1
00:00:03,120 --> 00:00:18,520
Speaker 1: Bloomberg Audio Studios, Podcasts, Radio News.

2
00:00:20,720 --> 00:00:24,760
Speaker 2: Hello and welcome to another episode of the Odd Lots Podcast.

3
00:00:24,840 --> 00:00:26,560
Speaker 2: I'm Joe Wattenthal.

4
00:00:26,200 --> 00:00:27,320
Speaker 3: And I'm Tracy Alloway.

5
00:00:27,760 --> 00:00:32,040
Speaker 2: Tracy, you know, there's a lot of like concerns about

6
00:00:32,640 --> 00:00:36,640
Speaker 2: AI obviously these days, and if anyone who's like reasonably

7
00:00:36,680 --> 00:00:39,760
Speaker 2: intelligent can like list tons like maybe like it's gonna

8
00:00:39,800 --> 00:00:42,559
Speaker 2: go rogue and be smarter than us or maybe whatever,

9
00:00:42,960 --> 00:00:45,120
Speaker 2: But I still think, like what or maybe there's just

10
00:00:45,159 --> 00:00:49,320
Speaker 2: gonna be this like flood of disinformation and deep fakes,

11
00:00:49,479 --> 00:00:51,600
Speaker 2: or maybe it's gonna put all journalists out of business,

12
00:00:51,640 --> 00:00:54,520
Speaker 2: which is certainly plausible, But I think, you know, like

13
00:00:54,560 --> 00:00:56,920
Speaker 2: something I think a lot about is just this idea

14
00:00:57,040 --> 00:01:00,680
Speaker 2: that regardless of what happens, like we're going to be

15
00:01:00,760 --> 00:01:04,280
Speaker 2: increasingly sort of trusting like a black box for answers

16
00:01:04,319 --> 00:01:08,080
Speaker 2: that we really have no idea where those answers, however

17
00:01:08,080 --> 00:01:09,360
Speaker 2: you want to describe that come from.

18
00:01:09,520 --> 00:01:12,080
Speaker 3: Yes, absolutely, And this is something that's come up on

19
00:01:12,120 --> 00:01:13,720
Speaker 3: the podcast a number of times.

20
00:01:13,760 --> 00:01:13,920
Speaker 4: Now.

21
00:01:13,959 --> 00:01:17,360
Speaker 3: I'm thinking way back to an episode we did that

22
00:01:17,520 --> 00:01:20,640
Speaker 3: was basically about the black box of algorithms and how

23
00:01:20,680 --> 00:01:23,880
Speaker 3: difficult it was to understand what goes into them and

24
00:01:23,920 --> 00:01:27,080
Speaker 3: then what comes out. And then of course we recently

25
00:01:27,080 --> 00:01:31,520
Speaker 3: did that episode on pricing and the idea of algorithmic pricing,

26
00:01:31,880 --> 00:01:36,440
Speaker 3: the idea of building proxy consumer profiles, And you're right,

27
00:01:36,680 --> 00:01:39,880
Speaker 3: the issue is we know that there's this new technology,

28
00:01:39,959 --> 00:01:42,840
Speaker 3: We know that there's all this data floating around, but

29
00:01:43,000 --> 00:01:46,920
Speaker 3: we don't entirely know how it is coming to the

30
00:01:47,000 --> 00:01:50,280
Speaker 3: conclusions or creating the output that it actually is.

31
00:01:50,680 --> 00:01:53,920
Speaker 2: You know, the pricing thing is interesting because you know,

32
00:01:54,640 --> 00:01:57,360
Speaker 2: in a market economy, you know, you could argue it's like, oh,

33
00:01:57,480 --> 00:02:01,040
Speaker 2: at any given moment, you know you're being served up

34
00:02:01,080 --> 00:02:03,520
Speaker 2: the like the optimal price, right. And in theory, even

35
00:02:03,560 --> 00:02:06,160
Speaker 2: with the most advanced algorithms and stuff like, maybe there's

36
00:02:06,160 --> 00:02:08,359
Speaker 2: some like this price is happening at the best price

37
00:02:08,440 --> 00:02:11,000
Speaker 2: for both the seller and the buyer, et cetera. But

38
00:02:11,080 --> 00:02:14,320
Speaker 2: I think like people just have a sort of deep

39
00:02:14,880 --> 00:02:18,480
Speaker 2: intuitive distrust about the fact that like, you know, you

40
00:02:18,520 --> 00:02:20,880
Speaker 2: can't go there and like touch it and verify it

41
00:02:20,919 --> 00:02:23,680
Speaker 2: and see like this is why this exists in the

42
00:02:23,720 --> 00:02:26,040
Speaker 2: state that it is. And I think it's going to

43
00:02:26,120 --> 00:02:29,839
Speaker 2: create a lot of I don't know, cultural apprehensions as

44
00:02:29,919 --> 00:02:32,840
Speaker 2: more and more decisions and more and more things that

45
00:02:32,919 --> 00:02:37,200
Speaker 2: are that affect our lives just seem to like emerge

46
00:02:37,240 --> 00:02:38,760
Speaker 2: spontaneously out of the box.

47
00:02:38,880 --> 00:02:41,840
Speaker 3: Absolutely. I'm thinking of all the people working at Chipotle

48
00:02:42,000 --> 00:02:44,480
Speaker 3: who are going to have to answer questions about not

49
00:02:44,560 --> 00:02:48,000
Speaker 3: just portion sizes now, but also questions from customers about

50
00:02:48,040 --> 00:02:49,360
Speaker 3: are they getting the best price.

51
00:02:49,600 --> 00:02:52,280
Speaker 2: Have you seen those awful videos that people are taking

52
00:02:52,360 --> 00:02:53,840
Speaker 2: of the Chipotle workers.

53
00:02:54,160 --> 00:02:57,080
Speaker 3: Yeah, I've seen some of us so vilely imagine, yes.

54
00:02:56,919 --> 00:03:00,040
Speaker 2: It's so vile anyway, that's a that's a that's a

55
00:03:00,080 --> 00:03:02,960
Speaker 2: separate thing. But yes, like all of these things, and

56
00:03:03,000 --> 00:03:06,440
Speaker 2: you know it's it's not just with AI obviously, and

57
00:03:06,520 --> 00:03:10,000
Speaker 2: like this sort of world exists in increasingly black boxes.

58
00:03:10,080 --> 00:03:12,280
Speaker 2: You put in the support ticket, you try to like

59
00:03:12,360 --> 00:03:15,160
Speaker 2: talk to someone in an embassy or a consulate or

60
00:03:15,200 --> 00:03:17,120
Speaker 2: anything you do when you sort of like send out

61
00:03:17,160 --> 00:03:21,440
Speaker 2: some requirement or some request to some bureaucracy or some company,

62
00:03:21,480 --> 00:03:24,000
Speaker 2: and then it moves around. I was on a I

63
00:03:24,000 --> 00:03:26,840
Speaker 2: had a flight recently that was delayed for nine hours,

64
00:03:26,880 --> 00:03:30,520
Speaker 2: and there's like this palpable frustration that everyone feels that

65
00:03:30,600 --> 00:03:33,360
Speaker 2: the person you know, standing at the gate like can't

66
00:03:33,400 --> 00:03:35,880
Speaker 2: answer their questions and they can't get anyone to answer

67
00:03:35,920 --> 00:03:37,720
Speaker 2: their questions, and it just sort of like, you know,

68
00:03:37,720 --> 00:03:41,200
Speaker 2: everyone explodes and everyone knows it's not the gate agent's fault,

69
00:03:41,600 --> 00:03:44,320
Speaker 2: but still, like, you know, there's just this frustration, like

70
00:03:44,440 --> 00:03:46,680
Speaker 2: where is the answer to what's going on?

71
00:03:47,080 --> 00:03:47,280
Speaker 1: Right?

72
00:03:47,320 --> 00:03:50,120
Speaker 3: And you can't ask a single person because that single

73
00:03:50,160 --> 00:03:52,840
Speaker 3: person doesn't, like the gate agent doesn't have the answers.

74
00:03:53,160 --> 00:03:56,960
Speaker 3: I think what's happening is like society has organized itself

75
00:03:57,040 --> 00:04:02,040
Speaker 3: in such a way as to devolve responsibility, and the

76
00:04:02,200 --> 00:04:07,280
Speaker 3: creation of all this new technology is basically going to

77
00:04:07,600 --> 00:04:10,560
Speaker 3: I guess, ramp all of that up, right, Like, so

78
00:04:10,960 --> 00:04:13,160
Speaker 3: it might not even be the gate agent that you're

79
00:04:13,200 --> 00:04:16,240
Speaker 3: asking in the future. It might be you trying to,

80
00:04:16,320 --> 00:04:19,839
Speaker 3: like I don't know, ask the algorithm, like why it

81
00:04:19,920 --> 00:04:23,680
Speaker 3: decided to bump you versus someone else or actually that

82
00:04:23,720 --> 00:04:26,400
Speaker 3: already happens, right, There is an algo that dictates like

83
00:04:26,640 --> 00:04:29,440
Speaker 3: who gets bumped from the plane and who doesn't. So yeah.

84
00:04:29,520 --> 00:04:31,480
Speaker 2: The other big one, of course is health insurance and

85
00:04:31,560 --> 00:04:34,599
Speaker 2: why some claims are suddenly denied and you never get

86
00:04:34,600 --> 00:04:39,680
Speaker 2: this answer anyway. The world is already filled with systems

87
00:04:40,040 --> 00:04:42,840
Speaker 2: in which we have some question and no one actually

88
00:04:42,880 --> 00:04:45,279
Speaker 2: can sort of you know give you the answer.

89
00:04:45,960 --> 00:04:49,360
Speaker 3: Yes, absolutely, and I think we might have the perfect guests.

90
00:04:49,480 --> 00:04:52,200
Speaker 2: We do have the perfect guest. It's someone we've talked

91
00:04:52,200 --> 00:04:55,920
Speaker 2: to multiple times on the podcast, one of the smartest

92
00:04:55,960 --> 00:04:59,360
Speaker 2: guys around, Always interesting, always worth paying attention to. We're

93
00:04:59,400 --> 00:05:01,480
Speaker 2: going to be speaking with Dan Davies. He is the

94
00:05:01,520 --> 00:05:05,720
Speaker 2: author of the new book The Unaccountability Machine, Why Big

95
00:05:05,800 --> 00:05:09,640
Speaker 2: Systems Make Terrible Decisions, and How the World Lost its Mind.

96
00:05:09,960 --> 00:05:12,520
Speaker 2: So this should be really fun. Dan, thank you so

97
00:05:12,600 --> 00:05:14,240
Speaker 2: much for coming back on the podcast.

98
00:05:14,480 --> 00:05:16,000
Speaker 4: Oh thanks very much for inviting me.

99
00:05:16,400 --> 00:05:19,200
Speaker 2: You know, before we get into the meat of your argument,

100
00:05:19,640 --> 00:05:23,040
Speaker 2: The Unaccountability Machine, I'm actually curious about the second half

101
00:05:23,320 --> 00:05:25,920
Speaker 2: of the title, how the World Lost its Mind, because

102
00:05:25,960 --> 00:05:28,479
Speaker 2: I certainly feel like the world lost its mind. But

103
00:05:28,640 --> 00:05:31,159
Speaker 2: I'm like a middle aged boomer, and I feel like,

104
00:05:31,520 --> 00:05:33,760
Speaker 2: you know, anytime you get to my age, you're like, oh,

105
00:05:33,800 --> 00:05:37,680
Speaker 2: the world's gone crazy, the world's gone mad, Like, you know,

106
00:05:37,800 --> 00:05:40,480
Speaker 2: why is everything so nuts these days? Has the world?

