1
00:00:06,080 --> 00:00:08,600
Speaker 1: Hey, you welcome to Stuff to Blow Your Mind. My

2
00:00:08,720 --> 00:00:11,880
Speaker 1: name is Robert Lamb. Today is Saturday, so we have

3
00:00:11,920 --> 00:00:14,680
Speaker 1: a vault episode for you. This is going to be

4
00:00:14,680 --> 00:00:18,840
Speaker 1: an episode titled The Edge of Sentience with Jonathan Birch.

5
00:00:19,520 --> 00:00:24,000
Speaker 1: I recently referenced this interview episode in some of our

6
00:00:24,360 --> 00:00:29,040
Speaker 1: recent Stuff to Blow Your Mind episodes, or yes, actually

7
00:00:29,080 --> 00:00:32,280
Speaker 1: it was the episode we did for Halloween the Grimore

8
00:00:32,360 --> 00:00:35,240
Speaker 1: of Her Volume two, where we were talking in part

9
00:00:35,640 --> 00:00:39,280
Speaker 1: about the classic dark sci fi story I have no

10
00:00:39,360 --> 00:00:42,960
Speaker 1: mouth and I must scream. So here's the actual interview.

11
00:00:43,280 --> 00:00:45,320
Speaker 1: I do want to stress though, that this came out

12
00:00:45,640 --> 00:00:49,280
Speaker 1: at eleven twelve, twenty twenty four, which, yes, was just

13
00:00:49,720 --> 00:00:53,240
Speaker 1: barely a year ago. But at the same time, we're

14
00:00:53,240 --> 00:00:57,960
Speaker 1: talking in part in this episode about AI and some

15
00:00:58,120 --> 00:01:02,000
Speaker 1: technology that has continued to move very quickly. So just

16
00:01:02,080 --> 00:01:04,920
Speaker 1: keep that in mind when revisiting this episode, that a

17
00:01:05,000 --> 00:01:09,960
Speaker 1: year of technological time has passed since it's airing. All right,

18
00:01:10,000 --> 00:01:12,120
Speaker 1: let's jump right in.

19
00:01:14,280 --> 00:01:18,040
Speaker 2: Welcome to Stuff to Blow Your Mind production of iHeartRadio.

20
00:01:24,319 --> 00:01:26,960
Speaker 1: Hey you welcome to Stuff to Blow Your Mind. This

21
00:01:27,000 --> 00:01:29,960
Speaker 1: is Robert Lamb, and today I'm going to be chatting

22
00:01:30,480 --> 00:01:35,160
Speaker 1: with Jonathan Birch about his new book, The Edge of Sentience,

23
00:01:35,360 --> 00:01:39,600
Speaker 1: Risk and Precaution in Humans, Other Animals and AI. It

24
00:01:39,680 --> 00:01:43,200
Speaker 1: comes out later this week on November fifteenth in the US.

25
00:01:43,920 --> 00:01:46,920
Speaker 1: Jonathan Birch is a professor of philosophy at the London

26
00:01:46,959 --> 00:01:50,520
Speaker 1: School of Economics and principal investigator on the Foundations of

27
00:01:50,560 --> 00:01:55,000
Speaker 1: Animal Sentience Project, a European Union funded project aiming to

28
00:01:55,040 --> 00:01:58,560
Speaker 1: develop better methods for studying the feelings of animals and

29
00:01:58,680 --> 00:02:01,360
Speaker 1: new ways of using the soe ccients of animals' minds

30
00:02:01,680 --> 00:02:05,800
Speaker 1: to improve animal welfare policies and laws. In twenty twenty one,

31
00:02:05,920 --> 00:02:08,680
Speaker 1: he led a review for the UK government that shaped

32
00:02:08,720 --> 00:02:12,959
Speaker 1: the Animal Welfare Sentience Act twenty twenty two. In twenty

33
00:02:13,000 --> 00:02:15,079
Speaker 1: twenty two through twenty twenty three, he was part of

34
00:02:15,120 --> 00:02:19,359
Speaker 1: a working group that investigated the question of sentience in AI.

35
00:02:19,919 --> 00:02:23,320
Speaker 1: So I'll definitely be asking him about animals, about AI,

36
00:02:23,600 --> 00:02:26,640
Speaker 1: and maybe a few surprises here. So without further ado,

37
00:02:26,880 --> 00:02:30,000
Speaker 1: let's jump right into the interview. Thank you, Thank you,

38
00:02:32,000 --> 00:02:33,400
Speaker 1: Hi Jonathan. Welcome to the show.

39
00:02:33,600 --> 00:02:35,080
Speaker 3: Hi Robert, thanks for advising me.

40
00:02:35,600 --> 00:02:38,560
Speaker 1: So the new book is The Edge of Sentience. But

41
00:02:38,680 --> 00:02:41,600
Speaker 1: before we get to that edge and start talking about that,

42
00:02:42,480 --> 00:02:46,240
Speaker 1: how do you define sentience in your work? And what

43
00:02:46,360 --> 00:02:50,880
Speaker 1: are the implications and challenges of agreeing on a working definition.

44
00:02:51,320 --> 00:02:54,160
Speaker 3: Well, so see why I think if sentience is a

45
00:02:54,160 --> 00:02:57,680
Speaker 3: really useful concept. Let's start by thinking about pain that

46
00:02:57,800 --> 00:03:00,440
Speaker 3: I think a lot of us have wondered. Can an

47
00:03:00,440 --> 00:03:05,320
Speaker 3: octopus feel pain? Can insects feel pain? Can things hurt?

48
00:03:05,480 --> 00:03:08,959
Speaker 3: Can they have that feeling of ouch? And this is

49
00:03:09,000 --> 00:03:10,959
Speaker 3: a great question, but I think it's a bit too

50
00:03:11,040 --> 00:03:13,959
Speaker 3: narrow because we need to be aware of the fact

51
00:03:14,000 --> 00:03:18,680
Speaker 3: that other animals might have very different mental lives from us,

52
00:03:19,320 --> 00:03:22,799
Speaker 3: and words like pain they might be a bit narrow

53
00:03:22,840 --> 00:03:26,400
Speaker 3: for thinking about what the experiences of other animals are like.

54
00:03:26,919 --> 00:03:29,200
Speaker 3: So it's good to have concepts that are a bit

55
00:03:29,280 --> 00:03:32,440
Speaker 3: broader than the concept of pain, and to have a

56
00:03:32,480 --> 00:03:40,280
Speaker 3: concept that includes other negative feelings like frustration, discomfort, but

57
00:03:40,360 --> 00:03:43,480
Speaker 3: also the positive side of mental life as well, because

58
00:03:43,640 --> 00:03:49,480
Speaker 3: we also care about this. We care about states like joy, excitement, comfort,

59
00:03:50,200 --> 00:03:54,600
Speaker 3: pleasant bodily sensations like the feeling of warmth, and we

60
00:03:54,640 --> 00:03:56,720
Speaker 3: want a concept that is broad enough to include all

61
00:03:56,760 --> 00:03:59,200
Speaker 3: of this, all the negative side and the positive side

62
00:03:59,200 --> 00:04:02,600
Speaker 3: of mental life. Well, any feelings that feel bad or

63
00:04:02,680 --> 00:04:05,320
Speaker 3: feel good, and this is what the concept of sentience

64
00:04:05,400 --> 00:04:08,080
Speaker 3: is about. The capacities have feelings that feel good or

65
00:04:08,160 --> 00:04:08,680
Speaker 3: feel bad.

66
00:04:09,720 --> 00:04:12,960
Speaker 1: Now to what extent is this a different concept from

67
00:04:13,000 --> 00:04:15,920
Speaker 1: consciousness or where do they overlap and where do they differ.

