you

338
00:19:00,520 --> 00:19:03,080
Speaker 4: need thousands of them in a big data center, multi

339
00:19:03,080 --> 00:19:05,399
Speaker 4: billion dollar data center, which takes a ton of power.

340
00:19:05,920 --> 00:19:08,760
Speaker 4: You do two things with them. You run inference, which

341
00:19:08,800 --> 00:19:10,560
Speaker 4: is the process of creating the output.

342
00:19:10,640 --> 00:19:11,840
Speaker 5: You ask chat.

343
00:19:11,560 --> 00:19:13,200
Speaker 4: GPT for a picture of Scooby Doo with a gun,

344
00:19:13,280 --> 00:19:15,399
Speaker 4: it spits out that's inference.

345
00:19:15,119 --> 00:19:18,080
Speaker 1: And then that does seem to be the only use

346
00:19:18,160 --> 00:19:20,520
Speaker 1: For most it's just the Scooby deal with a guy.

347
00:19:20,520 --> 00:19:21,520
Speaker 5: Scooby Doo with a gun.

348
00:19:21,680 --> 00:19:24,520
Speaker 4: But nevertheless, there's also training, which is both feeding the

349
00:19:24,560 --> 00:19:27,920
Speaker 4: model's information to build them and then training them after

350
00:19:27,960 --> 00:19:30,520
Speaker 4: the fact, so they spit out an output and you

351
00:19:30,520 --> 00:19:34,359
Speaker 4: say that's good or that's bad. Very simple. So the

352
00:19:34,400 --> 00:19:36,680
Speaker 4: reason that nvideo has been doing so well is everyone's

353
00:19:36,720 --> 00:19:42,000
Speaker 4: AI crazy, and all of the Magnificent seven, so primarily Meta, Google, Microsoft,

354
00:19:42,040 --> 00:19:45,639
Speaker 4: and Amazon have been buying billions of dollars of GPUs

355
00:19:45,720 --> 00:19:48,359
Speaker 4: hundreds of billions at this point, so that they can

356
00:19:48,400 --> 00:19:51,320
Speaker 4: build these AI data centers so that they hope that

357
00:19:51,400 --> 00:19:53,639
Speaker 4: something happens. I know, it sounds like a flattening of

358
00:19:53,680 --> 00:19:56,600
Speaker 4: the situation, but really it is a case of so

359
00:19:56,680 --> 00:19:59,639
Speaker 4: that they will have AI and AI will do something

360
00:19:59,640 --> 00:20:02,760
Speaker 4: and make that's vague because that's about as far as

361
00:20:02,760 --> 00:20:06,800
Speaker 4: they've gone. The majority of their AI revenue. Meta doesn't

362
00:20:06,840 --> 00:20:10,760
Speaker 4: really have AI revenue. They're screwed, Amazon, Google, Microsoft. Most

363
00:20:10,800 --> 00:20:14,600
Speaker 4: of their revenue is open AI and anthropic paying them

364
00:20:14,800 --> 00:20:19,119
Speaker 4: for to rent GPUs. So to be clear, the stock

365
00:20:19,200 --> 00:20:22,320
Speaker 4: market is really heavily weighted.

366
00:20:22,000 --> 00:20:24,600
Speaker 5: On Nvidia and in video's continued growth.

367
00:20:24,720 --> 00:20:27,840
Speaker 4: Then the reason that everyone's buying these GPUs is they

368
00:20:27,840 --> 00:20:31,199
Speaker 4: think that there's massive demand for AI. In reality, the

369
00:20:31,240 --> 00:20:34,240
Speaker 4: only real demand for AI is two companies, Open Ai

370
00:20:34,320 --> 00:20:37,080
Speaker 4: and Anthropic, who both lose.

371
00:20:36,920 --> 00:20:37,800
Speaker 5: Billions of dollars.

372
00:20:38,320 --> 00:20:41,560
Speaker 4: Doesn't there's no other The demand does not exist outside

373
00:20:41,600 --> 00:20:45,360
Speaker 4: of these companies. And it's extremely worrying to me because

374
00:20:46,240 --> 00:20:49,160
Speaker 4: all the data centers being built. You know that thing

375
00:20:49,160 --> 00:20:52,120
Speaker 4: you said earlier about how oh it's going to get

376
00:20:52,160 --> 00:20:54,760
Speaker 4: better is going to get better. The data centers are

377
00:20:54,760 --> 00:20:57,480
Speaker 4: going to are being built because people are saying, well,

378
00:20:57,480 --> 00:21:01,280
Speaker 4: there's going to be this demand, right, it's gonna be demand.

379
00:21:01,359 --> 00:21:06,919
Speaker 4: For sure, it doesn't exist. It's minuscule, and it's genuinely

380
00:21:07,000 --> 00:21:07,840
Speaker 4: quite worrying to me.

381
00:21:08,280 --> 00:21:12,720
Speaker 1: What's interesting about this problem, in my mind is that

382
00:21:13,440 --> 00:21:16,160
Speaker 1: it's different than the problem that everyone has been telling

383
00:21:16,200 --> 00:21:18,480
Speaker 1: me about. So the problem like, you know, I was

384
00:21:18,520 --> 00:21:21,560
Speaker 1: talking to a senator yesterday and he was saying like that,

385
00:21:21,920 --> 00:21:25,440
Speaker 1: unless we regulate AI, it's gonna take all the white

386
00:21:25,480 --> 00:21:28,080
Speaker 1: collar jobs, which is what we've heard again and again.

387
00:21:28,680 --> 00:21:33,720
Speaker 1: But there's this scenario where we never even get there

388
00:21:33,840 --> 00:21:37,879
Speaker 1: because it can't take all the I mean it can't

389
00:21:37,880 --> 00:21:40,199
Speaker 1: take all the jobs. It takes a very the you know,

390
00:21:40,280 --> 00:21:43,880
Speaker 1: it does the very low level software coding. But that's it.

391
00:21:44,560 --> 00:21:45,680
Speaker 5: Yeah, And that's the thing.

392
00:21:46,440 --> 00:21:49,800
Speaker 4: It frustrates me that the Bernie Sounder's piece in the

393
00:21:49,840 --> 00:21:51,879
Speaker 4: New York Times really frustrated me. Oh, we need to

394
00:21:51,920 --> 00:21:54,000
Speaker 4: do We need to do fifty percent so we can

395
00:21:54,040 --> 00:21:56,199
Speaker 4: share in the profits of AI. Let me know when

396
00:21:56,240 --> 00:21:59,520
Speaker 4: you find those, Bernie. Everyone has got themselves tied up

397
00:21:59,680 --> 00:22:02,399
Speaker 4: with this idea if the white collar apocalypse the white

398
00:22:02,400 --> 00:22:06,000
Speaker 4: collar job apocalypse. But that whole thing is a marketing

399
00:22:06,200 --> 00:22:10,680
Speaker 4: it's a marketing movement. It's not connected to any actual layoffs.

400
00:22:10,880 --> 00:22:13,360
Speaker 4: There was massive overhiring in twenty twenty one and twenty

401
00:22:13,440 --> 00:22:16,879
Speaker 4: twenty two across the economy. These layoffs are an example

402
00:22:17,320 --> 00:22:20,480
Speaker 4: of what happens when you overhire repeatedly. And they're also

403
00:22:20,560 --> 00:22:23,040
Speaker 4: hiring tons of people too. A lot of companies have

404
00:22:23,119 --> 00:22:25,639
Speaker 4: increased their hiring as software engineers. Why would they do

405
00:22:25,680 --> 00:22:27,920
Speaker 4: that when code is automated. Well, the answer is it's

406
00:22:27,960 --> 00:22:32,960
Speaker 4: not being automated. So the white collar job apocalypse is

407
00:22:33,119 --> 00:22:36,560
Speaker 4: entirely marketing. It is built to market the idea that

408
00:22:36,600 --> 00:22:40,440
Speaker 4: AI is inevitable because if they can't say that, then

409
00:22:40,480 --> 00:22:43,119
Speaker 4: they can't sell their stuff at the scale they need to.

410
00:22:43,359 --> 00:22:47,920
Speaker 4: And honestly, I'm disappointed but ozed by the government's response

411
00:22:48,000 --> 00:22:51,000
Speaker 4: of well, we've vaguely got to do something about this,

412
00:22:51,080 --> 00:22:53,280
Speaker 4: maybe we need to do we need to regulate this,

413
00:22:53,400 --> 00:22:57,080
Speaker 4: and then the regulation never actually seems to happen or

414
00:22:57,119 --> 00:22:59,760
Speaker 4: indeed apply to the things happening today.

415
00:23:00,400 --> 00:23:10,399
Speaker 1: Well, they have never regulated any technology like Facebook, Instagram, naa, YouTube,

416
00:23:10,760 --> 00:23:15,359
Speaker 1: Nothing is regulated. There hasn't been any regulation at all. Yeah,

417
00:23:15,400 --> 00:23:17,840
Speaker 1: so they're not going to do I mean, what are

418
00:23:17,880 --> 00:23:22,160
Speaker 1: the sort of smart things that their companies have figured

419
00:23:22,160 --> 00:23:26,520
Speaker 1: out is that they asked for regulation, and since Congress

420
00:23:26,600 --> 00:23:29,399
Speaker 1: is so incompetent, they won't ever get it anyway, and

421
00:23:29,480 --> 00:23:31,720
Speaker 1: if they do, it would be it would end up

422
00:23:31,760 --> 00:23:35,639
Speaker 1: being a barrier for small or companies anyway, because they

423
00:23:35,680 --> 00:23:37,520
Speaker 1: would likely set some of the regulation.

