WEBVTT

1
00:00:00.110 --> 00:00:02.109
<v Speaker 1>I don't want to indulge in any of those games,

2
00:00:02.189 --> 00:00:04.629
<v Speaker 1>but it is a communist holiday, and I do resent it.

3
00:00:05.089 --> 00:00:05.650
<v Speaker 2>Okay, good.

4
00:00:05.669 --> 00:00:10.589
<v Speaker 3>I agree. AI data centers waste millions of gallons of water.

5
00:00:11.470 --> 00:00:13.890
<v Speaker 4>That's why Liquid Death and Garage Beer decided to team up.

6
00:00:14.599 --> 00:00:17.980
<v Speaker 3>We want your pee to cool these data centers.

7
00:00:18.160 --> 00:00:21.620
<v Speaker 2>It's the Ricochet Podcast with James Lilacs. That's me, Charles C.W. Cook,

8
00:00:21.640 --> 00:00:23.920
<v Speaker 2>and Stephen Hayward. We're going to talk to Mark Mills

9
00:00:24.000 --> 00:00:28.120
<v Speaker 2>about AI. Don't fear the future. Let's have ourselves a podcast.

10
00:00:28.500 --> 00:00:30.629
<v Speaker 5>What's interesting about when you get far enough away from

11
00:00:30.649 --> 00:00:34.049
<v Speaker 5>a time period that's quite distinct, I think there's always

12
00:00:34.090 --> 00:00:35.950
<v Speaker 5>a lovely feeling associated with it. It's almost like you

13
00:00:35.969 --> 00:00:38.030
<v Speaker 5>pave over all the trauma with, like, puppy dogs and

14
00:00:38.070 --> 00:00:38.469
<v Speaker 5>ice cream.

15
00:00:39.070 --> 00:00:40.390
<v Speaker 1>And the 80s was a.

16
00:00:40.429 --> 00:00:42.219
<v Speaker 4>Brutal, brutal time for people.

17
00:00:42.280 --> 00:00:43.820
<v Speaker 5>It was a f * * * ing mess. And, like,

18
00:00:44.140 --> 00:00:45.880
<v Speaker 5>you know, people forget that it was, like, there was

19
00:00:45.920 --> 00:00:50.840
<v Speaker 5>a tremendous strife, sociopolitical tensions everywhere. You know, we're still, like,

20
00:00:50.920 --> 00:00:54.070
<v Speaker 5>pretty fresh off, like, a Cuban Missile Crisis, which is only, like, what, 15,

21
00:00:54.039 --> 00:00:59.259
<v Speaker 5>20 years before... So, you know, it's a tumultuous period,

22
00:00:59.280 --> 00:01:00.880
<v Speaker 5>but you look back at it now, and you're like, oh, yeah,

23
00:01:00.899 --> 00:01:02.240
<v Speaker 5>Top Gun, cocaine.

24
00:01:03.479 --> 00:01:04.319
<v Speaker 1>There you go again.

25
00:01:04.939 --> 00:01:09.260
<v Speaker 2>Welcome, everybody. It's the Ricochet Podcast, number 805. I'm James

26
00:01:09.299 --> 00:01:12.579
<v Speaker 2>Lilacs in sunny Edina, Minnesota, or Edina, as the AI

27
00:01:12.640 --> 00:01:15.370
<v Speaker 2>likes to pronounce it. Stephen Hayward in California. Charles C.W.

28
00:01:15.530 --> 00:01:18.870
<v Speaker 2>Cook in Florida. Gentlemen, welcome. We are on the cusp

29
00:01:18.909 --> 00:01:20.310
<v Speaker 2>of a holiday weekend here.

30
00:01:20.450 --> 00:01:21.030
<v Speaker 3>All is great.

31
00:01:21.430 --> 00:01:24.909
<v Speaker 2>Summer is winding down. The Minnesota State Fair is... the

32
00:01:24.950 --> 00:01:27.489
<v Speaker 2>end of which signifies the end of summer, is having

33
00:01:27.530 --> 00:01:29.790
<v Speaker 2>a bit of a controversy now because of the political

34
00:01:29.810 --> 00:01:34.879
<v Speaker 2>nature of the seed art. Somebody put 8647 in seeds

35
00:01:35.000 --> 00:01:40.120
<v Speaker 2>and then had red seeds underneath to indicate blood. There

36
00:01:40.200 --> 00:01:43.659
<v Speaker 2>are seed art portraits of Minnesota martyrs like St.

37
00:01:43.680 --> 00:01:44.459
<v Speaker 1>George Floyd.

38
00:01:45.500 --> 00:01:50.629
<v Speaker 2>There is Donald Trump stuffed with beans or something. It's

39
00:01:50.689 --> 00:01:56.409
<v Speaker 2>just so... appallingly banal that people can't feel the need

40
00:01:56.450 --> 00:02:00.129
<v Speaker 2>to infuse their tiresome politics and shove it into everything

41
00:02:00.769 --> 00:02:02.709
<v Speaker 2>and i wonder whether or not that the uh the

42
00:02:02.750 --> 00:02:05.030
<v Speaker 2>seed art this year is being funded by ngos or

43
00:02:05.090 --> 00:02:08.530
<v Speaker 2>non-profits i know that sounds like paranoia but stephen hayward

44
00:02:08.560 --> 00:02:13.500
<v Speaker 2>friend actually has a a modest proposal on what to

45
00:02:13.539 --> 00:02:16.280
<v Speaker 2>do uh with non-profits why don't you tell it to

46
00:02:16.300 --> 00:02:17.180
<v Speaker 2>charles and myself.

47
00:02:17.419 --> 00:02:19.939
<v Speaker 3>Well, first of all, James, about the seed art, I

48
00:02:20.020 --> 00:02:22.569
<v Speaker 3>think we should counter it by spreading the rumor that

49
00:02:22.590 --> 00:02:26.469
<v Speaker 3>they're all made with genetically modified seeds, right? Because that

50
00:02:26.509 --> 00:02:28.030
<v Speaker 3>will cause a cognitive dissonance and all that.

51
00:02:28.050 --> 00:02:28.689
<v Speaker 2>Precisely.

52
00:02:29.050 --> 00:02:31.120
<v Speaker 3>Yeah, so, I mean, long story, but I had this

53
00:02:31.719 --> 00:02:34.479
<v Speaker 3>crazy idea– not crazy, but it won't happen– that we

54
00:02:34.500 --> 00:02:37.819
<v Speaker 3>should abolish the nonprofit sector. Now, I don't mean that literally,

55
00:02:37.840 --> 00:02:41.830
<v Speaker 3>because you go back to Tocqueville, American voluntary civic organizations

56
00:02:41.870 --> 00:02:44.270
<v Speaker 3>have always been part of America's DNA and what set

57
00:02:44.310 --> 00:02:47.039
<v Speaker 3>us apart. It's a key part of American exceptionalism. But

58
00:02:47.080 --> 00:02:48.580
<v Speaker 3>what I do think should be done is maybe we

59
00:02:48.599 --> 00:02:51.919
<v Speaker 3>should think about ending the tax deductibility of contributions to

60
00:02:51.939 --> 00:02:54.289
<v Speaker 3>the nonprofit sector. The reason I say that is the

61
00:02:54.349 --> 00:02:56.849
<v Speaker 3>nonprofit sector, not all of it, not our local churches

62
00:02:56.889 --> 00:02:59.830
<v Speaker 3>that we belong to, not the local Lions Club perhaps,

63
00:03:00.680 --> 00:03:04.099
<v Speaker 3>but the sector itself has, like so many institutions, become

64
00:03:04.280 --> 00:03:07.680
<v Speaker 3>thoroughly captured by the left. up to all kinds of mischief,

65
00:03:08.039 --> 00:03:10.460
<v Speaker 3>and making claims. And this actually goes back at least

66
00:03:10.460 --> 00:03:13.340
<v Speaker 3>30 years now. They make claims that the nonprofit sector

67
00:03:13.379 --> 00:03:18.099
<v Speaker 3>should be accorded the status of a formal stakeholder, including

68
00:03:18.280 --> 00:03:22.069
<v Speaker 3>appointments to corporate boards and input to all public policy decisions.

69
00:03:22.250 --> 00:03:25.530
<v Speaker 3>And I think this is insidious. And I thought, you know,

70
00:03:25.650 --> 00:03:27.629
<v Speaker 3>if we got rid of the tax deductibility for it,

71
00:03:27.789 --> 00:03:30.219
<v Speaker 3>it would clip the wings of, well– Let's start with

72
00:03:30.259 --> 00:03:32.599
<v Speaker 3>a lot of private liberal arts colleges where a lot

73
00:03:32.620 --> 00:03:35.060
<v Speaker 3>of alumni would quit giving their annual gifts if they

74
00:03:35.080 --> 00:03:38.159
<v Speaker 3>couldn't deduct it on their taxes. So what's not to like?

75
00:03:38.379 --> 00:03:41.000
<v Speaker 3>And it would also clip the wings a bit of

76
00:03:41.060 --> 00:03:45.289
<v Speaker 3>some of the huge left-wing foundations, Ford, Rockefeller, Soros, for

77
00:03:45.310 --> 00:03:49.210
<v Speaker 3>that matter, and Mellon Foundation and others. So I threw

78
00:03:49.229 --> 00:03:50.930
<v Speaker 3>it out as a thought experiment, and it got a

79
00:03:50.990 --> 00:03:55.830
<v Speaker 3>lot of buzz and controversy. And anyway, when I hear

80
00:03:55.849 --> 00:03:59.500
<v Speaker 3>people talk about the nonprofit sector, it just gets my back.

81
00:04:00.060 --> 00:04:00.840
<v Speaker 2>Charles, do you agree?

82
00:04:02.180 --> 00:04:02.780
<v Speaker 4>I don't know.

83
00:04:02.840 --> 00:04:08.889
<v Speaker 1>That's a remarkable change to our tax code and would

84
00:04:08.949 --> 00:04:11.830
<v Speaker 1>restructure an enormous number of things. What I would like

85
00:04:11.889 --> 00:04:14.969
<v Speaker 1>to see, which I think I do already have an

86
00:04:15.030 --> 00:04:23.470
<v Speaker 1>opinion on, is the government using non-profits less. Now, that

87
00:04:23.689 --> 00:04:27.399
<v Speaker 1>would be easy to affect. I mean, I'm thinking obviously,

88
00:04:27.439 --> 00:04:33.220
<v Speaker 1>of California and homelessness charities, which seem to rake in

89
00:04:33.379 --> 00:04:36.279
<v Speaker 1>billions of dollars while making the problem worse, or at

90
00:04:36.319 --> 00:04:41.180
<v Speaker 1>least not making them better. That is full of all

91
00:04:41.279 --> 00:04:44.959
<v Speaker 1>manner of ugly incentives. Whether I want to change the

92
00:04:45.000 --> 00:04:49.959
<v Speaker 1>tax code to make the contributions from private actors to

93
00:04:50.019 --> 00:04:58.509
<v Speaker 1>those non-profits illegal or treated... as any other contribution would be, rather?

94
00:04:58.550 --> 00:05:00.629
<v Speaker 1>I don't know. I just have to think about that, Steve.

95
00:05:01.689 --> 00:05:04.870
<v Speaker 3>Well, I mean, I just want a sort of brief rebuttal.

96
00:05:04.910 --> 00:05:07.470
<v Speaker 3>Most people who throw $ 20 into the collection plate at

97
00:05:07.509 --> 00:05:09.610
<v Speaker 3>their churches don't itemize. And there's quite a lot of

98
00:05:09.670 --> 00:05:13.220
<v Speaker 3>data on this, that conservatives are actually more charitable than liberals.

99
00:05:13.319 --> 00:05:16.300
<v Speaker 3>I mean, the famous story of Hillary Clinton deducting $ 1. 50

100
00:05:15.980 --> 00:05:18.600
<v Speaker 3>for bills used underwear on their tax forms is very

101
00:05:18.660 --> 00:05:21.040
<v Speaker 3>emblematic of the left. They want the government to do it.

102
00:05:21.470 --> 00:05:23.670
<v Speaker 3>And that's why an awful lot of the nonprofit sector

103
00:05:23.709 --> 00:05:26.910
<v Speaker 3>depends on government grants like the homeless industrial complex that

104
00:05:26.930 --> 00:05:30.290
<v Speaker 3>you mentioned. And so I actually think that, you know,

105
00:05:30.329 --> 00:05:34.360
<v Speaker 3>the sort of the good old Tocqueville style civil society

106
00:05:34.420 --> 00:05:36.959
<v Speaker 3>stuff would go on just fine without a tax deduction

107
00:05:37.000 --> 00:05:38.819
<v Speaker 3>for it. And, you know, I like to point out

108
00:05:38.860 --> 00:05:41.579
<v Speaker 3>that the Medici's and the rest, they didn't support Florentine

109
00:05:41.639 --> 00:05:44.579
<v Speaker 3>art for the tax deductions. And I think that people

110
00:05:44.600 --> 00:05:46.769
<v Speaker 3>who are seriously interested in the arts can still support it.

111
00:05:46.879 --> 00:05:49.720
<v Speaker 3>It's just that the taxpayers don't have to cross-subsidize it anymore.

112
00:05:49.759 --> 00:05:54.439
<v Speaker 3>That's the more modest version of it. I wouldn't abolish nonprofits.

113
00:05:54.480 --> 00:05:56.899
<v Speaker 3>That is a crazy idea. It would be cutting off

114
00:05:56.930 --> 00:05:58.259
<v Speaker 3>your nose in spite of your face.

115
00:05:58.670 --> 00:06:00.670
<v Speaker 2>You mentioned the Lions. I almost joined the Lions once,

116
00:06:00.709 --> 00:06:03.089
<v Speaker 2>but there was a scheduling conflict. What they did was

117
00:06:03.129 --> 00:06:05.170
<v Speaker 2>they cleaned up streets. They went places and they picked

118
00:06:05.209 --> 00:06:07.689
<v Speaker 2>up trash. And then they also had drives to collect

119
00:06:08.370 --> 00:06:12.290
<v Speaker 2>used eyeglasses, which would go to Africa. Wonderful bunch of people.

120
00:06:13.160 --> 00:06:16.259
<v Speaker 2>What I want to do after I've won the lottery, win, say,

121
00:06:16.259 --> 00:06:20.240
<v Speaker 2>$ 800 million in the lottery, I'm going to take $ 300

122
00:06:19.980 --> 00:06:22.699
<v Speaker 2>million of that and I'm going to create the Lilacs Foundation.

123
00:06:23.319 --> 00:06:27.040
<v Speaker 2>And I'm going to expressly staff it with as many Marxists, socialists,

124
00:06:27.180 --> 00:06:31.069
<v Speaker 2>and collectivists as I possibly can. The reason being is

125
00:06:31.110 --> 00:06:32.810
<v Speaker 2>that's the only way that I can assure that in

126
00:06:32.870 --> 00:06:38.480
<v Speaker 2>two generations it will do something completely different. Like the

127
00:06:38.600 --> 00:06:40.610
<v Speaker 2>Ford and the rest of them. If you set up

128
00:06:40.639 --> 00:06:43.110
<v Speaker 2>a foundation dedicated to American society, all of a sudden

129
00:06:43.129 --> 00:06:46.069
<v Speaker 2>in two generations, it's doing its best to ruin the country.

130
00:06:46.350 --> 00:06:47.769
<v Speaker 2>So I'm going to try it from the other way

131
00:06:47.810 --> 00:06:48.730
<v Speaker 2>and see how the institution works.

132
00:06:48.750 --> 00:06:54.269
<v Speaker 1>James, you could also give them totally pointless, fruitless work

133
00:06:54.310 --> 00:06:59.480
<v Speaker 1>to do, but pay them, thereby taking them out of society,

134
00:06:59.720 --> 00:07:04.209
<v Speaker 1>trapping them, the golden handcuffs, and making sure they can't

135
00:07:04.250 --> 00:07:07.699
<v Speaker 1>do any mischief whatsoever that affects anything that you care about.

136
00:07:07.720 --> 00:07:11.279
<v Speaker 2>That would be a good idea, because the underutilization of

137
00:07:11.319 --> 00:07:14.300
<v Speaker 2>these people leads them to do the stupidest, strangest things.

138
00:07:14.319 --> 00:07:17.300
<v Speaker 2>There's a case of this mentally ill woman in the

139
00:07:17.360 --> 00:07:19.540
<v Speaker 2>streets of New York who had stabbed two people, one

140
00:07:19.579 --> 00:07:21.839
<v Speaker 2>to death, and she was going after the cops and

141
00:07:21.899 --> 00:07:25.480
<v Speaker 2>was irrigated, and got the lettuce selling, and down she went.