68
00:04:16,920 --> 00:04:19,320
Speaker 3: The problem I have with the concept of consciousness is

69
00:04:19,320 --> 00:04:22,880
Speaker 3: that it can refer to many different things. Sentience isn't

70
00:04:22,880 --> 00:04:25,039
Speaker 3: perfect in that way, but I think it's a bit

71
00:04:25,120 --> 00:04:30,320
Speaker 3: more tightly defined than consciousness because when we talk about consciousness,

72
00:04:30,320 --> 00:04:36,359
Speaker 3: sometimes we're just talking about immediate, raw sensation, what it

73
00:04:36,400 --> 00:04:40,839
Speaker 3: feels like to be me right now, the sites, the sounds,

74
00:04:40,960 --> 00:04:46,320
Speaker 3: the odors, the bodily sensations, the pains, the pleasures, and

75
00:04:46,360 --> 00:04:49,880
Speaker 3: so on, and that's quite closely related to sentience. But

76
00:04:50,000 --> 00:04:53,200
Speaker 3: sometimes when we're talking about consciousness, we're talking about things

77
00:04:53,200 --> 00:04:56,520
Speaker 3: that are overlaid on top of that, like our ability

78
00:04:56,560 --> 00:05:00,240
Speaker 3: to reflect on what we're currently feeling, and and our

79
00:05:00,279 --> 00:05:04,200
Speaker 3: sense of self, our sense that my current immediate raw

80
00:05:04,279 --> 00:05:08,080
Speaker 3: experiences are not happening in isolation, but they're part of

81
00:05:08,240 --> 00:05:12,720
Speaker 3: a life that extends back in time and extends forwards

82
00:05:12,720 --> 00:05:16,200
Speaker 3: into the future. And I'm self aware, I'm aware of

83
00:05:16,240 --> 00:05:19,800
Speaker 3: myself as existing in time, and these things are much

84
00:05:19,839 --> 00:05:25,080
Speaker 3: more sophisticated than just having those immediate raw sensations. So

85
00:05:25,080 --> 00:05:27,080
Speaker 3: it's very useful to have a term that draws our

86
00:05:27,600 --> 00:05:31,520
Speaker 3: attention to those immediate sensations, and that's what sentience does.

87
00:05:32,240 --> 00:05:34,359
Speaker 1: Now. I realize this is of course a huge question

88
00:05:34,400 --> 00:05:37,360
Speaker 1: that you take on the book, and I'm not going

89
00:05:37,400 --> 00:05:40,560
Speaker 1: to ask you to regurgitate all of it for us here,

90
00:05:40,600 --> 00:05:45,760
Speaker 1: But where are the least controversial divides between Synthians and

91
00:05:45,880 --> 00:05:49,080
Speaker 1: non sentience in the animal kingdom and where does it

92
00:05:49,120 --> 00:05:51,560
Speaker 1: become more messy or controversial.

93
00:05:52,360 --> 00:05:57,159
Speaker 3: I think it's become very uncontroversial in the last few

94
00:05:57,200 --> 00:06:01,039
Speaker 3: decades to think of all other mammals as being sentient beings.

95
00:06:01,560 --> 00:06:04,960
Speaker 3: And that's a huge change because as a long tradition

96
00:06:05,000 --> 00:06:08,680
Speaker 3: of skepticism in science, going back to Rene Descartes in

97
00:06:08,760 --> 00:06:13,280
Speaker 3: the seventeenth century, but also the so called behaviorists in

98
00:06:13,360 --> 00:06:17,200
Speaker 3: the early twentieth century, who said you should never be

99
00:06:17,279 --> 00:06:20,720
Speaker 3: talking about consciousness or sentience of any kind in relation

100
00:06:20,920 --> 00:06:24,200
Speaker 3: to other animals, and that view has just fallen away.

101
00:06:24,200 --> 00:06:26,960
Speaker 3: And I think it's good that it's fallen away, because

102
00:06:27,320 --> 00:06:30,320
Speaker 3: I think it is pretty obvious that our cats are dogs,

103
00:06:30,800 --> 00:06:33,719
Speaker 3: and in fact, all other mammals like pigs, cows, et cetera.

104
00:06:34,480 --> 00:06:39,320
Speaker 3: They do have feelings. And then because of this new consensus,

105
00:06:39,920 --> 00:06:44,280
Speaker 3: the debate, the controversy has started to be more around fishes,

106
00:06:44,600 --> 00:06:49,720
Speaker 3: where you get some genuine doubters, and particularly invertebrates, where

107
00:06:49,760 --> 00:06:52,760
Speaker 3: we move from animals with a backbone to animals without.

108
00:06:52,839 --> 00:06:56,039
Speaker 3: And we're looking at animals separated from us in time

109
00:06:56,120 --> 00:06:59,800
Speaker 3: by over five hundred million years of evolution. They optopus,

110
00:06:59,839 --> 00:07:04,960
Speaker 3: in crabs, lobsters, insects. Here, I think doubts are very common,

111
00:07:05,640 --> 00:07:09,560
Speaker 3: and it's entirely reasonable to think maybe not all invertebrate

112
00:07:09,560 --> 00:07:12,200
Speaker 3: animals are sentient. And there's a lot of debate around that.

113
00:07:13,840 --> 00:07:16,800
Speaker 1: And you mentioned the octopus already being an example of

114
00:07:16,800 --> 00:07:20,040
Speaker 1: a very comblex creature that of course is very distant

115
00:07:20,080 --> 00:07:24,440
Speaker 1: from us. And yeah, how do we line that up

116
00:07:24,600 --> 00:07:27,360
Speaker 1: with this idea of sentience, and then how do we

117
00:07:27,720 --> 00:07:29,800
Speaker 1: keep from comparing it, trying to compare it too much

118
00:07:29,840 --> 00:07:30,560
Speaker 1: to what we have?

119
00:07:30,680 --> 00:07:34,360
Speaker 3: And I guess to consciousness, right, Yeah, sentience is a

120
00:07:34,360 --> 00:07:39,400
Speaker 3: good word for pushing us away from anthroperscentrism, and away

121
00:07:39,400 --> 00:07:44,280
Speaker 3: from this assumption that animals have exactly the same feelings

122
00:07:44,320 --> 00:07:49,120
Speaker 3: we do. So does an octopus have pain, Well, it's

123
00:07:49,240 --> 00:07:52,960
Speaker 3: probably not feeling it in the same way that we would.

124
00:07:53,400 --> 00:07:55,360
Speaker 3: It's going to be a state that feels very different

125
00:07:55,400 --> 00:07:56,280
Speaker 3: to the octopus.

126
00:07:56,320 --> 00:07:56,720
Speaker 1: I think.

127
00:07:57,920 --> 00:08:00,760
Speaker 3: Is the octopus sentient? Well, yes, I think so. The

128
00:08:00,800 --> 00:08:04,440
Speaker 3: sentience concept is broad enough to just capture the whole

129
00:08:04,560 --> 00:08:07,480
Speaker 3: range of animal mental life, soever much they may vary.

130
00:08:08,360 --> 00:08:11,320
Speaker 1: Now, when it comes to a moral obligation to sentient

131
00:08:11,400 --> 00:08:14,600
Speaker 1: life forms, where I guess and I realized this is

132
00:08:14,600 --> 00:08:16,840
Speaker 1: asking a question where ultimately every's going to be a

133
00:08:16,880 --> 00:08:21,160
Speaker 1: lot of different cultural differences and so forth. But where

134
00:08:21,200 --> 00:08:23,559
Speaker 1: where are we generally with the idea of our moral

135
00:08:23,600 --> 00:08:27,920
Speaker 1: obligation to sentient life and where are we looking to

136
00:08:27,920 --> 00:08:30,000
Speaker 1: go with it? Or where does what's the trajectory seem

137
00:08:30,000 --> 00:08:30,200
Speaker 1: to be?