424
00:23:38,280 --> 00:23:41,119
Speaker 4: But what's funny funny you mentioned that one of the

425
00:23:41,160 --> 00:23:44,840
Speaker 4: other myths about all of this is that anthropic and

426
00:23:44,880 --> 00:23:49,400
Speaker 4: open aire even startups, they're not. They may be funded

427
00:23:49,440 --> 00:23:53,120
Speaker 4: like them, discussed like them. Their infrastructure, So the data

428
00:23:53,160 --> 00:23:56,600
Speaker 4: centers that run their services from the very beginning were

429
00:23:56,600 --> 00:24:00,320
Speaker 4: built by Microsoft, Google and Amazon. They are rich kids.

430
00:24:00,560 --> 00:24:02,720
Speaker 4: About one hundred and fifty billion. Our estimate is what

431
00:24:02,760 --> 00:24:05,760
Speaker 4: their actual infrastructure cost. No one talks about that because

432
00:24:05,800 --> 00:24:07,440
Speaker 4: everyone says, oh, they're plucky startups.

433
00:24:07,560 --> 00:24:08,000
Speaker 5: No they're not.

434
00:24:08,400 --> 00:24:10,800
Speaker 4: There's some series of the largest companies in the world,

435
00:24:10,880 --> 00:24:13,119
Speaker 4: and of course they ask for regulation. They do it

436
00:24:13,160 --> 00:24:16,800
Speaker 4: in this mealy mouthed way that is always marketing. Anthropic

437
00:24:16,880 --> 00:24:19,200
Speaker 4: mentioned the other day, just before getting banned by the

438
00:24:19,240 --> 00:24:21,399
Speaker 4: way that they think that the government should have the

439
00:24:21,480 --> 00:24:24,840
Speaker 4: option to pause a model in case it's too dangerous.

440
00:24:25,000 --> 00:24:27,360
Speaker 4: But you're also notice in that piece it said, oh,

441
00:24:27,520 --> 00:24:30,200
Speaker 4: but we're going to do recursive self improvement, which means

442
00:24:30,240 --> 00:24:33,760
Speaker 4: an AI that trains itself. It's all marketing. It's all

443
00:24:33,800 --> 00:24:36,800
Speaker 4: an attempt to deceive and make you think of a

444
00:24:36,840 --> 00:24:40,240
Speaker 4: technology that doesn't exist, but judge the one that does

445
00:24:40,280 --> 00:24:43,440
Speaker 4: today as if it's inevitable. It becomes that, as if

446
00:24:43,720 --> 00:24:47,680
Speaker 4: llms will magically become this autonomous, reliable software that they've

447
00:24:47,720 --> 00:24:50,120
Speaker 4: never ever been. There is no sign of it.

448
00:24:50,560 --> 00:24:52,720
Speaker 1: I have talked to like a lot of people where

449
00:24:52,760 --> 00:24:56,159
Speaker 1: they've said, you know, I've said, well, okay, so it

450
00:24:56,160 --> 00:24:58,800
Speaker 1: does coding, and it does writing, it does some writing.

451
00:24:59,400 --> 00:25:02,520
Speaker 1: What about for example, in the Times, there was some

452
00:25:02,640 --> 00:25:04,240
Speaker 1: AI writing that was very good.

453
00:25:06,080 --> 00:25:07,920
Speaker 5: I'm not familiar with the writing.

454
00:25:08,760 --> 00:25:10,280
Speaker 4: I'm the one that was to report.

455
00:25:10,560 --> 00:25:12,160
Speaker 5: I mean also most.

456
00:25:11,960 --> 00:25:15,520
Speaker 1: Of the writing is terrible. But I mean just you know,

457
00:25:16,680 --> 00:25:20,120
Speaker 1: I mean what like I've asked people before what can

458
00:25:20,119 --> 00:25:22,840
Speaker 1: AI do and they said, well, self driving.

459
00:25:22,560 --> 00:25:25,240
Speaker 4: Cars, which is not Just to be clear, those self

460
00:25:25,320 --> 00:25:29,280
Speaker 4: driving cars are not the technology behind open aionanthropic.

461
00:25:28,880 --> 00:25:32,439
Speaker 1: Right, And also aren't there a lot of many of

462
00:25:32,440 --> 00:25:34,960
Speaker 1: the self driving cars actually have somebody behind the wheel.

463
00:25:35,720 --> 00:25:38,120
Speaker 4: If they don't have they do in some cases. When

464
00:25:38,119 --> 00:25:41,520
Speaker 4: they don't, they have a kind of an overwatch that

465
00:25:41,600 --> 00:25:44,480
Speaker 4: can intervene. They can't intervene and like control that's that

466
00:25:44,600 --> 00:25:47,440
Speaker 4: is I believe in myth. They can't like remote control them,

467
00:25:47,440 --> 00:25:49,359
Speaker 4: but they can direct them. But at the same time

468
00:25:49,800 --> 00:25:52,520
Speaker 4: I find it autonomous cars interesting. But I was trying

469
00:25:52,560 --> 00:25:54,640
Speaker 4: to leave a casino in Vegas yesterday and I got

470
00:25:54,640 --> 00:25:57,040
Speaker 4: stuck behind three of the bloody things. Like they're really

471
00:25:57,119 --> 00:25:59,560
Speaker 4: get us that they get stuck. They get confused by

472
00:25:59,600 --> 00:26:01,000
Speaker 4: driving anyway.

473
00:26:01,040 --> 00:26:03,959
Speaker 1: You know. I was talking to an AI a person

474
00:26:03,960 --> 00:26:06,560
Speaker 1: who's you know, very into it, and said, you know,

475
00:26:06,600 --> 00:26:10,360
Speaker 1: they're going to be robots. How would that work well.

476
00:26:10,440 --> 00:26:14,359
Speaker 4: Robots are again a different technology to So, just to

477
00:26:14,359 --> 00:26:17,240
Speaker 4: be clear, the only thing that those data centers are

478
00:26:17,280 --> 00:26:19,600
Speaker 4: getting built for, the only thing that open AI and

479
00:26:19,640 --> 00:26:23,080
Speaker 4: Anthropic are doing is large language models. They are not

480
00:26:23,119 --> 00:26:25,879
Speaker 4: doing anything with robotics, cars or any of it.

481
00:26:25,920 --> 00:26:31,040
Speaker 1: And that's like you steal my book and train yes,

482
00:26:31,280 --> 00:26:33,560
Speaker 1: computers on my book and then.

483
00:26:33,560 --> 00:26:37,120
Speaker 4: Stuff from GitHub and code from online from the what is.

484
00:26:37,080 --> 00:26:40,560
Speaker 1: That and that does what? That trains the AI model

485
00:26:40,720 --> 00:26:45,520
Speaker 1: to write code sort of and write my books, but

486
00:26:45,720 --> 00:26:47,280
Speaker 1: not that well, so that's.

487
00:26:47,119 --> 00:26:48,040
Speaker 5: Called pre training.

488
00:26:48,200 --> 00:26:51,840
Speaker 4: So it's just you take a ton of GPUs, you

489
00:26:51,920 --> 00:26:54,760
Speaker 4: feed a bunch of data into the models, and then

490
00:26:54,800 --> 00:26:57,800
Speaker 4: you keep doing that and theoretically there is there are

491
00:26:57,800 --> 00:27:00,080
Speaker 4: some other processes involved, and theoretically at the end and

492
00:27:00,160 --> 00:27:02,639
Speaker 4: you have a better model. There is also other training.

493
00:27:02,680 --> 00:27:06,120
Speaker 4: But nevertheless, that's that's mostly what you're discussing. Yes, that

494
00:27:06,200 --> 00:27:09,879
Speaker 4: has absolutely nothing to do with robotics or autonomous cars.

495
00:27:10,000 --> 00:27:13,000
Speaker 4: To be clear, both of those are very data intensive,

496
00:27:13,720 --> 00:27:16,879
Speaker 4: but they are not the things when someone mentions AI

497
00:27:16,960 --> 00:27:19,919
Speaker 4: in the news right now, it's not nothing to do

498
00:27:20,000 --> 00:27:24,399
Speaker 4: with it. The reason that Altman and his cronies used

499
00:27:24,440 --> 00:27:27,800
Speaker 4: AI as the thing to do as the term is,

500
00:27:27,840 --> 00:27:31,919
Speaker 4: so that everyone would conflate all of these useful AI things,

501
00:27:31,960 --> 00:27:35,840
Speaker 4: these things that are actually innovative with this large language

502
00:27:35,840 --> 00:27:39,640
Speaker 4: model situation, which by the way, has been while there

503
00:27:39,640 --> 00:27:43,040
Speaker 4: have been improvements, has done basically the same thing for

504
00:27:43,080 --> 00:27:47,160
Speaker 4: the last few years. There is diminishing returns on feeding

505
00:27:47,240 --> 00:27:51,119
Speaker 4: them training data. It's a very strange situation, and I

506
00:27:51,240 --> 00:27:53,800
Speaker 4: think what it is. It's part of a larger problem

507
00:27:54,040 --> 00:27:58,280
Speaker 4: in the world of media literacy and now the authority,

508
00:27:58,600 --> 00:28:02,760
Speaker 4: the authority crisis we have where so much of our

509
00:28:02,800 --> 00:28:06,520
Speaker 4: economy is controlled by people that trust the last smart

510
00:28:06,520 --> 00:28:11,479
Speaker 4: sounding person they had, when actually this this llms are

511
00:28:11,560 --> 00:28:15,840
Speaker 4: a large scale exploitation of ignorance and of trust in

512
00:28:15,880 --> 00:28:16,920
Speaker 4: the wrong authorities.