107
00:05:40,480 --> 00:05:42,680
Speaker 2: Do we actually know that the world's gone mad? Or

108
00:05:42,720 --> 00:05:44,640
Speaker 2: is it just because like we're all sort of old

109
00:05:44,680 --> 00:05:47,279
Speaker 2: and cranky now everything seems like the world's gone mad.

110
00:05:47,680 --> 00:05:50,400
Speaker 4: Well, from your own perspective, you can never be sure. Yeah,

111
00:05:50,440 --> 00:05:54,640
Speaker 4: but I think there's actually reasonable objective ways that you

112
00:05:54,680 --> 00:05:58,520
Speaker 4: can check up on this, just by noticing that the

113
00:05:58,560 --> 00:06:02,680
Speaker 4: world gets more complicated as it gets bigger, and it

114
00:06:02,760 --> 00:06:05,840
Speaker 4: gets exponentially more complicated. And I mean that're in the

115
00:06:05,880 --> 00:06:10,240
Speaker 4: literal mathematical sense, because the number of connections grows faster

116
00:06:10,320 --> 00:06:15,000
Speaker 4: than the number of things, whereas our capability to understand

117
00:06:15,000 --> 00:06:19,880
Speaker 4: the world, manage it, and make decisions doesn't necessarily grow exponentially.

118
00:06:20,360 --> 00:06:24,479
Speaker 4: So this is the story, I would argue of economics,

119
00:06:24,480 --> 00:06:27,720
Speaker 4: It's the story of any management book that's worth reading,

120
00:06:28,040 --> 00:06:30,839
Speaker 4: because the central problem of management is the world is

121
00:06:30,839 --> 00:06:36,039
Speaker 4: getting faster, more complicated, faster than you can process that complexity.

122
00:06:36,400 --> 00:06:38,200
Speaker 4: What are you going to do about it? How are

123
00:06:38,200 --> 00:06:40,880
Speaker 4: you going to reorganize? And you know, we've just been

124
00:06:40,920 --> 00:06:44,800
Speaker 4: through a global financial crisis, we've just been through a

125
00:06:44,880 --> 00:06:48,440
Speaker 4: political what Aanam two is called poly crisis. I think

126
00:06:48,440 --> 00:06:50,680
Speaker 4: there's decent reasons to believe it's not just because we're

127
00:06:50,680 --> 00:06:55,120
Speaker 4: getting older, and it is actually a crisis of the

128
00:06:55,200 --> 00:06:59,479
Speaker 4: ability to make decisions matched up against the speed and

129
00:06:59,480 --> 00:07:01,719
Speaker 4: complexity see of the decisions that we're having to take.

130
00:07:02,560 --> 00:07:05,800
Speaker 3: So when we talk about the lack of accountability and

131
00:07:05,839 --> 00:07:10,600
Speaker 3: making bad decisions. Give us some concrete examples of things

132
00:07:10,640 --> 00:07:13,840
Speaker 3: that you have spoken about or written about in your book.

133
00:07:14,000 --> 00:07:15,280
Speaker 3: What are you thinking about here?

134
00:07:15,840 --> 00:07:19,120
Speaker 4: Well, I mean there are kind of there's there's little trivial,

135
00:07:19,120 --> 00:07:24,040
Speaker 4: funny examples, and there's big, huge, serious examples. So for example,

136
00:07:24,320 --> 00:07:28,640
Speaker 4: and apologize in advance because this is quite disgusting.

137
00:07:29,240 --> 00:07:30,520
Speaker 3: Oh is it the squirrels?

138
00:07:30,960 --> 00:07:32,040
Speaker 4: Would you rather I didn't talk?

139
00:07:32,720 --> 00:07:32,920
Speaker 1: Yeah?

140
00:07:33,080 --> 00:07:37,000
Speaker 3: You could. This is bad if there are children listening.

141
00:07:37,280 --> 00:07:39,120
Speaker 3: Maybe child.

142
00:07:41,080 --> 00:07:44,880
Speaker 4: At the start of this century there was a craze

143
00:07:45,160 --> 00:07:50,560
Speaker 4: for squirrels as pets in Europe, and squirrels were being

144
00:07:50,560 --> 00:07:53,560
Speaker 4: imported from North America in China to be pets, and

145
00:07:53,640 --> 00:07:58,040
Speaker 4: they had to have the right paperwork. And so one

146
00:07:58,120 --> 00:08:02,600
Speaker 4: day ad of four hundred of the poor little things

147
00:08:02,880 --> 00:08:07,040
Speaker 4: showed up in Chippell Airport in Amsterdam without any paperwork

148
00:08:07,200 --> 00:08:10,040
Speaker 4: and without a return address to send them back to.

149
00:08:10,880 --> 00:08:14,080
Speaker 4: And it's difficult to know what the airline should have done,

150
00:08:14,560 --> 00:08:17,000
Speaker 4: but you can't help thinking that there must have been

151
00:08:17,040 --> 00:08:20,400
Speaker 4: a better solution than what they actually did do, which

152
00:08:20,560 --> 00:08:24,440
Speaker 4: was that they threw all four hundred of them, except

153
00:08:24,480 --> 00:08:27,840
Speaker 4: for one or two that escaped into an industrial shred

154
00:08:29,240 --> 00:08:34,480
Speaker 4: and this caused an outrage. There were questions asked in

155
00:08:34,520 --> 00:08:38,199
Speaker 4: the Dutch Parliament and people immediately started asking how did

156
00:08:38,240 --> 00:08:43,040
Speaker 4: this happen? Who is responsible? And in fact, the press

157
00:08:43,120 --> 00:08:46,439
Speaker 4: release from the airline apologizing for this is studied as

158
00:08:46,480 --> 00:08:50,280
Speaker 4: a masterpiece of crisis pr in business schools. But when

159
00:08:50,280 --> 00:08:53,960
Speaker 4: they went back to inquire, they ended up realizing that

160
00:08:54,160 --> 00:08:58,000
Speaker 4: no one had ever really made the decision that that

161
00:08:58,160 --> 00:09:02,160
Speaker 4: was what they were going to do. The government's biosecurity

162
00:09:02,200 --> 00:09:07,280
Speaker 4: Ministry had set some standards for the importation of small animals,

163
00:09:08,200 --> 00:09:12,440
Speaker 4: the airline had set some standards for compliance with that policy.

164
00:09:13,040 --> 00:09:15,800
Speaker 4: The only people who were expected to make a decision

165
00:09:16,200 --> 00:09:19,920
Speaker 4: about whether this was grotesque and couldn't be done or

166
00:09:20,000 --> 00:09:23,520
Speaker 4: not were some low level employees in a shed at

167
00:09:23,559 --> 00:09:28,480
Speaker 4: Chippell Airport. And frankly, people who work in sheds aren't

168
00:09:28,600 --> 00:09:31,280
Speaker 4: usually going to be thinking that they're meant to be

169
00:09:31,320 --> 00:09:34,120
Speaker 4: second guessing the government. And so what happened is that

170
00:09:34,240 --> 00:09:37,600
Speaker 4: you had this phenomenon that turns up a lot of

171
00:09:37,640 --> 00:09:41,240
Speaker 4: the time at all levels of organization, which is that

172
00:09:41,320 --> 00:09:47,079
Speaker 4: something happened which nobody wanted, but was the predictable output

173
00:09:47,600 --> 00:09:49,280
Speaker 4: of the system that they had created.

174
00:10:05,080 --> 00:10:07,400
Speaker 2: Yeah, this is this is first of all that's grim,

175
00:10:07,440 --> 00:10:10,000
Speaker 2: but also like when you describe it that way, you

176
00:10:10,040 --> 00:10:12,600
Speaker 2: could see it because ultimately, like, all right, here's this

177
00:10:12,720 --> 00:10:16,600
Speaker 2: awful thing that happened to three hundred and ninety eight squirrels.

178
00:10:17,320 --> 00:10:20,800
Speaker 2: I guess in theory, someone had to I don't want

179
00:10:20,840 --> 00:10:23,680
Speaker 2: to get I don't know dumped the bag to the shretter.

180
00:10:24,400 --> 00:10:25,679
Speaker 4: I have not researched them.

181
00:10:25,760 --> 00:10:30,679
Speaker 2: Yeah, but that's not a very satisfying answer. I mean, yes, okay,

182
00:10:30,720 --> 00:10:33,440
Speaker 2: maybe there was someone who did the physical thing, and

183
00:10:33,520 --> 00:10:37,000
Speaker 2: I guess the entire you know, operations of the airline

184
00:10:37,080 --> 00:10:40,320
Speaker 2: and the port and the customs Bureau could like just

185
00:10:40,360 --> 00:10:43,880
Speaker 2: blame that one person. But that's not a very satisfying

186
00:10:44,120 --> 00:10:47,480
Speaker 2: conclusion I guess in terms of like how this actually happened.

187
00:10:47,520 --> 00:10:49,800
Speaker 2: It's funny I brought this documentary up. I was thinking

188
00:10:49,800 --> 00:10:52,400
Speaker 2: about this too, Like I'm with the destruction of the

189
00:10:52,440 --> 00:10:54,760
Speaker 2: old Penn station in New York, which it's like this

190
00:10:54,920 --> 00:10:58,720
Speaker 2: like extraordinary, like I guess, like Roman or Greek building,

191
00:10:58,760 --> 00:11:01,400
Speaker 2: and they just tore it down to like build the

192
00:11:01,960 --> 00:11:02,800
Speaker 2: pretty owes.

193
00:11:02,600 --> 00:11:04,840
Speaker 3: Something hideous, Yeah, something hideous.

194
00:11:04,480 --> 00:11:07,240
Speaker 2: In its place. And now the new Penn station is terrible,

195
00:11:07,280 --> 00:11:09,160
Speaker 2: but it feels like the same thing. It's like, how

196
00:11:09,200 --> 00:11:11,240
Speaker 2: did no one stop and say, like, wait, does this

197
00:11:11,360 --> 00:11:14,000
Speaker 2: make any sense in the long in the big picture.

198
00:11:14,520 --> 00:11:17,360
Speaker 4: Yeah. And the thing is they'd created a system on

199
00:11:17,440 --> 00:11:20,360
Speaker 4: the assumption that squirrels would show up in ones and

200
00:11:20,400 --> 00:11:24,280
Speaker 4: twos and they could be dealt with as individuals, and

201
00:11:24,320 --> 00:11:26,400
Speaker 4: that therefore you would never get into this sort of

202
00:11:26,440 --> 00:11:30,280
Speaker 4: situation because when you build a system, you're always building

203
00:11:30,320 --> 00:11:33,120
Speaker 4: a model of the world, and if something happens which

204
00:11:33,160 --> 00:11:35,800
Speaker 4: doesn't fit into your model in the world, your system

205
00:11:35,920 --> 00:11:40,000
Speaker 4: might do something awful. And there's a sort of symmetry

206
00:11:40,120 --> 00:11:44,120
Speaker 4: and a kind of resemblance here with much bigger and

207
00:11:44,200 --> 00:11:47,360
Speaker 4: more grim things like the Boeing seven three seven max,

208
00:11:47,600 --> 00:11:51,320
Speaker 4: like the light board scandal in financial markets. It's not

209
00:11:51,480 --> 00:11:55,720
Speaker 4: so much that anyone sat down and said, let's form

210
00:11:55,840 --> 00:11:59,520
Speaker 4: a conspiracy to manipulate interest rates, or let's build a

211
00:11:59,760 --> 00:12:03,800
Speaker 4: place that crashes under certain circumstances. It's just that no

212
00:12:03,840 --> 00:12:06,280
Speaker 4: one set things up so that that wouldn't happen.