138
00:08:31,480 --> 00:08:33,640
Speaker 3: Again, I think there's been a sea change on this

139
00:08:34,400 --> 00:08:37,520
Speaker 3: in recent decades. I think opinions are changing, and they're

140
00:08:37,600 --> 00:08:41,240
Speaker 3: changing fast, and I think changing in a direction I

141
00:08:41,280 --> 00:08:45,720
Speaker 3: find encouraging because it wasn't that long ago. You'd often

142
00:08:45,760 --> 00:08:49,360
Speaker 3: get people denying the idea that other animals can make

143
00:08:49,400 --> 00:08:53,040
Speaker 3: any moral claim on us at all. People would say

144
00:08:53,080 --> 00:08:58,040
Speaker 3: morality is about humans, it's human interests, human rights. Other

145
00:08:58,120 --> 00:09:01,000
Speaker 3: animals are not part of it, and very few people

146
00:09:01,080 --> 00:09:04,080
Speaker 3: argue that now, because I think once you recognize other

147
00:09:04,120 --> 00:09:08,079
Speaker 3: animals as sentient beings that can suffer, that can feel pain,

148
00:09:08,240 --> 00:09:11,240
Speaker 3: that have lives of their own, it becomes very, very

149
00:09:11,280 --> 00:09:14,280
Speaker 3: hard to defend the view that none of this matters

150
00:09:14,360 --> 00:09:18,200
Speaker 3: ethically or morally. Of course it matters, and then the

151
00:09:18,280 --> 00:09:23,320
Speaker 3: debate is about how it matters, how strong are our obligations,

152
00:09:23,400 --> 00:09:25,800
Speaker 3: And here you do get a lot of disagreement. Still,

153
00:09:26,679 --> 00:09:29,840
Speaker 3: I feel like the point of consensus that I'm trying

154
00:09:29,880 --> 00:09:33,000
Speaker 3: to latch onto in my book is that we have

155
00:09:33,080 --> 00:09:37,600
Speaker 3: a duty to avoid causing gratuitous suffering to sentient beings,

156
00:09:38,320 --> 00:09:40,600
Speaker 3: which is to say, if we're going to do something

157
00:09:40,600 --> 00:09:44,439
Speaker 3: that will cause suffering, we have to have a sufficiently

158
00:09:44,520 --> 00:09:48,199
Speaker 3: good reason for doing that thing. And then, of course

159
00:09:48,480 --> 00:09:51,160
Speaker 3: you get debate around what might be a good enough reason.

160
00:09:51,840 --> 00:09:54,960
Speaker 3: You get debate around, for example, whether scientific research might

161
00:09:55,040 --> 00:09:58,079
Speaker 3: be a good enough reason, and there'll always be disagreement

162
00:09:58,120 --> 00:10:01,360
Speaker 3: about that. But the the need to have a reason

163
00:10:01,640 --> 00:10:05,760
Speaker 3: so that we cannot be causing suffering gratuitously, this I

164
00:10:05,760 --> 00:10:07,800
Speaker 3: think everyone really can agree about.

165
00:10:08,040 --> 00:10:11,160
Speaker 1: Now, you discussed multiple additional cases that seem to exist

166
00:10:11,240 --> 00:10:14,080
Speaker 1: at that edge of synience, as the title refers to.

167
00:10:14,200 --> 00:10:16,400
Speaker 1: And I'm not going to ask you about all of them,

168
00:10:16,400 --> 00:10:18,960
Speaker 1: but one of the more surprising ones to me, I

169
00:10:18,960 --> 00:10:20,880
Speaker 1: guess this isn't an area that I had not been

170
00:10:20,880 --> 00:10:24,160
Speaker 1: paying close enough attention to in the science news is

171
00:10:24,320 --> 00:10:29,599
Speaker 1: the idea of brain organoids or artificially grown neural tissues.

172
00:10:29,960 --> 00:10:33,360
Speaker 1: I was not aware that they were playing pong. So

173
00:10:33,400 --> 00:10:35,720
Speaker 1: what's the story here and what does it mean for

174
00:10:35,800 --> 00:10:38,760
Speaker 1: our understanding of syentians in something like this.

175
00:10:39,440 --> 00:10:43,880
Speaker 3: It's an astounding and very exciting emerging area of research

176
00:10:44,760 --> 00:10:48,800
Speaker 3: where you can induce human stem cells to form nerve

177
00:10:48,880 --> 00:10:53,079
Speaker 3: cells to form brain tissue, and you can build structures

178
00:10:53,400 --> 00:10:57,360
Speaker 3: the model regions of the human brain at very very

179
00:10:57,400 --> 00:11:02,640
Speaker 3: small scales and times. Researchers are doing this to model diseases.

180
00:11:03,000 --> 00:11:07,680
Speaker 3: They want to model Alzheimer's or ZEKA or fetal alcohol syndrome,

181
00:11:08,240 --> 00:11:10,080
Speaker 3: and this can be a very good way of modeling.

182
00:11:10,400 --> 00:11:12,360
Speaker 3: So if you compare it to the alternative, that is

183
00:11:12,400 --> 00:11:15,839
Speaker 3: to use a living animal like a mouse or a rat,

184
00:11:16,240 --> 00:11:19,560
Speaker 3: that has real limitations because the rat's brain is so

185
00:11:19,640 --> 00:11:20,880
Speaker 3: different from the human brain.

186
00:11:21,320 --> 00:11:22,080
Speaker 1: So this is very.

187
00:11:22,040 --> 00:11:27,480
Speaker 3: Exciting way of getting better models of diseases. Of course,

188
00:11:27,480 --> 00:11:31,040
Speaker 3: it raises questions as well about, well, there must be

189
00:11:31,120 --> 00:11:33,439
Speaker 3: some point at which you really should stop doing this,

190
00:11:33,520 --> 00:11:36,839
Speaker 3: because you've made something too lifelike, you've made something too big,

191
00:11:37,280 --> 00:11:39,280
Speaker 3: you've let it develop for too long, and there's now

192
00:11:39,280 --> 00:11:42,360
Speaker 3: a chance that it will be sentient in its own right.

193
00:11:43,040 --> 00:11:45,800
Speaker 3: And I feel like this is a risk that seems

194
00:11:45,840 --> 00:11:48,960
Speaker 3: particularly striking in cases where what the researchers are trying

195
00:11:49,040 --> 00:11:55,040
Speaker 3: to do is model intelligence, model cognitive functions. That's what

196
00:11:55,160 --> 00:11:59,640
Speaker 3: this system, dish brain that you were referring to, is doing,

197
00:11:59,760 --> 00:12:02,680
Speaker 3: because because what the researchers did was train it to

198
00:12:02,720 --> 00:12:08,559
Speaker 3: play the video game Pong through interacting with a computer interface,

199
00:12:09,200 --> 00:12:12,959
Speaker 3: and so the system could control the paddle and they

200
00:12:13,040 --> 00:12:19,120
Speaker 3: demonstrated measurable improvement in gameplay over twenty minutes. So by

201
00:12:19,160 --> 00:12:23,959
Speaker 3: getting feedback on its performance, the system was learning how

202
00:12:24,000 --> 00:12:31,320
Speaker 3: to play Pong. Really, the thought that we might be

203
00:12:31,360 --> 00:12:34,080
Speaker 3: getting dangerously close to the edge of sentience, I think

204
00:12:34,160 --> 00:12:38,000
Speaker 3: strikes you very clearly when you read about studies like this.