513
00:28:17,640 --> 00:28:21,560
Speaker 1: It also seems like we're just being like this is

514
00:28:21,920 --> 00:28:24,160
Speaker 1: almost the same as the housing bubble.

515
00:28:24,400 --> 00:28:27,199
Speaker 4: It is, but it's just I really want to be

516
00:28:27,240 --> 00:28:30,199
Speaker 4: clear this. If anyone leaves this, this is not a

517
00:28:30,200 --> 00:28:33,720
Speaker 4: too big to fail situation. Too big to fail referred

518
00:28:33,760 --> 00:28:36,320
Speaker 4: to when the banks were going to collapse after the

519
00:28:36,320 --> 00:28:40,160
Speaker 4: Great Financial Crisis, and it's not top It was not

520
00:28:40,400 --> 00:28:43,880
Speaker 4: the rescue package. It was the repo window and the

521
00:28:43,920 --> 00:28:47,720
Speaker 4: Federal Reserve feeding hundreds of billions of dollars to the

522
00:28:47,720 --> 00:28:50,560
Speaker 4: banks every month, even might have even been every week,

523
00:28:50,800 --> 00:28:54,120
Speaker 4: because the underlying loan industry kind of would have collapsed

524
00:28:54,160 --> 00:28:57,600
Speaker 4: otherwise in this case, like the actual economics of this

525
00:28:57,640 --> 00:28:59,480
Speaker 4: situation are very strange.

526
00:28:59,720 --> 00:29:00,280
Speaker 5: In the up.

527
00:29:01,120 --> 00:29:04,000
Speaker 4: The money being spent, the money flowing into the economy

528
00:29:04,120 --> 00:29:08,400
Speaker 4: is almost exclusively building data centers and buying Gpusman video

529
00:29:08,680 --> 00:29:12,760
Speaker 4: and the associated hardware and the construction. It's not from

530
00:29:12,840 --> 00:29:16,040
Speaker 4: buying AI services. The actual revenues of these companies are

531
00:29:16,080 --> 00:29:18,760
Speaker 4: not that big. They are kind of a footnote of

532
00:29:18,800 --> 00:29:21,600
Speaker 4: the larger economy. They all lose money, but even their

533
00:29:21,640 --> 00:29:25,880
Speaker 4: revenues are kind of pathetic. And so the Great Financial

534
00:29:25,920 --> 00:29:29,120
Speaker 4: Crisis was horrifying in that a lot of people got

535
00:29:29,120 --> 00:29:30,760
Speaker 4: loans that they shouldn't have, but at least it was

536
00:29:30,760 --> 00:29:34,920
Speaker 4: for houses. At least it was for something with that, yeah, everything.

537
00:29:35,640 --> 00:29:37,920
Speaker 4: It was also surrounded by a consumer debt crisis. But

538
00:29:38,520 --> 00:29:42,760
Speaker 4: putting all that aside, in this case, the great financial

539
00:29:42,840 --> 00:29:46,680
Speaker 4: crisis that could be to come, the stock market bubble

540
00:29:46,720 --> 00:29:49,480
Speaker 4: will resemble the dot com bubble, but I think at

541
00:29:49,520 --> 00:29:51,720
Speaker 4: the side is going to be the collapse of these

542
00:29:51,800 --> 00:29:54,080
Speaker 4: data centers. Because think of it like this, These things

543
00:29:54,120 --> 00:29:56,520
Speaker 4: cost billions of dollars. They take billions of dollars of debt,

544
00:29:56,600 --> 00:29:58,760
Speaker 4: then they build the thing, and then they find out

545
00:29:58,760 --> 00:30:02,280
Speaker 4: that demand isn't there, that loan goes sideways and there

546
00:30:02,320 --> 00:30:03,800
Speaker 4: I think there was one hundred and seventy eight point

547
00:30:03,880 --> 00:30:05,920
Speaker 4: five billion dollars of these things in the US alone

548
00:30:06,000 --> 00:30:10,480
Speaker 4: last year. That is like a one percenter great housing crisis.

549
00:30:10,920 --> 00:30:14,440
Speaker 4: The problem is is that a great deal of that

550
00:30:14,600 --> 00:30:19,520
Speaker 4: was issued by private credit, and private credit is principally

551
00:30:19,520 --> 00:30:25,880
Speaker 4: funded by pensions, by insurance annuities, so teachers, pensions funds.

552
00:30:27,160 --> 00:30:29,520
Speaker 4: I mean, there's also cop One's Fireman Ones. All of

553
00:30:30,000 --> 00:30:34,160
Speaker 4: just a lot of innocent people's money is flowing into

554
00:30:34,240 --> 00:30:37,680
Speaker 4: this industry. And to be clear, you can't really bail

555
00:30:37,760 --> 00:30:40,600
Speaker 4: this out. You can't bail it, You can't bail out

556
00:30:40,640 --> 00:30:45,160
Speaker 4: everything because that you I just don't see how even

557
00:30:45,200 --> 00:30:49,440
Speaker 4: a right wing Congress could possibly clear green lighting hundreds

558
00:30:49,440 --> 00:30:52,680
Speaker 4: of billions of dollars for private credit. But then there's

559
00:30:52,680 --> 00:30:55,400
Speaker 4: the greater problem, which is all of this falls apart.

560
00:30:55,800 --> 00:30:58,000
Speaker 4: You can't bail out the stock market. You can't bail

561
00:30:58,040 --> 00:31:00,680
Speaker 4: out a stock price if in videos read new drops

562
00:31:00,680 --> 00:31:03,320
Speaker 4: because people on buying the GPS to put in data centers.

563
00:31:03,800 --> 00:31:06,440
Speaker 4: That's not a bailout situation. You don't bail that out.

564
00:31:06,800 --> 00:31:10,680
Speaker 4: Open AI runs out of money. Bailing them out doesn't

565
00:31:10,680 --> 00:31:13,600
Speaker 4: fix the greater problem with them constantly losing money, and

566
00:31:13,720 --> 00:31:18,760
Speaker 4: nationalizing them doesn't really make sense. Well, nationalize what exactly.

567
00:31:19,400 --> 00:31:22,200
Speaker 1: Ed. I'm we're way out of time, but I love

568
00:31:22,280 --> 00:31:24,960
Speaker 1: this conversation. I want you to come back. I would

569
00:31:25,000 --> 00:31:28,600
Speaker 1: love to say I continue to agree with this and

570
00:31:28,640 --> 00:31:30,720
Speaker 1: I and you know, it's one of these things where

571
00:31:30,720 --> 00:31:33,200
Speaker 1: it just doesn't make any sense to me. I mean,

572
00:31:33,240 --> 00:31:36,520
Speaker 1: there's definitely money to be made, but thank you, thank you,

573
00:31:36,600 --> 00:31:41,680
Speaker 1: thank you, Ed, thank you for having me. Ed Zitron

574
00:31:41,880 --> 00:31:47,040
Speaker 1: is the editor of Where's Your ed At and the

575
00:31:47,120 --> 00:31:49,720
Speaker 1: host of the Better Offline podcast.

576
00:31:50,040 --> 00:31:54,120
Speaker 6: Hey Hey, hey everyone, it's our latest Molly Harry mashup,

577
00:31:54,120 --> 00:31:57,640
Speaker 6: where Molly john Fast hits me with legal questions and

578
00:31:57,680 --> 00:32:02,280
Speaker 6: I return fire with questions about politics. Molly always great

579
00:32:02,320 --> 00:32:05,640
Speaker 6: to chat with you. I've been really all over this

580
00:32:06,040 --> 00:32:08,280
Speaker 6: slush fund, and I just wanted to ask you about

581
00:32:08,320 --> 00:32:10,640
Speaker 6: the political side of it, because it looked for a

582
00:32:10,720 --> 00:32:14,360
Speaker 6: time that there was a critical mass of Republicans ready

583
00:32:14,400 --> 00:32:17,360
Speaker 6: to push back against it. Then there were a couple

584
00:32:17,520 --> 00:32:21,400
Speaker 6: procedural votes during the vote rama where they came back

585
00:32:21,480 --> 00:32:22,360
Speaker 6: to the fold.

586
00:32:22,680 --> 00:32:23,520
Speaker 7: They haven't said.