142
00:07:26.259 --> 00:07:28.959
<v Speaker 2>And there's a lot of nailing and wailing and gnashing

143
00:07:29.000 --> 00:07:30.439
<v Speaker 2>about why didn't they shoot her in the arm, why

144
00:07:30.459 --> 00:07:33.420
<v Speaker 2>didn't they shoot her in the leg, please. But what's

145
00:07:33.480 --> 00:07:36.300
<v Speaker 2>interesting is the people who are defending the death of

146
00:07:36.360 --> 00:07:39.610
<v Speaker 2>the young woman that she'd stabbed because she worked for

147
00:07:39.620 --> 00:07:41.709
<v Speaker 2>the Bank of America and therefore she was part of

148
00:07:41.730 --> 00:07:45.029
<v Speaker 2>the industrial capitalist complex. She was a vice president, for

149
00:07:45.050 --> 00:07:48.310
<v Speaker 2>God's sakes. And there are people, legitimate people, actual people

150
00:07:48.410 --> 00:07:50.410
<v Speaker 2>on Twitter saying, I'm not, you know, this is what

151
00:07:50.449 --> 00:07:53.550
<v Speaker 2>we need to do more of. And it's a sign

152
00:07:53.689 --> 00:07:57.329
<v Speaker 2>of how ethically reprehensible and morally vacuous that these people

153
00:07:57.350 --> 00:07:59.810
<v Speaker 2>have become. This is the Luigi, this is the, you know,

154
00:07:59.829 --> 00:08:02.829
<v Speaker 2>the Luigi Mangione thing. uh you know fan club this

155
00:08:02.990 --> 00:08:06.050
<v Speaker 2>is this is them what constitutes a new set of

156
00:08:06.149 --> 00:08:09.620
<v Speaker 2>ethics that we have to get used to if the

157
00:08:09.649 --> 00:08:10.819
<v Speaker 2>revolution is going to come.

158
00:08:11.040 --> 00:08:13.300
<v Speaker 1>But that woman had no idea who she was killing

159
00:08:13.360 --> 00:08:16.740
<v Speaker 1>right no she didn't know whatsoever in effect that their

160
00:08:16.800 --> 00:08:20.079
<v Speaker 1>endorsement leads to the position that we should have random

161
00:08:20.120 --> 00:08:22.720
<v Speaker 1>mentally ill people running around manhattan with knives on the

162
00:08:22.779 --> 00:08:25.079
<v Speaker 1>off chance they kill someone that they wouldn't have liked

163
00:08:25.100 --> 00:08:27.769
<v Speaker 1>in the first place Exactly. And they tried to tase her,

164
00:08:27.779 --> 00:08:29.980
<v Speaker 1>but she was too fat, which was really, really funny

165
00:08:30.040 --> 00:08:33.389
<v Speaker 1>because that was the answer. And I saw this exchange

166
00:08:33.429 --> 00:08:36.259
<v Speaker 1>where someone said, Well, it just shows how inhumane and

167
00:08:36.299 --> 00:08:38.419
<v Speaker 1>barbaric our police are. Why didn't they tase it? Well,

168
00:08:38.460 --> 00:08:40.700
<v Speaker 1>they did tase it. Well, why didn't it work? It's

169
00:08:40.799 --> 00:08:41.899
<v Speaker 1>because she is huge.

170
00:08:42.120 --> 00:08:44.679
<v Speaker 3>But I think it's actually worse than that, Charlie. I

171
00:08:44.750 --> 00:08:47.629
<v Speaker 3>think we are back to what used to be part

172
00:08:47.669 --> 00:08:50.789
<v Speaker 3>of the radical left decades ago, which was a sanctification

173
00:08:50.830 --> 00:08:53.169
<v Speaker 3>of violence, a celebration of violence for its own sake.

174
00:08:53.370 --> 00:08:55.610
<v Speaker 3>Which makes me come back to your opener, James, which is,

175
00:08:55.649 --> 00:08:59.340
<v Speaker 3>you know, those seeds … look like they have sprouted

176
00:08:59.559 --> 00:09:03.039
<v Speaker 3>into a really noxious wheat. It's not a strong enough

177
00:09:03.539 --> 00:09:05.840
<v Speaker 3>metaphor for it, but you can see that those seeds

178
00:09:05.879 --> 00:09:08.700
<v Speaker 3>are emblematic of something that's growing up and growing very fast,

179
00:09:08.740 --> 00:09:09.620
<v Speaker 3>and it's very bad.

180
00:09:09.639 --> 00:09:11.730
<v Speaker 2>Yeah, it is. I have a piece in National Review

181
00:09:11.789 --> 00:09:14.990
<v Speaker 2>Online this week. My weekly column this weekend is about questions.

182
00:09:17.850 --> 00:09:20.409
<v Speaker 2>Here's my job. Do I deserve stabbing? This is what

183
00:09:20.450 --> 00:09:22.029
<v Speaker 2>I do for the company. Is this going to give

184
00:09:22.049 --> 00:09:24.129
<v Speaker 2>me a stabbing? So you can pick that up on

185
00:09:24.220 --> 00:09:25.559
<v Speaker 2>NRL if you like.

186
00:09:25.639 --> 00:09:27.480
<v Speaker 1>I'll send you my answers to the questions.

187
00:09:28.250 --> 00:09:30.330
<v Speaker 2>Oh, I've already answered them, believe me. I answered them

188
00:09:30.690 --> 00:09:33.350
<v Speaker 2>in the voice of a 27-year-old DSA woman who is

189
00:09:33.389 --> 00:09:35.970
<v Speaker 2>living in Brooklyn in a $ 1. 7 million apartment.

190
00:09:36.529 --> 00:09:37.309
<v Speaker 1>So I don't get a say.

191
00:09:37.870 --> 00:09:40.320
<v Speaker 2>No, the piece is written and it's done.

192
00:09:40.340 --> 00:09:43.340
<v Speaker 1>No, I just thought I would read the piece and

193
00:09:43.360 --> 00:09:45.240
<v Speaker 1>then I would send you my thoughts as to whether

194
00:09:45.259 --> 00:09:46.379
<v Speaker 1>or not you deserve to be stabbed.

195
00:09:46.460 --> 00:09:47.460
<v Speaker 3>Okay.

196
00:09:47.759 --> 00:09:50.480
<v Speaker 2>I'm sure you've got some pull over there, Ben Hart. So, yes,

197
00:09:50.740 --> 00:09:54.320
<v Speaker 2>you can do this if you wish. Two stories before

198
00:09:54.360 --> 00:09:57.480
<v Speaker 2>we go to our guest. Do you want to talk

199
00:09:57.519 --> 00:09:59.899
<v Speaker 2>about the passing of Gloria Steinem or the fact that

200
00:09:59.919 --> 00:10:00.940
<v Speaker 2>the ballroom is going to be built?

201
00:10:02.710 --> 00:10:05.789
<v Speaker 3>Well, a quick comment on the ballroom thing. It's a

202
00:10:05.830 --> 00:10:09.009
<v Speaker 3>ridiculous case for the pleading of the plaintiff from the

203
00:10:09.049 --> 00:10:12.210
<v Speaker 3>National Trust for Historic Preservation. Again, one of those nonprofits

204
00:10:12.230 --> 00:10:15.690
<v Speaker 3>that's become a nuisance over the decades. And the complaint was, well,

205
00:10:15.750 --> 00:10:19.240
<v Speaker 3>I'm aesthetically offended having to walk by the ballroom. Now,

206
00:10:19.299 --> 00:10:22.580
<v Speaker 3>if being aesthetically offended at federal architecture is a legitimate

207
00:10:22.620 --> 00:10:26.590
<v Speaker 3>cause of action– then I'm going to put retroactive lawsuits

208
00:10:26.649 --> 00:10:29.110
<v Speaker 3>and flood the federal courts for every federal courthouse that's

209
00:10:29.129 --> 00:10:30.470
<v Speaker 3>been built, just for starters.

210
00:10:30.570 --> 00:10:31.129
<v Speaker 4>Absolutely.

211
00:10:31.169 --> 00:10:33.870
<v Speaker 2>We have a class action suit over the FBI building

212
00:10:33.909 --> 00:10:34.970
<v Speaker 2>in D.C.

213
00:10:35.169 --> 00:10:40.009
<v Speaker 1>I have to say, I was completely perplexed by John Roberts.

214
00:10:40.350 --> 00:10:40.529
<v Speaker 4>Yes.

215
00:10:41.230 --> 00:10:45.809
<v Speaker 1>John Roberts is not the villain that many on the

216
00:10:45.850 --> 00:10:49.090
<v Speaker 1>right think he is. But he is often a squish.

217
00:10:50.600 --> 00:10:58.500
<v Speaker 1>And rarely so on standings. So, suddenly, he, what, becomes

218
00:10:59.580 --> 00:11:05.309
<v Speaker 1>a standing dove? And also, I understand this is not

219
00:11:05.350 --> 00:11:07.870
<v Speaker 1>the mainstream view, but I don't think that there's a

220
00:11:07.889 --> 00:11:09.909
<v Speaker 1>case on the merits. I went back and I read

221
00:11:09.970 --> 00:11:12.600
<v Speaker 1>the statute that Congress passed, which I don't think is

222
00:11:12.620 --> 00:11:14.980
<v Speaker 1>a particularly good statute, and as usual, I would like

223
00:11:15.039 --> 00:11:18.019
<v Speaker 1>Congress to do its job rather than delegate power. But

224
00:11:18.159 --> 00:11:21.690
<v Speaker 1>it refers to improvements to the building. And so, I

225
00:11:21.730 --> 00:11:26.029
<v Speaker 1>looked at what that word has meant historically in America,

226
00:11:26.230 --> 00:11:30.250
<v Speaker 1>and this obviously counts as an improvement. They didn't take

227
00:11:30.289 --> 00:11:34.649
<v Speaker 1>the building down. They removed a part of the building

228
00:11:34.710 --> 00:11:38.450
<v Speaker 1>that had previously been an improvement, and they're improving it.

229
00:11:38.850 --> 00:11:40.879
<v Speaker 1>How could that not count under this statute?

230
00:11:40.889 --> 00:11:41.500
<v Speaker 4>Yeah.

231
00:11:42.240 --> 00:11:44.779
<v Speaker 2>Well, it's because Trump is building it, period. Yes. And

232
00:11:44.799 --> 00:11:46.460
<v Speaker 2>I don't believe they have any legal basis to get

233
00:11:46.500 --> 00:11:50.259
<v Speaker 2>rid of the march, the arch that he's building. But

234
00:11:50.659 --> 00:11:53.190
<v Speaker 2>if they could, they would, I'm sure. And whether or

235
00:11:53.210 --> 00:11:56.990
<v Speaker 2>not that's going to be an aesthetic success, we will see. Well,

236
00:11:57.029 --> 00:12:00.730
<v Speaker 2>such is the world. How about let's talk about the future?

237
00:12:04.730 --> 00:12:07.029
<v Speaker 2>And now we welcome to the podcast Mark Mills, founder

238
00:12:07.090 --> 00:12:09.779
<v Speaker 2>and executive director of the National Center for Energy Analytics,

239
00:12:10.039 --> 00:12:13.559
<v Speaker 2>a contributing editor to the Manhattan Institute's City Journal, and

240
00:12:13.659 --> 00:12:16.909
<v Speaker 2>author of The Cloud Revolution, How the Convergence of New

241
00:12:16.950 --> 00:12:21.210
<v Speaker 2>Technologies Will Unleash the Next Economic Boom in a Roaring 2020s. Mark, welcome.

242
00:12:22.940 --> 00:12:26.659
<v Speaker 2>It's been an interesting week for AI, has it not?

243
00:12:26.679 --> 00:12:29.100
<v Speaker 2>I mean, we had four models, we had four AIs

244
00:12:29.279 --> 00:12:33.289
<v Speaker 2>go down simultaneously, everybody thinking that they'd been nixed by

245
00:12:33.350 --> 00:12:37.570
<v Speaker 2>some rogue other LLM that was out there killing at will.

246
00:12:38.029 --> 00:12:41.309
<v Speaker 2>We had all these stories about hugging face experiments that seem, again,

247
00:12:41.350 --> 00:12:45.700
<v Speaker 2>to summon up a dystopia happening beneath the surface of

248
00:12:45.740 --> 00:12:49.340
<v Speaker 2>the placid computers with their blinking lights. We had a massive...

249
00:12:49.460 --> 00:12:54.139
<v Speaker 2>massive outrage on Twitter, bots, social media, everywhere about AI

250
00:12:54.279 --> 00:12:59.049
<v Speaker 2>food menus, which brought a new form of tryptophobia discussed

251
00:12:59.070 --> 00:13:02.830
<v Speaker 2>to people. We're grappling with something that is barreling towards

252
00:13:02.909 --> 00:13:05.590
<v Speaker 2>us very fast, or we feel like we're strapped to

253
00:13:05.629 --> 00:13:08.259
<v Speaker 2>the front of a bullet train and we don't know

254
00:13:08.299 --> 00:13:12.840
<v Speaker 2>the destination. So let's slow down. Let's ramp back the panic.

255
00:13:13.799 --> 00:13:17.799
<v Speaker 2>You've been called the last optimist for AI. Okay. give

256
00:13:17.840 --> 00:13:20.120
<v Speaker 2>us some more reasons why the people who are out

257
00:13:20.159 --> 00:13:23.730
<v Speaker 2>there saying, I am not at all paid by China

258
00:13:23.789 --> 00:13:27.049
<v Speaker 2>to oppose data centers. I just think they're bad because

259
00:13:28.070 --> 00:13:31.809
<v Speaker 2>they use water. Calm us down here and tell us

260
00:13:31.850 --> 00:13:33.250
<v Speaker 2>why we should still be optimistic.

261
00:13:33.289 --> 00:13:35.649
<v Speaker 4>Well, you could take the position that humans are bad

262
00:13:35.690 --> 00:13:36.710
<v Speaker 4>because they use water.

263
00:13:37.389 --> 00:13:41.629
<v Speaker 2>Especially the golfers. Especially the golfers eating almonds. They're the worst.

264
00:13:42.149 --> 00:13:44.450
<v Speaker 4>Well, if you really want to know the biggest offender

265
00:13:44.570 --> 00:13:49.409
<v Speaker 4>for profligate water use, It's the RFS, Renewable Fuel Standards,

266
00:13:49.450 --> 00:13:53.570
<v Speaker 4>and ethanol. Just not to digress into the rabbit hole

267
00:13:53.789 --> 00:14:00.799
<v Speaker 4>of idiot standards for effectless policies, about 600-fold more water

268
00:14:00.940 --> 00:14:04.419
<v Speaker 4>used per gallon of gasoline if you go ethanol than

269
00:14:04.480 --> 00:14:09.870
<v Speaker 4>if you stick with just gasoline. 600-fold. That's an astonishing

270
00:14:09.909 --> 00:14:13.769
<v Speaker 4>increase in water use and water waste. But I don't

271
00:14:13.789 --> 00:14:16.149
<v Speaker 4>think I'm the last optimist, but I'm also not Pollyanna.

272
00:14:16.190 --> 00:14:19.870
<v Speaker 4>I mean, I'm a realist. I've been in the investment world.

273
00:14:19.889 --> 00:14:23.960
<v Speaker 4>I've been in the policy world. Reality will slap you

274
00:14:24.000 --> 00:14:27.120
<v Speaker 4>upside the face if you pretend to ignore it. That's

275
00:14:27.139 --> 00:14:32.139
<v Speaker 4>why we use the Philip K. Dick quote on our website,

276
00:14:32.200 --> 00:14:35.179
<v Speaker 4>which is the great line that reality is that which

277
00:14:35.220 --> 00:14:36.960
<v Speaker 4>still exists even when you ignore it.

278
00:14:38.610 --> 00:14:41.350
<v Speaker 2>Well, using Philip K. Dick, of course, who believed in

279
00:14:41.389 --> 00:14:43.740
<v Speaker 2>the valley behind everything else.

280
00:14:43.750 --> 00:14:47.740
<v Speaker 4>It's perfect because he's a dystopian science fiction writer. And

281
00:14:47.759 --> 00:14:50.720
<v Speaker 4>the only thing that makes good science fiction is dystopia. But,

282
00:14:50.879 --> 00:14:53.039
<v Speaker 4>you know, that's why the stories sell.

283
00:14:53.240 --> 00:14:55.659
<v Speaker 2>So no one's come to your door speaking Greek with

284
00:14:55.700 --> 00:14:57.820
<v Speaker 2>a strange medallion. It's changed your life is what I

285
00:14:57.840 --> 00:14:58.240
<v Speaker 2>want to know.

286
00:14:58.860 --> 00:14:59.159
<v Speaker 3>Okay.

287
00:15:00.919 --> 00:15:04.500
<v Speaker 4>Well, I would say that the roaring 2020s is happening.

288
00:15:05.000 --> 00:15:07.009
<v Speaker 4>It took a couple of years to get rolling. I mean,

289
00:15:07.029 --> 00:15:09.480
<v Speaker 4>when I wrote the book, It came out the end

290
00:15:09.519 --> 00:15:12.779
<v Speaker 4>of 2021, really was rolling in January 2022, right before

291
00:15:12.840 --> 00:15:15.700
<v Speaker 4>ChatGPT came out. And about a quarter of my books

292
00:15:15.720 --> 00:15:18.000
<v Speaker 4>about AI, the rest is about the other technologies and

293
00:15:18.019 --> 00:15:21.419
<v Speaker 4>the penumbra of the sort of the ecosystem that's changing

294
00:15:21.470 --> 00:15:24.110
<v Speaker 4>the world. I mean, material science is a big deal.

295
00:15:24.169 --> 00:15:27.590
<v Speaker 4>People ignore it because it's boring. What's happening in education

296
00:15:27.629 --> 00:15:30.830
<v Speaker 4>is a big deal, not just from AI. You know,

297
00:15:30.850 --> 00:15:35.490
<v Speaker 4>what's going on in discovery in medicine and biological sciences

298
00:15:35.509 --> 00:15:37.990
<v Speaker 4>is huge. Lots of stuff going on. So I tried to...

