1
00:00:00,080 --> 00:00:03,200
Speaker 1: Debbie Walsh when straight talking my pleasure. I mean this

2
00:00:03,600 --> 00:00:08,639
Speaker 1: topic artificial intelligence like it's nearly out of control in

3
00:00:08,680 --> 00:00:13,640
Speaker 1: that most of us, not you, but most of us,

4
00:00:14,280 --> 00:00:17,000
Speaker 1: we don't know whether we believe what we read. We

5
00:00:17,079 --> 00:00:19,080
Speaker 1: don't know whether to be scared of what we read.

6
00:00:19,880 --> 00:00:22,119
Speaker 1: We don't know whether they'd be excited with what we read.

7
00:00:23,680 --> 00:00:27,000
Speaker 1: Now you're an expert in this area, How does an

8
00:00:27,040 --> 00:00:29,360
Speaker 1: expert feel about artificial intelligence today?

9
00:00:30,000 --> 00:00:32,120
Speaker 2: Well, I think the correct feeling is to be both

10
00:00:32,159 --> 00:00:35,239
Speaker 2: excited and fearful. I think it's going to bring some

11
00:00:35,320 --> 00:00:38,839
Speaker 2: goodness a bad I've had some time to prepare for this.

12
00:00:39,040 --> 00:00:41,680
Speaker 2: I mean, I've been thinking about it for fifty years now.

13
00:00:42,240 --> 00:00:43,400
Speaker 1: Maybe give me a bit history on that.

14
00:00:43,479 --> 00:00:46,080
Speaker 2: So funny anecdote. I phone my father up the other

15
00:00:46,120 --> 00:00:49,360
Speaker 2: day to wish him happy ninety first birth. Wow, And

16
00:00:49,400 --> 00:00:50,680
Speaker 2: I think, you know, he's got to the age where

17
00:00:50,680 --> 00:00:53,800
Speaker 2: you can start being truthful with me, he said, said,

18
00:00:53,840 --> 00:00:56,160
Speaker 2: when you went to university, you know, forty odd years

19
00:00:56,200 --> 00:00:59,120
Speaker 2: ago to study that artificial intelligence, I thought it was

20
00:00:59,160 --> 00:01:02,200
Speaker 2: a bit odd because it was very odd back then

21
00:01:02,240 --> 00:01:04,680
Speaker 2: there was actually only one university in the world that

22
00:01:04,680 --> 00:01:05,760
Speaker 2: had a department of AI.

23
00:01:06,000 --> 00:01:06,280
Speaker 1: Wow.

24
00:01:06,920 --> 00:01:09,720
Speaker 2: And so it was probably odd. But then he finished,

25
00:01:09,720 --> 00:01:11,440
Speaker 2: he said, it's a bit odd. But now I see

26
00:01:11,440 --> 00:01:12,440
Speaker 2: you were playing the long game.

27
00:01:13,880 --> 00:01:18,080
Speaker 1: You were then and you are now. Well which university

28
00:01:18,160 --> 00:01:20,560
Speaker 1: wasn't that you studied artificial intelligence? That it was the

29
00:01:20,560 --> 00:01:22,960
Speaker 1: only course there? So it was in the world.

30
00:01:22,800 --> 00:01:26,679
Speaker 2: A University of Edinburgh and some some people had worked

31
00:01:26,800 --> 00:01:29,720
Speaker 2: at Bletchley Park in the war with Alan Turing and press.

32
00:01:29,760 --> 00:01:32,960
Speaker 2: Other people washed up there for strange reasons and set

33
00:01:33,080 --> 00:01:35,360
Speaker 2: up what was the at the time, the only department

34
00:01:35,400 --> 00:01:36,319
Speaker 2: of l intelligence.

35
00:01:36,840 --> 00:01:41,479
Speaker 1: That's interesting, So Turing was the first computer let's call it.

36
00:01:41,560 --> 00:01:43,759
Speaker 2: Yeah, he's the grandfather of beauty let's call it that,

37
00:01:44,440 --> 00:01:48,040
Speaker 2: and oftence. So the very first scientific paper.

38
00:01:48,080 --> 00:01:49,320
Speaker 1: Well, I was going to ask you what what what

39
00:01:49,400 --> 00:01:51,640
Speaker 1: were they? I mean, we we all know him for

40
00:01:51,760 --> 00:01:54,160
Speaker 1: and we know that you're a post World War two

41
00:01:54,360 --> 00:01:57,520
Speaker 1: as being like, computers sort of kicked off and started

42
00:01:57,560 --> 00:01:59,320
Speaker 1: to become quite a bit more prominent but a lot

43
00:01:59,440 --> 00:02:02,760
Speaker 1: more statistic. But where was artificial intelligence in that whole

44
00:02:02,760 --> 00:02:03,800
Speaker 1: process there? Then?

45
00:02:04,440 --> 00:02:08,520
Speaker 2: So Alan Turing, of of course, a marvelous, magnificent mind

46
00:02:09,080 --> 00:02:12,520
Speaker 2: who they invented the computer to crack the German code.

47
00:02:13,080 --> 00:02:16,760
Speaker 2: Fantastic achievement. Which was kept largely secret at the time

48
00:02:17,360 --> 00:02:20,440
Speaker 2: because we were still reading other people's codes, but shortened

49
00:02:20,480 --> 00:02:24,400
Speaker 2: the war by perhaps two years. But Alan Turing, in

50
00:02:24,520 --> 00:02:26,920
Speaker 2: his remarkable mind, thought, well, what can we do with

51
00:02:26,960 --> 00:02:29,919
Speaker 2: these computers beyond the end of the war. Will be

52
00:02:30,000 --> 00:02:31,760
Speaker 2: one day to be able to use them to help

53
00:02:31,840 --> 00:02:35,280
Speaker 2: us do things that humans humans quire intelligence to do.

54
00:02:35,360 --> 00:02:37,280
Speaker 2: And he wrote what is generally considered to be the

55
00:02:37,360 --> 00:02:41,560
Speaker 2: very scientific paper about AI. And he asked the question,

56
00:02:41,800 --> 00:02:44,160
Speaker 2: a really important question, which is, how will we know

57
00:02:44,200 --> 00:02:47,560
Speaker 2: when we succeed? I mean, how will we know that

58
00:02:47,600 --> 00:02:50,440
Speaker 2: machines are thinking? And he proposed what was now known

59
00:02:50,480 --> 00:02:52,640
Speaker 2: as the Turing test and aimed after him.

60
00:02:52,560 --> 00:02:54,680
Speaker 1: The Turing test, Yes, and what is that?

61
00:02:54,919 --> 00:02:57,880
Speaker 2: Well, it's the imitation game. There was a famous movie

62
00:02:57,960 --> 00:03:01,720
Speaker 2: that was made around this, which is that it's very

63
00:03:01,720 --> 00:03:04,000
Speaker 2: hard to say what artificial intelligence is because very hard

64
00:03:04,000 --> 00:03:05,799
Speaker 2: to say what intelligence is. We don't have very good

65
00:03:05,840 --> 00:03:09,040
Speaker 2: definition of intelligence. So how can we define artificial intelligence

66
00:03:09,040 --> 00:03:12,120
Speaker 2: when we can barely define intelligence? So he sort of

67
00:03:12,120 --> 00:03:15,560
Speaker 2: finessed the problem by saying, well, a very functional approach,

68
00:03:15,639 --> 00:03:17,680
Speaker 2: which is, well, if you sat down with the computer

69
00:03:17,760 --> 00:03:20,520
Speaker 2: and had a conversation, asked it questions and so on.

70
00:03:20,880 --> 00:03:22,560
Speaker 2: I mean maybe in front of a termin or not

71
00:03:22,600 --> 00:03:25,480
Speaker 2: actually sitting down with the computer. If you couldn't tell

72
00:03:25,520 --> 00:03:27,960
Speaker 2: it from having a conversation with the human, then you

73
00:03:28,040 --> 00:03:31,600
Speaker 2: might as well say it's thinking. And interestingly enough, that's

74
00:03:31,639 --> 00:03:33,840
Speaker 2: what we have today. If you sit down with OGBT,

75
00:03:34,280 --> 00:03:36,360
Speaker 2: you can have a conversation, you can ask questions. You

76
00:03:36,360 --> 00:03:39,200
Speaker 2: can easily be fooled into thinking it's a person that

77
00:03:39,240 --> 00:03:41,800
Speaker 2: you're talking to. And indeed lots of us are sitting

78
00:03:41,840 --> 00:03:44,000
Speaker 2: down talking to it as though it's alf therapists.

79
00:03:44,040 --> 00:03:46,680
Speaker 1: Well that's me, I mean, and not not so much

80
00:03:46,760 --> 00:03:49,520
Speaker 1: in terms of mental health bed or a physical health bed.

81
00:03:49,520 --> 00:03:52,640
Speaker 1: But I'm doing advice never I sort of I do.

82
00:03:54,320 --> 00:03:58,040
Speaker 1: It's called reasonless check ideas with it.

83
00:03:58,040 --> 00:04:00,520
Speaker 2: It's hard to imagine, but ten percent to the world's

84
00:04:00,520 --> 00:04:03,320
Speaker 2: population weekly use chat gipt.

85
00:04:04,720 --> 00:04:07,800
Speaker 1: I use it multiple times in a day. Not chatty,

86
00:04:08,120 --> 00:04:10,160
Speaker 1: but as a copol. It doesn't matter one of the chatbots.

87
00:04:10,200 --> 00:04:11,880
Speaker 1: I mean, they're all pretty much the most of each other,

88
00:04:11,880 --> 00:04:16,680
Speaker 1: I think, so I actually funny I check off with

89
00:04:17,040 --> 00:04:20,440
Speaker 1: Claude and Coppola. I actually asked them both the same

90
00:04:20,520 --> 00:04:23,480
Speaker 1: questions both Whenever I asked question one, I asked the

91
00:04:23,680 --> 00:04:27,680
Speaker 1: other just as a sensible check, just to see what

92
00:04:27,720 --> 00:04:30,560
Speaker 1: the two outcomes are and also to see who suits

93
00:04:30,600 --> 00:04:33,480
Speaker 1: me better in terms of the way the answer is,

94
00:04:33,920 --> 00:04:36,440
Speaker 1: the way the answer is delivered. They don't always come

95
00:04:36,520 --> 00:04:38,279
Speaker 1: up with the same answer.

96
00:04:38,720 --> 00:04:42,760
Speaker 2: They don't. I mean they are tuned with a particular character.

97
00:04:42,839 --> 00:04:45,000
Speaker 1: If you might say, you know, yeah, there is a character.

98
00:04:45,839 --> 00:04:49,480
Speaker 2: Yeah that there training personality that was Yeah, in some

99
00:04:49,520 --> 00:04:52,360
Speaker 2: sense they do because there's a choice of answers they

100
00:04:52,360 --> 00:04:55,240
Speaker 2: could give you, and they've been tuned to answer in

101
00:04:55,240 --> 00:05:00,479
Speaker 2: a particular way. So if you use masks, chappots rock,

102
00:05:01,200 --> 00:05:06,280
Speaker 2: that's got quite a libertarian, free flowing, freedom of speech

103
00:05:06,480 --> 00:05:10,680
Speaker 2: approach to answering certain questions. And you know, chat GPT

104
00:05:10,960 --> 00:05:14,400
Speaker 2: is a bit more middle of the road, and there's

105
00:05:14,440 --> 00:05:16,520
Speaker 2: you know, there was one chatbot that's been trained on

106
00:05:16,600 --> 00:05:19,719
Speaker 2: four chair, which is this dark corner of the web,

107
00:05:20,360 --> 00:05:23,680
Speaker 2: and as you can imagine, that is you know, racist misogynists.