587
00:32:23,360 --> 00:32:26,920
Speaker 6: A word about the IRS amnesty for Trump worth one

588
00:32:27,000 --> 00:32:30,680
Speaker 6: hundred million dollars or more, So, you know, having made

589
00:32:30,720 --> 00:32:34,120
Speaker 6: their little protest, are they now coming back into line

590
00:32:34,200 --> 00:32:36,120
Speaker 6: and going to let them get away with it?

591
00:32:36,440 --> 00:32:40,360
Speaker 1: I Mean, the thing with the slush fund is that,

592
00:32:40,840 --> 00:32:43,720
Speaker 1: just like with all of these other things, voters hated.

593
00:32:44,240 --> 00:32:49,920
Speaker 1: So if Republicans lose their will to push back, which

594
00:32:50,400 --> 00:32:55,000
Speaker 1: historically we've seen they have, if that happens, then that's

595
00:32:55,160 --> 00:32:58,360
Speaker 1: actually probably better for Democrats. I mean, I think a

596
00:32:58,400 --> 00:33:02,120
Speaker 1: slush fund is appalling, right. The idea that people who

597
00:33:02,120 --> 00:33:06,120
Speaker 1: beat up cops can be paid with tax dollars because

598
00:33:06,160 --> 00:33:10,080
Speaker 1: Trump wants to reward them for doing something illegal is

599
00:33:10,120 --> 00:33:14,720
Speaker 1: insane if it gets through and Democrats don't run on

600
00:33:14,760 --> 00:33:18,920
Speaker 1: that every day and don't and aren't like, I see

601
00:33:19,000 --> 00:33:23,160
Speaker 1: you voted for the slush fund. That is a payment fund.

602
00:33:23,240 --> 00:33:25,760
Speaker 1: I mean, that's the thing. It's like, it strikes me

603
00:33:26,080 --> 00:33:31,200
Speaker 1: that it is the kind of thing that if voters vote,

604
00:33:31,280 --> 00:33:33,480
Speaker 1: I mean, I just don't understand how they don't vote

605
00:33:33,520 --> 00:33:33,840
Speaker 1: on that.

606
00:33:34,360 --> 00:33:36,040
Speaker 7: And what about the IRS part?

607
00:33:36,040 --> 00:33:36,320
Speaker 2: Of it.

608
00:33:36,440 --> 00:33:38,240
Speaker 7: No, they not talked about up down the hill.

609
00:33:38,520 --> 00:33:45,240
Speaker 1: Right, you're a legal guy, you're a lawyer. It strikes

610
00:33:45,240 --> 00:33:49,240
Speaker 1: me that it's not legal an amnesty. I mean, is

611
00:33:49,280 --> 00:33:54,720
Speaker 1: there any legal precedent for giving someone legal amnesty forever

612
00:33:54,800 --> 00:33:57,600
Speaker 1: and ever and ever from the irs?

613
00:33:57,760 --> 00:33:59,560
Speaker 7: And that is that your question is that.

614
00:33:59,600 --> 00:34:01,440
Speaker 1: We are to question to you.

615
00:34:01,640 --> 00:34:05,040
Speaker 6: Yeah, bring it, you know, because it's a really good question,

616
00:34:05,360 --> 00:34:09,080
Speaker 6: Molly as usual, because you had Blant saying, oh, it's typical,

617
00:34:09,160 --> 00:34:11,359
Speaker 6: we do it all the time. What he means by

618
00:34:11,400 --> 00:34:14,480
Speaker 6: that is when there are real cases and that are

619
00:34:14,520 --> 00:34:16,560
Speaker 6: truly settled, brought.

620
00:34:16,280 --> 00:34:18,759
Speaker 7: By taxpayers who have some value there.

621
00:34:18,960 --> 00:34:22,160
Speaker 6: One of the things the DJ sometimes will do is

622
00:34:22,200 --> 00:34:25,279
Speaker 6: say other things that we're on the table now, like

623
00:34:25,920 --> 00:34:30,479
Speaker 6: possible tax liability, we will give you. So in that sense,

624
00:34:30,520 --> 00:34:34,440
Speaker 6: it was done. Little problem here. There is no votafied

625
00:34:34,600 --> 00:34:38,520
Speaker 6: lawsuit or settlement. It's all given up for either a

626
00:34:38,560 --> 00:34:43,399
Speaker 6: settlement agreement that is complete bogus, or just like Todd

627
00:34:43,400 --> 00:34:46,880
Speaker 6: Blant because he's feeling generous, serving it up to Trump.

628
00:34:46,960 --> 00:34:49,680
Speaker 7: So in that sense, no, I mean, there's no precedent.

629
00:34:49,880 --> 00:34:53,080
Speaker 1: Right, There's never been someone who's like who has been

630
00:34:53,239 --> 00:34:58,239
Speaker 1: granted protection from the irs forever and ever until they

631
00:34:58,280 --> 00:35:00,160
Speaker 1: die and their kids like, that's.

632
00:35:01,000 --> 00:35:01,680
Speaker 7: That's not a thing.

633
00:35:01,719 --> 00:35:04,440
Speaker 6: Although that's not what they're doing purporting to do here.

634
00:35:04,480 --> 00:35:08,040
Speaker 6: What they're saying is anything that exists now that we

635
00:35:08,160 --> 00:35:11,120
Speaker 6: could have gone into, we won't. That could be a

636
00:35:11,160 --> 00:35:14,200
Speaker 6: thing if you were actually settling a real case.

637
00:35:14,600 --> 00:35:17,320
Speaker 7: But it isn't a real case. And they gotta.

638
00:35:17,200 --> 00:35:20,560
Speaker 6: They're into between a rock and a hard place. They

639
00:35:20,640 --> 00:35:23,520
Speaker 6: got to choose if it goes to the course of this.

640
00:35:23,760 --> 00:35:23,960
Speaker 7: Yeah.

641
00:35:25,000 --> 00:35:28,760
Speaker 1: But also the kids too get amnesty, right, Yeah.

642
00:35:28,680 --> 00:35:32,200
Speaker 6: Because they're part of the plaintiffs here. But again, the

643
00:35:32,400 --> 00:35:35,759
Speaker 6: linchpin of it is we're giving you something, namely this

644
00:35:36,440 --> 00:35:40,840
Speaker 6: case we're settling, which is a worthless case in several

645
00:35:40,920 --> 00:35:41,560
Speaker 6: different ways.

646
00:35:41,840 --> 00:35:44,160
Speaker 7: That's the real rabbit in the hat.

647
00:35:44,440 --> 00:35:47,440
Speaker 1: The number of people who had their taxes leaked in

648
00:35:47,560 --> 00:35:50,600
Speaker 1: that group was a huge number of people. Yeah, So

649
00:35:51,320 --> 00:35:55,040
Speaker 1: I mean was hundreds of thousands of people. I mean

650
00:35:55,160 --> 00:35:59,799
Speaker 1: not everyone had their taxes published, but other people did

651
00:36:00,480 --> 00:36:03,440
Speaker 1: and were able to find settlements with the irs. So

652
00:36:03,600 --> 00:36:06,839
Speaker 1: my question is, if Trump gets this, what about all

653
00:36:06,840 --> 00:36:07,439
Speaker 1: the other people?

654
00:36:07,680 --> 00:36:10,480
Speaker 6: Yeah, And in fact it's gone the other way because

655
00:36:10,680 --> 00:36:14,040
Speaker 6: among he's got two problems even besides that it's collusive.

656
00:36:14,280 --> 00:36:17,680
Speaker 6: The first is that he did it too late. It's

657
00:36:17,800 --> 00:36:20,320
Speaker 6: time barred. But the second is the guy he sued,

658
00:36:20,640 --> 00:36:24,440
Speaker 6: the guy who did in fact leak tax information and

659
00:36:24,520 --> 00:36:27,360
Speaker 6: go to jail for it, wasn't an IRS employee. He

660
00:36:27,520 --> 00:36:30,960
Speaker 6: was a contractor and they and they successfully defended on

661
00:36:31,040 --> 00:36:31,560
Speaker 6: that basis.

662
00:36:31,640 --> 00:36:33,160
Speaker 7: That's why the IRS came.

663
00:36:33,080 --> 00:36:36,320
Speaker 6: In twenty five page memo and said, defend against this.

664
00:36:36,560 --> 00:36:39,880
Speaker 6: This is the kind of case that, of course the

665
00:36:39,960 --> 00:36:42,520
Speaker 6: DOJ will push back on, and that would have been

666
00:36:42,560 --> 00:36:46,200
Speaker 6: the professional determination and till it went up to mister

667
00:36:46,440 --> 00:36:50,759
Speaker 6: Todd nominated to be Attorney General Blanche, who just completely

668
00:36:51,560 --> 00:36:54,759
Speaker 6: greenlighted a totally different sort of deal for Trump. And

669
00:36:54,880 --> 00:36:57,480
Speaker 6: I just want to say, those thousands of thousands of people,

670
00:36:57,760 --> 00:37:00,879
Speaker 6: Trump is uniquely the guy who said I'm gonna give

671
00:37:00,920 --> 00:37:01,960
Speaker 6: you all my taxes.

672
00:37:02,000 --> 00:37:04,959
Speaker 7: You know, you would think his injury is really sort

673
00:37:05,000 --> 00:37:06,000
Speaker 7: of honest. Yeah.