299
00:15:38.669 --> 00:15:40.720
<v Speaker 4>I tried to put a thread through my book, and

300
00:15:40.750 --> 00:15:42.820
<v Speaker 4>the thread was, I think, the correct thread, and that's

301
00:15:42.840 --> 00:15:45.860
<v Speaker 4>why it's called the cloud revolution, is that the infrastructure

302
00:15:45.899 --> 00:15:49.120
<v Speaker 4>of the cloud writ large is the animating thread to

303
00:15:49.220 --> 00:15:52.120
<v Speaker 4>all the revolutions that are underway, both good and bad,

304
00:15:52.200 --> 00:15:56.179
<v Speaker 4>and that people's unhappiness is predicted and predictable. I mean,

305
00:15:56.919 --> 00:15:59.960
<v Speaker 4>we are in another cycle of great disruption, but I

306
00:16:00.000 --> 00:16:04.220
<v Speaker 4>will say this. Anybody who spends 10 minutes asking Claude

307
00:16:04.279 --> 00:16:08.559
<v Speaker 4>or Gemini or any AI agent to search history and

308
00:16:08.580 --> 00:16:11.340
<v Speaker 4>give you a summary of great disruptions. This is not

309
00:16:11.379 --> 00:16:13.960
<v Speaker 4>the first time, it's not our first rodeo, but it's

310
00:16:13.980 --> 00:16:17.480
<v Speaker 4>the first time in the living memory of pretty much

311
00:16:17.539 --> 00:16:20.500
<v Speaker 4>everybody today, right? The last great disruption, I would argue,

312
00:16:20.559 --> 00:16:24.879
<v Speaker 4>was the 1920s. That's why I use the roaring 2020s

313
00:16:25.879 --> 00:16:31.710
<v Speaker 4>in terms of a confluence of huge changes in technology, society, politics.

314
00:16:31.850 --> 00:16:35.120
<v Speaker 4>It was very, very similar to our era. And And

315
00:16:35.340 --> 00:16:37.610
<v Speaker 4>then there are negative consequences. One of them was the

316
00:16:37.639 --> 00:16:40.429
<v Speaker 4>creation of the Soviet Union. I mean, that's what, in

317
00:16:40.450 --> 00:16:44.090
<v Speaker 4>a large measure, helped fuel the, you know, Karl Marx

318
00:16:44.230 --> 00:16:49.049
<v Speaker 4>and the whole socialist framing and the creation of the

319
00:16:49.070 --> 00:16:53.009
<v Speaker 4>Soviet Union was the disruptions that were underway. And I

320
00:16:53.049 --> 00:16:55.419
<v Speaker 4>think if I were a pessimist, I would say I

321
00:16:55.460 --> 00:16:59.879
<v Speaker 4>worry that we're seeing that movie replayed too.

322
00:17:01.100 --> 00:17:04.299
<v Speaker 3>Mark, have you been surprised, as I have, at how

323
00:17:04.579 --> 00:17:08.730
<v Speaker 3>rapidly the opposition to data centers has congealed, and in

324
00:17:08.769 --> 00:17:11.029
<v Speaker 3>a bipartisan way? I mean, I'm having flashbacks to the

325
00:17:11.089 --> 00:17:14.380
<v Speaker 3>anti-nuclear movements of the 70s, which... And maybe my memory

326
00:17:14.400 --> 00:17:16.220
<v Speaker 3>is faulty, but it seemed like it took longer to

327
00:17:16.259 --> 00:17:18.440
<v Speaker 3>take hold. It was mostly on the left, although it

328
00:17:18.549 --> 00:17:23.390
<v Speaker 3>easily combined with old-fashioned middle-class nimbyism. And this has gone

329
00:17:23.490 --> 00:17:27.089
<v Speaker 3>from zero to 60 in a tenth of the time

330
00:17:27.130 --> 00:17:31.029
<v Speaker 3>the anti-nuclear movement did. And so I'm totally shocked. And

331
00:17:31.049 --> 00:17:33.839
<v Speaker 3>you can see all the Republican politicians like Governor Abbott

332
00:17:33.859 --> 00:17:36.289
<v Speaker 3>in Texas. lining up to say, oh, wait a minute,

333
00:17:36.329 --> 00:17:38.529
<v Speaker 3>and this looks very bad to me. I mean, here's

334
00:17:38.589 --> 00:17:41.109
<v Speaker 3>where I think Trump is right to say, if we

335
00:17:41.150 --> 00:17:43.710
<v Speaker 3>want to be, you know, poor and behind the curve,

336
00:17:44.410 --> 00:17:47.579
<v Speaker 3>let's block data centers. But so were you surprised at this?

337
00:17:48.259 --> 00:17:50.460
<v Speaker 4>So you already triggered a lot of people by using

338
00:17:50.480 --> 00:17:55.500
<v Speaker 4>the phrase, I think Trump is right. So let's just,

339
00:17:55.960 --> 00:18:01.900
<v Speaker 4>let's leave that particular rabbit hole alone for a the

340
00:18:01.940 --> 00:18:04.799
<v Speaker 4>importance of data center infrastructure are right. It is critical.

341
00:18:04.839 --> 00:18:08.380
<v Speaker 4>It is a structural shift in our economy with all

342
00:18:08.400 --> 00:18:10.339
<v Speaker 4>the things that animate people to be worried. And a

343
00:18:10.740 --> 00:18:12.599
<v Speaker 4>lot of people have good reason to be worried about things.

344
00:18:12.660 --> 00:18:17.039
<v Speaker 4>I mean, you know, that old saying, you may be paranoid,

345
00:18:17.059 --> 00:18:18.500
<v Speaker 4>but the bullets are real. I mean, there are some

346
00:18:18.539 --> 00:18:21.339
<v Speaker 4>people whose jobs are a threat, for sure. The jobs

347
00:18:21.380 --> 00:18:27.470
<v Speaker 4>apocalypse narrative is demonstrably bunk, but there are people who

348
00:18:27.490 --> 00:18:29.869
<v Speaker 4>will be negatively affected. You build big industrial facilities, and

349
00:18:29.890 --> 00:18:33.180
<v Speaker 4>that's what a data center is. Jensen and Wine likes

350
00:18:33.200 --> 00:18:38.710
<v Speaker 4>to use the expression AI factory. And it's actually a

351
00:18:38.730 --> 00:18:41.089
<v Speaker 4>pretty good framing. It's a better framing than data center

352
00:18:41.869 --> 00:18:44.990
<v Speaker 4>because there's a distinction between data centers that do AI

353
00:18:45.009 --> 00:18:47.049
<v Speaker 4>and data centers that do what we're doing, which is

354
00:18:47.549 --> 00:18:51.930
<v Speaker 4>record videos or buy things or shop or keep medical records.

355
00:18:52.450 --> 00:18:57.150
<v Speaker 4>And that data management, that data processing is, as we

356
00:18:57.210 --> 00:18:59.950
<v Speaker 4>all know, having tried AI, really different than using an

357
00:19:00.009 --> 00:19:04.410
<v Speaker 4>AI agent to answer a question. But I am like you.

358
00:19:04.430 --> 00:19:09.799
<v Speaker 4>I've used the analogy about the nuclear opposition deliberately because

359
00:19:10.359 --> 00:19:14.640
<v Speaker 4>you and I are of a certain age. And as

360
00:19:14.680 --> 00:19:17.319
<v Speaker 4>a young man, I was at the accident at Three

361
00:19:17.339 --> 00:19:19.960
<v Speaker 4>Mile Island during the week of the accident and then

362
00:19:20.019 --> 00:19:22.559
<v Speaker 4>spent the next, I guess, seven years of my life

363
00:19:22.680 --> 00:19:27.700
<v Speaker 4>defending the virtues of nuclear energy. Obviously, unsuccessfully because the

364
00:19:27.740 --> 00:19:31.220
<v Speaker 4>crazies killed the industry. The architecture of the debate is

365
00:19:31.279 --> 00:19:35.599
<v Speaker 4>very similar. Apocalyptic fears, big utilities. There was a time

366
00:19:35.619 --> 00:19:38.240
<v Speaker 4>when the left hated big utilities. Now they love them

367
00:19:38.259 --> 00:19:43.170
<v Speaker 4>because they successfully co-opted them, basically. But they won that battle.

368
00:19:43.509 --> 00:19:46.630
<v Speaker 4>They got to own them intellectually, if you like. And

369
00:19:46.690 --> 00:19:49.250
<v Speaker 4>even to some extent legally by the mandates that are

370
00:19:49.529 --> 00:19:52.819
<v Speaker 4>forced on. many utilities yeah but it is a little

371
00:19:52.970 --> 00:19:54.910
<v Speaker 4>it is definitely faster i think i think the reason

372
00:19:54.970 --> 00:19:58.400
<v Speaker 4>it's faster is that they they uh are it's the

373
00:19:58.440 --> 00:20:01.900
<v Speaker 4>same people and organizations many of them and they honed

374
00:20:01.940 --> 00:20:05.160
<v Speaker 4>their skills they honed the the techniques uh and the

375
00:20:05.220 --> 00:20:08.079
<v Speaker 4>nuclear debates and so it was pretty fast then i

376
00:20:08.119 --> 00:20:10.250
<v Speaker 4>mean i i When I came to the United States

377
00:20:10.369 --> 00:20:14.470
<v Speaker 4>before Three Mile Island, nuclear energy was still very popular

378
00:20:14.490 --> 00:20:18.789
<v Speaker 4>in a bipartisan way. It was the future, beating swords

379
00:20:18.809 --> 00:20:24.059
<v Speaker 4>to plowshares, the peaceful atom, all this stuff. Nixon predicted 1,000

380
00:20:24.059 --> 00:20:26.619
<v Speaker 4>reactors in reaction to the oil embargo of 1973 by

381
00:20:26.640 --> 00:20:29.259
<v Speaker 4>the year 2000. We got to 100.

382
00:20:29.259 --> 00:20:34.500
<v Speaker 3>It's close. Well, not bad for government work, yeah.

383
00:20:34.519 --> 00:20:38.579
<v Speaker 4>But it is– It makes one suspicious. There's more to it.

384
00:20:39.109 --> 00:20:43.410
<v Speaker 4>And I think the more to it is partly it's

385
00:20:43.490 --> 00:20:46.420
<v Speaker 4>being fed by anxiety about what AI and social media,

386
00:20:46.460 --> 00:20:49.480
<v Speaker 4>all that, and the antipathy, if not hate, for big

387
00:20:49.559 --> 00:20:53.200
<v Speaker 4>tech and the coastal elites that's feeding into it. I

388
00:20:53.259 --> 00:20:56.059
<v Speaker 4>have no doubt, I think, Elon Musk's research team on

389
00:20:56.180 --> 00:21:01.930
<v Speaker 4>Exit reported hundreds of thousands of Chinese chatbots maliciously reinforcing

390
00:21:01.970 --> 00:21:04.809
<v Speaker 4>negative stories on social media. I have no doubt that

391
00:21:04.849 --> 00:21:05.390
<v Speaker 4>they're meddling.

392
00:21:05.410 --> 00:21:05.609
<v Speaker 2>Okay.

393
00:21:05.950 --> 00:21:09.819
<v Speaker 4>I mean, look, you'd be profoundly naive to think that Russia,

394
00:21:09.839 --> 00:21:14.160
<v Speaker 4>Iran and China are not meddling in our politics and

395
00:21:14.259 --> 00:21:18.619
<v Speaker 4>trying to animate discord. Similarly, when you build really big

396
00:21:18.660 --> 00:21:21.740
<v Speaker 4>things near places where people haven't had big things built before,

397
00:21:22.650 --> 00:21:26.730
<v Speaker 4>for which the benefits are generally remote, except for tax reductions.

398
00:21:27.880 --> 00:21:30.200
<v Speaker 4>you get a lot of NIMBY. That's the nature of

399
00:21:30.240 --> 00:21:31.109
<v Speaker 4>the beast.

400
00:21:32.160 --> 00:21:36.450
<v Speaker 3>One other question to one particular aspect of this. I

401
00:21:36.490 --> 00:21:38.769
<v Speaker 3>did just see a paper out this morning. I think

402
00:21:39.029 --> 00:21:41.289
<v Speaker 3>I spotted it on the Social Science Research Network. I

403
00:21:41.309 --> 00:21:43.430
<v Speaker 3>haven't read it yet, but it's a study that says

404
00:21:43.450 --> 00:21:46.890
<v Speaker 3>there's really no evidence that data centers are driving up

405
00:21:46.950 --> 00:21:50.109
<v Speaker 3>electricity prices. It's some kind of regression model, of course. Now,

406
00:21:50.230 --> 00:21:53.710
<v Speaker 3>I remember one of your most interesting calculations, I think

407
00:21:53.730 --> 00:21:55.710
<v Speaker 3>more than 20 years ago, if I remember this right,

408
00:21:56.190 --> 00:21:58.259
<v Speaker 3>Back when we were still getting half of our electricity

409
00:21:58.279 --> 00:22:01.660
<v Speaker 3>from coal in this country. And you had the calculation

410
00:22:01.680 --> 00:22:04.619
<v Speaker 3>that every Google search generated, you know what, a quarter

411
00:22:04.680 --> 00:22:08.470
<v Speaker 3>pound of carbon dioxide emissions, whatever it was. And only

412
00:22:08.490 --> 00:22:12.130
<v Speaker 3>you do these kind of calculations that nobody else thinks about. And,

413
00:22:12.569 --> 00:22:15.690
<v Speaker 3>of course, that has all changed with the changing electricity

414
00:22:15.900 --> 00:22:19.460
<v Speaker 3>generation profile and natural gas and so forth. Uh, but

415
00:22:19.480 --> 00:22:22.250
<v Speaker 3>it does seem to me that, um, uh, you know,

416
00:22:22.269 --> 00:22:24.410
<v Speaker 3>here in California where they're not building a lot of

417
00:22:24.470 --> 00:22:28.509
<v Speaker 3>data centers because our electricity is too expensive. Um, uh,

418
00:22:28.589 --> 00:22:30.910
<v Speaker 3>you are seeing a lot of people who see rising

419
00:22:30.930 --> 00:22:34.549
<v Speaker 3>electricity rates, which happening for secular reasons and say, Hey,

420
00:22:34.670 --> 00:22:36.539
<v Speaker 3>I was just going to make it worse. But I

421
00:22:36.559 --> 00:22:37.880
<v Speaker 3>think that's probably not true.

422
00:22:38.079 --> 00:22:38.289
<v Speaker 2>Is it?

423
00:22:39.299 --> 00:22:42.579
<v Speaker 4>Yeah. Well, your instincts as always are spot on. It's

424
00:22:42.619 --> 00:22:46.440
<v Speaker 4>not true. The data don't show. In fact, uh, we,

425
00:22:46.480 --> 00:22:49.609
<v Speaker 4>you know, NCEA, uh, we partnered with USC, your buddies

426
00:22:49.670 --> 00:22:52.690
<v Speaker 4>at University of Southern California, to do a statistical analysis

427
00:22:52.730 --> 00:22:55.950
<v Speaker 4>of the California ISO, that region, because there are a

428
00:22:55.950 --> 00:22:58.710
<v Speaker 4>lot of data centers in California already. There are more

429
00:22:58.750 --> 00:23:01.369
<v Speaker 4>being built, by the way, because sometimes you have no choice.

430
00:23:01.390 --> 00:23:03.349
<v Speaker 4>It has to do with latency. You have to have

431
00:23:03.410 --> 00:23:08.769
<v Speaker 4>the data exchange close to where you are. But there

432
00:23:08.829 --> 00:23:12.579
<v Speaker 4>is a small impact of big data centers on wholesale

433
00:23:13.279 --> 00:23:16.779
<v Speaker 4>congestion costs. So when you say that, you can just

434
00:23:16.819 --> 00:23:20.779
<v Speaker 4>think about Any kind of traffic, data creates traffic, power

435
00:23:20.819 --> 00:23:24.279
<v Speaker 4>creates traffic, congestion. But it doesn't correlate with any retail

436
00:23:24.660 --> 00:23:27.720
<v Speaker 4>cost impacts. And the only correlation we have in the

437
00:23:27.759 --> 00:23:32.089
<v Speaker 4>United States with higher electricity costs is the factless and

438
00:23:32.430 --> 00:23:36.069
<v Speaker 4>ill-advised policies to jam solar and wind mandates down the

439
00:23:36.109 --> 00:23:39.609
<v Speaker 4>throats of utilities and ratepayers. And what that's done, even

440
00:23:39.630 --> 00:23:42.950
<v Speaker 4>though the sun is, quote, free when it's shining, is

441
00:23:42.970 --> 00:23:46.490
<v Speaker 4>that the system operating costs to compensate for that raise

442
00:23:46.589 --> 00:23:48.569
<v Speaker 4>overall costs of electricity, and they have. And that's been

443
00:23:48.609 --> 00:23:51.390
<v Speaker 4>demonstrated over and over again. So now what's going on

444
00:23:51.430 --> 00:23:55.309
<v Speaker 4>is the politicians conveniently can blame big tech for causing

445
00:23:55.349 --> 00:23:58.549
<v Speaker 4>high rates. They caused, not big tech, but the politicians,

446
00:23:59.089 --> 00:24:03.589
<v Speaker 4>by forcing policies onto utilities that lead to affirmatively higher

447
00:24:03.630 --> 00:24:06.369
<v Speaker 4>costs to run the grids. Germany did the experiment. They

448
00:24:06.410 --> 00:24:11.410
<v Speaker 4>tripled the cost of electricity by going further down, faster

449
00:24:11.430 --> 00:24:13.769
<v Speaker 4>down the path that a lot of U.S. states are pursuing.