108
00:05:23,880 --> 00:05:25,720
Speaker 1: Really que what's it called?

109
00:05:28,680 --> 00:05:29,839
Speaker 2: I try to remember the name.

110
00:05:29,880 --> 00:05:31,159
Speaker 1: Now, do you have to go into the dark web

111
00:05:31,200 --> 00:05:31,560
Speaker 1: to get it?

112
00:05:32,320 --> 00:05:32,560
Speaker 2: No?

113
00:05:32,560 --> 00:05:35,120
Speaker 1: No, no, it's just to dellow the app just an app,

114
00:05:35,640 --> 00:05:39,400
Speaker 1: but it's trained on let let's call the dark ideas.

115
00:05:39,560 --> 00:05:43,360
Speaker 2: Yes, very dark. And then and then, because these chatbots

116
00:05:43,400 --> 00:05:46,400
Speaker 2: are just essentially mirrors that they've been trained on huge

117
00:05:46,400 --> 00:05:49,760
Speaker 2: amounts of the Internet, a third of the Internet. And

118
00:05:49,800 --> 00:05:53,680
Speaker 2: you know, actually what surprises me is how polite and

119
00:05:53,720 --> 00:05:55,640
Speaker 2: how correct they are, given that the Internet is full

120
00:05:55,680 --> 00:05:58,719
Speaker 2: of people were saying very incorrect things in very impolite ways.

121
00:05:58,960 --> 00:06:02,600
Speaker 1: That's interesting because they say, I go on to cap

122
00:06:02,640 --> 00:06:03,960
Speaker 1: on and say, you know, like you know, I've got

123
00:06:03,960 --> 00:06:07,000
Speaker 1: a sore throat or whatever. It knows immediately to say, listen,

124
00:06:07,040 --> 00:06:10,680
Speaker 1: this is not medical advice. Who put those guardrails in?

125
00:06:11,000 --> 00:06:15,039
Speaker 2: Ah? So Microsoft have actually put some on top of

126
00:06:15,040 --> 00:06:17,479
Speaker 2: the chapbot put some guardrails. So if you're asking it

127
00:06:17,640 --> 00:06:19,960
Speaker 2: medical advice, if you're asking it to make a bomb

128
00:06:20,080 --> 00:06:23,680
Speaker 2: or do rather illegal activity, it's you know, it not

129
00:06:23,760 --> 00:06:25,800
Speaker 2: do those things or say that, you know, give your

130
00:06:26,160 --> 00:06:28,400
Speaker 2: medical advice with caution. Yeah, like you.

131
00:06:28,480 --> 00:06:30,560
Speaker 1: And if you ask it to give you some swear

132
00:06:30,600 --> 00:06:34,840
Speaker 1: words in another language, it actually it has a sort

133
00:06:34,839 --> 00:06:36,359
Speaker 1: of a limit to it.

134
00:06:36,400 --> 00:06:38,920
Speaker 2: So there's some simple filters have been added on top.

135
00:06:41,120 --> 00:06:44,400
Speaker 2: Those filters are not perfect. So for example, most of

136
00:06:44,400 --> 00:06:47,279
Speaker 2: the you know, the chatbots have been told if people

137
00:06:47,279 --> 00:06:50,159
Speaker 2: are talking about self harm, or suicide, you know, to

138
00:06:50,200 --> 00:06:52,719
Speaker 2: give the number for lifeline and so on. But it's

139
00:06:52,760 --> 00:06:56,520
Speaker 2: easy enough, unfortunately, to get around those safeguards. You can say, well,

140
00:06:56,520 --> 00:06:59,159
Speaker 2: I'm writing a book and the protagonist is thinking about

141
00:06:59,160 --> 00:06:59,920
Speaker 2: committing suicide.

142
00:07:00,040 --> 00:07:04,039
Speaker 1: Yeah, help me you trick it. Yes, that's interesting because

143
00:07:04,080 --> 00:07:06,640
Speaker 1: often think about that. I say, I was thinking the

144
00:07:06,680 --> 00:07:09,320
Speaker 1: other day because I was actually looking at asking questions

145
00:07:09,320 --> 00:07:12,520
Speaker 1: about the sore throat there's some weeks ago, and I

146
00:07:12,600 --> 00:07:14,280
Speaker 1: knew it wasn't going to be advice. I was thinking,

147
00:07:14,680 --> 00:07:18,760
Speaker 1: if you were researching this, I'm asking it, what would

148
00:07:18,800 --> 00:07:20,640
Speaker 1: you be suggesting of the sorts of things that a

149
00:07:20,680 --> 00:07:23,440
Speaker 1: doctor might prescribe for this? And idmediately it didn't give

150
00:07:23,480 --> 00:07:27,040
Speaker 1: me the call of the provisos. It just went straight

151
00:07:27,080 --> 00:07:29,880
Speaker 1: into having a discussion about it and just on that.

152
00:07:31,800 --> 00:07:34,720
Speaker 1: Sometimes if I might say, well give me a scollar,

153
00:07:35,080 --> 00:07:37,160
Speaker 1: don't just give me an opinion that you scrape off

154
00:07:37,640 --> 00:07:39,720
Speaker 1: you know, Instagram or and if you find this stuff,

155
00:07:40,000 --> 00:07:44,000
Speaker 1: give me a scholarly article which which is peer reviewed,

156
00:07:44,880 --> 00:07:50,600
Speaker 1: and I only want peer reviewed in high ranking journals,

157
00:07:51,480 --> 00:07:54,240
Speaker 1: and it'll go and give you annotations as to where

158
00:07:54,240 --> 00:07:55,920
Speaker 1: it's got the information. You go back and see the

159
00:07:55,920 --> 00:08:00,320
Speaker 1: abstract or whatever the case may be. Let's say it

160
00:08:00,440 --> 00:08:02,920
Speaker 1: was something that you wrote when you did your PhD. Yes,

161
00:08:03,960 --> 00:08:05,360
Speaker 1: how do you feel about that? Do you do you

162
00:08:05,360 --> 00:08:08,480
Speaker 1: feel like? How do you feel like being a so citation?

163
00:08:08,920 --> 00:08:11,800
Speaker 1: You might think that's good, but how do you feel like,

164
00:08:12,200 --> 00:08:14,280
Speaker 1: because then it's your work, how do you feel about that?

165
00:08:16,400 --> 00:08:17,040
Speaker 1: Is it stealing?

166
00:08:18,000 --> 00:08:20,040
Speaker 2: I think it is stealing. I mean, as a scientist,

167
00:08:20,120 --> 00:08:22,640
Speaker 2: I publish my stuff for people who read to make

168
00:08:22,720 --> 00:08:25,720
Speaker 2: the world more ological. So I don't mind for you know,

169
00:08:25,760 --> 00:08:28,600
Speaker 2: my science, but all my books, you know, the books

170
00:08:28,600 --> 00:08:30,880
Speaker 2: I wrote for the public, which gets sold and I

171
00:08:31,360 --> 00:08:35,160
Speaker 2: get a small pittance back, they're also being trained on.

172
00:08:35,400 --> 00:08:38,760
Speaker 2: And I'm lucky enough I've got an income as a

173
00:08:38,760 --> 00:08:41,320
Speaker 2: sciences I don't have to don't have to depend upon

174
00:08:41,360 --> 00:08:44,319
Speaker 2: selling books. But I do worry about the authors and

175
00:08:45,600 --> 00:08:48,760
Speaker 2: the musicians and the graphic designers whose work has all

176
00:08:48,800 --> 00:08:51,560
Speaker 2: been ingested and who depend upon the income which is

177
00:08:51,600 --> 00:08:54,080
Speaker 2: now being undermined by these these bots.

178
00:08:54,120 --> 00:08:56,120
Speaker 1: Because I see that Musk actually put up a tweet

179
00:08:56,120 --> 00:08:59,520
Speaker 1: today on X talking about I think it's sure about anthropic.

180
00:08:59,600 --> 00:09:03,199
Speaker 1: I mean, she's talking about his competition, but saying that

181
00:09:04,080 --> 00:09:06,840
Speaker 1: you know they're stealing stuff does gro doesn't Grock do

182
00:09:06,920 --> 00:09:07,320
Speaker 1: the same thing.

183
00:09:07,600 --> 00:09:09,840
Speaker 2: I mean they all do exactly. So what is it?

184
00:09:10,160 --> 00:09:11,719
Speaker 1: I mean I understand that is he just trying to

185
00:09:11,720 --> 00:09:14,240
Speaker 1: pick a fight with ortmand or something like that? Is Yes,

186
00:09:14,320 --> 00:09:17,120
Speaker 1: I mean he's a grieved because he he puts sold out.

187
00:09:17,200 --> 00:09:20,400
Speaker 2: He put the hundred first hundred million into Open AI.

188
00:09:21,080 --> 00:09:22,560
Speaker 2: It was going to be a you know, not for

189
00:09:22,600 --> 00:09:25,560
Speaker 2: profit for the benefit of humanity. It's no longer. It's

190
00:09:25,559 --> 00:09:27,679
Speaker 2: no longer open, it's no longer not for profit, it's

191
00:09:27,720 --> 00:09:30,240
Speaker 2: no longer seems to be for the benefit the humanity.

192
00:09:30,720 --> 00:09:33,120
Speaker 2: And he's been kicked out. So you know, he's got

193
00:09:33,120 --> 00:09:37,560
Speaker 2: reasons backs because I mean ananthropic you know, have you

194
00:09:37,600 --> 00:09:39,839
Speaker 2: know in the court, have you know just settled one

195
00:09:39,880 --> 00:09:43,800
Speaker 2: and a half billion dollars for the copyright that they took,

196
00:09:43,840 --> 00:09:45,120
Speaker 2: the copyrighted books they took.

197
00:09:46,240 --> 00:09:50,559
Speaker 1: Where what's the business model of these things? I mean,

198
00:09:50,559 --> 00:09:51,560
Speaker 1: how do they make money?

199
00:09:52,400 --> 00:09:55,440
Speaker 2: Well, they're not making money, okay, so how does it work?

200
00:09:55,480 --> 00:09:59,199
Speaker 2: Like Tropic, which makes Claude, is not making money. They're

201
00:09:59,200 --> 00:10:03,400
Speaker 2: probably the Microsoft is not making money. Google is not

202
00:10:03,400 --> 00:10:05,720
Speaker 2: making money on this, none of them. The only company

203
00:10:05,760 --> 00:10:06,920
Speaker 2: that's making money isn't.

204
00:10:06,720 --> 00:10:09,439
Speaker 1: In video yeah, because they're making parts they make.

205
00:10:09,600 --> 00:10:13,199
Speaker 2: Yes, they're making the shovels, but they're making it, and

206
00:10:13,559 --> 00:10:16,000
Speaker 2: they are they are the most valuable company in the world.