674
00:37:06,360 --> 00:37:10,719
Speaker 1: Yeah. I also wonder when we look at this, is

675
00:37:10,760 --> 00:37:14,080
Speaker 1: there anything that you could do at this moment to

676
00:37:14,200 --> 00:37:17,000
Speaker 1: prevent top Blanche from becoming ag.

677
00:37:17,360 --> 00:37:20,319
Speaker 6: That sounds like a litman to fast question, But I'll

678
00:37:20,360 --> 00:37:23,279
Speaker 6: tell you I haven't. I haven't found it closely. He

679
00:37:23,480 --> 00:37:27,120
Speaker 6: really blanched has the special contempt of anyone who was

680
00:37:27,160 --> 00:37:30,320
Speaker 6: ever in DJ because, unlike say Bondi.

681
00:37:30,200 --> 00:37:31,719
Speaker 7: He knows what he's doing.

682
00:37:31,880 --> 00:37:37,680
Speaker 6: He knows the total fundamental you know, Ethos Norms live

683
00:37:37,800 --> 00:37:41,799
Speaker 6: d OJ that he's just steamrolling every day. My best

684
00:37:41,880 --> 00:37:43,879
Speaker 6: sense is, yeah, it's a little bit up in the air.

685
00:37:43,920 --> 00:37:46,880
Speaker 6: I used to work on nominations some and there you know,

686
00:37:47,200 --> 00:37:50,560
Speaker 6: once it gains controversial, all bets are off. He lost

687
00:37:50,760 --> 00:37:53,480
Speaker 6: what could have been his best bet for a Democrat Fetterman,

688
00:37:53,880 --> 00:37:57,480
Speaker 6: till Us and Cornyan are in committee. If they don't

689
00:37:57,520 --> 00:37:59,760
Speaker 6: go with him, it doesn't even get out of committee.

690
00:38:00,000 --> 00:38:02,279
Speaker 6: And then there are the others that Trump brought down.

691
00:38:02,719 --> 00:38:05,680
Speaker 6: And remember you know when he goes to the Hill,

692
00:38:05,840 --> 00:38:10,960
Speaker 6: everything about Epstein, which Pam Bondi told the Congress, Oh

693
00:38:10,960 --> 00:38:14,040
Speaker 6: it was all, it was all Blanche is doing comes

694
00:38:14,120 --> 00:38:17,280
Speaker 6: back to the fore, everything about the slash fund, everything

695
00:38:17,360 --> 00:38:21,640
Speaker 6: about the reprisal prosecutions. So for my money, especially with

696
00:38:21,800 --> 00:38:24,680
Speaker 6: the dynamic that you mentioned, Mollie, like we're coming into

697
00:38:24,760 --> 00:38:29,480
Speaker 6: the midterms, and if the Dams adroitly enough like make

698
00:38:29,560 --> 00:38:32,080
Speaker 6: clear what the stakes are and what he's done. It's

699
00:38:32,080 --> 00:38:34,600
Speaker 6: always a little tricky, right, It's a little feels a

700
00:38:34,600 --> 00:38:35,479
Speaker 6: little pointy headed.

701
00:38:35,560 --> 00:38:37,799
Speaker 7: But if they can, Yeah, I think he I think

702
00:38:37,840 --> 00:38:39,120
Speaker 7: he's maybe in trouble.

703
00:38:39,400 --> 00:38:40,480
Speaker 1: Yeah during the years.

704
00:38:40,680 --> 00:38:44,000
Speaker 6: Okay, let's switch directions because for there's a little bit,

705
00:38:44,640 --> 00:38:48,160
Speaker 6: to quote Trump himself, a slight bit of intrigue going

706
00:38:48,239 --> 00:38:51,640
Speaker 6: on today on Capitol Hill where Trump is saying, oh,

707
00:38:51,800 --> 00:38:56,080
Speaker 6: not so fast on his nomination of Jay Clayton as

708
00:38:56,160 --> 00:38:59,200
Speaker 6: Director of National Intelligence. And it sure looks to me

709
00:38:59,760 --> 00:39:03,720
Speaker 6: like the whole rationale here is to get Bill Pulty,

710
00:39:03,880 --> 00:39:06,840
Speaker 6: whose job has been to be a sort of hatchet

711
00:39:06,960 --> 00:39:10,400
Speaker 6: man for Trump so far, doesn't even have a security clearance,

712
00:39:10,719 --> 00:39:13,920
Speaker 6: just a little time as acting director, and once he

713
00:39:14,040 --> 00:39:16,640
Speaker 6: gets the keys to the kingdom right and can do things,

714
00:39:16,840 --> 00:39:19,279
Speaker 6: you know, there's no sort of taking it back.

715
00:39:19,800 --> 00:39:21,680
Speaker 7: On our roundtable this week.

716
00:39:21,520 --> 00:39:23,719
Speaker 6: One of the guests that were that, you know, even

717
00:39:23,719 --> 00:39:27,160
Speaker 6: a little time in the office could turbo charge Pultie's

718
00:39:27,200 --> 00:39:32,200
Speaker 6: ability to target Trump's political opponents make mischief with the elections.

719
00:39:32,520 --> 00:39:33,600
Speaker 7: What is going on?

720
00:39:33,920 --> 00:39:36,600
Speaker 6: Is Trump like basically pulling a fast one here? And

721
00:39:36,760 --> 00:39:39,800
Speaker 6: can he get away with it with actually slowing down

722
00:39:40,120 --> 00:39:41,200
Speaker 6: Clayton's nomination.

723
00:39:41,560 --> 00:39:45,480
Speaker 1: Tom Cotton says, now, we'll see what happens. Buddy pushed

724
00:39:45,520 --> 00:39:48,160
Speaker 1: back on that It's not a great moment for Trump

725
00:39:48,280 --> 00:39:51,640
Speaker 1: to test to his friends in the senat are because

726
00:39:51,880 --> 00:39:55,400
Speaker 1: for obvious reasons, because of that Yolo Caucus, those three

727
00:39:56,000 --> 00:39:58,640
Speaker 1: senators who are loss of jobs.

728
00:39:58,719 --> 00:40:01,000
Speaker 6: Till Us, why do you call the yellow I like it.

729
00:40:00,960 --> 00:40:03,520
Speaker 1: Because they have you only live once, because they've all

730
00:40:03,760 --> 00:40:10,200
Speaker 1: lost their dogs. Trump, Yeah, Corny, Cassidy and Tillis.

731
00:40:09,920 --> 00:40:13,279
Speaker 6: Goodbye, mister bald I think so if you look at.

732
00:40:13,200 --> 00:40:15,919
Speaker 1: The if you look at the numbers, it's fifty three

733
00:40:16,080 --> 00:40:21,760
Speaker 1: forty seven, right, and what happens you need more than fifty?

734
00:40:22,120 --> 00:40:22,600
Speaker 7: I don't know.

735
00:40:22,880 --> 00:40:25,680
Speaker 1: I mean I think Trump is in is maybe in

736
00:40:25,760 --> 00:40:29,000
Speaker 1: hot water here. I also think like the whole idea

737
00:40:29,600 --> 00:40:32,879
Speaker 1: of the Senate ap proving Trump's guys was because, I mean,

738
00:40:32,960 --> 00:40:35,880
Speaker 1: even the Republicans who were not who were sort of

739
00:40:35,920 --> 00:40:38,440
Speaker 1: on the fence, would say, well, the president deserves to

740
00:40:38,480 --> 00:40:41,120
Speaker 1: have his people in there. Well, the president has had

741
00:40:41,160 --> 00:40:44,239
Speaker 1: his people in there. It's been a disaster, and there's

742
00:40:44,320 --> 00:40:47,520
Speaker 1: something for everyone to be mad at at this moment. Right,

743
00:40:47,800 --> 00:40:52,520
Speaker 1: the new nuclear deal is significantly worse than the old one.

744
00:40:53,000 --> 00:40:58,160
Speaker 1: The pro Israel Hawks, the people for whom Iran was

745
00:40:58,320 --> 00:41:01,800
Speaker 1: meant to be the beginning. That crew is mad because

746
00:41:01,920 --> 00:41:04,960
Speaker 1: they saw that they were getting some kind of ground war.

747
00:41:05,520 --> 00:41:08,560
Speaker 1: I'm not commenced that there's enough people who, I mean,

748
00:41:08,880 --> 00:41:11,000
Speaker 1: find a center who's not mad at Trump right now?

749
00:41:11,560 --> 00:41:14,200
Speaker 6: Yeah, But and my question he is the will they

750
00:41:14,320 --> 00:41:18,040
Speaker 6: kind of keep their anger under? You know? I thought

751
00:41:18,160 --> 00:41:20,360
Speaker 6: there was the beginning of a breakout moment in the

752
00:41:20,440 --> 00:41:23,120
Speaker 6: slush fund, but here they go tiptoeing back.

753
00:41:23,280 --> 00:41:24,960
Speaker 1: Do we know where the slush fund is?

754
00:41:25,920 --> 00:41:30,320
Speaker 6: Yeah, it's in courts now, but there's like standing issues.

755
00:41:30,840 --> 00:41:34,040
Speaker 6: There's it's down in a court in Florida about whether

756
00:41:34,160 --> 00:41:35,319
Speaker 6: they were defrauded.