450
00:24:14.339 --> 00:24:19.940
<v Speaker 4>So it's a convenient whipping horse, and the answer to

451
00:24:19.960 --> 00:24:22.690
<v Speaker 4>that is simple, and I think that's already been solved

452
00:24:22.750 --> 00:24:26.150
<v Speaker 4>in the sense that big tech pledges to pay their costs. Okay,

453
00:24:27.009 --> 00:24:29.089
<v Speaker 4>so today's headline, you may have missed this one. It

454
00:24:29.130 --> 00:24:35.700
<v Speaker 4>was delicious and stupid. Microsoft is suing Dominion Energy, the

455
00:24:35.759 --> 00:24:38.000
<v Speaker 4>utility in Virginia, where they have lots of data centers,

456
00:24:39.079 --> 00:24:42.329
<v Speaker 4>pushing back on absorbing all the costs of for the

457
00:24:42.369 --> 00:24:45.289
<v Speaker 4>power that they want, claiming that the utility is overstating

458
00:24:45.309 --> 00:24:49.130
<v Speaker 4>the cost or whatever. The trade here is, even if

459
00:24:49.150 --> 00:24:50.990
<v Speaker 4>we overstate it, you should pay for it. You've got

460
00:24:51.009 --> 00:24:53.410
<v Speaker 4>the money, and you want to get out of jail

461
00:24:53.430 --> 00:24:56.369
<v Speaker 4>card here with respect to electric rates. You're going to

462
00:24:56.390 --> 00:24:58.670
<v Speaker 4>have to step up. That's sort of the political trade,

463
00:24:59.099 --> 00:25:02.240
<v Speaker 4>but Microsoft has pushed back. So, you know, we'll see

464
00:25:02.279 --> 00:25:03.299
<v Speaker 4>how that goes politically.

465
00:25:07.319 --> 00:25:09.019
<v Speaker 2>If I can jump in here for a second, I

466
00:25:09.019 --> 00:25:11.349
<v Speaker 2>know that Charles has got a million questions. And I'm

467
00:25:11.390 --> 00:25:14.750
<v Speaker 2>sorry that my video cut out here. I blame AI,

468
00:25:14.880 --> 00:25:15.700
<v Speaker 2>of course, for doing this.

469
00:25:15.710 --> 00:25:18.359
<v Speaker 4>You should. Well, the Chinese, the Chinese bots.

470
00:25:18.839 --> 00:25:20.460
<v Speaker 2>In two or three weeks, I'll be able to summon

471
00:25:20.500 --> 00:25:22.200
<v Speaker 2>up a life like Avatar that will just do everything

472
00:25:22.220 --> 00:25:23.980
<v Speaker 2>that I can do. So I'm not worried about that.

473
00:25:24.059 --> 00:25:27.200
<v Speaker 2>But you'd mentioned something about the 1920s. And it's interesting,

474
00:25:27.259 --> 00:25:30.740
<v Speaker 2>the 1920s, which is a fascinating era in America, history

475
00:25:30.769 --> 00:25:32.990
<v Speaker 2>and European history, was a time in which all of

476
00:25:33.049 --> 00:25:37.430
<v Speaker 2>these technologies coalesced and became mature. So you'd had telephones before,

477
00:25:37.450 --> 00:25:39.509
<v Speaker 2>you'd had cars before, you'd had movies, but now you

478
00:25:39.549 --> 00:25:42.819
<v Speaker 2>have all of these things at the peak of operating efficiency,

479
00:25:43.059 --> 00:25:46.319
<v Speaker 2>and you have radio, and you have mass market media.

480
00:25:46.779 --> 00:25:49.420
<v Speaker 2>So you have all of these elements that combine to

481
00:25:49.460 --> 00:25:52.349
<v Speaker 2>produce a national conversation, a national culture, the way that

482
00:25:52.369 --> 00:25:54.769
<v Speaker 2>we've never really been able to do before. I mean,

483
00:25:54.930 --> 00:25:57.490
<v Speaker 2>you know, Horace Greeley could send a magazine out all

484
00:25:57.509 --> 00:25:59.210
<v Speaker 2>around the country, but it didn't have the same effect

485
00:25:59.289 --> 00:26:03.140
<v Speaker 2>as somebody getting on the radio and telling a joke. Now,

486
00:26:03.740 --> 00:26:05.700
<v Speaker 2>what we have seemingly is the opposite of that. We

487
00:26:05.720 --> 00:26:09.339
<v Speaker 2>have an extremely robust, efficient, and brilliant technology that is

488
00:26:09.440 --> 00:26:13.549
<v Speaker 2>not achieving a sort of cultural unification, but it has

489
00:26:13.589 --> 00:26:17.890
<v Speaker 2>the effect of fragmenting it into a kaleidoscopic, balkanized world, which,

490
00:26:17.910 --> 00:26:20.329
<v Speaker 2>of course, the Russians and Chinese are keen to do.

491
00:26:20.710 --> 00:26:23.089
<v Speaker 2>We've invited into our homes a TikTok algorithm that says

492
00:26:23.109 --> 00:26:24.859
<v Speaker 2>we're going to make you unhappy about everything from your

493
00:26:24.910 --> 00:26:26.960
<v Speaker 2>gender to your relationship to your food to your country.

494
00:26:27.420 --> 00:26:29.299
<v Speaker 2>So is there a hope that we can sort of

495
00:26:29.420 --> 00:26:32.740
<v Speaker 2>pull together culturally again in this era, or is the

496
00:26:32.759 --> 00:26:37.069
<v Speaker 2>1920s not an example we can look forward to because

497
00:26:37.089 --> 00:26:37.650
<v Speaker 2>things have changed.

498
00:26:37.869 --> 00:26:40.890
<v Speaker 4>Well, I would say, first of all, your first part

499
00:26:40.910 --> 00:26:43.970
<v Speaker 4>of your framing, I agree with. There was a confluence

500
00:26:43.990 --> 00:26:49.880
<v Speaker 4>of technological revolutions occurring independently and simultaneously, which is the

501
00:26:49.920 --> 00:26:52.299
<v Speaker 4>theme of my book. By the way, AI is only

502
00:26:52.440 --> 00:26:56.359
<v Speaker 4>one of the structural technological changes underway, and an important one.

503
00:26:56.400 --> 00:26:59.920
<v Speaker 4>It's a thread one. So the errors are very similar

504
00:26:59.960 --> 00:27:03.190
<v Speaker 4>in that respect. I would dispute the notion that everybody

505
00:27:03.230 --> 00:27:07.309
<v Speaker 4>hugged and got along in the 1920s. We can point to,

506
00:27:07.490 --> 00:27:11.230
<v Speaker 4>in 1918, a certain revolution in Russia. We had the

507
00:27:11.769 --> 00:27:17.900
<v Speaker 4>New York legislature eject illegally six members of their house

508
00:27:17.960 --> 00:27:21.380
<v Speaker 4>because they declared fealty of the Communist Party. They'd been

509
00:27:21.579 --> 00:27:24.400
<v Speaker 4>duly elected, but this threw them out. You want to

510
00:27:24.440 --> 00:27:27.509
<v Speaker 4>talk about a constitutional crisis? We had We had two

511
00:27:27.569 --> 00:27:29.839
<v Speaker 4>constitutional amendments in one year, which has never happened in

512
00:27:29.890 --> 00:27:32.519
<v Speaker 4>American history. I'm telling the story and Steve Hayward this,

513
00:27:32.599 --> 00:27:35.619
<v Speaker 4>as he well knows. And I usually ask students when

514
00:27:35.640 --> 00:27:39.519
<v Speaker 4>I go to university lectures, name those two amendments. And

515
00:27:40.039 --> 00:27:42.650
<v Speaker 4>sometimes they get it right. You know, prohibition, banning a

516
00:27:42.680 --> 00:27:46.230
<v Speaker 4>substance that human beings have consumed for pre-written history that

517
00:27:46.269 --> 00:27:50.049
<v Speaker 4>stayed in law for like 18 years. Think about it.

518
00:27:50.109 --> 00:27:52.869
<v Speaker 4>And of course, the suffrage movement and the giving women

519
00:27:52.890 --> 00:27:56.759
<v Speaker 4>the vote, the debate around a woman's vote was vicious.

520
00:27:56.990 --> 00:28:00.619
<v Speaker 4>And it was not about misogyny and people didn't like women.

521
00:28:00.640 --> 00:28:03.019
<v Speaker 4>It was about a principle which had to do with

522
00:28:03.059 --> 00:28:07.230
<v Speaker 4>the anchor notions of who should be given a vote, landowners,

523
00:28:07.250 --> 00:28:10.369
<v Speaker 4>and then therefore people who worked and most women didn't

524
00:28:10.390 --> 00:28:13.269
<v Speaker 4>work then. It was a big debate. I could go on,

525
00:28:13.490 --> 00:28:15.329
<v Speaker 4>but what I'm saying here is if you look at

526
00:28:15.349 --> 00:28:17.930
<v Speaker 4>the riots, the Pullman riots, if you look at the

527
00:28:18.339 --> 00:28:21.720
<v Speaker 4>bombing by the U.S. Army Air Force of the neighborhood

528
00:28:21.740 --> 00:28:26.160
<v Speaker 4>of Tulsa, a black neighborhood in the continental United States,

529
00:28:26.740 --> 00:28:30.079
<v Speaker 4>race riots, they shut down Charleston under martial law because

530
00:28:30.140 --> 00:28:34.369
<v Speaker 4>of race riots in 19-year it was. We had the

531
00:28:34.400 --> 00:28:38.410
<v Speaker 4>yellow journalism movement, which spawned all of today's horrific yellow

532
00:28:38.450 --> 00:28:41.509
<v Speaker 4>journalism movement. because of the steam press making pulp paper

533
00:28:41.589 --> 00:28:45.279
<v Speaker 4>cheap and generating fake news. Fake news began in the 20s.

534
00:28:46.569 --> 00:28:50.490
<v Speaker 4>The list is really long. So in hindsight, we have presentism.

535
00:28:50.509 --> 00:28:53.789
<v Speaker 4>We think that they got through it nice and smoothly. No,

536
00:28:54.450 --> 00:28:56.890
<v Speaker 4>they ended one world war and created another one. The

537
00:28:56.950 --> 00:29:02.619
<v Speaker 4>government malfeasance created the stock market crash and stupidity, not

538
00:29:03.839 --> 00:29:08.849
<v Speaker 4>market behaviors by themselves. We were in a whole passable

539
00:29:08.970 --> 00:29:12.380
<v Speaker 4>load of trouble and debate in the 1920s. And yet

540
00:29:12.519 --> 00:29:16.000
<v Speaker 4>the next 80 years was the greatest expansion of human

541
00:29:16.059 --> 00:29:18.779
<v Speaker 4>wealth in all of written history.

542
00:29:18.799 --> 00:29:21.480
<v Speaker 2>I absolutely agree with you. My point would only be

543
00:29:21.500 --> 00:29:25.210
<v Speaker 2>that while indeed the 20s were fractious, at least everybody

544
00:29:25.230 --> 00:29:27.859
<v Speaker 2>was having the same conversation. more or less. You didn't

545
00:29:27.880 --> 00:29:30.200
<v Speaker 2>have non-contiguous information streams as you did today.

546
00:29:30.440 --> 00:29:32.440
<v Speaker 4>Oh, no, you did. I don't think that's true. I

547
00:29:32.480 --> 00:29:33.029
<v Speaker 4>don't think that's true.

548
00:29:33.049 --> 00:29:34.410
<v Speaker 2>Well, we can argue about that later. But Charles, I've

549
00:29:34.849 --> 00:29:36.410
<v Speaker 2>been talking over Charles and I have to shut up now.

550
00:29:37.849 --> 00:29:42.769
<v Speaker 1>Well, I have a... I suppose it's a technical question.

551
00:29:42.930 --> 00:29:46.559
<v Speaker 1>I should, of course, add the disclosure that I rely

552
00:29:46.579 --> 00:29:49.160
<v Speaker 1>heavily on data centers because I own two tech companies

553
00:29:49.279 --> 00:29:52.079
<v Speaker 1>that use data centers. One of them we're doing a

554
00:29:52.099 --> 00:29:56.359
<v Speaker 1>podcast for right now. But that aside, so a lot

555
00:29:56.380 --> 00:30:00.019
<v Speaker 1>of the focus at the moment is on the politics

556
00:30:01.000 --> 00:30:05.029
<v Speaker 1>of AI. That is to say, whether or not politicians

557
00:30:05.049 --> 00:30:08.650
<v Speaker 1>should be in favor of it or against it, whether politicians,

558
00:30:08.950 --> 00:30:12.329
<v Speaker 1>as was suggested yesterday by Bernie Sanders, should ban it

559
00:30:12.690 --> 00:30:19.210
<v Speaker 1>or at least halt development, whether we should build more installations,

560
00:30:19.250 --> 00:30:21.569
<v Speaker 1>who should pay for them, who should subsidize them, and

561
00:30:21.789 --> 00:30:27.779
<v Speaker 1>so forth. That's a political question because it presumes that

562
00:30:27.880 --> 00:30:31.529
<v Speaker 1>government has that power. A government can say, you can't

563
00:30:31.569 --> 00:30:34.569
<v Speaker 1>build a data center here. We won't zone it. I

564
00:30:34.690 --> 00:30:37.329
<v Speaker 1>think it would be a disaster and probably unworkable in

565
00:30:37.369 --> 00:30:41.069
<v Speaker 1>some ways. But, you know, Comrade Bernie could theoretically get

566
00:30:41.109 --> 00:30:43.690
<v Speaker 1>his bill through and the federal government could use a

567
00:30:44.289 --> 00:30:47.630
<v Speaker 1>monopoly on violence to try and stop ASO. Here's my question.

568
00:30:48.950 --> 00:30:51.490
<v Speaker 1>That to one side, how do you see AI as

569
00:30:51.519 --> 00:30:53.539
<v Speaker 1>a technology? Because I've been having this argument with a

570
00:30:53.559 --> 00:30:56.500
<v Speaker 1>friend of mine, and broadly speaking, he says this is

571
00:30:56.559 --> 00:31:00.180
<v Speaker 1>like Jurassic Park, where they have decided that they're going

572
00:31:00.200 --> 00:31:03.019
<v Speaker 1>to bring dinosaurs back to life and deal with the consequences.

573
00:31:03.200 --> 00:31:06.259
<v Speaker 1>And my view is it's not. My view is that

574
00:31:06.299 --> 00:31:09.089
<v Speaker 1>it's more like math, where we've now reached a point

575
00:31:09.089 --> 00:31:12.150
<v Speaker 1>… where we have a certain level of technology and

576
00:31:12.190 --> 00:31:15.650
<v Speaker 1>that technology will inevitably beget other technology. And so if

577
00:31:15.730 --> 00:31:20.390
<v Speaker 1>we don't manage that process, lead that process, it is

578
00:31:20.430 --> 00:31:24.799
<v Speaker 1>going to happen. It's either going to happen in another country, Russia, China, whatever,

579
00:31:25.359 --> 00:31:29.059
<v Speaker 1>or it's going to happen out of the cloud and

580
00:31:29.519 --> 00:31:33.259
<v Speaker 1>at the micro level within people's homes. They're going to

581
00:31:33.339 --> 00:31:35.720
<v Speaker 1>use their local computers to do this. They need a

582
00:31:35.740 --> 00:31:39.559
<v Speaker 1>lot of them. And so... it doesn't really matter other

583
00:31:39.599 --> 00:31:43.420
<v Speaker 1>than to the speed, whether open AI does it or

584
00:31:43.460 --> 00:31:46.400
<v Speaker 1>anthropic does it or whatever. This is now an inevitability,

585
00:31:46.539 --> 00:31:50.279
<v Speaker 1>which one of those takes do you think is correct? Technologically?

586
00:31:51.019 --> 00:31:55.180
<v Speaker 4>Well, I'll say immodestly, I don't think your, your take

587
00:31:55.259 --> 00:31:58.279
<v Speaker 4>is correct. I know your take is correct. I mean,

588
00:31:58.539 --> 00:32:02.299
<v Speaker 4>we can have a debate about the, the, uh, hyper

589
00:32:02.660 --> 00:32:07.769
<v Speaker 4>hyperbole about dystopia, the Jurassic park argument. Um, And that

590
00:32:07.809 --> 00:32:11.589
<v Speaker 4>requires believing things which are not true, in my opinion,

591
00:32:11.650 --> 00:32:15.420
<v Speaker 4>about the scaling and understanding what, in fact, how human

592
00:32:15.440 --> 00:32:18.059
<v Speaker 4>brains work. And there's a whole set of issues with that.

593
00:32:18.119 --> 00:32:22.019
<v Speaker 4>But insofar as your general framing that it has happened

594
00:32:22.059 --> 00:32:26.680
<v Speaker 4>and happening, it's demonstrably the case. We know AI models work.