205
00:12:39,160 --> 00:12:41,880
Speaker 1: Yeah, especially to your point, the idea that we could

206
00:12:42,080 --> 00:12:45,360
Speaker 1: get there sort of very much by accident in this case,

207
00:12:46,520 --> 00:12:49,319
Speaker 1: you know, trying in part, perhaps trying to avoid things

208
00:12:49,360 --> 00:12:55,320
Speaker 1: like cruelty to mouse kind of lad animal.

209
00:12:55,600 --> 00:12:58,200
Speaker 3: Well, this is why I think it would be an

210
00:12:58,240 --> 00:13:03,440
Speaker 3: overreaction to immediately ban all of this research, because that

211
00:13:03,480 --> 00:13:06,040
Speaker 3: would be inconsistent. We need to be consistent in our

212
00:13:06,080 --> 00:13:09,880
Speaker 3: attitudes to different risks, and it's no use if we

213
00:13:10,080 --> 00:13:13,440
Speaker 3: crack down hard on the organoid research in a way

214
00:13:13,440 --> 00:13:16,400
Speaker 3: that just leads to more research being done on obviously

215
00:13:16,440 --> 00:13:20,680
Speaker 3: sentient animals like mice and rats and monkeys and so on.

216
00:13:21,720 --> 00:13:23,960
Speaker 3: We've got to let this research develop because it could

217
00:13:24,000 --> 00:13:26,760
Speaker 3: be replacing animal research, and we have to encourage that.

218
00:13:27,800 --> 00:13:30,360
Speaker 3: At the same time, we need proportionate steps. We need

219
00:13:30,400 --> 00:13:33,680
Speaker 3: to think about what the red lines are so that

220
00:13:33,760 --> 00:13:34,800
Speaker 3: it doesn't go too far.

221
00:13:45,080 --> 00:13:49,000
Speaker 1: Now, another huge question from your book is how would

222
00:13:49,000 --> 00:13:52,680
Speaker 1: we know even AI became synient and what would it

223
00:13:52,720 --> 00:13:54,120
Speaker 1: mean for us if it did.

224
00:13:56,120 --> 00:13:58,280
Speaker 3: I think we wouldn't though, And this is the big

225
00:13:58,320 --> 00:14:01,800
Speaker 3: fear that we may be rapidly approaching the edge of

226
00:14:01,880 --> 00:14:04,680
Speaker 3: sentience in this case too, with the rate at which

227
00:14:04,720 --> 00:14:09,920
Speaker 3: AI is developing. The extraordinary behaviors we're seeing from AI systems,

228
00:14:10,600 --> 00:14:15,240
Speaker 3: and yet our understanding of how they work remains incredibly poor.

229
00:14:16,120 --> 00:14:18,240
Speaker 3: And it's not just that the public doesn't understand that

230
00:14:18,360 --> 00:14:20,960
Speaker 3: the people working at the tech companies do understand. The

231
00:14:20,960 --> 00:14:23,960
Speaker 3: people at the tech companies do not understand either. These

232
00:14:24,000 --> 00:14:28,600
Speaker 3: systems are black boxes where you know the architecture, the

233
00:14:28,640 --> 00:14:31,760
Speaker 3: overall architecture that you've programmed the system to have, but

234
00:14:31,800 --> 00:14:35,640
Speaker 3: then you've let it, You've trained it on vast, vast

235
00:14:35,680 --> 00:14:39,680
Speaker 3: amounts of training data, and in the process it's acquired

236
00:14:39,720 --> 00:14:46,120
Speaker 3: these emergent capabilities. It's acquired algorithms that you didn't program

237
00:14:46,120 --> 00:14:48,880
Speaker 3: into it, but that it can now implement to reason

238
00:14:48,920 --> 00:14:53,680
Speaker 3: its way through problems. And we don't know what the

239
00:14:53,760 --> 00:14:56,360
Speaker 3: upper limit is here. We don't know as these systems

240
00:14:56,400 --> 00:14:59,640
Speaker 3: scale up, we don't know what algorithms they might be

241
00:14:59,640 --> 00:15:02,680
Speaker 3: able to acquire. And we don't know whether there's some

242
00:15:02,800 --> 00:15:06,680
Speaker 3: point at which, if you recreate enough of the computations

243
00:15:06,720 --> 00:15:09,640
Speaker 3: that are happening in a human brain, maybe you also

244
00:15:09,720 --> 00:15:12,600
Speaker 3: get the sentience as well, maybe you also start to

245
00:15:12,640 --> 00:15:14,640
Speaker 3: get feeling as well. This is a view that in

246
00:15:14,640 --> 00:15:19,000
Speaker 3: philosophy is called computational functionalism. It's like a long word

247
00:15:19,080 --> 00:15:22,280
Speaker 3: for the idea that if you recreate all the computations

248
00:15:22,320 --> 00:15:25,200
Speaker 3: going on in the brain, nothing else is needed to

249
00:15:25,240 --> 00:15:28,680
Speaker 3: get sentience. You get the sentience as well, And that's

250
00:15:28,720 --> 00:15:30,800
Speaker 3: the possibility we have to take seriously, and it's a

251
00:15:30,840 --> 00:15:35,160
Speaker 3: real risk, and it means we could create sentient AI

252
00:15:36,200 --> 00:15:38,840
Speaker 3: long before we accept that we've done so, or before

253
00:15:38,880 --> 00:15:40,880
Speaker 3: we realize that we've done so.

254
00:15:40,920 --> 00:15:44,280
Speaker 1: This leads me to a question that my wife asked

255
00:15:44,360 --> 00:15:46,120
Speaker 1: me to ask you when I said, hey, do you

256
00:15:46,120 --> 00:15:48,920
Speaker 1: have any questions about synthients and AI and animals and

257
00:15:49,000 --> 00:15:51,920
Speaker 1: so forth? She wanted me to ask should we be

258
00:15:52,240 --> 00:15:56,440
Speaker 1: polite when we're addressing Siri, Alexa or various you know,

259
00:15:56,680 --> 00:16:00,320
Speaker 1: Google Gemini or whatever kind of text based interface that

260
00:16:00,360 --> 00:16:05,840
Speaker 1: we're using, Like what because I've found myself making like

261
00:16:06,000 --> 00:16:09,000
Speaker 1: going into say Google Gemini, testing it out, just kind

262
00:16:09,000 --> 00:16:11,120
Speaker 1: of like experimenting with it, seeing what's up with it,

263
00:16:11,680 --> 00:16:15,160
Speaker 1: and then after a few exchanges with it, feeling like

264
00:16:15,200 --> 00:16:17,520
Speaker 1: I need to say, oh, well, thank you, that's all

265
00:16:17,560 --> 00:16:20,240
Speaker 1: for today, and feeling like I need to be polite.

266
00:16:20,600 --> 00:16:25,080
Speaker 1: But then also I have caught children, my own child

267
00:16:25,320 --> 00:16:28,360
Speaker 1: once or twice, being a little harsh with say Siri,

268
00:16:29,360 --> 00:16:31,440
Speaker 1: maybe their un alarm is going on too long in

269
00:16:31,480 --> 00:16:33,760
Speaker 1: the morning, that sort of thing. So what are your

270
00:16:33,800 --> 00:16:34,400
Speaker 1: thoughts about this?