757
00:41:35,840 --> 00:41:37,719
Speaker 7: But that's a tricky thing to do.

758
00:41:38,080 --> 00:41:40,880
Speaker 6: I think that the slush fund's been the best example

759
00:41:41,320 --> 00:41:45,239
Speaker 6: of the kind of one two punch of politics and law.

760
00:41:45,440 --> 00:41:48,920
Speaker 6: And if that, if the Republicans withdraw on that one fist,

761
00:41:49,440 --> 00:41:50,879
Speaker 6: I don't know if the courts.

762
00:41:50,600 --> 00:41:53,439
Speaker 7: Can kind of handle it all on their own.

763
00:41:53,760 --> 00:41:57,480
Speaker 6: And it's been really sally to me, how after getting

764
00:41:57,520 --> 00:42:00,360
Speaker 6: some steal in their spine. You're right, they're the Yolo crowd.

765
00:42:00,480 --> 00:42:03,399
Speaker 6: He's killed them, and yet I don't know, maybe it's

766
00:42:03,560 --> 00:42:06,400
Speaker 6: they want to be lobbyists, or they they're you know,

767
00:42:06,520 --> 00:42:10,040
Speaker 6: they don't want to look like total apostates. But it

768
00:42:10,160 --> 00:42:15,000
Speaker 6: seems as if the people who are totally incentivized to

769
00:42:16,000 --> 00:42:19,919
Speaker 6: vote their consciences now and depart from Trump are being

770
00:42:20,280 --> 00:42:22,560
Speaker 6: kind uh cowardly about it.

771
00:42:23,480 --> 00:42:25,719
Speaker 1: Republicans being cowardly. How is that?

772
00:42:25,840 --> 00:42:26,360
Speaker 5: Positives?

773
00:42:26,480 --> 00:42:35,160
Speaker 1: Just yeah, Okay, I'm curious about the Trump renovation fiasco.

774
00:42:35,920 --> 00:42:40,440
Speaker 1: So we have the fight over the Kennedy Center. So

775
00:42:40,680 --> 00:42:43,959
Speaker 1: far he has his name is off, but the tarp

776
00:42:44,120 --> 00:42:47,480
Speaker 1: is off, the tarp is up. We have the he

777
00:42:47,719 --> 00:42:51,640
Speaker 1: can't close it for two years is now in court.

778
00:42:52,120 --> 00:42:56,520
Speaker 1: We have the East Wing that's going to be legislated

779
00:42:56,840 --> 00:42:59,200
Speaker 1: to the hilt, right Like I think Trump thought of

780
00:42:59,320 --> 00:43:01,600
Speaker 1: he down the East Wing and he'd be able to

781
00:43:01,680 --> 00:43:04,560
Speaker 1: rebuild the East win. But that is not how any

782
00:43:04,600 --> 00:43:06,840
Speaker 1: of this is working. And I mean maybe that's the

783
00:43:06,880 --> 00:43:10,440
Speaker 1: difference between New York and DC. In New York, if

784
00:43:10,440 --> 00:43:13,200
Speaker 1: you tear something down, sometimes they'll let you just give

785
00:43:13,280 --> 00:43:15,600
Speaker 1: up on it. But in DC, now and then you

786
00:43:15,760 --> 00:43:19,879
Speaker 1: have the reflecting pool, which is just the no bid

787
00:43:20,120 --> 00:43:24,840
Speaker 1: contract to paint the bottom of the reflecting pool American

788
00:43:24,920 --> 00:43:29,880
Speaker 1: flag blue, which then encourages the growth of algae because

789
00:43:29,920 --> 00:43:33,400
Speaker 1: it's warmer and so we have American flag green. So

790
00:43:33,600 --> 00:43:37,240
Speaker 1: I would love you to just explain where Trump's legal

791
00:43:37,360 --> 00:43:40,279
Speaker 1: issues are here and the administration too.

792
00:43:40,719 --> 00:43:43,279
Speaker 6: Great because here's the stress point, mall I just did

793
00:43:43,360 --> 00:43:47,040
Speaker 6: a substack on this. It's all the same kind of

794
00:43:47,320 --> 00:43:51,000
Speaker 6: gambit with Trump. This he is weeding time and again

795
00:43:51,200 --> 00:43:56,160
Speaker 6: in Congress's gardens. So let's go to the Kennedy Center.

796
00:43:56,640 --> 00:44:00,799
Speaker 6: Judge Cooper just put it right out there the name.

797
00:44:01,000 --> 00:44:03,120
Speaker 7: Only Congress can change the name.

798
00:44:03,360 --> 00:44:06,200
Speaker 6: The big difference there is he had a really solid

799
00:44:06,360 --> 00:44:10,080
Speaker 6: case of someone who could bring it so called standing,

800
00:44:10,680 --> 00:44:12,960
Speaker 6: and that was Joyce Spady.

801
00:44:12,560 --> 00:44:16,040
Speaker 7: Who was on the board, and they just literally silenced

802
00:44:16,080 --> 00:44:16,719
Speaker 7: her at.

803
00:44:16,680 --> 00:44:21,200
Speaker 6: The meeting and voted at unanimous, didn't even record her vote.

804
00:44:21,239 --> 00:44:24,640
Speaker 6: So she had an injury. But their basic deal with

805
00:44:25,160 --> 00:44:26,400
Speaker 6: the same thing in the ballroom.

806
00:44:26,440 --> 00:44:27,440
Speaker 7: It's not up to them.

807
00:44:27,520 --> 00:44:30,440
Speaker 6: And by the way, all the information seems to be

808
00:44:30,560 --> 00:44:34,200
Speaker 6: Trump promised it would be completely funded by private funding.

809
00:44:34,280 --> 00:44:36,760
Speaker 7: Guess what it's not. It would only be four hundred

810
00:44:37,400 --> 00:44:39,200
Speaker 7: million dollars. Guess what it's not.

811
00:44:39,560 --> 00:44:42,840
Speaker 6: But all of these are about, hey, this is for

812
00:44:43,040 --> 00:44:45,040
Speaker 6: Congress to do it, and you know, you take a

813
00:44:45,120 --> 00:44:45,880
Speaker 6: step back.

814
00:44:46,000 --> 00:44:46,800
Speaker 7: The White House.

815
00:44:47,120 --> 00:44:51,000
Speaker 6: Man oh man, he really wants to completely just blow

816
00:44:51,080 --> 00:44:54,240
Speaker 6: this shit. The whole sort of skyline of DC.

817
00:44:54,680 --> 00:44:55,879
Speaker 7: And you've been there.

818
00:44:56,000 --> 00:45:00,200
Speaker 6: It's all really carefully designed, and you know, we all

819
00:45:00,280 --> 00:45:02,840
Speaker 6: go there in seventh grade. It's lovely, and he wants

820
00:45:02,920 --> 00:45:06,120
Speaker 6: to just you know, make it like an imperial monument

821
00:45:06,239 --> 00:45:09,879
Speaker 6: to himself. I've said this, I'm on record that even

822
00:45:09,960 --> 00:45:14,080
Speaker 6: though the legal claims are a little bit hard in terms.

823
00:45:13,880 --> 00:45:15,880
Speaker 7: Of standing, like who's going to be able to do it?

824
00:45:16,040 --> 00:45:17,720
Speaker 7: I just don't think this is going to happen.

825
00:45:17,840 --> 00:45:21,400
Speaker 6: I just think that that the judge who will actually

826
00:45:21,600 --> 00:45:26,400
Speaker 6: greenlight this grotesque Xana new mausoleum in the sky, you know,

827
00:45:26,680 --> 00:45:29,760
Speaker 6: it just it's just too much. It makes the American

828
00:45:29,880 --> 00:45:33,480
Speaker 6: people and so much of this stuff does barf. But

829
00:45:33,880 --> 00:45:36,880
Speaker 6: the legal the answer to your legal question is this

830
00:45:37,239 --> 00:45:41,560
Speaker 6: Trump has no business doing it. Congress has business doing it.

831
00:45:41,680 --> 00:45:45,799
Speaker 6: He tries ends around again and again, and the main

832
00:45:45,920 --> 00:45:48,600
Speaker 6: thing they lead with is, yeah, maybe it's illegal, but

833
00:45:48,719 --> 00:45:49,920
Speaker 6: nobody can challenge it.

834
00:45:50,239 --> 00:45:51,960
Speaker 7: That wasn't working with the ballroom.

835
00:45:52,120 --> 00:45:55,160
Speaker 6: Just today you have Cooper, he hadn't been asked before

836
00:45:55,320 --> 00:45:58,000
Speaker 6: saying uh uh, we need a plan to keep this

837
00:45:58,239 --> 00:46:02,040
Speaker 6: going for a couple years. That's the same Gidge who's

838
00:46:02,400 --> 00:46:04,560
Speaker 6: on the case there. You know, you have to have

839
00:46:04,680 --> 00:46:09,440
Speaker 6: something presented to the courts. But I think that combined

840
00:46:09,680 --> 00:46:12,399
Speaker 6: it always really matters. Even though it's a little bit

841
00:46:12,719 --> 00:46:16,920
Speaker 6: kind of a MorphOS the big popular pushback, it gives

842
00:46:17,040 --> 00:46:22,040
Speaker 6: some kind of tailwind to the opposition. It gives Republicans

843
00:46:22,120 --> 00:46:26,120
Speaker 6: a little bit of kind of gumption. Anyway, that's what

844
00:46:26,239 --> 00:46:29,759
Speaker 6: he's doing, and the green allergy and oh and then

845
00:46:29,880 --> 00:46:33,160
Speaker 6: the huge arch You know, it's none of it is

846
00:46:33,239 --> 00:46:36,640
Speaker 6: his business to do. But will the courts be able

847
00:46:36,760 --> 00:46:42,080
Speaker 6: to actually have a kind of leverage from some solid plaintiffence.