213
00:12:07,280 --> 00:12:10,840
Speaker 3: Do you think I'm afraid going for the squirrels are

214
00:12:10,840 --> 00:12:14,680
Speaker 3: probably going to be our archetypal example of this, But

215
00:12:14,920 --> 00:12:18,000
Speaker 3: do you think with the squirrels, for instance, the hyper

216
00:12:18,200 --> 00:12:22,440
Speaker 3: specificity of the goals or the job rolls of everyone

217
00:12:22,760 --> 00:12:26,360
Speaker 3: involved contributed to the outcome. So in the sense that

218
00:12:26,440 --> 00:12:29,960
Speaker 3: you have you know, I guess, the Dutch Wildlife Department

219
00:12:30,080 --> 00:12:34,400
Speaker 3: who is trying to protect Dutch wildlife. Then you have

220
00:12:34,480 --> 00:12:36,560
Speaker 3: like the guys at the airport who are charged with

221
00:12:36,760 --> 00:12:40,440
Speaker 3: actually carrying out these orders, and then you have the

222
00:12:40,520 --> 00:12:44,200
Speaker 3: airline which is charged with like looking at the paperwork,

223
00:12:44,679 --> 00:12:47,480
Speaker 3: do those tend to lead to when taken altogether, do

224
00:12:47,559 --> 00:12:49,599
Speaker 3: those tend to lead to worse outcomes?

225
00:12:50,000 --> 00:12:53,480
Speaker 4: Yeah? Absolutely. It's this kind of fragmentation of the decision

226
00:12:54,040 --> 00:12:57,640
Speaker 4: which is a result of the industrialization of the decision

227
00:12:57,679 --> 00:13:01,240
Speaker 4: and its industrialization literally in the atom Smith's sense that

228
00:13:01,679 --> 00:13:04,480
Speaker 4: you don't have anyone in the pin factory building an

229
00:13:04,640 --> 00:13:09,240
Speaker 4: entire pin. You have thirteen guys all performing one simple operation.

230
00:13:09,760 --> 00:13:12,440
Speaker 4: And that's a much more productive way to do things.

231
00:13:12,720 --> 00:13:15,480
Speaker 4: But then when you apply it to decision making, you

232
00:13:15,559 --> 00:13:19,280
Speaker 4: have the problem that everyone assumes that everyone else is

233
00:13:19,320 --> 00:13:24,400
Speaker 4: going to react when something unplanned happens. So this is,

234
00:13:24,520 --> 00:13:28,760
Speaker 4: like I say, it's the central problem of every good

235
00:13:28,840 --> 00:13:31,800
Speaker 4: management sex book. How do you deal with information? How

236
00:13:31,800 --> 00:13:33,959
Speaker 4: do you get a drink from a fire hose. How

237
00:13:33,960 --> 00:13:37,200
Speaker 4: do you stop yourself from being overwhelmed? The answer is

238
00:13:37,280 --> 00:13:40,079
Speaker 4: always in some way or another, you build a system

239
00:13:40,240 --> 00:13:43,160
Speaker 4: to make the decisions for you. But once you've built

240
00:13:43,200 --> 00:13:46,120
Speaker 4: that system to make the decisions for you, you no

241
00:13:46,200 --> 00:13:49,640
Speaker 4: longer feel ownership of that decision. Psychologically, you no longer

242
00:13:49,640 --> 00:13:52,920
Speaker 4: feel like you're accountable for the decision, because if you

243
00:13:52,960 --> 00:13:55,960
Speaker 4: were accountable, you might be able to change it. But

244
00:13:56,040 --> 00:13:58,680
Speaker 4: if you're going to be held accountable for this thing,

245
00:13:58,960 --> 00:14:01,840
Speaker 4: you haven't really moved that thing on, you haven't really

246
00:14:01,880 --> 00:14:03,160
Speaker 4: delegated it to the system.

247
00:14:04,080 --> 00:14:07,320
Speaker 2: You mentioned Boeing, And speaking of Boeing, there was a

248
00:14:07,360 --> 00:14:11,679
Speaker 2: great blog post back in April from Steve Randy Waldman

249
00:14:11,800 --> 00:14:14,840
Speaker 2: inter Fluidity, who I consider another one kind of in

250
00:14:14,880 --> 00:14:18,640
Speaker 2: your category of people who have basically been writing interesting

251
00:14:18,760 --> 00:14:21,280
Speaker 2: things on the internet for a very long time, and

252
00:14:21,320 --> 00:14:24,040
Speaker 2: he has the title was seeing like a CEO and

253
00:14:24,120 --> 00:14:27,600
Speaker 2: this idea that you know, when Boeing merged with McDonald douglas,

254
00:14:27,720 --> 00:14:30,360
Speaker 2: the McDonald douglas CEO became an outside hire and he

255
00:14:30,480 --> 00:14:35,040
Speaker 2: had to essentially gain some legibility into this new organization

256
00:14:35,120 --> 00:14:37,840
Speaker 2: that he inherited, and that was the cause for some

257
00:14:37,880 --> 00:14:42,120
Speaker 2: of this like streamlining and offshore or you know, offshoring,

258
00:14:42,680 --> 00:14:45,000
Speaker 2: things like that. Talk to us like about you know,

259
00:14:45,040 --> 00:14:49,040
Speaker 2: you mentioned Boeing and it's more you know, the crises

260
00:14:49,080 --> 00:14:51,480
Speaker 2: there that's been going on several years. It's a more severe,

261
00:14:51,560 --> 00:14:56,280
Speaker 2: serious issue than the squirrels. But how do you know,

262
00:14:56,280 --> 00:14:57,800
Speaker 2: how do you think about what happened there?

263
00:14:58,200 --> 00:15:01,360
Speaker 4: I think about it in very similar ways to Steve

264
00:15:01,440 --> 00:15:05,600
Speaker 4: Randy Wilburn because the and you see it in Boeing,

265
00:15:05,720 --> 00:15:08,600
Speaker 4: but it's visible to a greater or lesser extent in

266
00:15:08,960 --> 00:15:12,240
Speaker 4: very very many companies, probably the majority of companies today

267
00:15:12,720 --> 00:15:16,440
Speaker 4: that the C suite has an information environment which is

268
00:15:16,600 --> 00:15:22,560
Speaker 4: almost completely composed of financial numbers, because the financial numbers

269
00:15:23,080 --> 00:15:27,520
Speaker 4: are taken by them as objective facts. We can talk

270
00:15:28,160 --> 00:15:31,680
Speaker 4: long and hard with accountants about how objective those financial

271
00:15:31,760 --> 00:15:34,200
Speaker 4: numbers are and how many assumptions go into them, but

272
00:15:34,320 --> 00:15:38,280
Speaker 4: they arrive on a spreadsheet looking very much like objective

273
00:15:38,280 --> 00:15:43,200
Speaker 4: facts about the world. Things like engineering level, in principle

274
00:15:43,360 --> 00:15:45,560
Speaker 4: are objective facts, but you have to do a lot

275
00:15:45,640 --> 00:15:48,600
Speaker 4: more work to find them out and to know what's relevant.

276
00:15:49,200 --> 00:15:53,680
Speaker 4: And then you have issues like culture and kind of

277
00:15:53,720 --> 00:15:58,360
Speaker 4: the social environments which aren't even capable of being quantified,

278
00:15:58,760 --> 00:16:01,280
Speaker 4: So there's always this tender and see, if you're trying

279
00:16:01,320 --> 00:16:04,160
Speaker 4: to do that thing of manage your own capacity to

280
00:16:04,200 --> 00:16:08,400
Speaker 4: manage your information flow, that you're going to concentrate on

281
00:16:08,480 --> 00:16:12,560
Speaker 4: the things that look finite and look manageable, and that's

282
00:16:12,560 --> 00:16:14,960
Speaker 4: always going to be the financial numbers, which can be

283
00:16:15,120 --> 00:16:20,600
Speaker 4: a big problem because financial numbers can mislead. You know,

284
00:16:20,680 --> 00:16:23,520
Speaker 4: you can create illusions in an accounting system the same

285
00:16:23,520 --> 00:16:24,840
Speaker 4: way that you can with anything else.

286
00:16:25,720 --> 00:16:29,040
Speaker 3: How did we end up here? And I'm thinking specifically

287
00:16:29,360 --> 00:16:33,440
Speaker 3: to one particular development in the world of business, which

288
00:16:33,480 --> 00:16:36,840
Speaker 3: is the creation of the limited liability company. And I

289
00:16:36,840 --> 00:16:39,440
Speaker 3: guess the clues kind of in the name there, But

290
00:16:39,600 --> 00:16:43,600
Speaker 3: what were the decisions or the trends that sort of

291
00:16:43,640 --> 00:16:46,080
Speaker 3: came about in creating the current system.

292
00:16:46,880 --> 00:16:50,600
Speaker 4: Well, I mean, the limited liability company is certainly a

293
00:16:50,760 --> 00:16:54,280
Speaker 4: big kind of change if you think about these things

294
00:16:54,400 --> 00:16:58,960
Speaker 4: in feedback terms and information terms, and you know, one

295
00:16:59,000 --> 00:17:01,040
Speaker 4: of the big arguments of the BO is that we

296
00:17:01,080 --> 00:17:05,400
Speaker 4: should be looking to the mathematics of information theory rather

297
00:17:05,440 --> 00:17:09,359
Speaker 4: than the mathematics of optimization to explain and model some

298
00:17:09,440 --> 00:17:12,840
Speaker 4: of these things. But a limited liability company is an

299
00:17:12,840 --> 00:17:17,040
Speaker 4: information filter. It tells you that outcomes below a certain

300
00:17:17,080 --> 00:17:22,000
Speaker 4: amount aren't going to affect you anymore, and that changes

301
00:17:22,400 --> 00:17:25,600
Speaker 4: your information world, It changes what you care about. But

302
00:17:25,880 --> 00:17:31,199
Speaker 4: then what I think really started doing the damage, so

303
00:17:31,280 --> 00:17:34,720
Speaker 4: to speak, was the development of the leverage buyout in

304
00:17:34,840 --> 00:17:39,919
Speaker 4: the nineteen seventies and the shareholder value movement, as it

305
00:17:40,040 --> 00:17:44,040
Speaker 4: was kind of really kicked off by Milton Friedman's essay

306
00:17:44,040 --> 00:17:46,959
Speaker 4: on the social responsibility of a business to increase its

307
00:17:47,000 --> 00:17:50,720
Speaker 4: profits New York Times, nineteen seventy, but then built on

308
00:17:51,440 --> 00:17:55,439
Speaker 4: by just the entire kind of two decades of business

309
00:17:55,440 --> 00:17:59,680
Speaker 4: school research that followed from that. Because again, thinking about

310
00:17:59,720 --> 00:18:03,960
Speaker 4: it in information terms, a leftage buyout is a massive,

311
00:18:04,240 --> 00:18:08,919
Speaker 4: screaming signal. The requirement to make the payments on debts

312
00:18:09,280 --> 00:18:13,320
Speaker 4: becomes a signal that swamps anything else you might be

313
00:18:13,400 --> 00:18:16,640
Speaker 4: thinking of, because if you are a CEO and you've

314
00:18:16,680 --> 00:18:21,200
Speaker 4: got LBO levels of debt, that's your priority. You can't

315
00:18:21,240 --> 00:18:25,680
Speaker 4: think about anything that isn't related to servicing that debt.