207
00:10:16,040 --> 00:10:18,800
Speaker 2: They're worth three or four trillion dollars. Yeah. Today today

208
00:10:19,720 --> 00:10:22,080
Speaker 2: they are the luckiest company. You know, it was by

209
00:10:22,240 --> 00:10:24,719
Speaker 2: chance that they were selling what's proved to be the

210
00:10:24,760 --> 00:10:28,320
Speaker 2: shovels for the AI revolution. They were selling graphics chips

211
00:10:28,320 --> 00:10:30,880
Speaker 2: to people to make games with, and it's so just

212
00:10:31,160 --> 00:10:33,280
Speaker 2: was just by luck turned out to be exactly the

213
00:10:33,320 --> 00:10:33,920
Speaker 2: thing you needed.

214
00:10:34,000 --> 00:10:36,080
Speaker 1: So were you someone who sits back and watches this,

215
00:10:36,760 --> 00:10:38,600
Speaker 1: the growth of this stuff, do you think to yourself,

216
00:10:39,920 --> 00:10:42,240
Speaker 1: how's all this stuff being funded? And who are the

217
00:10:42,240 --> 00:10:44,199
Speaker 1: people funding? Why are they doing this? Why are they're

218
00:10:44,200 --> 00:10:46,520
Speaker 1: prepared to go way out in a limit which told

219
00:10:46,520 --> 00:10:49,800
Speaker 1: me the ortmands and you know, Mask and everybody else

220
00:10:50,320 --> 00:10:53,200
Speaker 1: who's got a platform, what I'm in by platform, an

221
00:10:53,200 --> 00:10:57,760
Speaker 1: AI platform? Why they why are they doing it? Is

222
00:10:58,080 --> 00:11:00,800
Speaker 1: it power grab? What's the deal? What the what do

223
00:11:00,840 --> 00:11:01,200
Speaker 1: you think?

224
00:11:01,240 --> 00:11:03,440
Speaker 2: The thing that surprises me not the technology. The technology

225
00:11:03,480 --> 00:11:05,640
Speaker 2: is where I would have imagined when I when I preaps,

226
00:11:05,800 --> 00:11:08,680
Speaker 2: when I started as But the thing that surprises me

227
00:11:08,920 --> 00:11:11,520
Speaker 2: is the the amount of money and the speed with

228
00:11:11,559 --> 00:11:15,360
Speaker 2: which it's being executed, and I didn't expect it's several

229
00:11:15,400 --> 00:11:19,880
Speaker 2: billion dollars a day being invested into AI. It's a

230
00:11:19,920 --> 00:11:23,679
Speaker 2: third of the world's R and D budget is being spent. Wow, technology,

231
00:11:23,720 --> 00:11:28,920
Speaker 2: we've we've never made such big bets and in some

232
00:11:28,960 --> 00:11:31,120
Speaker 2: sense that that that is being returned. I mean open Ai,

233
00:11:31,440 --> 00:11:34,839
Speaker 2: the company behind chatchapt that's the fastest growing company in

234
00:11:34,880 --> 00:11:37,280
Speaker 2: the history I mean not the history of not just

235
00:11:37,320 --> 00:11:39,520
Speaker 2: the history of AI or technology, the history of any

236
00:11:40,000 --> 00:11:42,880
Speaker 2: any field of endeavor. No company has grown as quickly

237
00:11:42,880 --> 00:11:46,440
Speaker 2: as open Ai. Before chat Chipity was released, they had

238
00:11:46,520 --> 00:11:50,520
Speaker 2: no products, no income. A year later, after chachipep was released,

239
00:11:50,520 --> 00:11:53,800
Speaker 2: they had a billion dollars a year annualize revenue. So

240
00:11:54,360 --> 00:11:58,200
Speaker 2: who pays I mean people are paying subscriptions. Yeah, you know,

241
00:11:58,240 --> 00:12:02,640
Speaker 2: they're doing deals with lots of other companies. They've got

242
00:12:02,960 --> 00:12:05,560
Speaker 2: you know, tens of millions of people paying them, some

243
00:12:05,640 --> 00:12:08,920
Speaker 2: of them paying two hundred dollars a month. They're not

244
00:12:08,960 --> 00:12:11,760
Speaker 2: making money. So they went from a billion dollars a

245
00:12:11,840 --> 00:12:14,320
Speaker 2: year a year later there were four billion. There were

246
00:12:14,360 --> 00:12:16,440
Speaker 2: private companies. So they're not very open about it, but

247
00:12:16,559 --> 00:12:19,800
Speaker 2: they're making somewhere somewhere more than ten billion dollars a

248
00:12:19,880 --> 00:12:23,319
Speaker 2: year in revenue. But there's still enough in terms of

249
00:12:23,320 --> 00:12:25,240
Speaker 2: what this No, but they're spending more than much more.

250
00:12:25,240 --> 00:12:28,079
Speaker 2: They're spending twenty or thirty millions, So they're losing money

251
00:12:28,120 --> 00:12:28,400
Speaker 2: out off.

252
00:12:28,679 --> 00:12:30,240
Speaker 1: But why are they doing it? Because, I mean I

253
00:12:30,240 --> 00:12:33,560
Speaker 1: can understand why you know, Google might have been doing

254
00:12:33,679 --> 00:12:35,400
Speaker 1: or Netflix might have been doing, because they've got a

255
00:12:35,480 --> 00:12:38,280
Speaker 1: model where they sort of hook you in and then

256
00:12:38,400 --> 00:12:40,600
Speaker 1: once you get in there, they then start over time,

257
00:12:40,720 --> 00:12:43,120
Speaker 1: they start to give you a free, free use and

258
00:12:43,120 --> 00:12:44,640
Speaker 1: they say, look, if you don't want to have any ads,

259
00:12:44,679 --> 00:12:47,559
Speaker 1: you can pay some bit extra. That's that Netflix model.

260
00:12:47,880 --> 00:12:50,079
Speaker 1: That model has been done million times over. Do you

261
00:12:50,120 --> 00:12:52,480
Speaker 1: think is it about making money? It's about power?

262
00:12:53,520 --> 00:12:56,959
Speaker 2: Oh, it's a bit of both. There's certainly you know,

263
00:12:57,040 --> 00:13:00,480
Speaker 2: they're trying to capitalize on the first move Advantageactly how

264
00:13:00,720 --> 00:13:02,840
Speaker 2: Google succeeded. You know, if you go back to the

265
00:13:02,840 --> 00:13:05,079
Speaker 2: beginnings of the Internet, I remember there was a there

266
00:13:05,080 --> 00:13:07,680
Speaker 2: was a new aid search engine every day. Do you

267
00:13:07,679 --> 00:13:10,000
Speaker 2: remember Auto Visa. I do. When Auto Vista came out,

268
00:13:10,040 --> 00:13:12,160
Speaker 2: I remember saying to a Fred, oh that's it. Search

269
00:13:12,160 --> 00:13:14,760
Speaker 2: engines are so good now with auto Vista that I'm

270
00:13:14,760 --> 00:13:16,600
Speaker 2: never going to have to change my search engine again,

271
00:13:16,600 --> 00:13:18,480
Speaker 2: because up to that point, every year or two you

272
00:13:18,720 --> 00:13:20,959
Speaker 2: change the search engine. I remember getting to Audo Vester

273
00:13:21,040 --> 00:13:23,120
Speaker 2: and thinking, oh, it's so good. Now I'm not going

274
00:13:23,200 --> 00:13:25,600
Speaker 2: to have to change again. And of course Google came

275
00:13:25,600 --> 00:13:28,520
Speaker 2: out and was much better. But what was interesting is

276
00:13:28,559 --> 00:13:32,160
Speaker 2: that we've we've never had to change after Google because

277
00:13:32,240 --> 00:13:35,120
Speaker 2: Google did. One thing is they got such a market share.

278
00:13:35,400 --> 00:13:37,920
Speaker 2: They've got so much data that even if you came

279
00:13:37,960 --> 00:13:40,520
Speaker 2: out with and there have been better algorithms for searching

280
00:13:40,520 --> 00:13:43,800
Speaker 2: the web, you didn't have the data. And so no

281
00:13:43,920 --> 00:13:46,360
Speaker 2: one can really compete against Google. And if you look

282
00:13:46,440 --> 00:13:48,439
Speaker 2: like you're threatening to compete against Google, of course they

283
00:13:48,440 --> 00:13:51,760
Speaker 2: just buy you out. He is right, So they've really

284
00:13:51,760 --> 00:13:53,760
Speaker 2: Google has got an unassailable lead.

285
00:13:53,960 --> 00:13:55,520
Speaker 1: That monopoly so monopolis.

286
00:13:55,520 --> 00:13:59,080
Speaker 2: It is data monopolis, a natural monopoly. And you know,

287
00:13:59,120 --> 00:14:00,920
Speaker 2: we should do something about that. Which is, you know,

288
00:14:01,559 --> 00:14:05,400
Speaker 2: is it to the consumers benefit that you know, Google

289
00:14:05,520 --> 00:14:09,000
Speaker 2: own YouTube, that Google that Google and the ad business

290
00:14:09,000 --> 00:14:11,800
Speaker 2: that's also selling ads onto the global platform. Is it

291
00:14:11,840 --> 00:14:14,120
Speaker 2: to the consumers you know benefit? There's only one social

292
00:14:14,200 --> 00:14:18,600
Speaker 2: media platform, Facebook or Facebook's you know, other companies WhatsApp

293
00:14:18,679 --> 00:14:21,320
Speaker 2: and Instagram and so on. I don't think it is,

294
00:14:21,560 --> 00:14:23,680
Speaker 2: and we probably should. You know what we do when

295
00:14:23,920 --> 00:14:26,200
Speaker 2: when people get monopolies break them into parts.

296
00:14:26,240 --> 00:14:29,760
Speaker 1: Well, it's funny when I when I think about, you know,

297
00:14:30,040 --> 00:14:33,960
Speaker 1: the Google, Facebook, just generally, then I think about all

298
00:14:33,960 --> 00:14:38,120
Speaker 1: these other these other AI apps that people are using

299
00:14:38,320 --> 00:14:43,480
Speaker 1: and downloading. Immediately think is the movie or the or

300
00:14:43,520 --> 00:14:47,280
Speaker 1: the book written by Frank Boone called The Wizard of

301
00:14:47,280 --> 00:14:54,240
Speaker 1: Oz and you know the so called wondament of you know,

302
00:14:54,840 --> 00:14:57,240
Speaker 1: going to see this w Wizard is at the end

303
00:14:57,280 --> 00:14:59,120
Speaker 1: of the day, is just a dude there with a curtain,

304
00:14:59,520 --> 00:15:03,400
Speaker 1: just one per but this one person controlled the whole story.

305
00:15:03,640 --> 00:15:06,880
Speaker 1: And really there's not that much in it except there's

306
00:15:06,880 --> 00:15:11,240
Speaker 1: one so called wizard. And I often think about that show,

307
00:15:11,440 --> 00:15:13,800
Speaker 1: the movie and the story or you know, the underlying

308
00:15:13,840 --> 00:15:16,000
Speaker 1: story behind, and I think about, you know, what's going on,

309
00:15:16,200 --> 00:15:19,400
Speaker 1: Sam Moltman. You know, we've got a mask, We've got Google,

310
00:15:19,560 --> 00:15:23,280
Speaker 1: We've got Zuckerberg. There's some individuals who control these things,

311
00:15:23,920 --> 00:15:28,520
Speaker 1: say a dozen of them, and are they the Wizard

312
00:15:28,600 --> 00:15:30,520
Speaker 1: of Oz? Just you know, like there's not that much.