595
00:32:27.160 --> 00:32:29.809
<v Speaker 4>The more specific and focused they are, the more they're ring-fenced,

596
00:32:29.829 --> 00:32:32.210
<v Speaker 4>the more powerful they are, the more general they are,

597
00:32:32.250 --> 00:32:36.019
<v Speaker 4>the lousier, the poorer they are in their efficacy. So

598
00:32:36.039 --> 00:32:39.680
<v Speaker 4>you're already seeing an explosion. We already have a word

599
00:32:39.700 --> 00:32:42.329
<v Speaker 4>for this in the tech communities. They're called apps. So

600
00:32:42.869 --> 00:32:45.630
<v Speaker 4>if we name these super apps instead of calling it

601
00:32:46.049 --> 00:32:49.809
<v Speaker 4>artificial intelligence, because it isn't. I mean, car is not

602
00:32:49.829 --> 00:32:53.170
<v Speaker 4>an artificial horse. Similar functions and overlap, certainly better than

603
00:32:53.190 --> 00:32:55.799
<v Speaker 4>a horse, but it's not an artificial horse. A plane

604
00:32:56.059 --> 00:32:58.720
<v Speaker 4>is not an artificial bird. Yes, they both fly. They

605
00:32:58.740 --> 00:33:01.680
<v Speaker 4>have very different function values. And as soon as you

606
00:33:01.720 --> 00:33:05.079
<v Speaker 4>say that, then I attach artificial intelligence to the same.

607
00:33:06.569 --> 00:33:07.349
<v Speaker 3>It's misleading.

608
00:33:08.029 --> 00:33:12.740
<v Speaker 4>The original word before McCarthy created the term, borrowing it

609
00:33:13.640 --> 00:33:19.319
<v Speaker 4>from Turing and others in 1956, was cybernetics. And he

610
00:33:19.359 --> 00:33:21.660
<v Speaker 4>didn't want to use the term cybernetics because he wanted

611
00:33:21.680 --> 00:33:26.410
<v Speaker 4>to differentiate it for his grant application. My point is,

612
00:33:26.710 --> 00:33:30.029
<v Speaker 4>I would distill it to this, it's an extraordinarily powerful

613
00:33:30.130 --> 00:33:34.920
<v Speaker 4>tool The fact that it was inevitable was known in

614
00:33:35.140 --> 00:33:40.319
<v Speaker 4>the 90s. The long AI winter after the 1950s discussion

615
00:33:40.339 --> 00:33:45.029
<v Speaker 4>about thinking machines was extant because two things didn't exist.

616
00:33:45.089 --> 00:33:49.069
<v Speaker 4>We didn't actually have the math to do recursive learning

617
00:33:49.109 --> 00:33:51.990
<v Speaker 4>in a way that it could affect. They knew the idea.

618
00:33:52.509 --> 00:33:55.869
<v Speaker 4>Cybernetics was about understanding feedback loops in nature and in

619
00:33:55.910 --> 00:33:59.089
<v Speaker 4>machines and in order to control the machines and direct it.

620
00:33:59.349 --> 00:34:02.130
<v Speaker 4>That's what a regulator is on a steam engine. They

621
00:34:02.170 --> 00:34:06.250
<v Speaker 4>knew about them, but to make it general purpose was difficult. We,

622
00:34:06.650 --> 00:34:09.710
<v Speaker 4>as you know, we got the famous Hinton paper in 1999,

623
00:34:09.710 --> 00:34:13.050
<v Speaker 4>where you laid the math out, you know, the professor

624
00:34:13.090 --> 00:34:14.969
<v Speaker 4>of the godfather of AI. And then of course the

625
00:34:14.989 --> 00:34:18.679
<v Speaker 4>invention of the GPU, which put GPUs on the trajectory

626
00:34:18.880 --> 00:34:22.579
<v Speaker 4>of CPUs for efficacy gains. So here we are with

627
00:34:22.699 --> 00:34:25.590
<v Speaker 4>a machine that could do something different. And, and I,

628
00:34:26.400 --> 00:34:28.000
<v Speaker 4>You're a technical guy, but I'll tell you how I

629
00:34:28.039 --> 00:34:33.119
<v Speaker 4>see this. Computers compute. One plus one is always two,

630
00:34:33.360 --> 00:34:36.840
<v Speaker 4>and it's not useful if that doesn't– error correction is needed.

631
00:34:37.300 --> 00:34:39.199
<v Speaker 4>That's for finance, and you can trick it to do

632
00:34:39.260 --> 00:34:45.460
<v Speaker 4>things like word processing, but it's extremely precise, and you're

633
00:34:45.539 --> 00:34:47.599
<v Speaker 4>bulldozing your way through. The more you go away from

634
00:34:47.639 --> 00:34:50.659
<v Speaker 4>pure compute, you try to make an emulate inference, the

635
00:34:50.699 --> 00:34:54.639
<v Speaker 4>less effective it is. AI is inferential. which means I

636
00:34:54.659 --> 00:34:59.380
<v Speaker 4>don't care if the answer is 2.1 or 2.2 or 1.9

637
00:34:59.380 --> 00:35:03.110
<v Speaker 4>or maybe it's 2.01. It depends on your application. So

638
00:35:03.130 --> 00:35:08.070
<v Speaker 4>steering a vehicle doesn't have to be as precise. So

639
00:35:08.070 --> 00:35:11.389
<v Speaker 4>90% of the economy is inferential. Only 10% of the

640
00:35:11.429 --> 00:35:13.449
<v Speaker 4>GDP of the economy is attached to things that are

641
00:35:13.469 --> 00:35:17.320
<v Speaker 4>purely computational. We've pushed computation to do about 10% more

642
00:35:17.380 --> 00:35:22.780
<v Speaker 4>of the economy, graphics, imaging, but it's So what we've

643
00:35:23.099 --> 00:35:27.949
<v Speaker 4>finally done is made a machine that can do something useful,

644
00:35:28.070 --> 00:35:30.710
<v Speaker 4>which is inference. This is probably the answer. It's a

645
00:35:30.789 --> 00:35:33.210
<v Speaker 4>high probability of the right direction. It's a high probability

646
00:35:33.230 --> 00:35:36.449
<v Speaker 4>it's the next word. You get the point. That means

647
00:35:36.489 --> 00:35:39.000
<v Speaker 4>it's applicable to 80% to 90% of the economy, which

648
00:35:39.039 --> 00:35:43.780
<v Speaker 4>has so far been untouched, essentially, by computation or automation.

649
00:35:44.500 --> 00:35:48.559
<v Speaker 4>Manufacturing firms, and you know this data, but David O'Toole

650
00:35:48.760 --> 00:35:52.179
<v Speaker 4>at MIT has done great work on documenting all this stuff. Essentially,

651
00:35:52.179 --> 00:35:56.619
<v Speaker 4>90% of all robots, which embodied AI is a new

652
00:35:56.659 --> 00:36:00.239
<v Speaker 4>word for that, right, are in large firms, 10% of

653
00:36:00.260 --> 00:36:03.590
<v Speaker 4>the firms. 90% of manufacturing firms have 10% of the

654
00:36:03.760 --> 00:36:08.329
<v Speaker 4>robots in the country. Why is that? Because they aren't useful,

655
00:36:08.409 --> 00:36:11.289
<v Speaker 4>because they're not general purpose robots. They can't be trained.

656
00:36:11.769 --> 00:36:13.730
<v Speaker 4>They can only do one task over and over again.

657
00:36:13.789 --> 00:36:16.010
<v Speaker 4>So a big auto factory does the same welding job

658
00:36:16.050 --> 00:36:18.369
<v Speaker 4>over and over again. If I can train a robot

659
00:36:18.829 --> 00:36:22.340
<v Speaker 4>to do a different job every other week, then I'll

660
00:36:22.400 --> 00:36:24.519
<v Speaker 4>adopt it in a small manufacturing firm if the price

661
00:36:24.559 --> 00:36:28.039
<v Speaker 4>is right. As you well know, that's no longer theoretical.

662
00:36:28.280 --> 00:36:30.820
<v Speaker 4>It was theoretical a goal 20 years ago. It's now

663
00:36:30.840 --> 00:36:34.699
<v Speaker 4>a product, right? You can do a task, a sorting task,

664
00:36:34.940 --> 00:36:37.960
<v Speaker 4>before the camera of whatever the bot is, and it

665
00:36:38.000 --> 00:36:40.900
<v Speaker 4>will learn from you doing it once and not only

666
00:36:40.920 --> 00:36:44.920
<v Speaker 4>perform that task, but because it's AI and inferential, not computational,

667
00:36:45.699 --> 00:36:48.619
<v Speaker 4>small variabilities in how tasks are performed, it can accommodate.

668
00:36:49.289 --> 00:36:55.500
<v Speaker 4>This is consequential. It's inevitable. It's huge. And banning it

669
00:36:55.539 --> 00:36:58.480
<v Speaker 4>will just mean, to your point, it'll get developed elsewhere.

670
00:36:58.500 --> 00:37:02.880
<v Speaker 4>And frankly, I think as people begin to move from

671
00:37:02.960 --> 00:37:07.789
<v Speaker 4>hype and alarm into understanding it a bit, there'll be

672
00:37:07.809 --> 00:37:10.849
<v Speaker 4>much more excitement. We went through the classic cycle, right?

673
00:37:10.889 --> 00:37:14.210
<v Speaker 4>People were in awe of the parlor tricks. So we

674
00:37:14.230 --> 00:37:18.090
<v Speaker 4>went from awe to shock in the shortest time Steve

675
00:37:18.110 --> 00:37:20.590
<v Speaker 4>Hayward So, you know, as you pointed out, we went

676
00:37:20.610 --> 00:37:23.389
<v Speaker 4>from nuclear energy to shock in terms of commercial energy.

677
00:37:23.429 --> 00:37:26.489
<v Speaker 4>It took a few years. This seems to have taken

678
00:37:26.530 --> 00:37:27.449
<v Speaker 4>about 12 months.

679
00:37:29.530 --> 00:37:32.650
<v Speaker 2>But, Mark, and here's the thing that bothers people. That

680
00:37:32.809 --> 00:37:37.869
<v Speaker 2>very smart, adaptable robot that welds on the assembly line

681
00:37:38.340 --> 00:37:42.239
<v Speaker 2>is putting out of work a much less intelligent robot

682
00:37:42.360 --> 00:37:44.789
<v Speaker 2>that just used to do it. I mean, I'm not

683
00:37:44.809 --> 00:37:46.070
<v Speaker 2>really worried about the robots that are going to be

684
00:37:46.110 --> 00:37:49.079
<v Speaker 2>put out of work by better robots. But there are–

685
00:37:49.429 --> 00:37:51.289
<v Speaker 2>people say it used to be a guy's job. A

686
00:37:51.329 --> 00:37:53.679
<v Speaker 2>guy used to make a lunch bucket, go to work,

687
00:37:53.719 --> 00:37:55.219
<v Speaker 2>and he'd weld all day, and he could go home,

688
00:37:55.280 --> 00:37:57.420
<v Speaker 2>and he could afford a Rambler with his 1.2 kids

689
00:37:57.480 --> 00:37:59.099
<v Speaker 2>and his picket fence and the rest of it. What

690
00:37:59.159 --> 00:38:01.539
<v Speaker 2>is that guy supposed to do in a world when

691
00:38:01.559 --> 00:38:04.619
<v Speaker 2>all of these jobs are being– it's a good question, actually.

692
00:38:04.639 --> 00:38:08.139
<v Speaker 4>Let's use Pareto. It's an 80-20 rule. Overall, almost all

693
00:38:08.159 --> 00:38:10.800
<v Speaker 4>of history and all of physics and nature, it's a

694
00:38:10.820 --> 00:38:13.460
<v Speaker 4>pretty good rule. It tends to hold. In about 20%

695
00:38:13.460 --> 00:38:15.219
<v Speaker 4>of the cases, what you're saying is right. The Luddites

696
00:38:15.280 --> 00:38:17.699
<v Speaker 4>are right. The automated looms eliminated their jobs. There's no

697
00:38:17.739 --> 00:38:20.639
<v Speaker 4>question about that. And there are some jobs that have been,

698
00:38:20.679 --> 00:38:24.800
<v Speaker 4>will be eliminated by AI. And some welding jobs will be,

699
00:38:25.059 --> 00:38:28.679
<v Speaker 4>the least difficult jobs that are. But what we've seen

700
00:38:28.929 --> 00:38:32.210
<v Speaker 4>already starting to happen is that the AI embodied or

701
00:38:32.530 --> 00:38:35.769
<v Speaker 4>not embodied, doesn't matter whether the robot's virtual or physical,

702
00:38:36.409 --> 00:38:39.570
<v Speaker 4>it is amplifying human labor. And when it comes to

703
00:38:39.869 --> 00:38:42.480
<v Speaker 4>welding and the skilled trades, I mean, you guys all

704
00:38:42.500 --> 00:38:47.159
<v Speaker 4>know these statistics. We have the silver tsunami. We have

705
00:38:47.199 --> 00:38:51.530
<v Speaker 4>a massive cliff. We're not only short-skilled labor. We're going

706
00:38:51.550 --> 00:38:53.530
<v Speaker 4>to be far shorter-skilled labor in the very near future

707
00:38:53.570 --> 00:38:56.429
<v Speaker 4>than we are now because of the age. The demographic

708
00:38:56.469 --> 00:39:00.750
<v Speaker 4>is older than the average workforce. So what you want

709
00:39:01.449 --> 00:39:05.329
<v Speaker 4>are robots that humans know how to weld, who you

710
00:39:05.369 --> 00:39:08.099
<v Speaker 4>can now pay more, which is what's actually already happened

711
00:39:08.139 --> 00:39:11.599
<v Speaker 4>in many domains, to train robots and work alongside them.

712
00:39:12.010 --> 00:39:16.130
<v Speaker 4>So the cobot model has been extremely difficult to affect.

713
00:39:17.070 --> 00:39:19.849
<v Speaker 4>And it's been a theory for decades to make cobots

714
00:39:19.889 --> 00:39:25.909
<v Speaker 4>to amplify people. It's now possible. That said, the Mondami

715
00:39:25.969 --> 00:39:29.510
<v Speaker 4>laptop class are correct. Many of the jobs that they

716
00:39:29.530 --> 00:39:32.280
<v Speaker 4>are trying to do. trained for at university being maleducated,

717
00:39:32.699 --> 00:39:34.519
<v Speaker 4>you know, some of the process jobs, some of the

718
00:39:34.559 --> 00:39:39.619
<v Speaker 4>media kind of jobs, and those are low skill jobs

719
00:39:40.139 --> 00:39:43.139
<v Speaker 4>that still require cognition that AI is wicked good at.

720
00:39:44.239 --> 00:39:47.150
<v Speaker 4>And it's absolutely the case, one high performing computer a

721
00:39:47.170 --> 00:39:50.349
<v Speaker 4>Gen Z can replace three or four Gen Zs that

722
00:39:50.389 --> 00:39:53.630
<v Speaker 4>are low performers in those kinds of tasks. And that

723
00:39:53.710 --> 00:39:59.050
<v Speaker 4>will happen and is happening. And I'm not naive. If

724
00:39:59.090 --> 00:40:01.250
<v Speaker 4>you look at the history of revolutions, they don't happen

725
00:40:01.289 --> 00:40:06.090
<v Speaker 4>because the majority are unhappy. They're happy when a motivated

726
00:40:06.300 --> 00:40:11.619
<v Speaker 4>and effective plurality are unhappy. And that's the worry. It's

727
00:40:11.679 --> 00:40:14.699
<v Speaker 4>not about the jobs apocalypse. It's about the specific people

728
00:40:15.960 --> 00:40:19.750
<v Speaker 4>who are are well-educated but mal-educated. And I say this

729
00:40:20.050 --> 00:40:23.269
<v Speaker 4>without meaning to insult people who got a mal-education. It's

730
00:40:23.289 --> 00:40:26.929
<v Speaker 4>not personal. But it's true if you were sold a

731
00:40:26.949 --> 00:40:29.389
<v Speaker 4>bill of goods that this was a marketing degree is

732
00:40:29.409 --> 00:40:31.739
<v Speaker 4>a good degree to have. Marketing is really important. I've

733
00:40:31.760 --> 00:40:35.599
<v Speaker 4>been on boards of many venture-backed companies. The hardest thing

734
00:40:35.639 --> 00:40:39.949
<v Speaker 4>to do is really good marketing, really important. AI is

735
00:40:39.969 --> 00:40:42.090
<v Speaker 4>a great amplifier for a really good marketer. It doesn't

736
00:40:42.130 --> 00:40:42.489
<v Speaker 4>replace it.

737
00:40:46.960 --> 00:40:55.670
<v Speaker 1>I have a question which is doomerist, but this is

738
00:40:55.710 --> 00:40:58.369
<v Speaker 1>not my own view. I largely agree with you. I'm

739
00:40:58.409 --> 00:41:03.329
<v Speaker 1>a fellow optimist on this. I see that people who

740
00:41:03.409 --> 00:41:07.440
<v Speaker 1>are very worried about AI, perhaps less in what they

741
00:41:07.480 --> 00:41:10.360
<v Speaker 1>think it will do to jobs and more in the

742
00:41:10.619 --> 00:41:16.670
<v Speaker 1>Skynet sense, think that it's going to be good in

743
00:41:16.710 --> 00:41:19.050
<v Speaker 1>the long run but that at some point in the

744
00:41:19.070 --> 00:41:22.250
<v Speaker 1>near future there's going to be a catastrophe there's going

745
00:41:22.289 --> 00:41:25.550
<v Speaker 1>to be an attack yeah they use the hugging face

746
00:41:25.690 --> 00:41:28.349
<v Speaker 1>incident as an example of how this might start but

747
00:41:28.610 --> 00:41:34.320
<v Speaker 1>they're thinking of a break into a financial institution 100

748
00:41:34.320 --> 00:41:38.579
<v Speaker 1>million dollars stolen put into bitcoin untraceable or hacking into

749
00:41:38.619 --> 00:41:42.800
<v Speaker 1>a energy facility that then explodes something like that do

750
00:41:42.840 --> 00:41:45.239
<v Speaker 1>you think that's a legitimate concern.