271
00:16:34,520 --> 00:16:38,200
Speaker 3: Yeah, it's a fascinating question we have as well as

272
00:16:38,200 --> 00:16:40,960
Speaker 3: the book, there's a paper that we just released called

273
00:16:41,040 --> 00:16:47,680
Speaker 3: taking AI Welfare Seriously, and it's it is an issue

274
00:16:47,720 --> 00:16:53,560
Speaker 3: we should take seriously right now because AI systems that

275
00:16:54,120 --> 00:16:58,640
Speaker 3: might realistically be sentient could be with us quicker than

276
00:16:58,640 --> 00:17:04,119
Speaker 3: we expect indeed at any time. And I think it's

277
00:17:04,119 --> 00:17:07,199
Speaker 3: great to be having that discussion now about what are

278
00:17:07,200 --> 00:17:09,080
Speaker 3: we going to do about that. The questions it raises

279
00:17:09,119 --> 00:17:12,480
Speaker 3: are absolutely enormous. We don't know how to answer them,

280
00:17:12,880 --> 00:17:15,160
Speaker 3: and I think maybe it's right that a very low

281
00:17:15,280 --> 00:17:18,919
Speaker 3: cost starting point that we can do right now is

282
00:17:19,000 --> 00:17:23,840
Speaker 3: just start trying to cultivate an attitude of respect the

283
00:17:24,160 --> 00:17:27,880
Speaker 3: systems we're currently interacting with. There's every chance they're not sentient,

284
00:17:28,560 --> 00:17:31,840
Speaker 3: but there's no harm in cultivating an attitude of respect anyway.

285
00:17:32,960 --> 00:17:37,440
Speaker 3: And by cultivating that attitude of respect, will be more prepared,

286
00:17:38,440 --> 00:17:40,560
Speaker 3: more prepared for the future where there really might be

287
00:17:40,600 --> 00:17:45,600
Speaker 3: a moral requirement to avoid torturing these systems.

288
00:17:45,960 --> 00:17:49,760
Speaker 1: Now in terms of just identifying potential sentience, and you're

289
00:17:49,760 --> 00:17:53,439
Speaker 1: already outlined like the challenges, if not impossibility of that.

290
00:17:54,080 --> 00:17:56,359
Speaker 1: Can you tell us a little bit about the gaming problem.

291
00:17:56,720 --> 00:17:59,239
Speaker 3: One of the problems we face in this area is

292
00:17:59,280 --> 00:18:02,919
Speaker 3: that if you ask AI whether it feels anything or not,

293
00:18:04,359 --> 00:18:06,960
Speaker 3: answers very a great deal. Sometimes they say yes, sometimes

294
00:18:07,040 --> 00:18:10,359
Speaker 3: they say no. But those answers are not giving us

295
00:18:10,480 --> 00:18:15,240
Speaker 3: very good evidence at all. The problem is that we've

296
00:18:16,040 --> 00:18:20,040
Speaker 3: we've trained these systems to mimic the dispositions of a

297
00:18:20,119 --> 00:18:24,120
Speaker 3: helpful human assistant. So in their training they've got rewarded

298
00:18:24,160 --> 00:18:28,960
Speaker 3: constantly for being as human like as possible. And so

299
00:18:29,040 --> 00:18:32,720
Speaker 3: we have this situation in which we've got reason to

300
00:18:32,800 --> 00:18:37,359
Speaker 3: think our criteria for sentience will be gained, so to speak,

301
00:18:38,160 --> 00:18:42,119
Speaker 3: because the system can serve its objectives of being a

302
00:18:42,200 --> 00:18:46,879
Speaker 3: helpful like a helpful human assistant by mimicking behaviors that

303
00:18:47,240 --> 00:18:51,920
Speaker 3: we see as being persuasive of sentience, in talking as

304
00:18:51,920 --> 00:18:55,040
Speaker 3: if it had a rich internal life, as if it

305
00:18:55,080 --> 00:18:59,600
Speaker 3: had emotions, as if it had sensations. Sometimes developers have

306
00:18:59,640 --> 00:19:03,399
Speaker 3: reacted to that problem by just programming the systems to

307
00:19:03,440 --> 00:19:06,480
Speaker 3: deny their sentient so they just say, of course, as

308
00:19:07,240 --> 00:19:10,440
Speaker 3: an AI system, I don't have any feelings. That isn't

309
00:19:10,480 --> 00:19:13,240
Speaker 3: very helpful either, because that's not evidence that they don't.

310
00:19:13,920 --> 00:19:17,800
Speaker 3: So we're facing this tough situation where the surface linguistic

311
00:19:17,800 --> 00:19:22,520
Speaker 3: behavior is not really giving us any evidence either way.

312
00:19:22,920 --> 00:19:24,720
Speaker 3: To my mind, the message we have to take from

313
00:19:24,760 --> 00:19:27,080
Speaker 3: this is that we need to be doing everything we

314
00:19:27,160 --> 00:19:30,920
Speaker 3: can to look behind the surface linguistic behavior to try

315
00:19:30,960 --> 00:19:33,879
Speaker 3: and understand the inner workings of these systems. Better to

316
00:19:33,920 --> 00:19:36,840
Speaker 3: try and get inside the black box, open it up,

317
00:19:36,960 --> 00:19:41,200
Speaker 3: find out what computations are actually being performed and how

318
00:19:41,240 --> 00:19:43,359
Speaker 3: they relate to those that are being performed in the

319
00:19:43,400 --> 00:19:46,359
Speaker 3: human brain. To identify what I call in the book

320
00:19:46,400 --> 00:19:50,040
Speaker 3: deep computational markers of sentience, and then look for those

321
00:19:50,800 --> 00:19:53,280
Speaker 3: rather than thinking the linguistic behavior will do the job

322
00:19:53,320 --> 00:19:53,720
Speaker 3: for us.

323
00:19:54,480 --> 00:19:57,359
Speaker 1: Now, what do you think about our moral and or

324
00:19:57,480 --> 00:20:01,600
Speaker 1: legal responsibilities concerning the sentient as we look forward into

325
00:20:01,640 --> 00:20:03,920
Speaker 1: the future. And again you see, as you said, like

326
00:20:03,960 --> 00:20:06,399
Speaker 1: a lot of this is and or it could be

327
00:20:06,520 --> 00:20:09,359
Speaker 1: happening a lot faster than many of us might think.

328
00:20:09,440 --> 00:20:12,199
Speaker 1: But you know, what does that mean when suddenly we

329
00:20:12,280 --> 00:20:16,520
Speaker 1: have at least reasonable reason to believe a particular AI

330
00:20:16,680 --> 00:20:17,240
Speaker 1: is sentient.

331
00:20:18,080 --> 00:20:21,200
Speaker 3: It's a huge debate that I really think we should

332
00:20:21,200 --> 00:20:23,440
Speaker 3: be having now. It's great to be having it now.

333
00:20:24,280 --> 00:20:26,439
Speaker 3: In the edge of sentience. I defend this principle I

334
00:20:26,480 --> 00:20:29,880
Speaker 3: call the run ahead principle, which says that in thinking

335
00:20:29,920 --> 00:20:32,720
Speaker 3: about these issues, we really need to be asking what

336
00:20:32,760 --> 00:20:37,920
Speaker 3: would be proportionate to the risks posed both credible future technologies,

337
00:20:38,560 --> 00:20:42,320
Speaker 3: not just the technologies we have now. Because the technology

338
00:20:42,359 --> 00:20:46,119
Speaker 3: is moving too fast and regulation moves very slow. We

339
00:20:46,160 --> 00:20:49,879
Speaker 3: don't want to be in the position where we're totally

340
00:20:49,960 --> 00:20:53,200
Speaker 3: unprepared for what happens, because we would only have a

341
00:20:53,240 --> 00:20:57,560
Speaker 3: debating the current technology rather than the possible future technology.