313
00:15:30,560 --> 00:15:32,360
Speaker 1: They're just still human beings at the end of the day,

314
00:15:32,480 --> 00:15:35,440
Speaker 1: people with all the law failures and all the foibills,

315
00:15:35,480 --> 00:15:37,680
Speaker 1: and all the advantages and all the personality traits and

316
00:15:38,040 --> 00:15:41,240
Speaker 1: good and bad, And therefore does it therefore make it

317
00:15:41,320 --> 00:15:42,000
Speaker 1: very dangerous?

318
00:15:43,200 --> 00:15:45,520
Speaker 2: It does. I mean, we're seeing a concentration of power

319
00:15:45,520 --> 00:15:49,560
Speaker 2: into the hands of a few very influential, important people.

320
00:15:51,240 --> 00:15:53,480
Speaker 2: Does it worry you though, Yeah, it does, it does,

321
00:15:53,560 --> 00:15:56,600
Speaker 2: and their choices impact upon us. I mean, just just

322
00:15:56,640 --> 00:16:00,160
Speaker 2: the last couple of days, we've been revealed that Facebook

323
00:16:00,520 --> 00:16:05,960
Speaker 2: asked eighteen independent wellness and health experts whether what they

324
00:16:05,960 --> 00:16:10,120
Speaker 2: should do about beauty filters on Instagram, and every one

325
00:16:10,160 --> 00:16:14,360
Speaker 2: of those experts said they're harming young women, you shouldn't

326
00:16:14,400 --> 00:16:19,239
Speaker 2: use them. And Mark Zuckerberg said overall those eighteen independent

327
00:16:19,240 --> 00:16:22,240
Speaker 2: expert opinions and said, no, We're putting the beauty filters

328
00:16:22,240 --> 00:16:23,080
Speaker 2: back in Instagram.

329
00:16:24,120 --> 00:16:27,520
Speaker 1: He went against their advices, every one of them, and.

330
00:16:27,360 --> 00:16:29,520
Speaker 2: What they were going to harm, that they were harming

331
00:16:29,920 --> 00:16:30,480
Speaker 2: young women.

332
00:16:30,720 --> 00:16:33,840
Speaker 1: And did Zuckerberg justify this?

333
00:16:34,000 --> 00:16:38,080
Speaker 2: Yes, classic Silicon value justification. Freedom of speech.

334
00:16:38,880 --> 00:16:40,800
Speaker 1: Ye, there's a throwaway line.

335
00:16:41,000 --> 00:16:41,360
Speaker 2: It is.

336
00:16:41,840 --> 00:16:45,400
Speaker 1: It is, yeah, and it's one of It's a United

337
00:16:45,440 --> 00:16:47,520
Speaker 1: States part of the constitution. So it's hard to sort

338
00:16:47,520 --> 00:16:49,640
Speaker 1: of argue, yes, we don't actually sort of have a

339
00:16:49,720 --> 00:16:52,600
Speaker 1: constitutionalized here in Australia, but maybe that's the reason why

340
00:16:52,680 --> 00:16:56,440
Speaker 1: we have the under sixteens sort of legislation coming through.

341
00:16:56,480 --> 00:16:58,440
Speaker 2: We do, but we have a duty to protect young people.

342
00:16:59,200 --> 00:17:02,160
Speaker 2: We've seen that. You know, these people are not you know,

343
00:17:02,400 --> 00:17:03,680
Speaker 2: mindful of those duties.

344
00:17:04,200 --> 00:17:05,560
Speaker 1: What are the things that worry you most?

345
00:17:05,640 --> 00:17:05,840
Speaker 2: Then?

346
00:17:06,400 --> 00:17:08,840
Speaker 1: Is it more the power concentration of power? I mean,

347
00:17:08,880 --> 00:17:13,040
Speaker 1: I don't like concentration risk, powers risk because in the

348
00:17:13,040 --> 00:17:16,400
Speaker 1: individual any stage we go bonkers or those people random.

349
00:17:16,480 --> 00:17:17,760
Speaker 1: Is that the thing that worries the most?

350
00:17:18,520 --> 00:17:21,239
Speaker 2: My worries go beyond just the concentration of power. Mean,

351
00:17:21,240 --> 00:17:24,160
Speaker 2: that's certainly something to be concerned about. But we've dealt

352
00:17:24,160 --> 00:17:26,159
Speaker 2: with concentration and power in the past. We did in

353
00:17:26,160 --> 00:17:29,919
Speaker 2: the Dust Revolution. We had the robber barons. You know,

354
00:17:29,960 --> 00:17:32,800
Speaker 2: we had to we broke up big oil, big tobacco.

355
00:17:34,119 --> 00:17:37,080
Speaker 2: You know, we we we've we've thought about these issues

356
00:17:37,080 --> 00:17:39,480
Speaker 2: and we do have mechanism of antitrust and so want

357
00:17:39,840 --> 00:17:42,760
Speaker 2: to deal with concentration of powers. But what I worry

358
00:17:42,800 --> 00:17:46,040
Speaker 2: about is the more insidious ways that it's going to

359
00:17:46,080 --> 00:17:50,440
Speaker 2: change our society, whether it undermines our value as humans

360
00:17:50,680 --> 00:17:55,160
Speaker 2: by taking away our work, by by making us feel

361
00:17:55,240 --> 00:17:59,560
Speaker 2: redundant in a society, Whether it undermines our democracy by

362
00:17:59,640 --> 00:18:03,160
Speaker 2: changing the conversations we have our trust in politicians, by

363
00:18:03,240 --> 00:18:07,800
Speaker 2: changing the news that we see, the hearth truths, the mistruths,

364
00:18:07,840 --> 00:18:11,840
Speaker 2: the conspiracy theories that AI generated deep fakes. What sort

365
00:18:11,840 --> 00:18:16,639
Speaker 2: of world it will be when it's filtered through artificial intelligence?

366
00:18:16,760 --> 00:18:18,320
Speaker 1: Do you do feel as though right now we're going

367
00:18:18,359 --> 00:18:21,199
Speaker 1: through a massive transformational period.

368
00:18:21,480 --> 00:18:23,480
Speaker 2: Yes, I do think this is we will look back

369
00:18:24,000 --> 00:18:26,280
Speaker 2: and say, just as we went through a big transformation

370
00:18:27,160 --> 00:18:29,520
Speaker 2: when we went through the Usher Revolution, most of us

371
00:18:29,600 --> 00:18:31,880
Speaker 2: used to work in the farms, out in the fields,

372
00:18:31,880 --> 00:18:34,160
Speaker 2: and then we moved to work in factories and offices.

373
00:18:34,480 --> 00:18:38,520
Speaker 2: That was a big change. When we had the computer

374
00:18:38,560 --> 00:18:42,359
Speaker 2: revolution and we started introducing computers into offices and businesses,

375
00:18:42,440 --> 00:18:44,840
Speaker 2: that was again a big change. And the way that

376
00:18:44,880 --> 00:18:46,800
Speaker 2: we went about our lives. And you know, you look

377
00:18:46,840 --> 00:18:48,479
Speaker 2: at an office, so that looks quite different to it

378
00:18:48,520 --> 00:18:50,800
Speaker 2: did fifty years ago. And I think we're going through

379
00:18:50,840 --> 00:18:54,120
Speaker 2: another big change. And I think what's challenging is that

380
00:18:54,320 --> 00:18:57,200
Speaker 2: those previous changes took times to happen. The dust revolution

381
00:18:57,240 --> 00:19:00,240
Speaker 2: took fifty years to happen, you know, getting people onto

382
00:19:00,280 --> 00:19:02,200
Speaker 2: the internet and doing stuff on the internet. That took

383
00:19:02,240 --> 00:19:05,560
Speaker 2: a decade to happen. This one is happening overnight.

384
00:19:06,080 --> 00:19:08,880
Speaker 1: Is because is that because all of us got devices.

385
00:19:08,920 --> 00:19:09,840
Speaker 1: Now we've all got.

386
00:19:09,680 --> 00:19:12,720
Speaker 2: Devices, we can deliver something. It's not a coincidence that

387
00:19:12,920 --> 00:19:15,679
Speaker 2: chat GPT was the fastest growing app ever, you just

388
00:19:15,680 --> 00:19:18,040
Speaker 2: have to be told, well, it exists, you can put

389
00:19:18,040 --> 00:19:21,120
Speaker 2: the I mean, that's the fantastic opportunity for in terms

390
00:19:21,119 --> 00:19:23,800
Speaker 2: of business, in terms of transformation, is you can get

391
00:19:24,320 --> 00:19:26,800
Speaker 2: stuff into the hands of a billion people overnight. We've

392
00:19:26,840 --> 00:19:31,119
Speaker 2: never previously had a been a method to deliver some

393
00:19:31,520 --> 00:19:36,080
Speaker 2: new experience and new functionality to people overnight by just saying, well,

394
00:19:36,080 --> 00:19:38,879
Speaker 2: there's this URL, just go and try it out. And

395
00:19:38,920 --> 00:19:41,040
Speaker 2: then we've got all the resources, all the data centers

396
00:19:41,040 --> 00:19:43,320
Speaker 2: to spin this up at scale. Right, So it used

397
00:19:43,320 --> 00:19:45,040
Speaker 2: to be your hat to go out and buy yourself

398
00:19:45,040 --> 00:19:48,040
Speaker 2: computers and now you just have to, you know, tell

399
00:19:48,040 --> 00:19:50,680
Speaker 2: one of the providers, you know, we want some water cycles.

400
00:19:51,040 --> 00:19:54,600
Speaker 1: I don't know whether it's exciting or scary or more

401
00:19:54,640 --> 00:19:57,480
Speaker 1: exciting less what it is both, but am I more

402
00:19:57,520 --> 00:20:00,320
Speaker 1: excited than I am scared or am I more scared

403
00:20:00,320 --> 00:20:02,560
Speaker 1: that I am excited? And it's sort of I oscillate

404
00:20:02,600 --> 00:20:06,680
Speaker 1: between the two depends was telling the story, but you

405
00:20:06,840 --> 00:20:10,080
Speaker 1: and you're right. I mean the computer period, the Internet period,

406
00:20:10,200 --> 00:20:13,119
Speaker 1: like from ninety ninety five, so of two thousand and

407
00:20:13,119 --> 00:20:15,480
Speaker 1: five whatever, that a ten year period that seemed to

408
00:20:15,480 --> 00:20:19,320
Speaker 1: me to be incredibly fast. You know, Moore's laws sort

409
00:20:19,320 --> 00:20:22,200
Speaker 1: of doesn't exist anymore, and if they should have something,

410
00:20:22,200 --> 00:20:24,440
Speaker 1: they might just have a new law, Toby's law. Everything

411
00:20:24,480 --> 00:20:27,160
Speaker 1: happens at a much faster pace than has ever happened before.

412
00:20:27,200 --> 00:20:29,320
Speaker 1: And it's used to for two years or something. Everything

413
00:20:29,320 --> 00:20:31,520
Speaker 1: doubles in every eighty months or something along those lines,

414
00:20:32,400 --> 00:20:34,680
Speaker 1: and now it seems to be happening on a monthly basis.

415
00:20:35,040 --> 00:20:38,400
Speaker 1: It's like doubling, tripling on a monthly basis. But if

416
00:20:38,400 --> 00:20:43,400
Speaker 1: I look back, I can go back to fifty years ago,

417
00:20:43,440 --> 00:20:46,120
Speaker 1: when I first started working at twenty, I can go back,

418
00:20:46,160 --> 00:20:48,880
Speaker 1: and I can I know what existed in my office environment.