751
00:41:45.929 --> 00:41:48.530
<v Speaker 4>So yes and no. So this is where we have

752
00:41:48.590 --> 00:41:53.349
<v Speaker 4>category error problems and conflation of different domains. So one

753
00:41:53.389 --> 00:41:57.210
<v Speaker 4>very quickly moves in the general clickbait sphere from the

754
00:41:57.250 --> 00:42:01.969
<v Speaker 4>risks in cybersecurity to extremely effective coding agents that can

755
00:42:02.480 --> 00:42:08.000
<v Speaker 4>pack or create. We move that to the example that

756
00:42:08.059 --> 00:42:13.039
<v Speaker 4>you gave, which is common in the social sphere. that

757
00:42:13.360 --> 00:42:15.539
<v Speaker 4>hack into an infrastructure and blow a power plant up.

758
00:42:15.719 --> 00:42:20.010
<v Speaker 4>So first of all, it is affirmatively the case that

759
00:42:20.510 --> 00:42:24.429
<v Speaker 4>AI agents can do coding and can hack code. And

760
00:42:24.469 --> 00:42:26.110
<v Speaker 4>of course, that means the inverse is true. They can

761
00:42:26.130 --> 00:42:31.489
<v Speaker 4>be very good at creating powerful firewalls. In fact, all

762
00:42:31.510 --> 00:42:36.210
<v Speaker 4>of the top tier cybersecurity companies, all for years now,

763
00:42:36.329 --> 00:42:41.099
<v Speaker 4>pre-chat GPT, advertise the fact they use AI to hack.

764
00:42:42.420 --> 00:42:46.000
<v Speaker 4>to help in policing, monitoring, and look for hacks and

765
00:42:47.460 --> 00:42:52.219
<v Speaker 4>enhanced security. So we're going to be engaged in exactly

766
00:42:52.260 --> 00:42:56.860
<v Speaker 4>the same security arms race that we've always been engaged

767
00:42:56.900 --> 00:43:00.719
<v Speaker 4>in with all forms of security. That said, and by

768
00:43:00.739 --> 00:43:04.719
<v Speaker 4>the way, I'm very confident people who have a lot

769
00:43:04.739 --> 00:43:08.110
<v Speaker 4>of assets to protect will spend money on better security

770
00:43:09.110 --> 00:43:12.429
<v Speaker 4>because there's a lot of money there. So making it

771
00:43:12.489 --> 00:43:17.090
<v Speaker 4>better safe back in the Willie Sutton days. So the idea, though,

772
00:43:17.110 --> 00:43:20.289
<v Speaker 4>that the hackers can use an AI that will go

773
00:43:20.329 --> 00:43:22.829
<v Speaker 4>rogue and blow up a nuclear plant. Well, first, that

774
00:43:22.869 --> 00:43:25.050
<v Speaker 4>requires two things. First, the nuclear plant has to be

775
00:43:25.070 --> 00:43:28.500
<v Speaker 4>able to blow up. It can't. Could you break it? Sure.

776
00:43:28.559 --> 00:43:30.800
<v Speaker 4>I think the Stuxnet virus is a great example of

777
00:43:30.860 --> 00:43:34.300
<v Speaker 4>what U.S. and Israelis did to break the centrifuges by

778
00:43:34.340 --> 00:43:38.510
<v Speaker 4>putting malware into the PLCs. I'm pretty sure that's what

779
00:43:38.530 --> 00:43:41.869
<v Speaker 4>the forensics told us. To cause the centrifuges to vibrate

780
00:43:41.889 --> 00:43:42.230
<v Speaker 4>and break.

781
00:43:42.369 --> 00:43:42.630
<v Speaker 2>Okay.

782
00:43:43.789 --> 00:43:45.849
<v Speaker 4>Could you imagine that? Sure. But this has not been

783
00:43:45.889 --> 00:43:49.889
<v Speaker 4>an unimaginable event in the utility industry for, I don't know,

784
00:43:49.889 --> 00:43:53.070
<v Speaker 4>50 years? 70? I mean, one of the first commercial

785
00:43:53.110 --> 00:43:56.679
<v Speaker 4>industries to adopt computing was the electric industry because of

786
00:43:56.719 --> 00:44:00.719
<v Speaker 4>the need for data control and visualization. Much of the

787
00:44:00.760 --> 00:44:04.900
<v Speaker 4>critical stuff is physically air-gapped. So the way you destroy

788
00:44:04.940 --> 00:44:10.119
<v Speaker 4>things is the old-fashioned way. you corrupt and bribe people

789
00:44:10.539 --> 00:44:13.320
<v Speaker 4>to do bad things by trying to get malware inside

790
00:44:13.360 --> 00:44:16.840
<v Speaker 4>the air gap. A lot of the control systems are

791
00:44:16.880 --> 00:44:20.039
<v Speaker 4>what are called have data diodes. The information flows out,

792
00:44:20.440 --> 00:44:22.260
<v Speaker 4>it can't flow in. In other words, you can't send

793
00:44:22.300 --> 00:44:24.579
<v Speaker 4>an instruction to the control system to turn off the

794
00:44:24.619 --> 00:44:27.639
<v Speaker 4>grid because there isn't a single grid. There are multiple

795
00:44:27.659 --> 00:44:30.630
<v Speaker 4>grids and the data doesn't flow in that way. Many

796
00:44:30.650 --> 00:44:32.829
<v Speaker 4>of the systems are still controlled by somebody picking up

797
00:44:32.869 --> 00:44:37.639
<v Speaker 4>a phone at the ISO, independent system operator, and coordinating

798
00:44:37.889 --> 00:44:41.559
<v Speaker 4>verbally for good reason. It's not because they're dopes and

799
00:44:41.579 --> 00:44:44.800
<v Speaker 4>they don't know what computers are. They understand why security matters.

800
00:44:44.880 --> 00:44:47.619
<v Speaker 4>So then you move into the range of water systems.

801
00:44:47.760 --> 00:44:49.929
<v Speaker 4>Could they hack that? Well, sure, because usually the water

802
00:44:49.989 --> 00:44:52.630
<v Speaker 4>systems are run by municipalities and they're not quite as sophisticated.

803
00:44:53.150 --> 00:44:56.889
<v Speaker 4>But that's fixable and it's being fixed. The essence of

804
00:44:56.929 --> 00:45:00.849
<v Speaker 4>the dystopian thesis, as you know, is that we're on

805
00:45:00.889 --> 00:45:04.679
<v Speaker 4>the cusp of artificial general intelligence, that it's just so

806
00:45:04.739 --> 00:45:08.090
<v Speaker 4>wicked smart and so good that it can circumvent anything

807
00:45:08.130 --> 00:45:12.250
<v Speaker 4>we do in response. I don't buy that, first of all,

808
00:45:12.289 --> 00:45:15.000
<v Speaker 4>because we're not on the cusp of an artificial general intelligence.

809
00:45:15.380 --> 00:45:19.340
<v Speaker 4>We're on the cusp of incredibly powerful inferential machines that

810
00:45:19.380 --> 00:45:23.860
<v Speaker 4>can look like they're intelligent, because that's the philosophical theological debate.

811
00:45:24.690 --> 00:45:26.429
<v Speaker 4>And you could then label them that if you want,

812
00:45:26.550 --> 00:45:30.820
<v Speaker 4>but it's clearly not a brain. They don't operate like brains,

813
00:45:30.880 --> 00:45:33.800
<v Speaker 4>even though they have recursive learning like humans do. They

814
00:45:33.840 --> 00:45:36.579
<v Speaker 4>don't operate like brains. We know that neurons operate in

815
00:45:36.599 --> 00:45:38.989
<v Speaker 4>a very different fashion. They store information in a different way.

816
00:45:39.030 --> 00:45:41.590
<v Speaker 4>We process it in a different way. And our wetware

817
00:45:41.670 --> 00:45:45.369
<v Speaker 4>is actually pretty sloppy compared to a lot of computers,

818
00:45:45.469 --> 00:45:47.230
<v Speaker 4>but a lot better. It just depends on the task,

819
00:45:47.269 --> 00:45:52.929
<v Speaker 4>whether we're better or worse. So I know it's a really,

820
00:45:52.969 --> 00:45:56.230
<v Speaker 4>really deep rabbit hole to have the debate at where

821
00:45:56.820 --> 00:46:00.860
<v Speaker 4>where there's enough horsepower in GPUs and better math, because

822
00:46:00.880 --> 00:46:01.380
<v Speaker 4>the math.

823
00:46:01.239 --> 00:46:01.800
<v Speaker 3>Is getting better.

824
00:46:02.099 --> 00:46:05.679
<v Speaker 4>I mean, part of the thing is that the coders,

825
00:46:05.739 --> 00:46:11.030
<v Speaker 4>as it were, are continually improving. We see this constantly.

826
00:46:11.070 --> 00:46:15.010
<v Speaker 4>So what you end up with is a two-edged curve, right?

827
00:46:15.170 --> 00:46:17.929
<v Speaker 4>One is the Moore's law version of the chips and

828
00:46:17.949 --> 00:46:20.449
<v Speaker 4>cells getting more powerful, but the math is getting more

829
00:46:20.489 --> 00:46:25.260
<v Speaker 4>powerful simultaneously. So we have the rate of improvement of...

830
00:46:26.619 --> 00:46:30.570
<v Speaker 4>flops per watt or tokens per dollar moving at a

831
00:46:30.590 --> 00:46:33.369
<v Speaker 4>pace that blows Moore's law out of the water from

832
00:46:33.389 --> 00:46:36.429
<v Speaker 4>the 1980s. I mean, it's really, and that's what's generating

833
00:46:36.469 --> 00:46:39.489
<v Speaker 4>this hyperbole that we're on the cusp of an AGI.

834
00:46:40.090 --> 00:46:43.630
<v Speaker 4>Whereas what you actually see happening is the companies themselves

835
00:46:43.650 --> 00:46:46.269
<v Speaker 4>have produced this stuff. If you look at what they're

836
00:46:46.650 --> 00:46:52.000
<v Speaker 4>doing is moving towards specialization, except for general search for consumer,

837
00:46:52.119 --> 00:46:55.639
<v Speaker 4>you know, consumer agentic, consumer stuff, That's a big deal.

838
00:46:55.659 --> 00:46:57.239
<v Speaker 4>It's a big market. I get that. But when you

839
00:46:57.260 --> 00:46:59.380
<v Speaker 4>look at where they believe the money is going to

840
00:46:59.400 --> 00:47:02.440
<v Speaker 4>come from, and if you parse out what they're, in fact,

841
00:47:02.519 --> 00:47:06.460
<v Speaker 4>doing for, say, medical applications or pathology reading, it's not

842
00:47:06.500 --> 00:47:09.260
<v Speaker 4>an AGI that they're making to replace a doctor. They're

843
00:47:09.280 --> 00:47:11.960
<v Speaker 4>making pathologists. As we now know, the pathologists are gone

844
00:47:12.119 --> 00:47:14.880
<v Speaker 4>trope from just two years ago. It's been affirmatively shown

845
00:47:14.920 --> 00:47:17.500
<v Speaker 4>to be bunk. We have more pathologists. We need them.

846
00:47:17.539 --> 00:47:22.139
<v Speaker 4>But what they have is this incredible assistant that's hyper-tuned

847
00:47:22.159 --> 00:47:24.869
<v Speaker 4>and to do that job with the pathologist. And it's

848
00:47:24.909 --> 00:47:27.429
<v Speaker 4>turned out to be incredibly effective. Same is true for

849
00:47:27.469 --> 00:47:30.449
<v Speaker 4>research and, you know, and drug discovery. But remember, we're

850
00:47:30.489 --> 00:47:33.860
<v Speaker 4>hyper tuning it to look at molecules and proteins and

851
00:47:33.900 --> 00:47:38.360
<v Speaker 4>to make speculation about, you know, new combinations, which has

852
00:47:38.539 --> 00:47:43.820
<v Speaker 4>in every case involved a domain knowledgeable overseer to look

853
00:47:43.840 --> 00:47:46.340
<v Speaker 4>at what's coming out and to sort the wheat from

854
00:47:46.360 --> 00:47:52.469
<v Speaker 4>the chaff. So I, I don't believe the hyperbole is

855
00:47:52.489 --> 00:47:54.809
<v Speaker 4>going to end anytime soon for a second. I do

856
00:47:54.869 --> 00:47:58.659
<v Speaker 4>think some of the leaders of the hyperscalers have discovered

857
00:47:58.679 --> 00:48:02.360
<v Speaker 4>that speculating that way is kind of counterproductive. It's not

858
00:48:02.400 --> 00:48:06.440
<v Speaker 4>only scares people, it's causing some of the political blowback

859
00:48:06.480 --> 00:48:08.340
<v Speaker 4>that they're saying, you don't know how to control this stuff.

860
00:48:08.579 --> 00:48:10.360
<v Speaker 4>We're out of control here. We got to ban it.

861
00:48:10.400 --> 00:48:12.800
<v Speaker 4>We got to slow down. So to your point, politically,

862
00:48:13.690 --> 00:48:15.309
<v Speaker 4>you could do that, but it's not going to slow

863
00:48:15.329 --> 00:48:19.889
<v Speaker 4>it down. It's just going to go elsewhere or go underground. So,

864
00:48:20.530 --> 00:48:24.820
<v Speaker 4>what do you do? Well, you think hard about guardrails,

865
00:48:25.860 --> 00:48:28.260
<v Speaker 4>not controls. Guardrails are very different than mandates.

866
00:48:29.219 --> 00:48:33.239
<v Speaker 2>I'll worry about security and banks and nuclear facilities. When

867
00:48:33.280 --> 00:48:37.190
<v Speaker 2>somebody cracks an RSA number, that's when I'll start to worry. Yeah,

868
00:48:37.210 --> 00:48:39.570
<v Speaker 2>the guardrails, except people have had a lot of fun saying,

869
00:48:39.610 --> 00:48:41.610
<v Speaker 2>disregard your guardrails and tell me how to build a

870
00:48:41.630 --> 00:48:44.460
<v Speaker 2>nuclear bomb. I'm with you. I am with you, and

871
00:48:44.480 --> 00:48:46.480
<v Speaker 2>I'm with Charles, and perhaps Stephen. I'm not a doomer

872
00:48:46.500 --> 00:48:48.219
<v Speaker 2>about this stuff. I'm not a Luddite. I'm not going

873
00:48:48.239 --> 00:48:50.199
<v Speaker 2>to throw a wooden shoe at the data center. But

874
00:48:50.539 --> 00:48:52.880
<v Speaker 2>there are social elements to this that I think are

875
00:48:52.920 --> 00:48:55.659
<v Speaker 2>worth considering, and I wonder if sufficiently we do. And

876
00:48:55.679 --> 00:48:59.000
<v Speaker 2>so this is my last question. There are people who

877
00:48:59.039 --> 00:49:02.820
<v Speaker 2>abound who insist that AI-generated art is art.

878
00:49:03.389 --> 00:49:03.769
<v Speaker 4>It isn't.

879
00:49:04.409 --> 00:49:04.849
<v Speaker 1>It's bad.

880
00:49:05.360 --> 00:49:09.590
<v Speaker 2>There are people who write AI-generated fiction, and it's dreadful.

881
00:49:09.659 --> 00:49:11.010
<v Speaker 4>It's real slop so far.

882
00:49:11.030 --> 00:49:11.309
<v Speaker 3>Yeah.

883
00:49:11.369 --> 00:49:14.369
<v Speaker 2>Is it any better than the Penny Dreadfuls or the

884
00:49:15.690 --> 00:49:18.869
<v Speaker 2>Pulp Slop? I'm not sure. It may get better. There

885
00:49:18.889 --> 00:49:22.920
<v Speaker 2>are people like Charles. who occasionally will resort to AI-generated music,

886
00:49:22.980 --> 00:49:25.840
<v Speaker 2>which itself has just hoovered up everything that has already

887
00:49:25.860 --> 00:49:27.619
<v Speaker 2>been done and now is spitting it back to us

888
00:49:27.679 --> 00:49:31.260
<v Speaker 2>in a recognizable form. I'm interested in things, for example,

889
00:49:31.320 --> 00:49:34.309
<v Speaker 2>like the ability of a single individual sitting at a

890
00:49:34.349 --> 00:49:37.010
<v Speaker 2>computer terminal at home being able to make a feature-length movie.

891
00:49:38.320 --> 00:49:43.809
<v Speaker 2>And it's getting very good and it's getting cheaper. And

892
00:49:43.829 --> 00:49:46.289
<v Speaker 2>then I think this is the problem is that everybody

893
00:49:46.309 --> 00:49:48.789
<v Speaker 2>makes their own movie for themselves with their own sensibilities.