316
00:18:26,160 --> 00:18:28,040
Speaker 2: Let's go back and talk about the sort of big

317
00:18:28,080 --> 00:18:30,919
Speaker 2: picture ideas. In your book, you talk a lot about

318
00:18:30,920 --> 00:18:36,720
Speaker 2: this field called cybernetics, and cybernetics is a name sounds

319
00:18:36,760 --> 00:18:38,480
Speaker 2: like something that they would have come up with in

320
00:18:38,520 --> 00:18:41,640
Speaker 2: the early nineties, like the Wired magazine people like would

321
00:18:41,680 --> 00:18:44,880
Speaker 2: get into like cybernetics in ninety one. But actually this

322
00:18:44,920 --> 00:18:48,720
Speaker 2: field has been around at least since the nineteen forties.

323
00:18:48,960 --> 00:18:51,199
Speaker 2: I'm surprised they even had a word like cybernetics in

324
00:18:51,200 --> 00:18:55,240
Speaker 2: the nineteen forties. But what is cybernetics? And talk to

325
00:18:55,320 --> 00:18:58,080
Speaker 2: us about the sort of general framework you use as

326
00:18:58,080 --> 00:19:00,359
Speaker 2: a sort of to start talking about this stories in

327
00:19:00,400 --> 00:19:00,800
Speaker 2: your book.

328
00:19:00,880 --> 00:19:03,359
Speaker 4: Sure, I mean, you're right, it was its Second World

329
00:19:03,400 --> 00:19:08,119
Speaker 4: War kind of talk. It's originally from a word meaning

330
00:19:08,200 --> 00:19:11,280
Speaker 4: the man who steers the boat, So it's a cybernetics

331
00:19:11,280 --> 00:19:14,640
Speaker 4: in that sense. And the first guy to use it

332
00:19:15,400 --> 00:19:18,960
Speaker 4: was a scientist working on creating an automated gun site

333
00:19:19,640 --> 00:19:24,720
Speaker 4: for the United States Air Force. And the idea here

334
00:19:25,080 --> 00:19:29,600
Speaker 4: is that there is some quantity that is preserved in

335
00:19:29,640 --> 00:19:34,920
Speaker 4: an automated gun site between the operator, the radar, the server,

336
00:19:35,040 --> 00:19:38,160
Speaker 4: motors and that and all the components of that system,

337
00:19:38,280 --> 00:19:43,080
Speaker 4: and that component is information. And so at the time

338
00:19:43,119 --> 00:19:46,800
Speaker 4: this was Norbert Viena. I'm talking about the scientist in

339
00:19:46,840 --> 00:19:50,439
Speaker 4: the automated gun sites. Another guy working in the same field.

340
00:19:50,600 --> 00:19:53,320
Speaker 4: You might have heard of was Claude Shannon at Bell Labs,

341
00:19:53,760 --> 00:19:58,240
Speaker 4: who was inventing information theory at Bell Labs and has

342
00:19:58,240 --> 00:20:02,240
Speaker 4: some fundamental theorems there, And in many ways, science of

343
00:20:02,280 --> 00:20:09,280
Speaker 4: cybernetics is information theory applied to control. So you might

344
00:20:09,400 --> 00:20:11,919
Speaker 4: have an information theory kind of pie the piece of

345
00:20:11,960 --> 00:20:15,479
Speaker 4: maths that tells you how much bandwidth you need to

346
00:20:15,520 --> 00:20:19,959
Speaker 4: transmit a given signal, and the cybernetic interpretation of that

347
00:20:20,040 --> 00:20:24,360
Speaker 4: math would be that it's telling you how much capacity

348
00:20:24,520 --> 00:20:29,680
Speaker 4: you need to manage a system that's of similar kind

349
00:20:29,680 --> 00:20:34,800
Speaker 4: of noise. So this was all made huge use of

350
00:20:35,600 --> 00:20:39,080
Speaker 4: in controlling things where you have access to the whole

351
00:20:39,320 --> 00:20:42,720
Speaker 4: information environment. So a lot of that early maths from

352
00:20:42,760 --> 00:20:46,760
Speaker 4: the first cyberneticians has just kind of stayed with us

353
00:20:47,920 --> 00:20:50,760
Speaker 4: through the invention of the electronic computer, and a lot

354
00:20:50,760 --> 00:20:55,960
Speaker 4: of it is actually at work in really modern artificial intelligence.

355
00:20:56,560 --> 00:21:01,879
Speaker 4: Meragai called ben Recht who works on recommendation algorithms, and

356
00:21:02,040 --> 00:21:05,480
Speaker 4: he's quite upfront that a lot of his fundamental mathematical

357
00:21:05,520 --> 00:21:08,640
Speaker 4: techniques are from the four seas and fifties just being

358
00:21:08,680 --> 00:21:13,280
Speaker 4: applied in the context of massively more computing power. What

359
00:21:13,320 --> 00:21:16,280
Speaker 4: I'm interested in is where you take those kinds of

360
00:21:16,840 --> 00:21:22,320
Speaker 4: theorems and apply them in a slightly more unrigorous, slightly

361
00:21:22,359 --> 00:21:28,280
Speaker 4: more metaphorical sense to situations of management and organization where

362
00:21:28,320 --> 00:21:31,359
Speaker 4: you don't have access to the full information environment and

363
00:21:31,440 --> 00:21:34,879
Speaker 4: you just have to say, we're going to think about

364
00:21:34,920 --> 00:21:40,399
Speaker 4: this not in an optimizing economics kind of neoclassical economic sense,

365
00:21:40,680 --> 00:21:42,680
Speaker 4: but we're going to think about this as a system

366
00:21:43,000 --> 00:21:46,159
Speaker 4: that has to be kept under control, and we're going

367
00:21:46,240 --> 00:21:49,320
Speaker 4: to say, well, how much resource do we need in

368
00:21:49,400 --> 00:21:53,040
Speaker 4: order to stabilize this system as a system, Which is

369
00:21:53,040 --> 00:21:54,879
Speaker 4: a kind of abstract way of looking at it, but

370
00:21:54,960 --> 00:21:58,199
Speaker 4: it's the same fundamental problem of management. How do you

371
00:21:58,200 --> 00:22:01,119
Speaker 4: get a drink from a firehouse? How do you match

372
00:22:01,200 --> 00:22:04,480
Speaker 4: your own capacity to manage to the complexity of the

373
00:22:04,520 --> 00:22:05,560
Speaker 4: thing that you're in charge of.

374
00:22:06,040 --> 00:22:09,720
Speaker 3: Wait, can you give us a practical example of the

375
00:22:09,760 --> 00:22:13,159
Speaker 3: application of cybernetics. I guess because it does sound to

376
00:22:13,200 --> 00:22:15,639
Speaker 3: your point earlier, it sounds a little bit abstract in

377
00:22:15,760 --> 00:22:16,280
Speaker 3: my mind.

378
00:22:17,359 --> 00:22:20,840
Speaker 4: Well, I mean practical example I think is the history

379
00:22:20,920 --> 00:22:24,200
Speaker 4: of the development of the corporation. So from the first

380
00:22:24,280 --> 00:22:28,399
Speaker 4: days of the American railroads, which were probably the first

381
00:22:28,640 --> 00:22:32,919
Speaker 4: really big corporate structures, the world ever saw. You have

382
00:22:33,160 --> 00:22:37,159
Speaker 4: this problem that as the network builds out, it gets

383
00:22:37,280 --> 00:22:41,600
Speaker 4: more complicated, and the ability of the head office to

384
00:22:41,680 --> 00:22:46,520
Speaker 4: manage it doesn't grow faster. And you can try and

385
00:22:46,560 --> 00:22:50,439
Speaker 4: solve that by adding more people. You can get a

386
00:22:50,480 --> 00:22:55,000
Speaker 4: great improvement by adding wireless telegraphs. But fundamentally, at some

387
00:22:55,200 --> 00:22:59,040
Speaker 4: point this railroad is going to grow big enough that

388
00:22:59,119 --> 00:23:03,400
Speaker 4: you have to devolve. You have to split it into branches,

389
00:23:03,800 --> 00:23:07,160
Speaker 4: and you have to give autonomy to some of the subsidiaries,

390
00:23:07,480 --> 00:23:11,720
Speaker 4: because that's the only way that you can match the

391
00:23:11,760 --> 00:23:16,440
Speaker 4: bandwidth of the management to the bandwidth of the control problem.

392
00:23:16,520 --> 00:23:19,720
Speaker 4: So I'd say that what we always see in any

393
00:23:19,720 --> 00:23:24,560
Speaker 4: big organization is that it grows, it gets more complicated,

394
00:23:25,080 --> 00:23:27,720
Speaker 4: It tries to deal with that by adding more resources

395
00:23:27,720 --> 00:23:31,320
Speaker 4: at head office, it ends up not being able to

396
00:23:31,400 --> 00:23:36,800
Speaker 4: keep up, and then it reorganizes. And the reorganizations almost

397
00:23:36,840 --> 00:23:42,320
Speaker 4: always either involve pushing responsibility down to the shop floor

398
00:23:42,480 --> 00:23:45,800
Speaker 4: or down to the branches, or they involve spinning off

399
00:23:45,840 --> 00:23:49,119
Speaker 4: parts of the business into a separate organization and giving

400
00:23:49,160 --> 00:23:51,520
Speaker 4: up the task of controlling it at all.

401
00:24:07,240 --> 00:24:09,399
Speaker 2: What are accountability sinks?

402
00:24:10,920 --> 00:24:13,800
Speaker 4: The accountability sink is just it's a name for a

403
00:24:13,840 --> 00:24:18,400
Speaker 4: particular move of cybern ethics that I notice a lot

404
00:24:18,440 --> 00:24:25,040
Speaker 4: of these days, which is when you consciously break the

405
00:24:25,080 --> 00:24:30,320
Speaker 4: feedback links from the subjects of a particular decision to

406
00:24:31,160 --> 00:24:34,560
Speaker 4: yourself or to the unit that's kind of meant to

407
00:24:34,600 --> 00:24:37,119
Speaker 4: be making it. So your gate agent Joe is just

408
00:24:37,160 --> 00:24:41,040
Speaker 4: a classic example of the accountability sink, because they talk

409
00:24:41,160 --> 00:24:45,800
Speaker 4: to you with the voice of a corporation and they

410
00:24:45,840 --> 00:24:47,960
Speaker 4: say that this is the policy, there's nothing I can

411
00:24:47,960 --> 00:24:50,920
Speaker 4: do to change it, and then you only able to

412
00:24:50,960 --> 00:24:53,640
Speaker 4: talk back to them as a human being like yourself,

413
00:24:53,960 --> 00:24:56,040
Speaker 4: so you can't get mad at them because it's not

414
00:24:56,119 --> 00:24:58,720
Speaker 4: their decision they get Just to.

415
00:24:58,680 --> 00:24:59,919
Speaker 2: Be clear, I did not get mad.

416
00:25:01,480 --> 00:25:04,000
Speaker 3: There's a video of Joe out there.

417
00:25:04,800 --> 00:25:07,919
Speaker 2: I just sat there and I closed my eyes, and

418
00:25:08,000 --> 00:25:10,480
Speaker 2: I did stand around because I was sort of curious

419
00:25:10,520 --> 00:25:12,360
Speaker 2: the gate banter. But I did not get bad.