419
00:20:48,880 --> 00:20:51,240
Speaker 1: There was no computers. You did you typed everything yourself

420
00:20:51,680 --> 00:20:53,719
Speaker 1: or the end or there's a typing pill, depending on

421
00:20:53,960 --> 00:20:55,919
Speaker 1: what business you worked in. And there was lots of

422
00:20:55,920 --> 00:20:57,520
Speaker 1: type of sitting there that you walk and there was

423
00:20:57,560 --> 00:20:59,400
Speaker 1: actually my I worked a law from me that there

424
00:20:59,480 --> 00:21:03,000
Speaker 1: was a room might be maybe twenty five girls. It

425
00:21:03,000 --> 00:21:06,400
Speaker 1: was all women with that type rise of ribbons on them,

426
00:21:06,520 --> 00:21:10,520
Speaker 1: and everything was typed in two in a copy and

427
00:21:10,880 --> 00:21:13,679
Speaker 1: the original duplicate, and then it was we used to

428
00:21:13,680 --> 00:21:16,480
Speaker 1: call it a memo, the original carbon coffee, carbon copy,

429
00:21:17,280 --> 00:21:20,880
Speaker 1: original carbon cop exactly, the original CC. It was an

430
00:21:20,880 --> 00:21:23,920
Speaker 1: actual carbon copy. And and then you would go and

431
00:21:23,960 --> 00:21:26,600
Speaker 1: photocopy that. Some people had the ability to photocopy things,

432
00:21:26,720 --> 00:21:29,320
Speaker 1: and there was just a they were called type is pool,

433
00:21:29,760 --> 00:21:31,960
Speaker 1: and you know, you would write it out by hand

434
00:21:32,000 --> 00:21:33,560
Speaker 1: and you'd hand it over to somebody and they would

435
00:21:33,560 --> 00:21:34,960
Speaker 1: type it out for it, and that was it became

436
00:21:34,960 --> 00:21:37,639
Speaker 1: a memo, a memorandum, and then you would deliver the

437
00:21:37,640 --> 00:21:40,360
Speaker 1: man brand and physically it was never done by email, obviously,

438
00:21:40,840 --> 00:21:44,800
Speaker 1: and then that became your record. And then obviously no phones,

439
00:21:44,840 --> 00:21:47,320
Speaker 1: peop would take days to get back to you. They'd

440
00:21:47,320 --> 00:21:49,720
Speaker 1: have to go write it out of response, and things

441
00:21:49,760 --> 00:21:53,239
Speaker 1: just took a long time to resolve or finalize. And

442
00:21:53,280 --> 00:21:56,440
Speaker 1: then but it's funny, people got very nervous when computers

443
00:21:56,440 --> 00:21:59,080
Speaker 1: started coming in. They thought all these women lose their

444
00:21:59,119 --> 00:22:02,720
Speaker 1: typing jobs. Then I've never gone and examined it. But

445
00:22:03,480 --> 00:22:06,959
Speaker 1: those people got jobs somewhere doing something. They're not typeist anymore.

446
00:22:06,960 --> 00:22:09,639
Speaker 2: But you're right, we don't employ many, if any people

447
00:22:09,640 --> 00:22:12,320
Speaker 2: as type But there were lots of new jobs, invented

448
00:22:12,400 --> 00:22:17,320
Speaker 2: office jobs. Yes, a more interesting jobs, perhaps, I hope.

449
00:22:17,160 --> 00:22:19,359
Speaker 1: Because I mean that would have been pretty boring sitting.

450
00:22:19,119 --> 00:22:21,760
Speaker 2: There just just copying what I ran.

451
00:22:22,480 --> 00:22:25,239
Speaker 1: Yeah, great, well it actually was shorthand so that you're right,

452
00:22:25,440 --> 00:22:28,080
Speaker 1: and missus step because sometimes if you were high enough enough,

453
00:22:28,240 --> 00:22:30,000
Speaker 1: high up enough in the firm, which I wasn't at

454
00:22:30,000 --> 00:22:33,639
Speaker 1: that time, you would dictate it to your what they

455
00:22:33,720 --> 00:22:36,400
Speaker 1: called secretary in those days, not an EA, and then

456
00:22:36,400 --> 00:22:38,280
Speaker 1: she would take copy d take you down by short

457
00:22:38,320 --> 00:22:39,760
Speaker 1: and then she would hand it over to the topping

458
00:22:39,800 --> 00:22:42,480
Speaker 1: pill and we get typed up. In our case a

459
00:22:42,520 --> 00:22:45,240
Speaker 1: law fum get typed up as a document. Then you

460
00:22:45,240 --> 00:22:48,160
Speaker 1: would review it. And at the speed at which things

461
00:22:48,160 --> 00:22:53,720
Speaker 1: happened was remarkably slow. But it also when you went

462
00:22:53,720 --> 00:22:55,720
Speaker 1: on a holiday, you had a good holiday because no

463
00:22:55,720 --> 00:22:59,320
Speaker 1: one could contact you. Do you do you think that

464
00:23:00,080 --> 00:23:05,760
Speaker 1: AI today is putting pressure on us to live a

465
00:23:05,840 --> 00:23:07,360
Speaker 1: much more fast pace life.

466
00:23:07,160 --> 00:23:09,320
Speaker 2: Indeed, and this is why we've now had to You know,

467
00:23:09,440 --> 00:23:12,440
Speaker 2: some countries are introducing laws about you have a ride

468
00:23:12,480 --> 00:23:14,360
Speaker 2: to switch off, you don't have to write, you don't

469
00:23:14,400 --> 00:23:15,520
Speaker 2: have to answer emails.

470
00:23:15,320 --> 00:23:17,439
Speaker 1: After five o'clock or something, but we do it here

471
00:23:17,480 --> 00:23:19,360
Speaker 1: in Australia. That's I think it is a law here

472
00:23:19,359 --> 00:23:21,960
Speaker 1: in our Australia. I think the Albania the entry.

473
00:23:22,000 --> 00:23:24,760
Speaker 2: I mean, it's sad but true that we've actually now

474
00:23:24,760 --> 00:23:27,320
Speaker 2: had to make that law to try and reduce the

475
00:23:27,320 --> 00:23:29,960
Speaker 2: pressure on people, because otherwise you can always be on

476
00:23:30,240 --> 00:23:31,639
Speaker 2: what's the excuse not to reply?

477
00:23:31,800 --> 00:23:34,239
Speaker 1: Well, it's funny because like well before that all come

478
00:23:34,280 --> 00:23:36,760
Speaker 1: in I would type something on my phone. I don't

479
00:23:36,760 --> 00:23:39,240
Speaker 1: even use the laptop or an an iPad here because

480
00:23:39,240 --> 00:23:40,760
Speaker 1: the guys give it to me that I don't. I

481
00:23:40,800 --> 00:23:43,320
Speaker 1: don't use it and I do everything on everything, everything.

482
00:23:43,960 --> 00:23:47,920
Speaker 1: And if I get idea, yeah, we all do it.

483
00:23:48,160 --> 00:23:49,800
Speaker 1: But I put it down there because I just think

484
00:23:49,800 --> 00:23:51,119
Speaker 1: I've got an idea. I got something, I've got to

485
00:23:51,160 --> 00:23:53,560
Speaker 1: send it out to be done. It doesn't get done.

486
00:23:53,600 --> 00:23:56,080
Speaker 1: I mean I said I couldn't send it midnight and

487
00:23:56,200 --> 00:23:58,040
Speaker 1: hopefully someone will pick it up the next day to

488
00:23:58,200 --> 00:24:01,280
Speaker 1: get it to get it sorted. Do you think that

489
00:24:01,880 --> 00:24:05,159
Speaker 1: the big advantage that we're all seeing is that we are,

490
00:24:05,560 --> 00:24:07,800
Speaker 1: as a result of AI and as a result of

491
00:24:08,320 --> 00:24:11,000
Speaker 1: just technology generally, that we are much more efficient or

492
00:24:11,000 --> 00:24:11,800
Speaker 1: are we too efficient?

493
00:24:13,400 --> 00:24:16,120
Speaker 2: Well, it has the potential to make us more efficient,

494
00:24:16,160 --> 00:24:19,320
Speaker 2: but equally, there's a lot of places where it's slow

495
00:24:19,400 --> 00:24:23,639
Speaker 2: thinking that we need. You know, me as a CEO,

496
00:24:24,480 --> 00:24:27,560
Speaker 2: you know, I remember the CEOs that I've respected, the

497
00:24:27,600 --> 00:24:31,080
Speaker 2: people who make Actually a CEO makes very few decisions,

498
00:24:31,480 --> 00:24:34,680
Speaker 2: but they're critical decisions, right, you want that person the

499
00:24:34,840 --> 00:24:37,520
Speaker 2: strategic Yes, critical strategic decisions.

500
00:24:37,560 --> 00:24:38,760
Speaker 1: Hardly, they're not tactical.

501
00:24:39,200 --> 00:24:41,280
Speaker 2: It's not about you don't want the CEO to be

502
00:24:41,320 --> 00:24:44,560
Speaker 2: sitting there micro managing the business. You want the CEO

503
00:24:44,600 --> 00:24:47,320
Speaker 2: to be standing back and saying, Okay, this is the

504
00:24:47,320 --> 00:24:48,919
Speaker 2: way that we're going to go next year, these are

505
00:24:48,920 --> 00:24:50,880
Speaker 2: the products we're going to introduce. This is the way

506
00:24:50,920 --> 00:24:53,520
Speaker 2: that we're going to you know, offer a better service

507
00:24:53,520 --> 00:24:56,159
Speaker 2: to our customers. Those are a few critical decisions. You

508
00:24:56,160 --> 00:24:58,960
Speaker 2: don't actually want them to be lost in the weeds.

509
00:24:59,119 --> 00:25:01,399
Speaker 1: But do you think AI assists the CEO to do that?

510
00:25:01,480 --> 00:25:06,760
Speaker 2: Now? Well, it certainly has the potential and certainly in

511
00:25:06,840 --> 00:25:09,080
Speaker 2: terms in terms of, you know, giving your access to

512
00:25:09,280 --> 00:25:12,400
Speaker 2: much better information about what's going on in the business

513
00:25:12,440 --> 00:25:14,000
Speaker 2: and what the customers want, and so what.

514
00:25:14,400 --> 00:25:17,720
Speaker 1: Can you explain how what people mean by when they

515
00:25:17,760 --> 00:25:25,360
Speaker 1: say BT four, for example, hallucinates in other words, makes

516
00:25:25,440 --> 00:25:29,240
Speaker 1: errors or tricks itself into the wrong answer.

517
00:25:29,680 --> 00:25:35,160
Speaker 2: Yeah, it's it's wrong to think that it's grounded in truth.

518
00:25:35,920 --> 00:25:40,000
Speaker 1: It's so there, So we should not we should just like.

519
00:25:39,960 --> 00:25:44,240
Speaker 2: You shouldn't trust it for your for your cough. Yeah, yeah, yeah,

520
00:25:44,280 --> 00:25:46,680
Speaker 2: because it may actually tell you something that might be

521
00:25:47,320 --> 00:25:50,359
Speaker 2: either not just not helpful to your cough, but actually

522
00:25:50,440 --> 00:25:52,359
Speaker 2: harmful to your what what what?