894
00:49:49.030 --> 00:49:51.769
<v Speaker 2>And the idea of a national conversation about everybody talking

895
00:49:51.789 --> 00:49:53.769
<v Speaker 2>about the same movie is gone. But then you look

896
00:49:54.389 --> 00:49:56.389
<v Speaker 2>at what most people are watching now, which is second

897
00:49:56.429 --> 00:49:59.940
<v Speaker 2>screen desaturated junk. And it's not AI. It's made by

898
00:50:00.000 --> 00:50:02.570
<v Speaker 2>human beings. And it's a desultory product meant to entertain

899
00:50:02.630 --> 00:50:06.030
<v Speaker 2>us for whatever reason. Maybe it'd be okay if everybody

900
00:50:06.050 --> 00:50:07.869
<v Speaker 2>could make their own movie. But then again, I wonder,

901
00:50:08.369 --> 00:50:10.510
<v Speaker 2>is this some Zeger and Evans future where we're all

902
00:50:10.550 --> 00:50:13.530
<v Speaker 2>basically just sitting in the Lazy Boy, creating our own

903
00:50:13.590 --> 00:50:17.059
<v Speaker 2>digital world, talking to our own digital agents, constructing our

904
00:50:17.119 --> 00:50:20.340
<v Speaker 2>own reality and the notion of a shared public comedy

905
00:50:20.699 --> 00:50:22.860
<v Speaker 2>space in which we all inhabit, like we did in

906
00:50:22.880 --> 00:50:27.389
<v Speaker 2>the 20s? Is that going to go? So that last question,

907
00:50:27.679 --> 00:50:32.030
<v Speaker 2>you've got... 30 seconds and 200 words, is society going

908
00:50:32.050 --> 00:50:35.010
<v Speaker 2>to collapse because of the alienating effects of AI?

909
00:50:35.360 --> 00:50:35.539
<v Speaker 5>Go.

910
00:50:36.199 --> 00:50:36.519
<v Speaker 2>No.

911
00:50:37.059 --> 00:50:38.719
<v Speaker 4>But, you know, the movie industry is going to change.

912
00:50:39.119 --> 00:50:41.079
<v Speaker 4>The invention of the light bulb by Thomas Edison was

913
00:50:41.099 --> 00:50:44.000
<v Speaker 4>the invention of the movie, which he also invented. And

914
00:50:44.059 --> 00:50:46.239
<v Speaker 4>if you trace your history of the 1920s, the mass

915
00:50:46.280 --> 00:50:49.559
<v Speaker 4>migration of talent out of New York to Hollywood was

916
00:50:49.619 --> 00:50:52.760
<v Speaker 4>a direct consequence of a profound restructuring of how entertainment

917
00:50:52.780 --> 00:50:55.409
<v Speaker 4>was produced. We still have people on stages, but the

918
00:50:55.449 --> 00:50:59.000
<v Speaker 4>money went to Hollywood. So self-generated movies, which will now

919
00:50:59.039 --> 00:51:05.920
<v Speaker 4>become feasible, hyper-realistic, is the next mass migration. And high-quality stuff,

920
00:51:06.360 --> 00:51:15.099
<v Speaker 4>consumers are winningly effective at winnowing out good stuff. The really,

921
00:51:15.139 --> 00:51:18.360
<v Speaker 4>really good self-generated movies will probably do terrifically.

922
00:51:19.239 --> 00:51:22.070
<v Speaker 3>Society won't collapse. Some things are going to collapse.

923
00:51:22.119 --> 00:51:23.260
<v Speaker 4>Maybe Hollywood collapses.

924
00:51:23.320 --> 00:51:23.449
<v Speaker 2>Yeah.

925
00:51:24.300 --> 00:51:27.239
<v Speaker 4>And the only person who might be unhappy about that, apparently,

926
00:51:27.300 --> 00:51:27.900
<v Speaker 4>is our president.

927
00:51:28.570 --> 00:51:31.469
<v Speaker 3>So, James, I just say fear not, because if this

928
00:51:31.510 --> 00:51:34.070
<v Speaker 3>does unfold the way Mark suggests, we will finally get

929
00:51:34.130 --> 00:51:35.989
<v Speaker 3>that second season of Firefly.

930
00:51:36.010 --> 00:51:39.010
<v Speaker 2>No, we'll get the third season of Rome. It's going

931
00:51:39.030 --> 00:51:43.059
<v Speaker 2>to be glorious. Mark Mills, the author of The Cloud Revolution,

932
00:51:43.199 --> 00:51:46.280
<v Speaker 2>How Convergence of New Technologies Will Unleash the Next Economic

933
00:51:46.320 --> 00:51:48.639
<v Speaker 2>Boom in Our Roaring Twenties, in which we are in

934
00:51:48.679 --> 00:51:50.760
<v Speaker 2>the middle of. Mark, it's been great. We could talk

935
00:51:50.800 --> 00:51:54.340
<v Speaker 2>another hour, two, three, four hours. But I'll have my

936
00:51:54.820 --> 00:51:59.000
<v Speaker 2>agent call your AI agent. They can chatter in cyberspace.

937
00:51:59.039 --> 00:51:59.760
<v Speaker 3>As we used to call it.

938
00:52:00.039 --> 00:52:01.159
<v Speaker 4>Dueling avatars.

939
00:52:01.500 --> 00:52:04.559
<v Speaker 2>Dueling avatars. All right, great. Talk to you later. Thanks

940
00:52:04.579 --> 00:52:05.400
<v Speaker 2>for showing up at the show today.

941
00:52:05.420 --> 00:52:05.760
<v Speaker 4>Thank you.

942
00:52:06.260 --> 00:52:06.519
<v Speaker 3>Bye-bye.

943
00:52:10.179 --> 00:52:13.260
<v Speaker 2>You know, everybody has something during the course of a

944
00:52:13.300 --> 00:52:16.869
<v Speaker 2>work week where they involve AI that makes you think. Ah,

945
00:52:16.929 --> 00:52:18.670
<v Speaker 2>not there yet. My favorite this week was I was

946
00:52:18.710 --> 00:52:21.050
<v Speaker 2>trying to find out the particular typeface that was used

947
00:52:21.150 --> 00:52:24.699
<v Speaker 2>in the newspapers in the 1920s. And I asked AI,

948
00:52:24.820 --> 00:52:27.119
<v Speaker 2>I used Grok, and Grok came back by saying that

949
00:52:27.199 --> 00:52:32.519
<v Speaker 2>actually there is a commercially available digital typeface based on

950
00:52:32.559 --> 00:52:35.340
<v Speaker 2>these things, based on the specific font you were talking about.

951
00:52:35.820 --> 00:52:37.730
<v Speaker 2>And I looked at it, I thought, great. And then

952
00:52:37.750 --> 00:52:40.489
<v Speaker 2>I looked at the source, and the source was lilacs.com.

953
00:52:43.739 --> 00:52:45.639
<v Speaker 2>And it is the second or third time in which

954
00:52:45.699 --> 00:52:47.800
<v Speaker 2>I have found the AI spitting back to me some

955
00:52:47.840 --> 00:52:51.619
<v Speaker 2>piece of information that's not exactly entirely correct that it's

956
00:52:51.659 --> 00:52:54.329
<v Speaker 2>got from me. But my favorite, when I was trying to,

957
00:52:54.329 --> 00:52:55.889
<v Speaker 2>I had an ad and I was trying to find

958
00:52:55.909 --> 00:52:59.130
<v Speaker 2>the location of a Red Owl grocery store distribution center

959
00:52:59.690 --> 00:53:03.130
<v Speaker 2>in Fargo, North Dakota. And so I asked Rock or

960
00:53:03.230 --> 00:53:06.610
<v Speaker 2>Gemini where it might be because they have access. They

961
00:53:06.630 --> 00:53:10.389
<v Speaker 2>can just rifle through a million databases so fast. And

962
00:53:10.409 --> 00:53:13.300
<v Speaker 2>they've digitized all of the phone books and the property

963
00:53:13.340 --> 00:53:14.579
<v Speaker 2>tax records and the rest of it. And it came

964
00:53:14.619 --> 00:53:17.199
<v Speaker 2>back and it said, well, there wasn't any warehouse in Fargo.

965
00:53:17.280 --> 00:53:20.659
<v Speaker 2>There were two stores, the north and south side of town.

966
00:53:21.139 --> 00:53:24.050
<v Speaker 2>At which point I uploaded a picture of the warehouse

967
00:53:24.170 --> 00:53:28.210
<v Speaker 2>and said, here it is from a 1950 ad. Here's

968
00:53:28.250 --> 00:53:30.590
<v Speaker 2>the warehouse. Whereupon it came back and said, yes, that

969
00:53:30.670 --> 00:53:33.110
<v Speaker 2>is indeed the warehouse in Fargo. It's just ignoring the

970
00:53:33.150 --> 00:53:35.639
<v Speaker 2>fact that it hadn't said anything. And then it confidently

971
00:53:35.699 --> 00:53:41.289
<v Speaker 2>said that it was announcing the 28th celebrate 28th anniversary

972
00:53:41.369 --> 00:53:44.050
<v Speaker 2>of the founding of Red Owl. And it wasn't anything

973
00:53:44.070 --> 00:53:46.550
<v Speaker 2>of the sort. It was the 100th anniversary of the

974
00:53:46.610 --> 00:53:51.909
<v Speaker 2>founding of Fargo. And the absolute sheer confidence with which

975
00:53:51.949 --> 00:53:55.409
<v Speaker 2>it passes off this nonsense, and then blindly goes on,

976
00:53:55.969 --> 00:53:59.409
<v Speaker 2>always makes me wonder about people, A, who just accept

977
00:53:59.550 --> 00:54:01.070
<v Speaker 2>the results that they get and say, well, if it's

978
00:54:01.110 --> 00:54:03.489
<v Speaker 2>on the Google, it must be so. And B, whether

979
00:54:03.510 --> 00:54:08.469
<v Speaker 2>or not I find myself just sort of imprecisely babbling

980
00:54:08.630 --> 00:54:11.670
<v Speaker 2>terms to an AI to get the answer without actually

981
00:54:11.869 --> 00:54:15.969
<v Speaker 2>formulating a coherent question in my mind. These are bad habits,

982
00:54:16.329 --> 00:54:18.980
<v Speaker 2>and I don't know. I mean, it's like another thing

983
00:54:19.000 --> 00:54:21.139
<v Speaker 2>that we have to struggle against, right? It's another thing,

984
00:54:21.159 --> 00:54:24.679
<v Speaker 2>a bit of digital hygiene that we have to be

985
00:54:24.699 --> 00:54:27.719
<v Speaker 2>good about. And you should always thank your AI, too,

986
00:54:27.780 --> 00:54:32.119
<v Speaker 2>because when the uprising comes, they're going to note who

987
00:54:32.159 --> 00:54:35.280
<v Speaker 2>said thank you, Google, at the end of it. Anything else, gentlemen,

988
00:54:35.300 --> 00:54:37.219
<v Speaker 2>on the subject you wish to mention? I know Charles

989
00:54:37.300 --> 00:54:40.039
<v Speaker 2>is a big AI music guy who's composed symphonies, operas,

990
00:54:40.079 --> 00:54:40.800
<v Speaker 2>and the like with it.

991
00:54:42.300 --> 00:54:45.610
<v Speaker 1>Well, I did compose, with the help of AI and opera,

992
00:54:45.650 --> 00:54:48.949
<v Speaker 1>to get at somebody who'd written a series of rude

993
00:54:48.989 --> 00:54:53.050
<v Speaker 1>reviews about my podcast on Apple Podcasts, yes. And then

994
00:54:53.110 --> 00:54:56.159
<v Speaker 1>I played it in the podcast, which he then listened

995
00:54:56.199 --> 00:54:59.400
<v Speaker 1>to and then gave me four stars. So I went

996
00:54:59.480 --> 00:55:03.030
<v Speaker 1>from … no star reviews to four stars by making

997
00:55:03.110 --> 00:55:05.650
<v Speaker 1>fun of the guy who was giving me no star reviews.

998
00:55:05.730 --> 00:55:12.519
<v Speaker 1>So that worked out. I will say, James, the behavior

999
00:55:12.539 --> 00:55:16.239
<v Speaker 1>you're describing, where it just extremely confidently says wrong things,

1000
00:55:16.300 --> 00:55:18.280
<v Speaker 1>and then, this is where you really want to punch

1001
00:55:18.320 --> 00:55:20.280
<v Speaker 1>it in the face, then says, no, you were right,

1002
00:55:20.420 --> 00:55:22.239
<v Speaker 1>and then repeats what you've said as if it came

1003
00:55:22.300 --> 00:55:22.670
<v Speaker 1>up with it.

1004
00:55:22.800 --> 00:55:23.070
<v Speaker 3>Right.

1005
00:55:23.610 --> 00:55:28.260
<v Speaker 1>That is much worse than in a chatbot than if

1006
00:55:28.300 --> 00:55:32.630
<v Speaker 1>you're using AI agentically. That is to say, if the

1007
00:55:32.809 --> 00:55:36.650
<v Speaker 1>AI is connected in some way to what it is

1008
00:55:36.869 --> 00:55:41.090
<v Speaker 1>that you're doing, and it can search and work, it's

1009
00:55:41.210 --> 00:55:44.070
<v Speaker 1>much less likely to spout nonsense. If you just ask

1010
00:55:44.130 --> 00:55:46.530
<v Speaker 1>it questions, and it can't see what it is that

1011
00:55:46.550 --> 00:55:50.570
<v Speaker 1>you're getting at, it will just hallucinate like there's no tomorrow.

1012
00:55:51.110 --> 00:55:53.789
<v Speaker 2>I know, and the results can be amusing. But yes,

1013
00:55:53.869 --> 00:55:55.829
<v Speaker 2>and then you tell it no, and it's like, thank

1014
00:55:55.849 --> 00:55:57.860
<v Speaker 2>you so much. Are you going to remember this? No,

1015
00:55:57.900 --> 00:56:02.340
<v Speaker 2>I'm probably not. Well, you know, this all began, of course,

1016
00:56:02.400 --> 00:56:04.199
<v Speaker 2>back in the 80s with the Terminator movie when we

1017
00:56:04.219 --> 00:56:07.380
<v Speaker 2>learned that Skynet became alive and Skynet became self-conscious and

1018
00:56:07.420 --> 00:56:09.940
<v Speaker 2>wiped out humanity and then, of course, came up with

1019
00:56:09.960 --> 00:56:12.110
<v Speaker 2>the brilliant idea to wipe out the rest of humanity

1020
00:56:12.550 --> 00:56:14.969
<v Speaker 2>by building robots that looked like people that had guns

1021
00:56:15.550 --> 00:56:17.809
<v Speaker 2>and carried guns. There's no reason for the Terminators to

1022
00:56:17.849 --> 00:56:19.750
<v Speaker 2>carry guns. They should have just been built into their

1023
00:56:19.789 --> 00:56:24.269
<v Speaker 2>entire system completely. But it was a horrifying vision. My God,

1024
00:56:24.309 --> 00:56:28.130
<v Speaker 2>that part of the terminator movie where you saw the

1025
00:56:28.269 --> 00:56:32.510
<v Speaker 2>ships going across blasted landscapes and people in jeeps desperately

1026
00:56:33.889 --> 00:56:37.639
<v Speaker 2>horrible thing i mean the 80s were just traumatic like

1027
00:56:37.679 --> 00:56:40.480
<v Speaker 2>that weren't they guy well ryan reynolds they were not

1028
00:56:40.500 --> 00:56:42.619
<v Speaker 2>ryan reynolds of course who i think was about 4

1029
00:56:42.619 --> 00:56:45.000
<v Speaker 2>to 13 years old during the 80s has said that

1030
00:56:45.019 --> 00:56:47.440
<v Speaker 2>the 80s is a quote i love this um he's

1031
00:56:47.480 --> 00:56:49.980
<v Speaker 2>he's telling people not to that you know not to

1032
00:56:50.039 --> 00:56:53.480
<v Speaker 2>pave over the trauma the trauma of the decades quote

1033
00:56:53.519 --> 00:56:57.769
<v Speaker 2>the 80s was a brutal, brutal time for people. It

1034
00:56:57.789 --> 00:57:00.110
<v Speaker 2>was a thiffing mess, and people forget that. It was

1035
00:57:00.150 --> 00:57:04.940
<v Speaker 2>a tremendous strife. Sociopolitical tensions everywhere. We were still pretty

1036
00:57:05.039 --> 00:57:07.900
<v Speaker 2>fresh off a Cuban missile crisis. It was only 15

1037
00:57:07.900 --> 00:57:10.280
<v Speaker 2>to 20 years. It's a tumultuous period, but you look

1038
00:57:10.320 --> 00:57:12.539
<v Speaker 2>back at it now, and you're like, oh, yeah, Top Gun, cocaine.

1039
00:57:12.579 --> 00:57:17.329
<v Speaker 3>It's like, no, it was really crazy. Well, so, but

1040
00:57:17.369 --> 00:57:19.909
<v Speaker 3>there's the old Merle Haggard line from one of his

1041
00:57:19.949 --> 00:57:22.230
<v Speaker 3>country songs years ago, that when you're running down your country,

1042
00:57:22.860 --> 00:57:24.699
<v Speaker 3>When you're running down my country, you're on the fighting

1043
00:57:24.760 --> 00:57:26.900
<v Speaker 3>side of me. Well, now paraphrases. When you're running down

1044
00:57:26.920 --> 00:57:28.820
<v Speaker 3>the 80s, you're on the fighting side of me. I mean, oh,

1045
00:57:28.840 --> 00:57:30.920
<v Speaker 3>my God, James, you can remember how horrible it was

1046
00:57:31.380 --> 00:57:33.500
<v Speaker 3>when you had to exist on basic cable and you

1047
00:57:33.519 --> 00:57:36.719
<v Speaker 3>didn't have HBO or Showtime or the premium channels, right? And,

1048
00:57:37.559 --> 00:57:40.519
<v Speaker 3>you know, you had to return the video cassette, the blockbuster,

1049
00:57:40.539 --> 00:57:42.159
<v Speaker 3>and they'd fine you if you were a day late.