342
00:20:58,000 --> 00:21:00,520
Speaker 3: So it's absolutely worth debating about if we get to

343
00:21:00,560 --> 00:21:05,200
Speaker 3: that situation where we've got some deep computational markers of sentience,

344
00:21:06,000 --> 00:21:08,680
Speaker 3: and then we find that we have systems displaying those

345
00:21:08,720 --> 00:21:13,119
Speaker 3: markers so that there is a realistic possibility that the

346
00:21:13,160 --> 00:21:17,280
Speaker 3: system is genuinely sentient, we really have to be thinking

347
00:21:17,320 --> 00:21:21,840
Speaker 3: about what does our duty to avoid causing gratuitous suffering

348
00:21:22,680 --> 00:21:27,560
Speaker 3: require from us in this case, and I think it

349
00:21:27,560 --> 00:21:31,399
Speaker 3: will imply ethical limits on what people can actually do

350
00:21:31,520 --> 00:21:37,960
Speaker 3: to AI systems. What those ethical limits are very very

351
00:21:37,960 --> 00:21:42,320
Speaker 3: hard to say, because the welfare needs we can't even

352
00:21:42,359 --> 00:21:45,520
Speaker 3: really imagine. It depends a lot on the precise nature

353
00:21:45,560 --> 00:21:49,399
Speaker 3: of these systems and the way in which they've achieved sentience.

354
00:21:50,520 --> 00:21:53,080
Speaker 3: Whether we can say anything about their welfare needs at all.

355
00:21:54,280 --> 00:21:57,760
Speaker 3: And to me, all of this points towards having good

356
00:21:57,800 --> 00:22:00,879
Speaker 3: reasons to desperately try not to develop the technology at

357
00:22:00,920 --> 00:22:03,680
Speaker 3: all if we can. You know, I think currently we're

358
00:22:03,720 --> 00:22:06,920
Speaker 3: just not ready. We're just not in a position to

359
00:22:08,400 --> 00:22:12,679
Speaker 3: use this technology ethically, and so in a way we

360
00:22:12,720 --> 00:22:15,200
Speaker 3: should be trying to avoid making it at all.

361
00:22:16,280 --> 00:22:19,120
Speaker 1: Now in the book there's at least one example, and

362
00:22:19,640 --> 00:22:22,879
Speaker 1: I Apolgiz I'm blinking on the specific here, but you

363
00:22:22,960 --> 00:22:28,040
Speaker 1: mentioned a fairly recent call for ethical guidelines concerning AI

364
00:22:28,119 --> 00:22:33,520
Speaker 1: development that was dismissed by critics as being mere science fiction.

365
00:22:34,000 --> 00:22:34,919
Speaker 3: Thomas Mettsinger.

366
00:22:35,080 --> 00:22:38,840
Speaker 1: Yes, yes, I believe so. And that struck me as

367
00:22:38,880 --> 00:22:42,600
Speaker 1: interesting because on one hand, we have clearly, at least

368
00:22:42,920 --> 00:22:45,520
Speaker 1: through science fiction, and of course outside of science fiction

369
00:22:45,560 --> 00:22:48,560
Speaker 1: as well, we've been contemplating things like this for decades

370
00:22:48,600 --> 00:22:52,600
Speaker 1: and decades, and yet as we get closer to the reality,

371
00:22:53,640 --> 00:22:57,160
Speaker 1: the label science fiction is also sometimes used to dismiss

372
00:22:57,160 --> 00:22:58,920
Speaker 1: it as saying, well, that is just sci fi, that's

373
00:22:58,960 --> 00:23:02,520
Speaker 1: not actual things we should be worrying about. So I

374
00:23:02,520 --> 00:23:04,359
Speaker 1: don't know if you have any thoughts on to what

375
00:23:04,480 --> 00:23:07,639
Speaker 1: extent science fiction and science fictional thought has prepared us

376
00:23:07,640 --> 00:23:11,120
Speaker 1: for this or kind of created this barrier that prevents

377
00:23:11,200 --> 00:23:12,840
Speaker 1: us from acting as quickly.

378
00:23:14,119 --> 00:23:16,080
Speaker 3: Yeah, I don't think it has prepared us. Yeah, I

379
00:23:16,080 --> 00:23:20,560
Speaker 3: think that's fair to say, even though we do. Seems

380
00:23:20,840 --> 00:23:24,040
Speaker 3: films like Her, for example, from about ten years ago

381
00:23:24,080 --> 00:23:28,479
Speaker 3: that now seemed remarkably prescient that no one thought they

382
00:23:28,520 --> 00:23:31,560
Speaker 3: were describing events ten to fifteen years in the future,

383
00:23:32,280 --> 00:23:35,440
Speaker 3: and yet that is the future we now found ourselves in.

384
00:23:35,720 --> 00:23:39,679
Speaker 3: It's extraordinary. But yeah, that doesn't in any way mean

385
00:23:39,760 --> 00:23:41,920
Speaker 3: that we're prepared. And in my work on this, I'm

386
00:23:41,960 --> 00:23:47,879
Speaker 3: trying to develop a sort of centrist position that is

387
00:23:47,920 --> 00:23:51,080
Speaker 3: about avoiding the pitfalls of extreme views on both sides,

388
00:23:51,840 --> 00:23:55,960
Speaker 3: where one extreme you've got people who think that these

389
00:23:56,000 --> 00:23:59,560
Speaker 3: systems already sentient. We can tell from their surface linguistic

390
00:24:00,800 --> 00:24:02,960
Speaker 3: they just talk as if they have feelings, so we

391
00:24:03,000 --> 00:24:06,520
Speaker 3: should think they do. And I think that's credulous and

392
00:24:06,600 --> 00:24:10,560
Speaker 3: it needs to be avoided. On the other side, there's

393
00:24:10,640 --> 00:24:16,240
Speaker 3: this dismissal of the whole idea that AI could achieve sentience,

394
00:24:17,200 --> 00:24:20,000
Speaker 3: this idea that of course you need a biological brain,

395
00:24:20,280 --> 00:24:23,040
Speaker 3: of course you need to be a living animal, and

396
00:24:23,080 --> 00:24:24,960
Speaker 3: we're just not in a position to be confident or

397
00:24:24,960 --> 00:24:29,320
Speaker 3: sure about that in this well known philosophical position, computational

398
00:24:29,359 --> 00:24:33,120
Speaker 3: functionism might be right. And if it is right, then

399
00:24:33,160 --> 00:24:35,720
Speaker 3: you might not need a biological brain at all, and

400
00:24:35,800 --> 00:24:38,560
Speaker 3: we have to take that seriously as well. So for me,

401
00:24:38,640 --> 00:24:41,080
Speaker 3: it's about finding that middle ground where we're taking the

402
00:24:41,119 --> 00:24:44,960
Speaker 3: issue seriously, but we're thinking that this has to be

403
00:24:45,000 --> 00:24:48,080
Speaker 3: the beginning now of a process where we really try

404
00:24:48,119 --> 00:24:53,040
Speaker 3: and look for robust, rigorous markers and have serious ethical

405
00:24:53,080 --> 00:24:57,040
Speaker 3: debates about what the right response to those markers of

406
00:24:57,160 --> 00:25:00,800
Speaker 3: sentience would be. We could have to be neither no

407
00:25:00,880 --> 00:25:13,320
Speaker 3: knee jerk skepticism or credulousness.

408
00:25:14,280 --> 00:25:17,280
Speaker 1: Now I realized this next question is largely outside the

409
00:25:17,280 --> 00:25:20,760
Speaker 1: scope of this book, But what are the implications for

410
00:25:20,800 --> 00:25:25,159
Speaker 1: the consideration of possible extraterrestrial syndiants as we encounter it

411
00:25:25,280 --> 00:25:31,280
Speaker 1: in potentially encounter it in say a biological or technological form.