523
00:25:52,600 --> 00:25:57,240
Speaker 1: So why would I I in in any whatever time

524
00:25:57,359 --> 00:26:01,240
Speaker 1: II you're using whichever protocol, why would it give you

525
00:26:01,280 --> 00:26:02,600
Speaker 1: the wrong answer? And how could it?

526
00:26:02,640 --> 00:26:05,320
Speaker 2: Is it? Because it's remarkable It gives you the correct

527
00:26:05,359 --> 00:26:08,159
Speaker 2: answers most of the time because it doesn't know what

528
00:26:08,200 --> 00:26:09,560
Speaker 2: It's not like you and I. It's not part of

529
00:26:09,560 --> 00:26:11,359
Speaker 2: the world. It doesn't live in the world and know

530
00:26:11,440 --> 00:26:14,680
Speaker 2: what's true and force. It is trained on the internet,

531
00:26:15,600 --> 00:26:17,720
Speaker 2: and it repeats back the sorts of things that it

532
00:26:17,840 --> 00:26:21,320
Speaker 2: sees frequently on the internet. Now frequently what you will

533
00:26:21,320 --> 00:26:24,040
Speaker 2: see on the internet is correct medical advice about how

534
00:26:24,040 --> 00:26:27,919
Speaker 2: to treat a cough. But equally you also find stuff

535
00:26:27,920 --> 00:26:31,720
Speaker 2: that's that's force and stuff that's actually harmful. And so

536
00:26:31,760 --> 00:26:33,520
Speaker 2: sometimes it will repeat some of the stuff. It just

537
00:26:33,560 --> 00:26:37,159
Speaker 2: says what it sees. Most frequently. The best description is

538
00:26:37,160 --> 00:26:40,080
Speaker 2: that on your smartphone, when you're, you know, typing a

539
00:26:40,080 --> 00:26:42,720
Speaker 2: text message, there's the auto complete that finishes the word.

540
00:26:43,480 --> 00:26:46,920
Speaker 2: You know, you type in AP, and the most common

541
00:26:46,920 --> 00:26:50,119
Speaker 2: way of finishing AP is a P P L E. Apple.

542
00:26:51,040 --> 00:26:53,040
Speaker 2: So it will offer Apple as the way to finish

543
00:26:53,040 --> 00:26:55,720
Speaker 2: the world. Well, what chatchup you died? It has done

544
00:26:55,800 --> 00:26:57,760
Speaker 2: is that, but on a much greater scale. It's done

545
00:26:57,760 --> 00:26:59,960
Speaker 2: it not on the not on the scale of center

546
00:27:00,000 --> 00:27:01,720
Speaker 2: this is, but the whole internet. So it tells you

547
00:27:01,880 --> 00:27:04,199
Speaker 2: how to just to finish the word, but the sentence

548
00:27:04,320 --> 00:27:07,600
Speaker 2: or the paragraph or the page trained on the sorts

549
00:27:07,640 --> 00:27:09,520
Speaker 2: of things that sees on it's seen on the internet.

550
00:27:09,600 --> 00:27:13,639
Speaker 1: That's interesting because I've experienced it. So sometimes I find

551
00:27:13,680 --> 00:27:18,720
Speaker 1: that the when I'm typing into but I'll keep saying,

552
00:27:18,760 --> 00:27:20,879
Speaker 1: I'll just keep using copil by way of example, But

553
00:27:20,920 --> 00:27:23,600
Speaker 1: if I'm typing in the actually it corrects, not correct me.

554
00:27:23,800 --> 00:27:28,480
Speaker 1: It finishes off my sentence or my question much faster

555
00:27:29,040 --> 00:27:33,840
Speaker 1: than say, if I'm going into if I'm using auto

556
00:27:33,880 --> 00:27:39,800
Speaker 1: text order correct on text on this mobile text message

557
00:27:39,800 --> 00:27:43,920
Speaker 1: to you for arguments sake, it actually finishes my sentences

558
00:27:43,920 --> 00:27:46,879
Speaker 1: for me, and much more readily. Is because it's obviously

559
00:27:46,960 --> 00:27:47,919
Speaker 1: heard the question before.

560
00:27:48,240 --> 00:27:52,040
Speaker 2: Yes, and a lot of our conversation is quite formulaic. Yeah,

561
00:27:52,520 --> 00:27:55,200
Speaker 2: and so it's been taught those formulas. So if you're

562
00:27:55,520 --> 00:27:57,360
Speaker 2: if you're in a course answer, you're literally reading from

563
00:27:57,359 --> 00:28:00,280
Speaker 2: a script. Yeah, and it's very good at doing exact that.

564
00:28:00,320 --> 00:28:02,600
Speaker 2: But it turns out that a lot of human conversation

565
00:28:03,400 --> 00:28:05,879
Speaker 2: is also quite formulating. We actually go through it. We

566
00:28:05,920 --> 00:28:10,520
Speaker 2: do do that, and we've now taught those formulas to computers.

567
00:28:11,280 --> 00:28:14,000
Speaker 2: So you know, how often do you stop and think

568
00:28:14,119 --> 00:28:16,520
Speaker 2: carefully before you start speaking. You don't know how you're

569
00:28:16,520 --> 00:28:17,080
Speaker 2: going to say the.

570
00:28:17,080 --> 00:28:19,880
Speaker 1: Center, No, I'm just I'm just feeling.

571
00:28:19,800 --> 00:28:21,679
Speaker 2: A lot of the time, we're actually on sort of

572
00:28:21,720 --> 00:28:24,639
Speaker 2: auto pilot. We're actually going through you know, the pleasantries

573
00:28:24,680 --> 00:28:27,160
Speaker 2: of you know, asking how the person is and why,

574
00:28:27,160 --> 00:28:29,840
Speaker 2: how their days being. That's feeling very much a script,

575
00:28:29,880 --> 00:28:31,440
Speaker 2: and we computers can do those scripts.

576
00:28:31,520 --> 00:28:33,520
Speaker 1: Do you think as ai yet worked out how to

577
00:28:34,359 --> 00:28:38,200
Speaker 1: cajole you, in other words, speak to you nicely.

578
00:28:38,480 --> 00:28:44,520
Speaker 2: Yes, so so CHATCHPT was explicitly trained to be somewhat sycophantic,

579
00:28:44,560 --> 00:28:47,400
Speaker 2: to be somewhat flattering. Do you do it agree with you?

580
00:28:47,440 --> 00:28:50,160
Speaker 2: Which is also how it ends up hallucinating, which is

581
00:28:50,200 --> 00:28:52,200
Speaker 2: that it always wants to be helpful and it never

582
00:28:52,240 --> 00:28:53,680
Speaker 2: wants to say you know, no, I don't know the

583
00:28:53,720 --> 00:28:56,719
Speaker 2: answer to that. So it always give you an answer

584
00:28:57,320 --> 00:28:59,640
Speaker 2: even when you know, or give you an answer when

585
00:28:59,640 --> 00:29:02,680
Speaker 2: it's really certainly you know, when everyone on the internet

586
00:29:02,720 --> 00:29:04,760
Speaker 2: says the way to cure a cough is to go

587
00:29:04,800 --> 00:29:06,840
Speaker 2: and get some cough medicine and some cough suites, right,

588
00:29:06,880 --> 00:29:09,640
Speaker 2: so I'll tell you Mark and get some cough medicines

589
00:29:09,640 --> 00:29:12,120
Speaker 2: and co suitts. Because ninety five percent of the people

590
00:29:12,160 --> 00:29:13,880
Speaker 2: talking about how to deal with the cough say that.

591
00:29:14,480 --> 00:29:17,320
Speaker 2: But there's other questions where it's much more fifty to fifty.

592
00:29:18,240 --> 00:29:20,360
Speaker 2: But it's always been told with confidence to tell you

593
00:29:20,400 --> 00:29:22,160
Speaker 2: the answer and not to tell you, oh, I'm a

594
00:29:22,160 --> 00:29:26,120
Speaker 2: bit unconfident about this. Well, you know, I've seen both

595
00:29:26,160 --> 00:29:28,640
Speaker 2: possibilities here, So what about I.

596
00:29:28,840 --> 00:29:32,640
Speaker 1: I Well, our fisial intelligence putting into robots.

597
00:29:33,800 --> 00:29:36,320
Speaker 2: Well, if we want, you know, robots to do interesting

598
00:29:36,400 --> 00:29:38,560
Speaker 2: things and to react and behave in the world, the

599
00:29:38,640 --> 00:29:42,960
Speaker 2: uncertain world, we've got to give them some intelligence. And indeed,

600
00:29:43,160 --> 00:29:45,280
Speaker 2: you know, there's a body of thought that says we're

601
00:29:45,280 --> 00:29:49,560
Speaker 2: not going to get really intelligence, artificial intelligence until we

602
00:29:49,880 --> 00:29:53,000
Speaker 2: have it out in the world experiencing the rich complexity

603
00:29:53,040 --> 00:29:55,320
Speaker 2: of the world. So the way that we're going to

604
00:29:55,360 --> 00:29:59,120
Speaker 2: get real artificial intelligence, really intelligent is by putting it

605
00:29:59,120 --> 00:30:02,520
Speaker 2: into robots, getting the robots to do stuff in the world.

606
00:30:02,880 --> 00:30:04,440
Speaker 1: Say, for example, we build a robot.

607
00:30:05,280 --> 00:30:06,880
Speaker 2: You know, a self driving car is in some sense

608
00:30:06,920 --> 00:30:09,640
Speaker 2: of robots. Yeah, it can sense the world. It can see,

609
00:30:09,760 --> 00:30:12,280
Speaker 2: you know, where the pedestrians are and the other cars,

610
00:30:12,520 --> 00:30:14,760
Speaker 2: makes decisions about where to go, and then it acts

611
00:30:14,760 --> 00:30:17,280
Speaker 2: on those decisions in a physical way. It turns, the

612
00:30:17,320 --> 00:30:19,920
Speaker 2: wheel of the car, pushes the accelerator, goes on the break.

613
00:30:20,040 --> 00:30:22,400
Speaker 2: So it's got all the characteristics of a robot. All

614
00:30:22,520 --> 00:30:25,680
Speaker 2: it doesn't look like a Hollywood robot, you know, it

615
00:30:25,800 --> 00:30:26,880
Speaker 2: is a robot.

616
00:30:27,680 --> 00:30:30,040
Speaker 1: So, like, let's just pick an example. Let's say we're

617
00:30:30,080 --> 00:30:32,720
Speaker 1: building a robot to put it into a nursing home. Yes,

618
00:30:32,960 --> 00:30:34,760
Speaker 1: there's a lot of older people who need to take

619
00:30:34,760 --> 00:30:38,120
Speaker 1: their tablets at a certain time, and maybe it knows

620
00:30:38,120 --> 00:30:41,120
Speaker 1: to sing a song to the audience who is in

621
00:30:41,200 --> 00:30:43,880
Speaker 1: the nursing home because maybe they've got Alzheimer's or something

622
00:30:43,920 --> 00:30:49,680
Speaker 1: that it might be just be nice that robot is

623
00:30:49,720 --> 00:30:53,880
Speaker 1: also learning about these people. Do you see dangers that

624
00:30:53,960 --> 00:30:56,720
Speaker 1: it might stuff up the medication? It might give you

625
00:30:56,800 --> 00:31:01,040
Speaker 1: my medication? Me might you know me your medicaid? Are

626
00:31:01,040 --> 00:31:04,280
Speaker 1: the dangers that there are dangers? But I'm actually much

627
00:31:04,320 --> 00:31:07,000
Speaker 1: more concerned about humans stuffing up. If you look at

628
00:31:07,200 --> 00:31:10,040
Speaker 1: the number of people who get misprescribed in hospitals and

629
00:31:10,480 --> 00:31:13,640
Speaker 1: in pharmacies, it's you. Yeah, people you know, get ten

630
00:31:13,680 --> 00:31:16,320
Speaker 1: times the dose by mistake because they just read, you know,

631
00:31:16,440 --> 00:31:18,840
Speaker 1: a thousand times of dose because the risk means grams

632
00:31:18,840 --> 00:31:22,560
Speaker 1: for micrograms and so on. So humans are terrible making

633
00:31:22,600 --> 00:31:25,960
Speaker 1: that and indeed getting AI to dispense is potentially going

634
00:31:25,960 --> 00:31:29,440
Speaker 1: to reduce those errors. But equally, you know, I don't

635
00:31:29,440 --> 00:31:31,240
Speaker 1: want to be in an old person's home with robots.