420
00:25:12,400 --> 00:25:14,880
Speaker 4: I just want to establish Yeah, but then you might

421
00:25:14,960 --> 00:25:17,280
Speaker 4: ask them, and I will confess I've done this. I've

422
00:25:17,320 --> 00:25:21,320
Speaker 4: asked them someone politely for the phone number of someone

423
00:25:21,359 --> 00:25:24,320
Speaker 4: who I can call up who is responsible for that decision,

424
00:25:24,920 --> 00:25:27,600
Speaker 4: and you know it's not the policy. You can't get

425
00:25:27,600 --> 00:25:30,440
Speaker 4: that phone number. The whole point of this was to

426
00:25:30,480 --> 00:25:36,959
Speaker 4: create a sink into which unpleasant feedback can be poured

427
00:25:37,400 --> 00:25:41,640
Speaker 4: and does dissipated harmlessly. And when you start thinking about

428
00:25:41,680 --> 00:25:44,959
Speaker 4: these things in terms of accountability sinks, you start seeing

429
00:25:44,960 --> 00:25:49,720
Speaker 4: them everywhere. Because everywhere that there's a policy that can't

430
00:25:49,720 --> 00:25:53,760
Speaker 4: be broken and no feedback to the person who could

431
00:25:53,800 --> 00:25:57,080
Speaker 4: get the policy changed, that's an accountability sink. That's a

432
00:25:57,119 --> 00:26:01,040
Speaker 4: way that someone has protected themselves from the consequences of

433
00:26:01,080 --> 00:26:05,399
Speaker 4: their decisions, possibly at huge costs to the organization that

434
00:26:05,440 --> 00:26:07,199
Speaker 4: they're working for, but possibly not.

435
00:26:09,640 --> 00:26:12,800
Speaker 3: I guess I'll ask the obvious question, which is how

436
00:26:12,800 --> 00:26:16,880
Speaker 3: do we break out of accountability sinks? And I think

437
00:26:16,920 --> 00:26:19,520
Speaker 3: the frustration with everyone is that, you know, you feel

438
00:26:19,520 --> 00:26:21,960
Speaker 3: powerless when you're caught in one, when you can't get

439
00:26:22,280 --> 00:26:24,480
Speaker 3: the answer that you want, or when you can't speak

440
00:26:24,520 --> 00:26:26,720
Speaker 3: to the decision maker and try to reason with them

441
00:26:26,800 --> 00:26:28,960
Speaker 3: or explain why this might be a one off or

442
00:26:29,000 --> 00:26:32,240
Speaker 3: a peculiar situation. And then it just feels like the

443
00:26:32,320 --> 00:26:35,640
Speaker 3: idea of actually starting to break apart some of these

444
00:26:35,680 --> 00:26:39,760
Speaker 3: sinks and move to more of an era of personal

445
00:26:39,800 --> 00:26:44,200
Speaker 3: accountability the book stops here and all of that, Yeah,

446
00:26:44,200 --> 00:26:47,439
Speaker 3: there we go. It just seems further and further away.

447
00:26:48,119 --> 00:26:50,600
Speaker 4: Well, it is. And the horrible answer to your question,

448
00:26:50,720 --> 00:26:55,080
Speaker 4: Tracy is that maybe we can't, or maybe as individuals

449
00:26:55,119 --> 00:26:58,840
Speaker 4: we can't. And that's actually, in my view, potentially very

450
00:26:58,880 --> 00:27:02,800
Speaker 4: bad news for society because all of this sink, you know,

451
00:27:02,880 --> 00:27:05,280
Speaker 4: all of this negative emotion from people about the way

452
00:27:05,280 --> 00:27:08,840
Speaker 4: the world is, goes into the sinks. But like any sink,

453
00:27:08,960 --> 00:27:11,800
Speaker 4: it piles up, and it piles up, and then after

454
00:27:11,880 --> 00:27:15,399
Speaker 4: a while it all spills out, and then suddenly we

455
00:27:15,480 --> 00:27:19,600
Speaker 4: get things like Brexit in the UK, or kind of

456
00:27:19,800 --> 00:27:22,680
Speaker 4: first go of Donald Trump in the USA. We get

457
00:27:22,720 --> 00:27:27,920
Speaker 4: people who have used to being decided upon and used

458
00:27:27,960 --> 00:27:33,080
Speaker 4: to being ignored, getting steadily more and more dissatisfied with

459
00:27:33,119 --> 00:27:35,879
Speaker 4: the system, and then finally they start to use the

460
00:27:35,920 --> 00:27:40,639
Speaker 4: only power left to them to just make use their

461
00:27:40,720 --> 00:27:43,520
Speaker 4: votes in a way that says, I am no longer

462
00:27:43,560 --> 00:27:46,000
Speaker 4: satisfied with this, this is no longer tolerable to me.

463
00:27:46,320 --> 00:27:48,560
Speaker 4: I'm going to use my vote to tear the system apart.

464
00:27:49,200 --> 00:27:53,359
Speaker 4: And so with all these things, can you can divert

465
00:27:53,400 --> 00:27:55,359
Speaker 4: these things for a while, but it's at the cost

466
00:27:55,400 --> 00:27:56,600
Speaker 4: of building up fragility.

467
00:27:57,359 --> 00:28:00,439
Speaker 2: Yeah, I have to say, you know, I I didn't

468
00:28:00,720 --> 00:28:04,119
Speaker 2: finish your book I read. I'm about halfway through, but

469
00:28:04,240 --> 00:28:09,800
Speaker 2: it did leave me fairly nihilistic or pessimistic that it's like,

470
00:28:10,119 --> 00:28:13,639
Speaker 2: these are these inexorable centrifugal or centripetal forces. I can

471
00:28:13,680 --> 00:28:16,840
Speaker 2: never remember which is which, and they're pulling us into

472
00:28:16,920 --> 00:28:21,160
Speaker 2: all of these sort of high stress decisions and it's

473
00:28:21,200 --> 00:28:24,000
Speaker 2: really bad, and things are going to things are gonna

474
00:28:24,040 --> 00:28:26,880
Speaker 2: keep breaking, and we're going to get angry or anger.

475
00:28:27,600 --> 00:28:29,960
Speaker 2: We talked about AI in the beginning, and a sort

476
00:28:30,000 --> 00:28:32,560
Speaker 2: of provocative idea that you talk about in your book

477
00:28:32,640 --> 00:28:36,239
Speaker 2: is the idea of the corporation as it exists and

478
00:28:36,320 --> 00:28:39,400
Speaker 2: as we've known about it as already a proto AI.

479
00:28:39,520 --> 00:28:41,160
Speaker 2: So you go to chat GPT and you put in

480
00:28:41,200 --> 00:28:43,800
Speaker 2: a request and something spits out and it's impressive whatever,

481
00:28:44,120 --> 00:28:46,560
Speaker 2: but that actually this is just sort of a specific

482
00:28:46,640 --> 00:28:50,240
Speaker 2: example of what the corporation has been for a long time.

483
00:28:50,720 --> 00:28:54,200
Speaker 4: Absolutely, I mean, and this is the beauty of abstract mass.

484
00:28:54,200 --> 00:28:56,600
Speaker 4: It describes things without you needing to know what they

485
00:28:56,640 --> 00:29:00,320
Speaker 4: actually are. These are all just decision making systems. I

486
00:29:00,400 --> 00:29:03,520
Speaker 4: had a conversation with someone at the European Commission the

487
00:29:03,600 --> 00:29:07,320
Speaker 4: year before last, because in Europe they passed in Act,

488
00:29:07,800 --> 00:29:11,240
Speaker 4: saying that if a decision is made affecting you, like

489
00:29:11,360 --> 00:29:14,240
Speaker 4: to turn you down for health insurance, then if that's

490
00:29:14,240 --> 00:29:17,360
Speaker 4: made by an algorithm, you have a right of explainability,

491
00:29:17,800 --> 00:29:19,760
Speaker 4: So you have a right that someone can explain to

492
00:29:19,840 --> 00:29:23,320
Speaker 4: you why that algorithm made that decision for you. And

493
00:29:23,440 --> 00:29:25,480
Speaker 4: you know, I thought that's quite good, But it's kind

494
00:29:25,480 --> 00:29:28,320
Speaker 4: of ironic that this decision is coming from the European Commission.

495
00:29:28,680 --> 00:29:30,720
Speaker 4: So I asked the guy who works there, well, you know,

496
00:29:30,800 --> 00:29:33,280
Speaker 4: when you make a decision like that, what right of

497
00:29:33,320 --> 00:29:36,200
Speaker 4: explainability do I have from you? And the answer is

498
00:29:36,200 --> 00:29:40,320
Speaker 4: ha hah, no, none at all. M All these things

499
00:29:40,560 --> 00:29:43,479
Speaker 4: are basically working in the same way. The AI is

500
00:29:43,600 --> 00:29:46,760
Speaker 4: working like the corporation, which is working like the government,

501
00:29:47,400 --> 00:29:52,600
Speaker 4: and the same problems of information managements affect them all.

502
00:29:53,000 --> 00:29:55,640
Speaker 4: But it's not as nihilistic as you think, in my view,

503
00:29:55,760 --> 00:30:00,400
Speaker 4: because that means that these things can be so subject

504
00:30:00,400 --> 00:30:03,240
Speaker 4: to the same kinds of solutions. You know, if we

505
00:30:03,280 --> 00:30:06,560
Speaker 4: think about the original reason for building the accountability sink,

506
00:30:07,080 --> 00:30:11,400
Speaker 4: it was that someone felt overwhelm us by information and

507
00:30:11,480 --> 00:30:14,320
Speaker 4: so didn't feel that they were responsible for the decision

508
00:30:14,760 --> 00:30:18,640
Speaker 4: and so wanted to cut the link of accountability. If

509
00:30:18,680 --> 00:30:21,240
Speaker 4: you can put the AI in the loop in such

510
00:30:21,280 --> 00:30:25,360
Speaker 4: a way that the decision maker is more able to

511
00:30:25,440 --> 00:30:28,480
Speaker 4: manage their information flow, then they don't have so much

512
00:30:28,600 --> 00:30:32,240
Speaker 4: need to break the feedback links because they've got more

513
00:30:32,760 --> 00:30:35,440
Speaker 4: functional ways to deal with them.

514
00:30:36,040 --> 00:30:39,640
Speaker 3: I have a theoretical I have a theoretical question, which is,

515
00:30:39,760 --> 00:30:43,760
Speaker 3: what does accountability actually look like for a decision taken

516
00:30:43,840 --> 00:30:48,320
Speaker 3: by an algorithm? Is it that, like we understand the

517
00:30:48,640 --> 00:30:53,280
Speaker 3: factors that went into the model, and like the decisions

518
00:30:53,320 --> 00:30:56,800
Speaker 3: within the model that spat out a particular outcome, or

519
00:30:56,960 --> 00:31:00,200
Speaker 3: is it that the person who is using the ourg go,

520
00:31:00,800 --> 00:31:04,400
Speaker 3: you know, decides to think more thoughtfully about how they're

521
00:31:04,520 --> 00:31:04,920
Speaker 3: using it.