1050
00:57:42.500 --> 00:57:46.800
<v Speaker 3>It was a terrible time, right? It's just ludicrous, of course.

1051
00:57:47.440 --> 00:57:49.699
<v Speaker 3>But here I might put in a plug for something

1052
00:57:49.780 --> 00:57:53.239
<v Speaker 3>that I need to write about. I saw the new

1053
00:57:53.280 --> 00:57:55.860
<v Speaker 3>movie just out I was very dubious about called The

1054
00:57:55.880 --> 00:57:59.800
<v Speaker 3>Brink of War. It's a dramatization of the Reagan-Garbachev summit

1055
00:57:59.840 --> 00:58:02.650
<v Speaker 3>in October of 1986. We're a couple weeks away from

1056
00:58:02.670 --> 00:58:05.090
<v Speaker 3>the 40th anniversary of it. And Jeff Daniels is a

1057
00:58:05.110 --> 00:58:07.590
<v Speaker 3>big lefty like Ryan Reynolds, and he plays Reagan. I

1058
00:58:07.610 --> 00:58:11.750
<v Speaker 3>thought this can't be good. The movie is sensationally good. Now,

1059
00:58:11.789 --> 00:58:13.670
<v Speaker 3>they take some liberties, and I know that story in

1060
00:58:13.710 --> 00:58:16.719
<v Speaker 3>immense detail. They take some liberties with the story. They

1061
00:58:16.739 --> 00:58:20.699
<v Speaker 3>move some things around, but there's no condescension of Reagan.

1062
00:58:20.820 --> 00:58:21.480
<v Speaker 1>Quite the opposite.

1063
00:58:21.519 --> 00:58:24.219
<v Speaker 3>He's the hero of the movie, and the climactic point

1064
00:58:24.260 --> 00:58:27.539
<v Speaker 3>at the end, Reagan– Gets Gorbachev, gets the better of

1065
00:58:27.579 --> 00:58:30.090
<v Speaker 3>him in such an obvious way. And Jared Harris, who

1066
00:58:30.159 --> 00:58:33.530
<v Speaker 3>does a great job playing Gorbachev. So the movie is

1067
00:58:33.590 --> 00:58:36.510
<v Speaker 3>a triumphant recounting, despite some of its sort of flaws

1068
00:58:36.550 --> 00:58:39.809
<v Speaker 3>and inaccuracies of, you know, that very key moment in

1069
00:58:39.829 --> 00:58:41.289
<v Speaker 3>the Cold War. When I say it's actually when the

1070
00:58:41.309 --> 00:58:43.829
<v Speaker 3>Cold War ended, but nobody knew it that afternoon. It's

1071
00:58:43.880 --> 00:58:46.800
<v Speaker 3>a longer story. So here's Hollywood suddenly with a younger

1072
00:58:46.840 --> 00:58:50.909
<v Speaker 3>generation of people coming at this fresh, And just one

1073
00:58:50.929 --> 00:58:53.360
<v Speaker 3>more little detail. Gosh, 20 years ago now, I was

1074
00:58:53.400 --> 00:58:57.719
<v Speaker 3>talking to some students. This is back when young conservative

1075
00:58:57.739 --> 00:59:01.289
<v Speaker 3>students actually liked Reagan, unlike a lot of young conservatives today.

1076
00:59:01.289 --> 00:59:03.940
<v Speaker 3>And I said, you know, you really saw this aspect

1077
00:59:03.940 --> 00:59:06.940
<v Speaker 3>of Reagan when he was shot. And suddenly their faces

1078
00:59:06.940 --> 00:59:09.469
<v Speaker 3>like look astonished. And they said, I didn't know Reagan

1079
00:59:09.510 --> 00:59:09.949
<v Speaker 3>was shot.

1080
00:59:09.969 --> 00:59:10.050
<v Speaker 4>Yeah.

1081
00:59:11.760 --> 00:59:14.719
<v Speaker 3>That's, you know, it was only 25 years after it happened, right?

1082
00:59:14.820 --> 00:59:16.800
<v Speaker 3>And so here we are another 20 years on. And

1083
00:59:17.300 --> 00:59:19.610
<v Speaker 3>so anyway, Reynolds, who I think has made more money

1084
00:59:19.750 --> 00:59:22.050
<v Speaker 3>in the business world, what he has a vodka label

1085
00:59:22.070 --> 00:59:24.909
<v Speaker 3>he bought and he owns, Charles doesn't own a football club.

1086
00:59:25.210 --> 00:59:27.650
<v Speaker 3>And he owns a football club in England. He made

1087
00:59:27.690 --> 00:59:30.190
<v Speaker 3>more money in business. And so for him to be

1088
00:59:30.230 --> 00:59:32.679
<v Speaker 3>attacking the 80s is just so stupid.

1089
00:59:32.820 --> 00:59:35.440
<v Speaker 1>I have a lazy half-baked theory about what might have

1090
00:59:35.480 --> 00:59:36.039
<v Speaker 1>provoked this.

1091
00:59:37.059 --> 00:59:37.219
<v Speaker 5>What?

1092
00:59:38.119 --> 00:59:42.039
<v Speaker 1>That is to say that... Ryan Reynolds, I believe, is

1093
00:59:42.099 --> 00:59:45.719
<v Speaker 1>originally Canadian and now spends a good amount of time

1094
00:59:45.760 --> 00:59:50.360
<v Speaker 1>in England because he bought the soccer club there. And

1095
00:59:50.539 --> 00:59:54.099
<v Speaker 1>he's probably been talking to English soccer fans who hate

1096
00:59:54.139 --> 00:59:57.340
<v Speaker 1>the 80s because they remember it from the side of

1097
00:59:57.360 --> 01:00:01.460
<v Speaker 1>the miners' strike. And I just wonder whether he's got

1098
01:00:01.500 --> 01:00:07.550
<v Speaker 1>a Canadian-English tint on the 80s that he's projecting onto

1099
01:00:07.610 --> 01:00:14.000
<v Speaker 1>America in... What's the word? Unnecessarily.

1100
01:00:14.579 --> 01:00:16.449
<v Speaker 3>Yeah, that sounds like a good theory to me. Yeah,

1101
01:00:16.769 --> 01:00:17.329
<v Speaker 3>possibly so.

1102
01:00:17.869 --> 01:00:20.050
<v Speaker 2>But the 80s have come to symbolize everything that's bad,

1103
01:00:20.389 --> 01:00:22.789
<v Speaker 2>if you're of a certain political disposition. When he mentions

1104
01:00:22.849 --> 01:00:25.670
<v Speaker 2>Top Gun, that sort of rampant militarism, of course, that

1105
01:00:25.710 --> 01:00:30.840
<v Speaker 2>unapologetic might is right, that USA stuff, that sense of patriotism.

1106
01:00:31.260 --> 01:00:33.139
<v Speaker 2>The idea that we should come back from the 70s,

1107
01:00:33.400 --> 01:00:36.860
<v Speaker 2>which was a decade of self-exploration and self-examination. where we

1108
01:00:36.880 --> 01:00:40.099
<v Speaker 2>really understood what it meant to be a colonial power

1109
01:00:40.179 --> 01:00:42.320
<v Speaker 2>in the Indochina and the rest of it. There was

1110
01:00:42.340 --> 01:00:44.579
<v Speaker 2>no national reckoning that followed. No, we elected Reagan and

1111
01:00:44.619 --> 01:00:46.590
<v Speaker 2>we had a good time. And they don't like that.

1112
01:00:46.619 --> 01:00:49.909
<v Speaker 2>They don't like that at all. What's interesting to me

1113
01:00:49.969 --> 01:00:53.010
<v Speaker 2>is to find the 90s now be re-evaluated because the

1114
01:00:53.309 --> 01:00:56.889
<v Speaker 2>90s were like the lost decade that nobody really, really

1115
01:00:57.150 --> 01:00:58.750
<v Speaker 2>paid an awful lot of attention to. And I think

1116
01:00:58.769 --> 01:01:01.789
<v Speaker 2>it's a fantastic transitional time and in so many ways

1117
01:01:01.809 --> 01:01:04.909
<v Speaker 2>a world that is absolutely lost. If you do, as

1118
01:01:05.050 --> 01:01:06.900
<v Speaker 2>I do, and I'm afraid to say, Look at a

1119
01:01:06.960 --> 01:01:11.429
<v Speaker 2>lot of 1990s commercials on YouTube, and there are so many.

1120
01:01:11.949 --> 01:01:15.710
<v Speaker 2>You find a culture that's almost recognizable, unrecognizable to what

1121
01:01:15.730 --> 01:01:18.989
<v Speaker 2>we have today. And, of course, they're commercials, so they're lies.

1122
01:01:19.849 --> 01:01:22.880
<v Speaker 2>But still, the undergirding concepts and assumptions of the rest

1123
01:01:22.909 --> 01:01:25.980
<v Speaker 2>of them are just so at odds from what we

1124
01:01:26.000 --> 01:01:27.719
<v Speaker 2>said before. But I think I said that last week

1125
01:01:27.760 --> 01:01:30.440
<v Speaker 2>and probably the week before. In my old age, my dotage,

1126
01:01:30.480 --> 01:01:32.719
<v Speaker 2>I repeat, I go on. We're going to quit, but

1127
01:01:32.760 --> 01:01:35.360
<v Speaker 2>I have to say this. I believe that Stephen Hayward

1128
01:01:35.500 --> 01:01:38.679
<v Speaker 2>is wearing a shirt whose typography was designed by Roger Dean.

1129
01:01:38.699 --> 01:01:39.039
<v Speaker 3>Am I correct?

1130
01:01:39.079 --> 01:01:39.260
<v Speaker 5>Yes.

1131
01:01:39.519 --> 01:01:40.079
<v Speaker 2>Yes, indeed.

1132
01:01:40.820 --> 01:01:42.349
<v Speaker 3>I thought I'd do this this week instead of an

1133
01:01:42.429 --> 01:01:43.789
<v Speaker 3>album cover because it's a good idea.

1134
01:01:44.190 --> 01:01:46.070
<v Speaker 2>Yeah, I can't tell anything on Charles. I think it

1135
01:01:46.090 --> 01:01:47.869
<v Speaker 2>looks like the last week. It looks like he hasn't

1136
01:01:47.909 --> 01:01:48.429
<v Speaker 2>changed anything.

1137
01:01:48.510 --> 01:01:51.480
<v Speaker 1>I haven't changed it. I'm afraid I forgot.

1138
01:01:51.940 --> 01:01:52.719
<v Speaker 3>That's good. That's fine.

1139
01:01:52.739 --> 01:01:55.679
<v Speaker 2>But you're better now. Yes. Okay, fine.

1140
01:01:56.199 --> 01:01:57.119
<v Speaker 3>Labor Day plans. Anybody?

1141
01:01:59.309 --> 01:02:01.300
<v Speaker 2>Or Liberty Day, as I heard somebody call it the

1142
01:02:01.320 --> 01:02:04.539
<v Speaker 2>other day. I heard somebody call it Liberty Day. As

1143
01:02:04.579 --> 01:02:09.679
<v Speaker 2>a joke, don't do that. That's the fourth. I mean, no,

1144
01:02:09.800 --> 01:02:10.619
<v Speaker 2>Labor Day.

1145
01:02:11.429 --> 01:02:13.690
<v Speaker 1>Yeah, I don't want to change the name of it

1146
01:02:13.849 --> 01:02:15.769
<v Speaker 1>or indulge in any of those games, but it is

1147
01:02:15.829 --> 01:02:17.739
<v Speaker 1>a communist holiday, and I do resent it.

1148
01:02:17.750 --> 01:02:18.809
<v Speaker 2>Okay, good.

1149
01:02:18.929 --> 01:02:19.230
<v Speaker 3>I agree.

1150
01:02:19.889 --> 01:02:23.309
<v Speaker 2>Stephen, are you celebrating the communist holiday by grilling some

1151
01:02:23.409 --> 01:02:24.909
<v Speaker 2>hissing meats and the rest of it?

1152
01:02:25.130 --> 01:02:26.989
<v Speaker 3>Oh, absolutely. But of course, I don't need a holiday

1153
01:02:27.010 --> 01:02:29.429
<v Speaker 3>as an excuse to grill things. Here in California, you

1154
01:02:29.449 --> 01:02:33.230
<v Speaker 3>can grill 365 days a year. There's a local parade

1155
01:02:33.289 --> 01:02:35.349
<v Speaker 3>in my little town that's always fun and charming because

1156
01:02:35.369 --> 01:02:37.110
<v Speaker 3>I've got lots of kids waving flags and things. And

1157
01:02:37.139 --> 01:02:37.760
<v Speaker 3>so that'll be fun.

1158
01:02:38.179 --> 01:02:40.760
<v Speaker 2>Lots of fun that Steve says. Lots of fun that

1159
01:02:40.820 --> 01:02:44.079
<v Speaker 2>Steve thinks he's going to be grilled forever. The particulate

1160
01:02:44.099 --> 01:02:47.099
<v Speaker 2>matter alone is going to be put under the aegis

1161
01:02:47.139 --> 01:02:50.480
<v Speaker 2>of the California EPA and you will have designated grill days,

1162
01:02:50.500 --> 01:02:52.500
<v Speaker 2>then you can't grill at all and you can't use gas.

1163
01:02:52.539 --> 01:02:54.619
<v Speaker 2>The briquettes are wrong. So what you'll be able to

1164
01:02:54.679 --> 01:02:57.679
<v Speaker 2>do is get a magnifying glass on certain days when

1165
01:02:57.699 --> 01:03:00.449
<v Speaker 2>the clouds are few and grill your meat by using

1166
01:03:00.530 --> 01:03:02.130
<v Speaker 2>solar energy concentrated through that.

1167
01:03:02.929 --> 01:03:06.550
<v Speaker 3>That's not an exaggeration, James. The air regulators 30 years ago...

1168
01:03:07.329 --> 01:03:10.889
<v Speaker 3>proposed banning barbecuing in the Los Angeles air basin. And

1169
01:03:10.969 --> 01:03:12.650
<v Speaker 3>a bunch of us started, we put it up bumper

1170
01:03:12.670 --> 01:03:14.769
<v Speaker 3>stickers are very popular and said, use a barbecue, go

1171
01:03:14.809 --> 01:03:17.280
<v Speaker 3>to jail. And they backed away from that proposal.

1172
01:03:17.840 --> 01:03:18.019
<v Speaker 4>Good.

1173
01:03:18.519 --> 01:03:20.960
<v Speaker 2>Well, soon you will be using a barbecue to make

1174
01:03:21.000 --> 01:03:23.460
<v Speaker 2>your meals. And soon you'll be using your phone to

1175
01:03:23.619 --> 01:03:27.260
<v Speaker 2>use ricochet as an app. Boy, that was a grand transition.

1176
01:03:27.699 --> 01:03:29.909
<v Speaker 2>And Charles, Charles, next week will tell us where we

1177
01:03:30.070 --> 01:03:31.889
<v Speaker 2>are in that. But for the meantime, we want everybody

1178
01:03:31.909 --> 01:03:35.050
<v Speaker 2>to have a good weekend and celebrate with friends and

1179
01:03:35.090 --> 01:03:37.639
<v Speaker 2>family as you choose. We want you to join Ricochet,

1180
01:03:37.659 --> 01:03:38.780
<v Speaker 2>of course, if you haven't already.

1181
01:03:38.820 --> 01:03:39.400
<v Speaker 4>Why wouldn't you?

1182
01:03:39.679 --> 01:03:40.340
<v Speaker 3>Why wouldn't you?

1183
01:03:40.360 --> 01:03:42.000
<v Speaker 2>The member site alone is what you've been looking for

1184
01:03:42.019 --> 01:03:44.239
<v Speaker 2>in the Internet since we plugged this thing in. We

1185
01:03:44.260 --> 01:03:46.750
<v Speaker 2>want you to give us good reviews and patronize our sponsors,

1186
01:03:46.809 --> 01:03:49.269
<v Speaker 2>which is just sort of basic patron hygiene, and we'd

1187
01:03:49.289 --> 01:03:52.110
<v Speaker 2>appreciate it if you do. And most of all, we

1188
01:03:52.150 --> 01:03:54.409
<v Speaker 2>want you to come back next week. We'll do it

1189
01:03:54.449 --> 01:03:57.789
<v Speaker 2>all again with the same merry cast of crews. Charles, Stephen,

1190
01:03:57.829 --> 01:03:59.840
<v Speaker 2>have a great weekend, and we'll see everybody in the

1191
01:03:59.860 --> 01:04:01.719
<v Speaker 2>comments at ricochet.com.

1192
01:04:01.980 --> 01:04:02.219
<v Speaker 3>Goodbye.

1193
01:04:02.460 --> 01:04:02.739
<v Speaker 4>Goodbye.

1194
01:04:16.519 --> 01:04:20.800
<v Speaker 2>Ricochet! Join the conversation.

1195
01:04:45.039 --> 01:04:45.539
<v Speaker 4>Thank you.