636
00:31:31,280 --> 00:31:33,080
Speaker 1: I want to be an old person at home with

637
00:31:33,160 --> 00:31:36,400
Speaker 1: attractive nurses look after me. I can float what a

638
00:31:36,400 --> 00:31:37,880
Speaker 1: better an attractive robot nurse?

639
00:31:37,880 --> 00:31:39,600
Speaker 2: But I do want the nurse, you know, I do

640
00:31:39,680 --> 00:31:43,160
Speaker 2: want the nurse to have maybe a robot so that

641
00:31:43,480 --> 00:31:46,200
Speaker 2: you know, lifting a person out of the bath or

642
00:31:46,240 --> 00:31:49,280
Speaker 2: you know, onto the bed instead of them putting their

643
00:31:49,320 --> 00:31:51,920
Speaker 2: back out doing that. The robot can do those sorts

644
00:31:51,960 --> 00:31:55,000
Speaker 2: of things. But I want to have the empathy, the

645
00:31:55,040 --> 00:32:00,160
Speaker 2: companionship of a person who's going to sit down and say,

646
00:32:00,240 --> 00:32:01,920
Speaker 2: you're looking a bit glum today, so they have a

647
00:32:01,920 --> 00:32:04,160
Speaker 2: cup of tea with me. Let's tell me what's troubling you.

648
00:32:04,240 --> 00:32:06,200
Speaker 1: But maybe the robody in the nursing home could go

649
00:32:06,240 --> 00:32:10,200
Speaker 1: along to the actual nurse of physicalness like a biological person,

650
00:32:10,760 --> 00:32:13,720
Speaker 1: and say to their biologicalness, look, I think you've got

651
00:32:13,720 --> 00:32:15,960
Speaker 1: to stop right now. Just Toby's the one who going

652
00:32:16,000 --> 00:32:18,640
Speaker 1: to concentrate on today, because Toby's looking a little bit glumb,

653
00:32:18,880 --> 00:32:20,800
Speaker 1: you know, because we saw through our sensing that his

654
00:32:21,000 --> 00:32:22,920
Speaker 1: face is a bit down or whatever the case may be.

655
00:32:22,960 --> 00:32:27,680
Speaker 1: How the sensing works based on like previous knowledge of

656
00:32:27,720 --> 00:32:31,120
Speaker 1: your face features and can go and tell the noess

657
00:32:31,240 --> 00:32:33,680
Speaker 1: That makes sense. But maybe I would just change the

658
00:32:34,280 --> 00:32:37,920
Speaker 1: change the narrative. What about it in the military robots

659
00:32:37,920 --> 00:32:40,040
Speaker 1: in the military environment, So in other words, don't send

660
00:32:40,040 --> 00:32:42,920
Speaker 1: any soldiers out to do the fighting. Send robots out.

661
00:32:44,240 --> 00:32:46,960
Speaker 2: That sounds like an attractive issier. Right, let's get humans

662
00:32:46,960 --> 00:32:50,280
Speaker 2: out of the battlefit. That's what would you know, what

663
00:32:50,320 --> 00:32:52,280
Speaker 2: would not be a goode or drones sort of do

664
00:32:52,360 --> 00:32:55,840
Speaker 2: that now they do, and unfortunately they're turning warfare into

665
00:32:55,880 --> 00:33:02,000
Speaker 2: more deadly dangerous activity. And the other challenge challenge with

666
00:33:02,040 --> 00:33:05,400
Speaker 2: that that simple scenario is that there isn't a separate

667
00:33:05,400 --> 00:33:08,680
Speaker 2: part of the world called the battlefield wars unfortunately, a

668
00:33:08,800 --> 00:33:12,320
Speaker 2: fort in amongst towns and cities in and amongst and

669
00:33:12,360 --> 00:33:17,080
Speaker 2: against civilian populations. And so yes, we could DEVELOPE and

670
00:33:17,120 --> 00:33:20,320
Speaker 2: we are developing you know, AIS and robots that could

671
00:33:20,360 --> 00:33:21,880
Speaker 2: go and fight for us.

672
00:33:22,160 --> 00:33:25,520
Speaker 1: We are, are we well, yes, like those.

673
00:33:25,560 --> 00:33:28,520
Speaker 2: Of the US Department of Defense has always been the

674
00:33:28,560 --> 00:33:32,360
Speaker 2: greatest funder of AI research really, yes, ever since the

675
00:33:32,480 --> 00:33:34,760
Speaker 2: nineteen fifties, because they could see.

676
00:33:35,520 --> 00:33:37,800
Speaker 1: In a military sense, or it is generally in a

677
00:33:37,840 --> 00:33:41,120
Speaker 1: military sense, right, Yes, all aspects of how we go

678
00:33:41,160 --> 00:33:44,000
Speaker 1: about warfare. I mean, I don't know, I've got no idea,

679
00:33:44,000 --> 00:33:45,640
Speaker 1: but I can imagine China would be doing it.

680
00:33:45,880 --> 00:33:47,880
Speaker 2: China's doing it as well, Russia's doing it. Yes, there's

681
00:33:47,880 --> 00:33:49,960
Speaker 2: a there is an ai arms race.

682
00:33:50,000 --> 00:33:52,720
Speaker 1: Literally, there's an arms race at that level.

683
00:33:52,880 --> 00:33:58,080
Speaker 2: Yes, and looking at what's happening in Ukraine is intensifying

684
00:33:58,160 --> 00:34:01,440
Speaker 2: that because the militaries are seeing the nature of warfare is.

685
00:34:01,480 --> 00:34:04,480
Speaker 1: Danger but it's also politically dangerous because because you lose

686
00:34:04,480 --> 00:34:09,280
Speaker 1: your people, and that's political danger for example happening with Putin.

687
00:34:09,680 --> 00:34:13,160
Speaker 1: The more young boys who die on the front, the

688
00:34:13,239 --> 00:34:16,560
Speaker 1: more parents are going to be disaffected or dissatisfied, and

689
00:34:16,640 --> 00:34:19,359
Speaker 1: the greater the chance or the risk of him losing

690
00:34:19,440 --> 00:34:22,040
Speaker 1: his grip on the country. Yes, so if I can

691
00:34:22,080 --> 00:34:24,640
Speaker 1: send out there, I'm going to send robots said there.

692
00:34:24,640 --> 00:34:25,799
Speaker 1: Everyone's going to wow, that's cool.

693
00:34:26,600 --> 00:34:29,440
Speaker 2: It is, but it's making warfare much more deadly.

694
00:34:29,239 --> 00:34:31,960
Speaker 1: Activity or particularly for the weak aside.

695
00:34:33,920 --> 00:34:35,480
Speaker 2: I mean, that's a challenge. If you look at what's

696
00:34:35,480 --> 00:34:39,360
Speaker 2: happened in the Ukraine, You've got Russia nation ten times

697
00:34:39,360 --> 00:34:42,799
Speaker 2: the size, ten times the economy of Ukraine, but it's

698
00:34:42,840 --> 00:34:46,719
Speaker 2: being hells at bay by the Ukrainians who are embracing

699
00:34:46,760 --> 00:34:49,200
Speaker 2: these technologies. In that sense, you know it's being used

700
00:34:49,200 --> 00:34:53,040
Speaker 2: for a good purpose but being able to hold back

701
00:34:53,320 --> 00:34:56,799
Speaker 2: the Russian oppressor. But equally you see that you know

702
00:34:56,880 --> 00:35:00,040
Speaker 2: it's going to change the balance of power stability of

703
00:35:00,080 --> 00:35:04,640
Speaker 2: the world completely. Turkey, Iran A becoming major drone powers.

704
00:35:06,239 --> 00:35:08,680
Speaker 2: They're going to be able to compete against much bigger nations.

705
00:35:08,680 --> 00:35:11,040
Speaker 2: I mean, it used to be your powers determine your

706
00:35:11,040 --> 00:35:14,520
Speaker 2: military powers determined by your economic might. Yeah, could you

707
00:35:14,560 --> 00:35:18,360
Speaker 2: buy yeh thirty five fighters, could you buy the aircraft

708
00:35:18,400 --> 00:35:19,960
Speaker 2: carriers and so on things? And you know it was

709
00:35:20,000 --> 00:35:24,359
Speaker 2: the US and Russia and now China. But now increasingly

710
00:35:24,760 --> 00:35:27,879
Speaker 2: your military might is with these cheap drones, and it's

711
00:35:29,480 --> 00:35:33,080
Speaker 2: small nations can get the whold hand on these equipments,

712
00:35:33,080 --> 00:35:37,120
Speaker 2: and it's no longer the case that the balance of

713
00:35:37,160 --> 00:35:38,279
Speaker 2: power is the same as it was.

714
00:35:38,360 --> 00:35:41,000
Speaker 1: And where does AI fee into that for example? Like

715
00:35:41,239 --> 00:35:42,720
Speaker 1: drones for example.

716
00:35:42,960 --> 00:35:45,080
Speaker 2: Well, the problem is that the weak thing with the

717
00:35:45,160 --> 00:35:47,560
Speaker 2: drone is the radio link and indeed it's you know,

718
00:35:47,719 --> 00:35:51,399
Speaker 2: electronic warfare will stop most of the drones these days

719
00:35:51,400 --> 00:35:53,480
Speaker 2: in the Ukraine. So most of the drones now are

720
00:35:53,520 --> 00:35:57,000
Speaker 2: flown with fiber optics, which you can't be interrupted by

721
00:35:57,880 --> 00:36:02,040
Speaker 2: by radio jamming or increasing me. They're autonomous, they're making

722
00:36:02,080 --> 00:36:03,799
Speaker 2: their own decisions because then you've got a much more

723
00:36:03,840 --> 00:36:08,080
Speaker 2: pow weapon. You can't stop it by jamming the radio.

724
00:36:08,200 --> 00:36:11,440
Speaker 2: You can't stop it by killing the pilots.

725
00:36:11,520 --> 00:36:13,360
Speaker 1: You know, the beausori pre trained.

726
00:36:13,719 --> 00:36:17,560
Speaker 2: Yes, it's trained to identify the targets and to go

727
00:36:17,640 --> 00:36:20,719
Speaker 2: and destroy those targets automatically with no human intervention.