522
00:31:06,720 --> 00:31:09,560
Speaker 4: It's a that's a really interesting question which I'm going

523
00:31:09,640 --> 00:31:13,360
Speaker 4: to think for two seconds, delight and waffle before answering,

524
00:31:14,160 --> 00:31:19,000
Speaker 4: which is that I think the basic definition of accountability

525
00:31:19,040 --> 00:31:21,400
Speaker 4: in the decision sense, in my view, is that you're

526
00:31:21,440 --> 00:31:25,800
Speaker 4: accountable for a decision exactly to the extent that you

527
00:31:25,880 --> 00:31:28,600
Speaker 4: are able to change it. So, in terms of a

528
00:31:28,640 --> 00:31:33,680
Speaker 4: decision made by an AI, it is accountable if it

529
00:31:33,720 --> 00:31:37,800
Speaker 4: could be made to make a different decision by new

530
00:31:37,880 --> 00:31:41,800
Speaker 4: information being provided to it. So the crucial thing is

531
00:31:42,560 --> 00:31:45,800
Speaker 4: not so much having someone to point at and attribute

532
00:31:45,800 --> 00:31:49,760
Speaker 4: moral responsibility to. The crucial thing is to have some

533
00:31:50,280 --> 00:31:53,600
Speaker 4: link between the subject of the decision. Who can just say,

534
00:31:53,840 --> 00:31:57,760
Speaker 4: let's review this, let's have a court of appeal. And

535
00:31:58,280 --> 00:32:01,040
Speaker 4: you know, if the algo still thinks that this decision

536
00:32:01,080 --> 00:32:03,400
Speaker 4: needs to be done, if the elgo still thinks I'm

537
00:32:03,440 --> 00:32:07,240
Speaker 4: not an insurable risk, then maybe I've been Maybe I

538
00:32:07,240 --> 00:32:09,280
Speaker 4: still don't agree with that, but at least I know

539
00:32:09,480 --> 00:32:12,800
Speaker 4: I've been heard. It's not coming to me just simply

540
00:32:13,080 --> 00:32:15,120
Speaker 4: as a one way communication channel.

541
00:32:15,800 --> 00:32:20,680
Speaker 2: You know, making a more optimistic view. So you said

542
00:32:20,680 --> 00:32:25,360
Speaker 2: something interesting or important, which is that for a company,

543
00:32:25,200 --> 00:32:29,040
Speaker 2: the financial numbers are the closest thing, the closest form

544
00:32:29,080 --> 00:32:33,680
Speaker 2: of information that is at least like objective in some sense.

545
00:32:33,720 --> 00:32:36,280
Speaker 2: But then there's all these other things like how how

546
00:32:36,320 --> 00:32:39,680
Speaker 2: well is your engineering team working together? How is the

547
00:32:39,720 --> 00:32:43,160
Speaker 2: culture that are just inherently much more difficult. And they

548
00:32:43,200 --> 00:32:45,880
Speaker 2: are all kinds of like consultants and other companies that

549
00:32:45,920 --> 00:32:48,760
Speaker 2: try to like answer this for executives and you know,

550
00:32:48,880 --> 00:32:53,200
Speaker 2: rank employees different things. Aaron Levy, you know, he's the

551
00:32:53,240 --> 00:32:55,520
Speaker 2: CEO of Box and he's one of the few like

552
00:32:55,880 --> 00:32:58,960
Speaker 2: tech CEOs who tweets some interesting stuff from time to time.

553
00:32:59,160 --> 00:33:01,600
Speaker 2: But he said he had this recent thread about how

554
00:33:01,600 --> 00:33:03,560
Speaker 2: his company is using AI, and he said, you know,

555
00:33:03,600 --> 00:33:06,840
Speaker 2: the exciting thing is, you know, we have some data,

556
00:33:07,480 --> 00:33:09,880
Speaker 2: but then we have all this unstructured data that exists

557
00:33:09,880 --> 00:33:11,640
Speaker 2: in our company, and we've never been able to do

558
00:33:11,720 --> 00:33:15,160
Speaker 2: with anything, probably like chat logs from customer support and

559
00:33:15,200 --> 00:33:18,000
Speaker 2: all this. And he said, the exciting thing for them

560
00:33:18,200 --> 00:33:21,840
Speaker 2: is the prospect of turning all of this sort of unstructured,

561
00:33:22,000 --> 00:33:25,800
Speaker 2: unusable information that the company has into something that can

562
00:33:25,840 --> 00:33:30,720
Speaker 2: be essentially searchable and that insights can be gleaned from it.

563
00:33:30,760 --> 00:33:32,880
Speaker 2: Is there a story or is there a path in

564
00:33:32,920 --> 00:33:36,720
Speaker 2: your view where artificial intelligence can actually make some of

565
00:33:36,760 --> 00:33:41,400
Speaker 2: these other parts of a system more legible and more

566
00:33:41,440 --> 00:33:45,560
Speaker 2: interactive and more concrete to the executives in a way

567
00:33:45,600 --> 00:33:48,800
Speaker 2: that brings it those that other data on par with

568
00:33:48,880 --> 00:33:49,800
Speaker 2: the financial data.

569
00:33:50,200 --> 00:33:52,040
Speaker 4: I mean, I really hope. So, I mean, like kind

570
00:33:52,080 --> 00:33:53,880
Speaker 4: of one thing that Aaron Levy could do that would

571
00:33:53,880 --> 00:33:57,200
Speaker 4: be really radical would be to open up as much

572
00:33:57,200 --> 00:34:00,920
Speaker 4: of that data as possible to the investors and let

573
00:34:00,960 --> 00:34:04,120
Speaker 4: them pass it and have that as a main channel

574
00:34:04,160 --> 00:34:10,000
Speaker 4: of communication of corporate performance. Rather than generally accepted accounting principles,

575
00:34:10,040 --> 00:34:12,880
Speaker 4: because if you think about that phrase, that's an accountability

576
00:34:12,880 --> 00:34:17,040
Speaker 4: sink right there. What are these accounting principles? They're generally accepted.

577
00:34:17,960 --> 00:34:20,680
Speaker 4: Can I change them? No, that would not be generally accepted.

578
00:34:20,960 --> 00:34:24,160
Speaker 4: What if this is completely irrelevant set of metrics to

579
00:34:24,280 --> 00:34:27,399
Speaker 4: my business? Well, that you still have to do it

580
00:34:27,680 --> 00:34:31,480
Speaker 4: in exactly this way, even if it's not presenting what

581
00:34:31,520 --> 00:34:34,400
Speaker 4: you think is actually generating value. And in a world

582
00:34:34,480 --> 00:34:39,239
Speaker 4: where we've got better ways of processing bulk information like that,

583
00:34:39,840 --> 00:34:42,080
Speaker 4: then I think there's a real question about whether gap

584
00:34:42,520 --> 00:34:46,200
Speaker 4: is something that we should be so fixated on, whether

585
00:34:46,239 --> 00:34:49,280
Speaker 4: we should be thinking that the only way to report

586
00:34:49,680 --> 00:34:54,000
Speaker 4: corporate performance is in a way that's optimized. Basically, for

587
00:34:54,080 --> 00:34:56,720
Speaker 4: a guy with a greener eyeshade sitting at a desk

588
00:34:57,000 --> 00:35:01,000
Speaker 4: in the beginning of the twentieth century, flip through printed

589
00:35:01,480 --> 00:35:03,560
Speaker 4: reports and accounts, you know, that's not the way that

590
00:35:03,600 --> 00:35:06,799
Speaker 4: we process information anymore. And so maybe that shouldn't be

591
00:35:07,360 --> 00:35:11,960
Speaker 4: the way that we report information anymore, because, as you say,

592
00:35:12,880 --> 00:35:17,799
Speaker 4: these assumptions that go into gap earnings start driving decisions

593
00:35:18,000 --> 00:35:20,120
Speaker 4: and they were never meant to drive decisions.

594
00:35:20,840 --> 00:35:23,960
Speaker 2: It's interesting Tracy now thinking about it in this financial sense,

595
00:35:24,000 --> 00:35:28,640
Speaker 2: how many of these accountabilities sinks? Like even like performance benchmarks, right,

596
00:35:28,680 --> 00:35:30,560
Speaker 2: it's like, oh, we beat the S and P or whatever.

597
00:35:30,600 --> 00:35:33,080
Speaker 2: It's like, why the S and P etcetera. Well, you

598
00:35:33,120 --> 00:35:36,160
Speaker 2: know it's there, right, we could point to it and

599
00:35:36,200 --> 00:35:38,000
Speaker 2: we could say it. But like once you start thinking

600
00:35:38,000 --> 00:35:41,080
Speaker 2: of them in all of the indices and measures we cite,

601
00:35:41,080 --> 00:35:43,760
Speaker 2: you could see, Tracy, how they like serve that purpose

602
00:35:43,800 --> 00:35:45,880
Speaker 2: of just like yeah, look this is what we measure against.

603
00:35:45,920 --> 00:35:48,680
Speaker 3: Oh yeah, of course. I mean incentives matter, right, Like

604
00:35:48,760 --> 00:35:50,880
Speaker 3: that's something that we say over and over again on

605
00:35:50,920 --> 00:35:53,880
Speaker 3: this podcast and when it comes to accounting. Okay, just

606
00:35:53,920 --> 00:35:55,640
Speaker 3: to push back a little bit, but like there is

607
00:35:55,680 --> 00:35:59,480
Speaker 3: an argument to have standardized accounting rules so that we

608
00:35:59,560 --> 00:36:03,280
Speaker 3: don't always end up with companies running off and creating

609
00:36:03,320 --> 00:36:07,759
Speaker 3: community adjusted ebit dah and things like that. But on

610
00:36:07,800 --> 00:36:11,120
Speaker 3: the subject of incentives, I wanted to ask, you know,

611
00:36:11,160 --> 00:36:14,719
Speaker 3: I read David Graber's Bull Jobs this year and it's

612
00:36:14,760 --> 00:36:17,120
Speaker 3: still sort of looming large in my memory, and I

613
00:36:17,120 --> 00:36:21,160
Speaker 3: guess my question is how much overlap is there between

614
00:36:21,400 --> 00:36:25,480
Speaker 3: the accountability issues that you describe in your book and

615
00:36:25,960 --> 00:36:31,080
Speaker 3: the way specific jobs, especially middle management are structured. And

616
00:36:31,120 --> 00:36:32,719
Speaker 3: I don't mean to trigger you because I know that

617
00:36:32,760 --> 00:36:35,080
Speaker 3: you mentioned in the book that you got into it

618
00:36:35,120 --> 00:36:37,520
Speaker 3: a little bit with Graber over a different subject. But

619
00:36:37,560 --> 00:36:38,839
Speaker 3: if you want to talk about that too.