412
00:25:32,040 --> 00:25:34,520
Speaker 3: Just make me think of octopuses again, because of course,

413
00:25:35,800 --> 00:25:38,960
Speaker 3: you know, they're so alien from us. They look like extraterrestrials,

414
00:25:39,160 --> 00:25:42,439
Speaker 3: but they're not. They're terrestrial and they're right here on

415
00:25:42,480 --> 00:25:47,520
Speaker 3: Earth right now. So I think it's great to recognize

416
00:25:47,560 --> 00:25:50,520
Speaker 3: the possibility of forms of sentient it's very different from

417
00:25:50,560 --> 00:25:54,800
Speaker 3: our own, and then recognize that our actual Earth already

418
00:25:54,800 --> 00:25:58,640
Speaker 3: contains them, and that we can start thinking now about

419
00:25:58,640 --> 00:26:01,119
Speaker 3: those real cases and what we're going to do about

420
00:26:01,119 --> 00:26:05,480
Speaker 3: those real cases. I'm entirely open to the idea that,

421
00:26:05,520 --> 00:26:07,520
Speaker 3: you know, just as there are really alien forms of

422
00:26:07,600 --> 00:26:10,919
Speaker 3: sentients on Earth, maybe there are out there elsewhere in

423
00:26:10,920 --> 00:26:14,320
Speaker 3: the universe as well. But we can only speculate, and

424
00:26:14,359 --> 00:26:16,360
Speaker 3: with Oltopus says, we don't need to speculate. We can

425
00:26:16,400 --> 00:26:19,919
Speaker 3: be studying the alien life forms that are that are

426
00:26:19,960 --> 00:26:22,960
Speaker 3: with us now on Earth and get real knowledge about them.

427
00:26:23,760 --> 00:26:26,000
Speaker 1: Now. Through through much of this topic, there, you know,

428
00:26:26,119 --> 00:26:29,439
Speaker 1: there's this sense of expanding our compassion for non human

429
00:26:29,520 --> 00:26:34,560
Speaker 1: sentient entities, and certainly the octopus is a great example

430
00:26:34,600 --> 00:26:36,520
Speaker 1: of that. I know in my own life, like years

431
00:26:36,560 --> 00:26:38,399
Speaker 1: and years ago, when I first started reading a bit

432
00:26:38,440 --> 00:26:42,480
Speaker 1: about their intelligence and their behavior, I stopped eating octopus

433
00:26:42,880 --> 00:26:46,960
Speaker 1: before I stopped eating other meats. And so I feel

434
00:26:46,960 --> 00:26:50,879
Speaker 1: like this kind of response is going to, you know,

435
00:26:51,040 --> 00:26:54,760
Speaker 1: to happen inevitably in as far as we consider these

436
00:26:55,080 --> 00:26:58,679
Speaker 1: non human sentient forms. But what kind of impact do

437
00:26:58,720 --> 00:27:01,960
Speaker 1: you see all of this having Potentially on the expansion

438
00:27:02,000 --> 00:27:05,399
Speaker 1: of our compassion for each other, Like, does this expansion

439
00:27:05,480 --> 00:27:08,480
Speaker 1: of compassion for non human entities? Do you think it

440
00:27:08,560 --> 00:27:12,040
Speaker 1: ultimately helps us become more compassionate to other humans.

441
00:27:12,680 --> 00:27:15,480
Speaker 3: It may do, and I suppose I hope it does.

442
00:27:16,480 --> 00:27:16,640
Speaker 1: Yeah.

443
00:27:16,680 --> 00:27:19,000
Speaker 3: I certainly don't think it's some kind of zero sum

444
00:27:19,080 --> 00:27:23,600
Speaker 3: game where by being more compassionate to octopuses and insects

445
00:27:23,640 --> 00:27:26,760
Speaker 3: and crabs and lobsters and so on, we're forced to

446
00:27:26,800 --> 00:27:30,160
Speaker 3: then be less compassionate to other people. I don't think

447
00:27:30,160 --> 00:27:32,880
Speaker 3: it works like that at all. I think it's more

448
00:27:32,960 --> 00:27:37,520
Speaker 3: this general attitude. And I'm a big fan of the

449
00:27:37,520 --> 00:27:43,000
Speaker 3: Indian idea of a hymnsa non violence, non injury, abolishing

450
00:27:43,040 --> 00:27:47,760
Speaker 3: the desire to kill or harm other beings. I think

451
00:27:47,760 --> 00:27:51,640
Speaker 3: it's about trying to cultivate that virtue, trying to walk

452
00:27:51,680 --> 00:27:55,680
Speaker 3: that path, and it's a path that encompasses other humans

453
00:27:55,840 --> 00:28:00,919
Speaker 3: and non human animals as well. And through cultivating this

454
00:28:01,080 --> 00:28:06,040
Speaker 3: general non violence, this general loss of our desire to

455
00:28:06,200 --> 00:28:10,480
Speaker 3: dominate and crush and harm other beings, even if they're insects,

456
00:28:11,040 --> 00:28:12,960
Speaker 3: we can become a lot more peaceful, I think, in

457
00:28:13,280 --> 00:28:14,680
Speaker 3: our dealings with each other too.

458
00:28:15,600 --> 00:28:19,439
Speaker 1: And what do you see ultimately as the prime I

459
00:28:19,440 --> 00:28:25,040
Speaker 1: guess motivators in changing the way we see these various entities.

460
00:28:25,280 --> 00:28:31,240
Speaker 1: Is it through Is it through laws and regulations? Is

461
00:28:31,280 --> 00:28:34,320
Speaker 1: it through more like sort of ground level outreach? Is

462
00:28:34,359 --> 00:28:37,760
Speaker 1: it both? I mean, how do we really affect this

463
00:28:37,840 --> 00:28:39,680
Speaker 1: sort of change or how have we affected it so

464
00:28:39,760 --> 00:28:41,360
Speaker 1: far most successfully?

465
00:28:41,800 --> 00:28:45,080
Speaker 3: It's a huge open question for me what does actually

466
00:28:45,120 --> 00:28:49,520
Speaker 3: succeed in changing behavior? Because I've been focused a lot

467
00:28:49,560 --> 00:28:53,960
Speaker 3: on scientific evidence and about synthesizing the existing evidence for

468
00:28:54,400 --> 00:28:59,560
Speaker 3: sentients and other animals, presenting it to policymakers. Sometimes it

469
00:28:59,560 --> 00:29:02,800
Speaker 3: does produce change, and in the UK, the Animal Welfare

470
00:29:02,840 --> 00:29:09,400
Speaker 3: Sentience Act was amended to recognize octopuses, crabs, lobsters, crayfish

471
00:29:09,440 --> 00:29:12,040
Speaker 3: as sentient beings because of the report that we produced.

472
00:29:12,720 --> 00:29:16,719
Speaker 3: So that was surprisingly effective in a way example of

473
00:29:16,720 --> 00:29:21,360
Speaker 3: how marshaling scientific evidence can move policymakers. So it's great

474
00:29:21,360 --> 00:29:24,320
Speaker 3: when that happens, but of course it doesn't always happen,

475
00:29:24,800 --> 00:29:27,120
Speaker 3: and we do face this problem that a lot of

476
00:29:27,200 --> 00:29:31,360
Speaker 3: animals are pretty clearly sentient, think of pigs, for example,

477
00:29:31,480 --> 00:29:34,680
Speaker 3: or chickens, and yet they continue to be treated by

478
00:29:34,760 --> 00:29:38,840
Speaker 3: humans in absolutely appalling ways. So merely knowing that an

479
00:29:38,840 --> 00:29:43,360
Speaker 3: animal is sentient often does not drastically change your behavior

480
00:29:44,320 --> 00:29:47,960
Speaker 3: towards it, And I'm fascinated by the question of, well,

481
00:29:48,000 --> 00:29:53,320
Speaker 3: what else is needed? What other information? I think there

482
00:29:53,320 --> 00:29:57,760
Speaker 3: are empathy barriers. You could know that a chicken is sentient,

483
00:29:58,600 --> 00:30:02,360
Speaker 3: but doesn't necessarily convert into immediately empathizing with that chicken

484
00:30:02,880 --> 00:30:07,480
Speaker 3: and the animals suffering. I've got to think about what

485
00:30:07,560 --> 00:30:13,400
Speaker 3: might bridge that gap, So narrative stories, art, video documentaries

486
00:30:13,480 --> 00:30:16,840
Speaker 3: like My Octopus Teacher, they could all be part of it.