848
00:46:42,200 --> 00:46:44,040
Speaker 7: Okay, I want to go back to what you just

849
00:46:44,200 --> 00:46:45,920
Speaker 7: mentioned about Iran.

850
00:46:46,320 --> 00:46:48,880
Speaker 6: You know, it does seem like, as you put it,

851
00:46:48,960 --> 00:46:54,200
Speaker 6: a pretty powerful contingent of Republicans are very unhappy with it.

852
00:46:54,560 --> 00:46:57,960
Speaker 6: As you say, the deal as it relates to ability

853
00:46:58,120 --> 00:47:01,200
Speaker 6: to have nuclear weapons seems worse that the one Obaba

854
00:47:01,280 --> 00:47:05,880
Speaker 6: got Livesy grab seems hot and bothered. But you also

855
00:47:06,000 --> 00:47:10,719
Speaker 6: see the same native hesitation about criticizing Trump. So what

856
00:47:11,239 --> 00:47:14,600
Speaker 6: is a little grumbling the most we can expect from

857
00:47:15,000 --> 00:47:19,239
Speaker 6: the Republican opponents here or are they really ready to

858
00:47:19,400 --> 00:47:22,080
Speaker 6: take him on on a pretty crappy.

859
00:47:21,840 --> 00:47:24,800
Speaker 7: Deal to the in the opinion of most people I respect.

860
00:47:25,360 --> 00:47:28,680
Speaker 1: Here's the thing. Trump is worried about the midterms. Gas

861
00:47:28,800 --> 00:47:32,400
Speaker 1: price or high. He knows this is fucking up the economy.

862
00:47:32,719 --> 00:47:36,399
Speaker 1: So the question is can he get the genie back

863
00:47:36,480 --> 00:47:39,160
Speaker 1: in the bottle before the midterms. Can he get gassed

864
00:47:39,239 --> 00:47:42,600
Speaker 1: down to read something? Can he get the straights moving again?

865
00:47:43,000 --> 00:47:47,080
Speaker 1: Can he soothe it and smooth it again? I think

866
00:47:47,480 --> 00:47:50,560
Speaker 1: there are big The big players here are the Hawks,

867
00:47:50,640 --> 00:47:54,399
Speaker 1: the Republican Hawks, and then Nan Yahoo. We have seen

868
00:47:54,560 --> 00:47:59,320
Speaker 1: now with both Trump and with Biden that Nat Yahoo

869
00:47:59,680 --> 00:48:03,759
Speaker 1: is not a controllable figure. So the question is this

870
00:48:03,960 --> 00:48:06,720
Speaker 1: is not a good deal for Israel, and it treats

871
00:48:06,760 --> 00:48:10,640
Speaker 1: them as a vassal steak. That's pretty bad. Trump is

872
00:48:10,719 --> 00:48:14,120
Speaker 1: funded by a lot of israel Hawks, like a lot

873
00:48:14,239 --> 00:48:17,239
Speaker 1: of the big money in Trump world. Are people for

874
00:48:17,400 --> 00:48:20,480
Speaker 1: whom Israel is the reason they are voting for Trump.

875
00:48:20,880 --> 00:48:24,760
Speaker 1: So the question I think is does Israel say okay,

876
00:48:25,040 --> 00:48:28,000
Speaker 1: it matters more? And by Israel I mean net Yahoo

877
00:48:28,440 --> 00:48:31,359
Speaker 1: or the people who are you know, the people who

878
00:48:31,360 --> 00:48:35,319
Speaker 1: are funding Trump because Israel is the number one right,

879
00:48:35,800 --> 00:48:38,960
Speaker 1: is there a number one issue? Do those people say

880
00:48:39,239 --> 00:48:41,719
Speaker 1: it's more important for Republicans who win the midterms. So

881
00:48:41,960 --> 00:48:45,760
Speaker 1: let's just let him do this deal? Or do they say,

882
00:48:46,920 --> 00:48:50,799
Speaker 1: you know, this is bullshit? I mean, I don't know. Look,

883
00:48:51,080 --> 00:48:55,840
Speaker 1: so far we have not seen nen Yahoo pause his

884
00:48:55,960 --> 00:49:00,479
Speaker 1: attacks on Lebanon. So if this is a ceasefire where

885
00:49:00,920 --> 00:49:04,200
Speaker 1: there are still firing, then I'm not convinced that that's

886
00:49:04,280 --> 00:49:06,520
Speaker 1: going to be helpful for anyone. I don't know how

887
00:49:06,600 --> 00:49:07,440
Speaker 1: long Trump goes.

888
00:49:07,360 --> 00:49:09,239
Speaker 7: On that it can I just ask you, you know,

889
00:49:09,960 --> 00:49:13,160
Speaker 7: how does it work concretely? And that's right? They don't

890
00:49:13,239 --> 00:49:14,600
Speaker 7: do it. It's not a great deal.

891
00:49:15,239 --> 00:49:18,520
Speaker 6: What's the sort of form of the pushback from Graham

892
00:49:18,719 --> 00:49:19,400
Speaker 6: and company?

893
00:49:19,640 --> 00:49:21,719
Speaker 7: And does it actually defeat them?

894
00:49:21,920 --> 00:49:25,879
Speaker 1: I think the question is does Trump like a lot

895
00:49:25,960 --> 00:49:28,719
Speaker 1: of those people are donors, Like if you're thinking about

896
00:49:28,760 --> 00:49:31,680
Speaker 1: a lot of the Republican donors who are Israel donors

897
00:49:31,800 --> 00:49:35,000
Speaker 1: who donate largely on that and I'm thinking about the

898
00:49:35,120 --> 00:49:39,600
Speaker 1: Sheldon Atolsen's he's dead but Mary and Natolson, that crew,

899
00:49:40,280 --> 00:49:44,920
Speaker 1: the APAC world, the lobbyists and donors who are Israel donors.

900
00:49:45,200 --> 00:49:49,239
Speaker 1: So the question is what do those people do with

901
00:49:49,480 --> 00:49:52,560
Speaker 1: this deal? And that I don't know. I mean, they

902
00:49:52,640 --> 00:49:55,480
Speaker 1: can threaten to not give him money. They can threaten

903
00:49:55,600 --> 00:49:59,400
Speaker 1: to not give them Republicans money. They can look, Trump

904
00:49:59,560 --> 00:50:03,600
Speaker 1: had Elon Musk, who is the richest trillionaire, and he

905
00:50:03,719 --> 00:50:07,920
Speaker 1: seems pretty committed to giving money to whatever Republicans want.

906
00:50:08,000 --> 00:50:10,440
Speaker 1: So maybe it doesn't matter. Maybe he can make up

907
00:50:10,440 --> 00:50:13,480
Speaker 1: the cash somewhere else. But his interest in Israel, I

908
00:50:13,520 --> 00:50:16,880
Speaker 1: think has been largely transactional up until this moment. So

909
00:50:17,080 --> 00:50:18,640
Speaker 1: if that money goes away, I don't know.

910
00:50:19,120 --> 00:50:21,879
Speaker 8: Okay, your served a lot of stuff that Trump world

911
00:50:21,960 --> 00:50:25,360
Speaker 8: does right now, he loses in court because it's illegal.

912
00:50:25,680 --> 00:50:29,120
Speaker 8: So I wonder, I mean, right, the judges are like,

913
00:50:29,239 --> 00:50:31,879
Speaker 8: and usually it's so illegal that the judges.

914
00:50:31,520 --> 00:50:34,000
Speaker 1: Are like, oh, you can't fucking do that. So I

915
00:50:34,160 --> 00:50:37,760
Speaker 1: wonder is there a world in which and there's certainly

916
00:50:38,280 --> 00:50:40,400
Speaker 1: Trump is doing a lot of illegal staff like self

917
00:50:40,480 --> 00:50:43,919
Speaker 1: dealing and making money, and his kids are making money,

918
00:50:44,120 --> 00:50:47,600
Speaker 1: and his friends are making money, and this one's making

919
00:50:47,640 --> 00:50:50,120
Speaker 1: money and that one's making money. And again, the Supreme

920
00:50:50,160 --> 00:50:52,719
Speaker 1: Court has said that Donald Trump is a king and

921
00:50:52,920 --> 00:50:56,759
Speaker 1: that he cannot face any legal challenge. Right basically has

922
00:50:56,840 --> 00:50:59,720
Speaker 1: blanket immunity for whatever he wants to do because he's Trump,

923
00:51:00,200 --> 00:51:03,040
Speaker 1: and Trump has been happy to pardon anyone you know

924
00:51:03,120 --> 00:51:05,239
Speaker 1: who wants it. But I wonder what you think the

925
00:51:05,360 --> 00:51:10,160
Speaker 1: legal culpability is for the outside circle, for everyone else,

926
00:51:10,520 --> 00:51:14,120
Speaker 1: For you know, even if you're pardoned, are you still responsible?