620
00:36:39,160 --> 00:36:41,560
Speaker 4: Oh, I missed David so much because we used to

621
00:36:41,560 --> 00:36:47,520
Speaker 4: wind each other up so badly, and I got into

622
00:36:47,560 --> 00:36:49,960
Speaker 4: a different argument that's not mentioned into the book over

623
00:36:50,000 --> 00:36:53,120
Speaker 4: bull jobs. Because it's just like, if you're saying that

624
00:36:53,400 --> 00:36:56,480
Speaker 4: middle management is a bulk job, then you're saying that

625
00:36:56,480 --> 00:37:01,760
Speaker 4: the Sarahbellum is a bull's organ The middle management exists

626
00:37:01,960 --> 00:37:06,239
Speaker 4: precisely because of all the metrics, and all the financial

627
00:37:06,280 --> 00:37:10,799
Speaker 4: and non financial metrics on the chief executive's dashboard are

628
00:37:11,560 --> 00:37:15,640
Speaker 4: massive information reducing filters. The middle managers are the people

629
00:37:15,640 --> 00:37:19,160
Speaker 4: who carry the knowledge of the ways in which those

630
00:37:19,200 --> 00:37:24,760
Speaker 4: metrics can misrepresent reality and how to cure the problems

631
00:37:24,760 --> 00:37:28,560
Speaker 4: which arise when they are. You know, when someone's in

632
00:37:28,640 --> 00:37:31,319
Speaker 4: danger of making a decision. The first thing a bad

633
00:37:31,360 --> 00:37:34,640
Speaker 4: company does before it creates something like the Libuil scandal

634
00:37:34,719 --> 00:37:36,920
Speaker 4: or the seven three seven mags is thin out the

635
00:37:37,000 --> 00:37:42,000
Speaker 4: ranks of its middle management. And this particularly without wanting

636
00:37:42,040 --> 00:37:45,520
Speaker 4: to relate the gate arguments with David Graeber, particularly since

637
00:37:45,520 --> 00:37:48,640
Speaker 4: he's not around any more to answer back. In his

638
00:37:48,920 --> 00:37:53,880
Speaker 4: day job as an anthropologist, David was so subtle and

639
00:37:54,000 --> 00:37:59,279
Speaker 4: intelligent about the ceremonial roles of elders and people who

640
00:37:59,440 --> 00:38:04,240
Speaker 4: built census among hunter gatherers to decide on what bands

641
00:38:04,239 --> 00:38:06,760
Speaker 4: they would do. And then when you have those exact

642
00:38:06,840 --> 00:38:11,360
Speaker 4: same problem solving and dispute resolving jobs happening, you know,

643
00:38:11,440 --> 00:38:14,280
Speaker 4: in the offices at Bloomberger at a law firm, suddenly

644
00:38:14,320 --> 00:38:19,239
Speaker 4: he thinks that they're bulk jobs. So that's a bit

645
00:38:19,239 --> 00:38:23,359
Speaker 4: of a personal hobby horse. But basically all those or

646
00:38:23,480 --> 00:38:27,279
Speaker 4: very many of those roles are actually the preservation of

647
00:38:27,320 --> 00:38:31,120
Speaker 4: the information systems and the memory of organizations.

648
00:38:32,160 --> 00:38:36,560
Speaker 2: Who is Stafford Beer and why why does he loom

649
00:38:36,640 --> 00:38:38,520
Speaker 2: so large in the story you tell about the world.

650
00:38:38,800 --> 00:38:42,279
Speaker 4: Well, Stafford Bit was the father of management sebernetics. He

651
00:38:42,440 --> 00:38:45,400
Speaker 4: was the guy who first said you can take the

652
00:38:45,640 --> 00:38:49,600
Speaker 4: mathematics of information theory and apply it to industrial organization.

653
00:38:50,400 --> 00:38:57,560
Speaker 4: He was also David Bowie's favorite management consultants. He was very,

654
00:38:57,680 --> 00:39:01,280
Speaker 4: very influential on Brian Eno and the divilment of ambient music.

655
00:39:02,080 --> 00:39:07,200
Speaker 4: He was a hippie. He tried to it's not clear

656
00:39:07,239 --> 00:39:09,399
Speaker 4: that this was a joke, but he did try to

657
00:39:09,440 --> 00:39:12,839
Speaker 4: invent a computing pond where the growth of algae would

658
00:39:12,880 --> 00:39:15,280
Speaker 4: correspond to the solutions of differential equations.

659
00:39:15,760 --> 00:39:18,000
Speaker 3: He's going to be my summer project. I got a

660
00:39:18,040 --> 00:39:19,120
Speaker 3: pond with algae.

661
00:39:19,360 --> 00:39:21,400
Speaker 4: Yeah. The problem was that he used to feed them

662
00:39:21,400 --> 00:39:23,839
Speaker 4: iron filings to make them grow, and after a while

663
00:39:23,880 --> 00:39:27,520
Speaker 4: the entire pond became magnetic. But he was just this crazy,

664
00:39:27,840 --> 00:39:34,240
Speaker 4: larger than life figure who did these incredibly successful management

665
00:39:34,280 --> 00:39:39,920
Speaker 4: consulting assignments, but just somehow never quite was able to

666
00:39:39,960 --> 00:39:43,440
Speaker 4: get on with enough people in the corporate world to

667
00:39:43,560 --> 00:39:46,799
Speaker 4: really get his ideas across. And then he ended up

668
00:39:46,880 --> 00:39:52,759
Speaker 4: in Chile in nineteen seventy two with this incredibly romantic

669
00:39:53,200 --> 00:39:59,040
Speaker 4: but ultimately doomed project to reinvent socialism for the twenty

670
00:39:59,080 --> 00:40:04,480
Speaker 4: first century under the ide governments, which I mean realistically

671
00:40:04,520 --> 00:40:06,960
Speaker 4: it was never going to work because the confuting resources

672
00:40:07,000 --> 00:40:10,000
Speaker 4: were completely disproportionate to the task. But it never got

673
00:40:10,040 --> 00:40:13,080
Speaker 4: a fair trial, obviously, because the Pinochet coup happened about

674
00:40:13,120 --> 00:40:15,040
Speaker 4: eleven months after he started the project.

675
00:40:15,640 --> 00:40:17,880
Speaker 2: So why do we talk a little bit more about

676
00:40:17,920 --> 00:40:21,799
Speaker 2: like the current day? You know, I started in the

677
00:40:22,520 --> 00:40:27,120
Speaker 2: conversation like kind of not disagreeing, but questioning the premise,

678
00:40:27,239 --> 00:40:30,080
Speaker 2: like the world has lost its Has the world really

679
00:40:30,120 --> 00:40:32,480
Speaker 2: lost its mind? Or just the three of us in

680
00:40:32,520 --> 00:40:36,560
Speaker 2: this conversation getting old and angry? What do you see

681
00:40:36,600 --> 00:40:38,360
Speaker 2: like when you like think about like applying. You know,

682
00:40:38,400 --> 00:40:41,720
Speaker 2: we talked about Boeing and you mentioned the live war scandal.

683
00:40:41,760 --> 00:40:43,840
Speaker 2: But how are these things when you look around the

684
00:40:43,840 --> 00:40:46,880
Speaker 2: world today and you look around whatever apparent world losing

685
00:40:46,920 --> 00:40:48,120
Speaker 2: its mind? What are you seeing?

686
00:40:49,000 --> 00:40:51,040
Speaker 4: I think the number one thing I'm seeing, And this

687
00:40:51,200 --> 00:40:53,719
Speaker 4: is a point where David Graeber had it absolutely right,

688
00:40:54,520 --> 00:40:58,560
Speaker 4: is debt. You know, we have so many cases at

689
00:40:58,600 --> 00:41:03,960
Speaker 4: presence of companies where you have plenty of people who

690
00:41:04,040 --> 00:41:08,000
Speaker 4: know exactly what they need to do, what investments they

691
00:41:08,120 --> 00:41:12,560
Speaker 4: want to make, but they can't have any plans which

692
00:41:12,600 --> 00:41:16,560
Speaker 4: stretch out any further than the next debt repayment. And

693
00:41:16,760 --> 00:41:21,479
Speaker 4: to a large extent, that's because of leverage buyouts and

694
00:41:22,000 --> 00:41:26,000
Speaker 4: managements acting in anticipation of the risk of a leverage buyout.

695
00:41:26,560 --> 00:41:32,960
Speaker 4: But this really is a degradation of the higher functions,

696
00:41:33,000 --> 00:41:38,080
Speaker 4: the brain functions of the corporations of the anglosphere world.

697
00:41:38,440 --> 00:41:42,440
Speaker 4: The practice of firing middle managers because you don't know

698
00:41:42,480 --> 00:41:46,760
Speaker 4: what they do is also demonstrably, in my view, making

699
00:41:46,840 --> 00:41:51,080
Speaker 4: corporations stupid. It might have been that at the start

700
00:41:51,120 --> 00:41:53,680
Speaker 4: of the LBO boom in the seventies there were too

701
00:41:53,719 --> 00:41:56,759
Speaker 4: many of these guys on soft jobs with country club

702
00:41:56,800 --> 00:41:59,920
Speaker 4: memberships and private jets and whatnot, but we've clearly gone

703
00:41:59,920 --> 00:42:02,600
Speaker 4: too far in the other direction in my view. And

704
00:42:02,719 --> 00:42:09,400
Speaker 4: then we've got the frightening tendency of government organizations to

705
00:42:09,719 --> 00:42:14,920
Speaker 4: outsource absolutely critical functions, which means that all of the

706
00:42:15,000 --> 00:42:18,799
Speaker 4: knowledge of the systems that they're meant to be regulating

707
00:42:19,120 --> 00:42:24,839
Speaker 4: and dealing with. What's an example of that, right, best

708
00:42:24,880 --> 00:42:31,439
Speaker 4: example currently thinking, I think m just the question of

709
00:42:31,640 --> 00:42:35,480
Speaker 4: infrastructure building, for example in the UK, where we have

710
00:42:36,560 --> 00:42:40,480
Speaker 4: there's a river crossing in to the east of London

711
00:42:41,000 --> 00:42:45,239
Speaker 4: which is being responsible for generating the largest pile of

712
00:42:45,400 --> 00:42:48,719
Speaker 4: paper ever brought together in one place in the history

713
00:42:48,719 --> 00:42:54,000
Speaker 4: of humanity. And the reason for that is that the

714
00:42:54,040 --> 00:42:57,400
Speaker 4: people who are meant to be deciding what an appropriate

715
00:42:57,480 --> 00:43:03,720
Speaker 4: level of consultation is don't know anything about environmental impact

716
00:43:03,800 --> 00:43:08,080
Speaker 4: studies and building bridges anymore. So they commission results reports,

717
00:43:08,400 --> 00:43:13,600
Speaker 4: the commission reports from professional services firms and professional services

718
00:43:13,640 --> 00:43:17,839
Speaker 4: firms want to generate repeat business, and so you've got

719
00:43:17,840 --> 00:43:21,000
Speaker 4: this situation where the people who are meant to be

720
00:43:21,080 --> 00:43:25,640
Speaker 4: seeing the whole system as a system don't really understand

721
00:43:25,680 --> 00:43:29,440
Speaker 4: it anymore because they've outsourced all of their engineering knowledge,

722
00:43:29,480 --> 00:43:33,120
Speaker 4: all of their economic and environmental knowledge. And as a result,

723
00:43:33,800 --> 00:43:38,600
Speaker 4: the UK or the London Department for Transports are going

724
00:43:38,640 --> 00:43:41,120
Speaker 4: to be paying as much as ten times as much

725
00:43:41,400 --> 00:43:44,399
Speaker 4: as it should reasonably cost to build a bridge over

726
00:43:44,400 --> 00:43:45,080
Speaker 4: the River Thames.

727
00:43:45,920 --> 00:43:49,279
Speaker 3: It's interesting that you single out debt as a sort

728
00:43:49,280 --> 00:43:52,719
Speaker 3: of deciding factor versus share price, because this is the

729
00:43:52,800 --> 00:43:55,160
Speaker 3: one we hear a lot about in the context of

730
00:43:55,560 --> 00:44:00,680
Speaker 3: corporate short termism and people usually trying to hit specific

731
00:44:00,840 --> 00:44:03,600
Speaker 3: share price metric that may or may not be tied

732
00:44:03,640 --> 00:44:04,640
Speaker 3: into their compensation.