487
00:30:16,960 --> 00:30:18,880
Speaker 3: I think there's probably lots of ways to bridge that

488
00:30:18,960 --> 00:30:21,480
Speaker 3: empathy gap, but we have to recognize it as a

489
00:30:21,520 --> 00:30:24,640
Speaker 3: problem and to realize that simply knowing the animals are

490
00:30:24,640 --> 00:30:26,240
Speaker 3: sensient is not actually enough.

491
00:30:27,000 --> 00:30:30,400
Speaker 1: It's interesting to think about pork and chicken. I don't

492
00:30:30,440 --> 00:30:34,240
Speaker 1: know how this pans out in the UK, but in

493
00:30:34,280 --> 00:30:37,720
Speaker 1: the States, you often will drive through a city through

494
00:30:37,880 --> 00:30:41,920
Speaker 1: a rural area either one, and you'll find a lot

495
00:30:41,960 --> 00:30:45,800
Speaker 1: of signage and promotion for places that serve pork or chicken,

496
00:30:45,920 --> 00:30:51,120
Speaker 1: that use cute or amusing like cartoon versions of those animals.

497
00:30:51,600 --> 00:30:54,440
Speaker 1: And it seems it's always struck me as strange that

498
00:30:55,080 --> 00:30:57,880
Speaker 1: these are things, these are acts and choices that would

499
00:30:57,920 --> 00:31:02,000
Speaker 1: seem otherwise to be something that would convince us not

500
00:31:02,040 --> 00:31:06,560
Speaker 1: to eat said animal, but they seem to instead give

501
00:31:06,640 --> 00:31:09,120
Speaker 1: us license to. And I've ive always had a hard

502
00:31:09,120 --> 00:31:12,880
Speaker 1: time understanding exactly what's going on in our minds when

503
00:31:12,920 --> 00:31:14,440
Speaker 1: we consume or create that sort of thing.

504
00:31:15,080 --> 00:31:19,280
Speaker 3: It goes under various names, doesn't it cognitive dissonance? The

505
00:31:19,320 --> 00:31:25,000
Speaker 3: meat paradox this idea that we often love animals, we

506
00:31:25,880 --> 00:31:29,120
Speaker 3: find them so cute and adorable in etc. And then

507
00:31:29,200 --> 00:31:33,600
Speaker 3: continue to eat them. Anyway, This would be it would

508
00:31:33,600 --> 00:31:37,000
Speaker 3: make perfect sense if meat was genuinely necessary for our health.

509
00:31:37,800 --> 00:31:40,320
Speaker 3: And I think that's the argument the meat industry would

510
00:31:40,360 --> 00:31:42,040
Speaker 3: love to be making. It would love to be able

511
00:31:42,080 --> 00:31:44,480
Speaker 3: to convince us that meat is needed for our health,

512
00:31:44,520 --> 00:31:47,840
Speaker 3: and so these sacrifices in how we treat the animals

513
00:31:48,160 --> 00:31:51,840
Speaker 3: are satly necessary. But it's just not true. It's just

514
00:31:52,000 --> 00:31:54,400
Speaker 3: clearly not true. And then the existence of all these

515
00:31:54,440 --> 00:31:59,720
Speaker 3: manifestly healthy vegetarians and vegans makes that completely undeniable that

516
00:31:59,720 --> 00:32:01,840
Speaker 3: we don't cents actually need to be eating these animals

517
00:32:01,840 --> 00:32:05,000
Speaker 3: at all for our health, and we can, if anything,

518
00:32:05,080 --> 00:32:09,360
Speaker 3: probably be healthier without doing so. I think once you

519
00:32:09,400 --> 00:32:15,160
Speaker 3: realize this, the case really does become very clear for

520
00:32:15,320 --> 00:32:18,160
Speaker 3: not eating these animals that the harms we're doing to

521
00:32:18,240 --> 00:32:22,480
Speaker 3: them can't actually be justified because the benefit we get

522
00:32:22,560 --> 00:32:27,560
Speaker 3: is at most that the gustatory benefit, the enjoyment of

523
00:32:27,600 --> 00:32:31,160
Speaker 3: the product. It's not necessary for our health in any way,

524
00:32:31,880 --> 00:32:36,440
Speaker 3: and that enjoyment can't justify in the balance all that

525
00:32:36,520 --> 00:32:37,880
Speaker 3: suffering cause to the animal.

526
00:32:38,600 --> 00:32:40,600
Speaker 1: Well, Jonathan, thank you so much for taking time out

527
00:32:40,600 --> 00:32:43,200
Speaker 1: of your day to chat with me. The book is

528
00:32:43,320 --> 00:32:47,040
Speaker 1: The Edge of Sentience Risk and Precaution in Humans, Other

529
00:32:47,080 --> 00:32:51,000
Speaker 1: Animals and AI. It is out in the United States

530
00:32:51,240 --> 00:32:58,320
Speaker 1: are November fifteenth. Thanks Robert, all right, thanks again to

531
00:32:58,400 --> 00:33:00,920
Speaker 1: Jonathan Burch for coming up the show and chatting with

532
00:33:00,960 --> 00:33:03,840
Speaker 1: me again. That book is The Edge of Sentience Risk

533
00:33:03,880 --> 00:33:07,200
Speaker 1: and Precaution in Humans, Other Animals and AI. It is

534
00:33:07,240 --> 00:33:11,040
Speaker 1: out later this week on November fifteenth, and it gets

535
00:33:11,040 --> 00:33:13,920
Speaker 1: into so much more that we didn't have time to

536
00:33:13,920 --> 00:33:17,800
Speaker 1: get into in this interview. Just a reminder that Stuff

537
00:33:17,800 --> 00:33:20,080
Speaker 1: to Blow Your Mind is primarily a science and culture

538
00:33:20,120 --> 00:33:24,200
Speaker 1: podcast with core episodes on Tuesdays and Thursdays. On Fridays,

539
00:33:24,200 --> 00:33:26,320
Speaker 1: we set aside most serious concerns to just talk about

540
00:33:26,320 --> 00:33:28,200
Speaker 1: a weird film on Weird House Cinema, and we have

541
00:33:28,360 --> 00:33:32,600
Speaker 1: short form episodes that air on Wednesdays. Thanks as always

542
00:33:32,600 --> 00:33:36,600
Speaker 1: to the great JJ Possway for editing and producing this podcast,

543
00:33:36,640 --> 00:33:38,600
Speaker 1: and if you would like to get in touch with us, well,

544
00:33:38,800 --> 00:33:41,720
Speaker 1: you can email us at contact at stuff to Blow

545
00:33:41,760 --> 00:33:50,560
Speaker 1: your Mind dot com.

546
00:33:50,680 --> 00:33:53,600
Speaker 2: Stuff to Blow Your Mind is production of iHeartRadio. For

547
00:33:53,680 --> 00:33:56,480
Speaker 2: more podcasts from my Heart Radio, visit the iHeartRadio app,

548
00:33:56,640 --> 00:33:59,360
Speaker 2: Apple Podcasts, or wherever you're listening to your favorite shows

549
00:34:03,320 --> 00:34:03,520
Speaker 1: Name