927
00:51:14,160 --> 00:51:16,120
Speaker 1: You know, like, for example, let's I'm going to give

928
00:51:16,120 --> 00:51:19,600
Speaker 1: you an example. His kids are making money every which way.

929
00:51:19,840 --> 00:51:22,680
Speaker 1: A lot of this clearly it's you know, and the

930
00:51:22,800 --> 00:51:26,400
Speaker 1: idea that you'd be protected by presidential and you know,

931
00:51:26,960 --> 00:51:31,439
Speaker 1: some kind of privilege seems insane. Say Trump pardons them,

932
00:51:31,640 --> 00:51:34,359
Speaker 1: can they still are there any ways in which they

933
00:51:34,920 --> 00:51:38,200
Speaker 1: will become, you know, legal defendants for the rest of

934
00:51:38,200 --> 00:51:38,760
Speaker 1: their lives.

935
00:51:39,120 --> 00:51:42,600
Speaker 6: Yeah, and by the way, what an avalanche of pardners

936
00:51:42,640 --> 00:51:45,440
Speaker 6: we're expecting, probably at the end of his term. So

937
00:51:45,480 --> 00:51:48,239
Speaker 6: the story answer is yeah, But it's all tricky, and

938
00:51:48,520 --> 00:51:50,360
Speaker 6: I think it goes to politics a little.

939
00:51:50,440 --> 00:51:51,880
Speaker 7: Will the devs have the stomach?

940
00:51:51,920 --> 00:51:55,719
Speaker 6: I'm sure there'll be really strong currents of opinion in there,

941
00:51:55,960 --> 00:51:59,239
Speaker 6: like shit, just let it go already, so we can

942
00:51:59,680 --> 00:52:02,880
Speaker 6: try to consolidate power if they once they gained. But

943
00:52:03,000 --> 00:52:07,480
Speaker 6: a few things to mention here. First, there's always state prosecutions.

944
00:52:07,560 --> 00:52:10,400
Speaker 6: We saw those before. That's why Trump is a fellon

945
00:52:10,520 --> 00:52:12,719
Speaker 6: thirty four times over because of New York State. But

946
00:52:12,800 --> 00:52:15,600
Speaker 6: there's a lot of ways people criticize the fact that

947
00:52:15,800 --> 00:52:21,399
Speaker 6: state and federal prosecutions could overlap, and nothing he can

948
00:52:21,480 --> 00:52:24,759
Speaker 6: do with his parts about state. One second, even with

949
00:52:25,160 --> 00:52:28,600
Speaker 6: the king part, as you said, it doesn't reach where

950
00:52:28,719 --> 00:52:33,000
Speaker 6: he has been not acting officially. So you know, we

951
00:52:33,400 --> 00:52:36,319
Speaker 6: think about the slush fund. It's just Donald Trump, who

952
00:52:36,440 --> 00:52:38,799
Speaker 6: when he wasn't president, brought this case. If he did

953
00:52:38,840 --> 00:52:43,280
Speaker 6: something fraudulent or whatever, it should be something you can reach. Finally,

954
00:52:43,560 --> 00:52:46,480
Speaker 6: as to his kids, Molly, it strikes me that they

955
00:52:46,600 --> 00:52:50,400
Speaker 6: got no specific protection. I think you may see and

956
00:52:50,840 --> 00:52:54,680
Speaker 6: he'll cite the bided precedent. You may see pardons for

957
00:52:55,040 --> 00:52:58,480
Speaker 6: that reason. But there's going to be playing the joys

958
00:52:58,560 --> 00:53:01,240
Speaker 6: because he will try to do a so called general

959
00:53:01,320 --> 00:53:04,800
Speaker 6: part it. Anything they've ever done, you know, including bubblegum

960
00:53:04,960 --> 00:53:07,759
Speaker 6: on the street, they get a pass for pardons don't

961
00:53:07,840 --> 00:53:09,960
Speaker 6: necessarily work that way.

962
00:53:10,480 --> 00:53:12,680
Speaker 7: And then actually the final point I should raise is

963
00:53:12,920 --> 00:53:13,640
Speaker 7: you're so right.

964
00:53:13,880 --> 00:53:16,520
Speaker 6: I mean, we've had this kind of rain on the treasury,

965
00:53:16,640 --> 00:53:19,680
Speaker 6: especially in the last few months since he's been president.

966
00:53:19,760 --> 00:53:22,640
Speaker 6: The way you get at that though, is impeachment. What

967
00:53:22,880 --> 00:53:26,000
Speaker 6: will the Dems kind of stomach for that be. So

968
00:53:26,120 --> 00:53:30,360
Speaker 6: there's plenty of stuff to do, including oversight and you know,

969
00:53:30,520 --> 00:53:32,880
Speaker 6: haling people to down to the Congress.

970
00:53:32,960 --> 00:53:33,440
Speaker 2: What did you do?

971
00:53:33,719 --> 00:53:34,320
Speaker 7: X Y and Z.

972
00:53:34,600 --> 00:53:37,040
Speaker 6: And I think in the first instance, it's going to

973
00:53:37,120 --> 00:53:42,160
Speaker 6: be a kind of political internal debate with for the

974
00:53:42,320 --> 00:53:45,640
Speaker 6: Dems if should they let's say, take the House and

975
00:53:45,800 --> 00:53:48,200
Speaker 6: not the Senate, or even daily take the Senate.

976
00:53:48,400 --> 00:53:52,000
Speaker 7: My amateur opinion, Hey, I think that does it for us.

977
00:53:52,080 --> 00:53:54,680
Speaker 6: But I just want to recommend to FA heads everywhere.

978
00:53:54,920 --> 00:53:58,719
Speaker 6: You're really lovely and heartrending piece. If I could about

979
00:53:58,760 --> 00:54:01,719
Speaker 6: your famous moment title I Forgave My Mother, but it

980
00:54:01,960 --> 00:54:04,439
Speaker 6: was too late in the New York Times last month.

981
00:54:04,800 --> 00:54:09,200
Speaker 6: It's really a very lovely, earnest, thoughtful piece. Muzzle tough

982
00:54:09,280 --> 00:54:13,240
Speaker 6: to you and look forward to our next mashup.

983
00:54:14,680 --> 00:54:16,640
Speaker 2: The moment.

984
00:54:18,280 --> 00:54:19,240
Speaker 1: Jesse Cannon.

985
00:54:19,719 --> 00:54:23,000
Speaker 3: So, Mami, the Justice Department is going to dismiss air

986
00:54:23,080 --> 00:54:27,560
Speaker 3: pollutions lawsuit against XAI. If you've seen more perfect Unions

987
00:54:27,640 --> 00:54:31,839
Speaker 3: reporting on this suit, it is absolutely disgusting what they've

988
00:54:31,880 --> 00:54:35,240
Speaker 3: done to this community. They've basically ruined the surrounding area

989
00:54:35,320 --> 00:54:36,200
Speaker 3: with this data center.

990
00:54:36,600 --> 00:54:41,759
Speaker 1: Yeah, I mean, this is Elon Musk using his government contacts.

991
00:54:42,040 --> 00:54:45,880
Speaker 1: Remember he was a special government employee. The Justice Department

992
00:54:46,080 --> 00:54:49,920
Speaker 1: is supposed to protect the American people. What it is

993
00:54:50,040 --> 00:54:54,080
Speaker 1: doing is protecting Trump and his cronies and making it

994
00:54:54,200 --> 00:54:56,239
Speaker 1: so they can do what they want to do. And

995
00:54:56,520 --> 00:54:59,920
Speaker 1: so here we are. They want to intervene in the

996
00:55:00,160 --> 00:55:04,360
Speaker 1: case and dismiss the lawsuit, arguing the plant needs power

997
00:55:04,760 --> 00:55:07,320
Speaker 1: because it's critical to the economy. I mean, just this

998
00:55:07,520 --> 00:55:10,040
Speaker 1: is like the same thinking that was like Donald Trump

999
00:55:10,040 --> 00:55:13,759
Speaker 1: should be able to award a no bid contract to

1000
00:55:14,400 --> 00:55:19,640
Speaker 1: paint the reflecting pool because it's so critical. You'll remember

1001
00:55:19,719 --> 00:55:24,640
Speaker 1: the reflecting pool. It is now covered in algae. That's

1002
00:55:24,800 --> 00:55:31,120
Speaker 1: it for this episode of Fast Politics. Tune in every Monday, Wednesday,

1003
00:55:31,440 --> 00:55:35,840
Speaker 1: Thursday and Saturday to hear the best minds and politics

1004
00:55:35,960 --> 00:55:40,200
Speaker 1: make sense of all this chaos. If you enjoy this podcast,

1005
00:55:40,640 --> 00:55:44,000
Speaker 1: please send it to a friend and keep the conversation going.

1006
00:55:44,520 --> 00:55:45,600
Speaker 1: Thanks for listening.