733
00:44:05,880 --> 00:44:09,160
Speaker 4: Yeah. I think I'm right on that. I know that

734
00:44:09,160 --> 00:44:11,719
Speaker 4: people disagree with me, but you know, we had share

735
00:44:11,760 --> 00:44:15,040
Speaker 4: prices in the fifties and sixties and we didn't have

736
00:44:15,120 --> 00:44:18,680
Speaker 4: this kind of problem of short termism. The difference is

737
00:44:19,160 --> 00:44:25,120
Speaker 4: that now we've got takeovers, particularly private equity and LBOs,

738
00:44:25,200 --> 00:44:28,840
Speaker 4: but also in general the use of debt in takeovers,

739
00:44:29,440 --> 00:44:32,799
Speaker 4: and that makes the share price more salient because any

740
00:44:32,840 --> 00:44:36,200
Speaker 4: incumbent management knows that if the share price falls, it

741
00:44:36,320 --> 00:44:41,320
Speaker 4: makes them vulnerable to a takeover. So to my mind,

742
00:44:41,880 --> 00:44:44,240
Speaker 4: I think it's not so much you know, the share

743
00:44:44,280 --> 00:44:50,360
Speaker 4: price as the exaggerated importance of short term financial metrics,

744
00:44:50,840 --> 00:44:53,640
Speaker 4: which is partly through the share price, but much more

745
00:44:54,120 --> 00:44:55,560
Speaker 4: just simply because of leverage.

746
00:44:56,239 --> 00:44:58,680
Speaker 2: One of the things that you know, you mentioned the

747
00:44:58,800 --> 00:45:01,680
Speaker 2: UK bridge and it ten times more expensive then needs

748
00:45:01,680 --> 00:45:03,279
Speaker 2: to be in a pile of paperwork. I mean, this

749
00:45:03,480 --> 00:45:06,759
Speaker 2: is probably the number one thing that like, you know,

750
00:45:06,960 --> 00:45:09,879
Speaker 2: people I follow on Twitter talk about all the time,

751
00:45:09,920 --> 00:45:12,480
Speaker 2: which is just how hard it is to build anything

752
00:45:12,600 --> 00:45:16,200
Speaker 2: in the United States and the interlocking systems of environmental

753
00:45:16,239 --> 00:45:20,000
Speaker 2: regulations and nimbi's and everything else, and it's the big

754
00:45:20,080 --> 00:45:22,759
Speaker 2: challenge of the IRA and it feels like listening to this,

755
00:45:23,239 --> 00:45:27,040
Speaker 2: it's just accountability sink. After accountability sink is to blame.

756
00:45:27,160 --> 00:45:31,080
Speaker 4: It's absolutely and it's accountability sinks being put up because

757
00:45:31,120 --> 00:45:34,360
Speaker 4: the people who were meant to be taking the decisions,

758
00:45:34,680 --> 00:45:37,480
Speaker 4: and who in the nineteen fifties and sixties did take

759
00:45:37,520 --> 00:45:41,279
Speaker 4: the decisions, are no longer really able to They've not

760
00:45:41,400 --> 00:45:44,200
Speaker 4: got the confidence of their decisions. They're not sure they're

761
00:45:44,239 --> 00:45:46,600
Speaker 4: going to be able to defend them in litigation, and

762
00:45:46,640 --> 00:45:51,880
Speaker 4: it's mainly because they've lost their executive functions with successive

763
00:45:51,920 --> 00:45:55,480
Speaker 4: staff cuts and retirements and outsourcing contracts.

764
00:45:55,880 --> 00:45:58,000
Speaker 2: Dan Davies, It's so great to catch up with you.

765
00:45:58,080 --> 00:46:01,560
Speaker 2: It's fascinating conversation, fascination way of thinking about the world.

766
00:46:01,920 --> 00:46:05,520
Speaker 2: Highly recommend everyone check out your book, The Unaccountability Machine.

767
00:46:06,000 --> 00:46:07,680
Speaker 2: Thank you so much for coming back on outlag.

768
00:46:08,200 --> 00:46:09,319
Speaker 4: Oh, thanks to so much.

769
00:46:09,440 --> 00:46:26,200
Speaker 2: Oh is a pleasure, Tracy, I really I love talking

770
00:46:26,200 --> 00:46:27,960
Speaker 2: to Dan. First of all, it's always I just like

771
00:46:28,000 --> 00:46:30,480
Speaker 2: hearing his voice. You know this, I do think like

772
00:46:30,560 --> 00:46:33,120
Speaker 2: accountability sinks is now going to be one of those

773
00:46:33,200 --> 00:46:35,840
Speaker 2: phrases that I'm just going to now start seeing everywhere.

774
00:46:36,040 --> 00:46:38,279
Speaker 2: And you know there's whole industry like McKinsey right, Like

775
00:46:38,360 --> 00:46:40,239
Speaker 2: that's like, you know, it's like again someone else to

776
00:46:40,560 --> 00:46:42,560
Speaker 2: lay off your workers, et cetera. Like you just start

777
00:46:42,600 --> 00:46:44,240
Speaker 2: seeing how big that is everywhere.

778
00:46:44,280 --> 00:46:46,279
Speaker 3: Well, this is what I was thinking. You know, you

779
00:46:46,320 --> 00:46:49,399
Speaker 3: brought in the US example of building infrastructure or other

780
00:46:49,520 --> 00:46:52,520
Speaker 3: energy products, and I kept thinking back to Jiggershaw and

781
00:46:52,600 --> 00:46:56,920
Speaker 3: his point about the lack of institutional memory of how

782
00:46:56,960 --> 00:47:00,640
Speaker 3: to build nuclear power plants. Right, it's not necessarily that

783
00:47:01,239 --> 00:47:05,359
Speaker 3: it's so complicated to get environmental permits and things like that,

784
00:47:05,400 --> 00:47:08,080
Speaker 3: although that is certainly part of it. But it's also

785
00:47:08,239 --> 00:47:11,640
Speaker 3: that the people who used to do this haven't been

786
00:47:11,680 --> 00:47:15,000
Speaker 3: doing it for a long time or are no longer around.

787
00:47:15,080 --> 00:47:18,040
Speaker 3: And that kind of goes to Down's point about middle

788
00:47:18,080 --> 00:47:22,600
Speaker 3: management being the sort of like what's the word I'm.

789
00:47:22,520 --> 00:47:23,840
Speaker 2: Thinking connective tissue.

790
00:47:24,080 --> 00:47:27,120
Speaker 3: Connective tissue is a good one of like institutional memory.

791
00:47:27,280 --> 00:47:30,120
Speaker 2: Yeah, no, it's only true, you know, in defense of

792
00:47:30,239 --> 00:47:33,440
Speaker 2: the the Nimbi's. You know, I keep mentioning this New

793
00:47:33,520 --> 00:47:35,600
Speaker 2: York documentary and watching and I got to the yeah,

794
00:47:35,600 --> 00:47:36,880
Speaker 2: I want to watch that, you gotta watch it. But

795
00:47:36,920 --> 00:47:39,359
Speaker 2: I got to the episode where like really talks about

796
00:47:39,400 --> 00:47:42,520
Speaker 2: like Robert Moses and just like plowing these big highways

797
00:47:42,520 --> 00:47:45,960
Speaker 2: through neighborhoods and putting up these like terrible, like terrible

798
00:47:46,080 --> 00:47:48,000
Speaker 2: housing projects that are like, you know.

799
00:47:48,120 --> 00:47:52,080
Speaker 3: Sort of right, continuously prioritizing highways.

800
00:47:52,239 --> 00:47:55,360
Speaker 2: That is someone who did not have the problem of

801
00:47:56,080 --> 00:47:59,960
Speaker 2: Nimbi's or a million different interlocking constraints on him, Like

802
00:48:00,080 --> 00:48:02,799
Speaker 2: there are our drawbacks to when someone has like too

803
00:48:02,920 --> 00:48:06,960
Speaker 2: much autonomoy and it's sort of like, yeah, but now

804
00:48:07,000 --> 00:48:09,520
Speaker 2: it does seem arguably we've gone too far in the

805
00:48:09,560 --> 00:48:13,000
Speaker 2: other direction, in which everyone just clings to their accountability

806
00:48:13,000 --> 00:48:14,520
Speaker 2: sink and can't get it right done.

807
00:48:14,560 --> 00:48:20,040
Speaker 3: Everything is a collective decision therefore can be held responsible. Yeah,

808
00:48:20,040 --> 00:48:23,640
Speaker 3: I feel like there must be a reasonable middle ground.

809
00:48:24,200 --> 00:48:26,560
Speaker 3: And yet I don't know. I'm trying to think if

810
00:48:26,560 --> 00:48:30,040
Speaker 3: I know of, like any organizations that have completely cracked

811
00:48:30,320 --> 00:48:33,799
Speaker 3: like the nut just for a little while. Yeah, collectivism

812
00:48:33,880 --> 00:48:38,000
Speaker 3: versus individual responsibility, I don't know. Well, on that note,

813
00:48:38,000 --> 00:48:38,759
Speaker 3: shall we leave it there?

814
00:48:38,840 --> 00:48:39,520
Speaker 2: Let's leave it there.

815
00:48:39,640 --> 00:48:42,320
Speaker 3: This has been another episode of the All Thoughts podcast.

816
00:48:42,360 --> 00:48:45,640
Speaker 3: I'm Tracy Alloway. You can follow me at Tracy Alloway and.

817
00:48:45,560 --> 00:48:48,280
Speaker 2: I'm Joe Wisenthal. You can follow me at The Stalwart.

818
00:48:48,600 --> 00:48:51,200
Speaker 2: Follow our guest Dan Davies. He's the author of the

819
00:48:51,200 --> 00:48:55,000
Speaker 2: book The Unaccountability Machine, Why Big Systems Make Terrible Decisions

820
00:48:55,000 --> 00:48:56,319
Speaker 2: and How the World Lost its Mind.

821
00:48:56,360 --> 00:48:57,040
Speaker 3: Go check it out.

822
00:48:57,239 --> 00:49:01,200
Speaker 2: His handle is at d squared Digest. Follow our producers

823
00:49:01,239 --> 00:49:04,640
Speaker 2: Carmen Rodriguez at Carmen Arman, dash Ol Bennett at Dashbot

824
00:49:04,680 --> 00:49:08,280
Speaker 2: and Kilbrooks at Kilbrooks. Thank you to our producer Moses Ondem.

825
00:49:08,560 --> 00:49:11,160
Speaker 2: For more Oddlags content, go to Bloomberg dot com slash

826
00:49:11,200 --> 00:49:13,279
Speaker 2: odd Lots, where we have transcripts, a blog, and a

827
00:49:13,320 --> 00:49:15,720
Speaker 2: newsletter and you can chat about all of these topics

828
00:49:15,760 --> 00:49:18,920
Speaker 2: twenty four to seven in our discord Discord dot gg

829
00:49:19,080 --> 00:49:20,839
Speaker 2: slash odd lots.

830
00:49:20,480 --> 00:49:24,240
Speaker 3: And speaking of personal accountability. If you like odd Lots,

831
00:49:24,360 --> 00:49:28,160
Speaker 3: please leave us a positive review on your favorite podcast platform.

832
00:49:28,360 --> 00:49:31,080
Speaker 3: And remember, if you are a Bloomberg subscriber, you can

833
00:49:31,160 --> 00:49:34,399
Speaker 3: listen to all of our episodes absolutely ad free. All

834
00:49:34,440 --> 00:49:36,840
Speaker 3: you need to do is connect your Bloomberg account with

835
00:49:36,960 --> 00:49:39,680
Speaker 3: Apple Podcasts. In order to do that, you can find

836
00:49:39,719 --> 00:49:43,359
Speaker 3: the Bloomberg channel on Apple Podcasts and follow the instructions there.

837
00:49:43,640 --> 00:49:44,480
Speaker 3: Thanks for listening.