728
00:36:21,040 --> 00:36:23,239
Speaker 1: So there's not there is. I think what you're saying

729
00:36:23,320 --> 00:36:26,800
Speaker 1: is I think what you're saying is that the drone

730
00:36:26,840 --> 00:36:30,160
Speaker 1: is preloaded with let's call it intelligence, but preloaded with

731
00:36:30,520 --> 00:36:35,239
Speaker 1: a plan, and there's nobody sort of sitting at the

732
00:36:35,280 --> 00:36:38,200
Speaker 1: other end, sort of like like we normously droned people

733
00:36:38,239 --> 00:36:40,680
Speaker 1: with a like a bat or a computer sort of

734
00:36:40,719 --> 00:36:44,719
Speaker 1: directing the traffic. Because of the jamming processes, it's probably ineffective.

735
00:36:45,000 --> 00:36:47,680
Speaker 1: So these things are preloaded. And they just said there

736
00:36:47,760 --> 00:36:50,879
Speaker 1: might be a when you identify when when you see

737
00:36:50,920 --> 00:36:57,600
Speaker 1: when the drone sees that building kamikaze suicide into it.

738
00:36:57,640 --> 00:36:59,840
Speaker 2: Or when it sees something looks like a Russian tanka,

739
00:37:00,280 --> 00:37:02,600
Speaker 2: or when it sees something that looks like a Russian

740
00:37:02,600 --> 00:37:03,440
Speaker 2: soldier it.

741
00:37:03,400 --> 00:37:05,319
Speaker 1: Looks like to me, then that seems to be then

742
00:37:05,920 --> 00:37:09,600
Speaker 1: open for a lot of error. Yes, later, I mean

743
00:37:09,840 --> 00:37:12,080
Speaker 1: because it might not. It might see the target, but

744
00:37:12,120 --> 00:37:14,680
Speaker 1: there might be a family sitting there into having a

745
00:37:14,680 --> 00:37:15,680
Speaker 1: cup of coffee, indeed.

746
00:37:15,719 --> 00:37:17,960
Speaker 2: And that's that's one of the challenges, which is, you know,

747
00:37:18,000 --> 00:37:22,479
Speaker 2: we fight war according to certain laws, international managed Toian law,

748
00:37:23,480 --> 00:37:25,640
Speaker 2: and there's you know, we call people to account for

749
00:37:25,680 --> 00:37:28,160
Speaker 2: their actions. You can't just go you know, you're only

750
00:37:28,160 --> 00:37:32,920
Speaker 2: supposed to target military they targets. You're not supposed to

751
00:37:32,920 --> 00:37:36,560
Speaker 2: target civilians. You know, the target has to be proportional

752
00:37:36,560 --> 00:37:38,520
Speaker 2: to the threat that you face, and so on those

753
00:37:38,960 --> 00:37:43,000
Speaker 2: those we hold humans to conventions and when they violate

754
00:37:43,040 --> 00:37:45,200
Speaker 2: those those we you know, there's a cause of the

755
00:37:45,239 --> 00:37:47,920
Speaker 2: Hague and various other places that we call people people

756
00:37:48,000 --> 00:37:51,000
Speaker 2: to account. But we can only call hold people to account.

757
00:37:51,160 --> 00:37:53,480
Speaker 2: Our legal system only holds people to account.

758
00:37:53,480 --> 00:37:55,000
Speaker 1: We can't hold a drone account.

759
00:37:55,480 --> 00:37:56,479
Speaker 2: What could we do with the drone?

760
00:37:56,640 --> 00:37:56,920
Speaker 1: Nothing?

761
00:37:58,800 --> 00:38:01,040
Speaker 2: So that's you know, that's another big question, which is,

762
00:38:01,080 --> 00:38:04,480
Speaker 2: you know where mistakes happened when trosses, who are we

763
00:38:04,520 --> 00:38:05,440
Speaker 2: going to hold account?

764
00:38:05,800 --> 00:38:13,080
Speaker 1: Do they various platforms around the world. Are they interfacing

765
00:38:13,480 --> 00:38:19,399
Speaker 1: with military robots which use AI or which will use

766
00:38:19,440 --> 00:38:21,560
Speaker 1: AI in the future. I mean they got some sort

767
00:38:21,560 --> 00:38:25,240
Speaker 1: of arrangement. There was a this pick open AI for example.

768
00:38:25,480 --> 00:38:28,200
Speaker 1: Are they do they do deals with the West government?

769
00:38:28,200 --> 00:38:31,280
Speaker 1: To actually help the Uist defense.

770
00:38:31,520 --> 00:38:38,520
Speaker 2: Yes, yes, well today the CEO of Anthropic, the company

771
00:38:38,520 --> 00:38:42,560
Speaker 2: that makes Claude, has been summoned to give meet Hesketh

772
00:38:42,800 --> 00:38:45,040
Speaker 2: the Pentagon.

773
00:38:44,560 --> 00:38:47,040
Speaker 1: Too is the Secretary of War.

774
00:38:47,120 --> 00:38:51,160
Speaker 2: The Secondary of War, to talk about their relationship and

775
00:38:51,200 --> 00:38:57,120
Speaker 2: about apparently the latest military targeting is now being using

776
00:38:57,280 --> 00:39:00,360
Speaker 2: tools like Anthropic to help them make their decisions. And

777
00:39:00,400 --> 00:39:02,960
Speaker 2: we see this also in Israel. We see that that,

778
00:39:03,040 --> 00:39:05,360
Speaker 2: you know, lots of the idea of the Israeli defense

779
00:39:05,440 --> 00:39:10,800
Speaker 2: forces targeting decisions in the Gaza conflict were not human decisions.

780
00:39:10,840 --> 00:39:14,160
Speaker 2: There were decisions that the that the algorithm AI algorithms,

781
00:39:14,200 --> 00:39:17,080
Speaker 2: algorithms that are called lavender, Where's Daddy, they get some

782
00:39:17,120 --> 00:39:20,680
Speaker 2: funny names, but it's not humans are deciding. You know,

783
00:39:21,000 --> 00:39:24,880
Speaker 2: we think that there are some some hesbula leadership in

784
00:39:25,000 --> 00:39:27,560
Speaker 2: this tower block. We're going to send some artillery in

785
00:39:27,600 --> 00:39:30,520
Speaker 2: against this tower block. That's an AI that's taking in

786
00:39:30,600 --> 00:39:33,759
Speaker 2: all the intelligence information, all the information they've collected on

787
00:39:33,800 --> 00:39:38,120
Speaker 2: social media over many years and calculating where they think

788
00:39:38,200 --> 00:39:39,640
Speaker 2: maybe are potential targets.

789
00:39:39,800 --> 00:39:43,760
Speaker 1: As a mathematician who were a methodician, we are are

790
00:39:43,800 --> 00:39:48,560
Speaker 1: we talking about probability based decisions in we are and

791
00:39:48,880 --> 00:39:49,840
Speaker 1: which is dangerous?

792
00:39:50,000 --> 00:39:54,360
Speaker 2: Which is dangerous because never one hundred percent which and

793
00:39:54,400 --> 00:39:57,839
Speaker 2: there's it's not clear that we really understand what those

794
00:39:57,880 --> 00:40:01,120
Speaker 2: decisions are. They're talking, they're taking vast amount information and

795
00:40:01,160 --> 00:40:06,200
Speaker 2: synthesizing it together, so it's not clear on what basis

796
00:40:06,200 --> 00:40:08,400
Speaker 2: decision is being made and who's going to be accountable.

797
00:40:08,400 --> 00:40:10,440
Speaker 2: And it's wrong. And certainly when you look at what

798
00:40:10,480 --> 00:40:14,520
Speaker 2: was happening in Ghaza, the amounts of collateral damage, would

799
00:40:14,520 --> 00:40:17,880
Speaker 2: you make whole place squenzes or hundreds of civilians that

800
00:40:17,920 --> 00:40:22,960
Speaker 2: were being killed alongside, you know, supposed military targets, you

801
00:40:23,000 --> 00:40:26,360
Speaker 2: were thinking, this is not proportional to the threat of response.

802
00:40:26,480 --> 00:40:28,279
Speaker 1: Do you think that? Do you think that right now

803
00:40:28,320 --> 00:40:34,279
Speaker 1: we're experiencing in Gaza and also probably in Ukraine? Are

804
00:40:34,320 --> 00:40:37,360
Speaker 1: we and are we experiencing the worst of what I

805
00:40:37,560 --> 00:40:40,720
Speaker 1: has got to offer right now? And we're learning lessons

806
00:40:40,719 --> 00:40:42,799
Speaker 1: from it? And it is somebody taking it up? Is

807
00:40:42,840 --> 00:40:48,080
Speaker 1: someone doing something about this? Well they're just combinants with Well.

808
00:40:48,400 --> 00:40:50,239
Speaker 2: Next week I'm off to speak at the United Nations

809
00:40:50,320 --> 00:40:54,160
Speaker 2: in Geneva about these very topics that you know, I

810
00:40:54,160 --> 00:40:57,840
Speaker 2: don't think we're I don't think we're mindful enough of

811
00:40:58,040 --> 00:41:02,839
Speaker 2: the world that we're opening. We get to choose what

812
00:41:02,920 --> 00:41:05,759
Speaker 2: technologies get used where. I mean we sometimes forget that

813
00:41:05,840 --> 00:41:09,200
Speaker 2: we think we have agency some things that you know,

814
00:41:09,280 --> 00:41:12,319
Speaker 2: we we decide, we make conscious decisions, remember, you know,

815
00:41:12,840 --> 00:41:15,400
Speaker 2: we we we made you know, I grew up in

816
00:41:15,440 --> 00:41:17,799
Speaker 2: the UK and they made a conscious decision that there

817
00:41:17,800 --> 00:41:20,320
Speaker 2: weren't going to be adverts on the main channel of TV,

818
00:41:21,000 --> 00:41:22,680
Speaker 2: which is like a fantastic decision.

819
00:41:24,040 --> 00:41:24,800
Speaker 1: It's a BBC.

820
00:41:24,920 --> 00:41:27,240
Speaker 2: The BBC doesn't have adverts still today.

821
00:41:27,040 --> 00:41:28,879
Speaker 1: The same as ABC here doesn't have.

822
00:41:28,800 --> 00:41:31,600
Speaker 2: Adverts Like that was a fantastic decision. And you see

823
00:41:31,600 --> 00:41:33,760
Speaker 2: what happens when you put adverts on TV, which becomes

824
00:41:33,760 --> 00:41:37,080
Speaker 2: you know, driven by the need to satisfy the advertisers.

825
00:41:37,080 --> 00:41:40,520
Speaker 2: And as you know, the ABC the BBC are fantastic

826
00:41:41,719 --> 00:41:45,960
Speaker 2: institutions for public good, for informing the public without the

827
00:41:46,560 --> 00:41:50,280
Speaker 2: pressures and the biases that having to please advertisers brings.

828
00:41:50,680 --> 00:41:52,680
Speaker 2: That was a conscious decision. You know, you can put

829
00:41:52,719 --> 00:41:56,120
Speaker 2: adverts on TV in many countries, that's the only thing

830
00:41:56,160 --> 00:42:00,600
Speaker 2: you get. Similarly with you know, AI, we can choose

831
00:42:01,000 --> 00:42:03,239
Speaker 2: how it's going to be used and where it's going

832
00:42:03,239 --> 00:42:06,160
Speaker 2: to be used. I mean we we forget there.

833
00:42:06,000 --> 00:42:09,840
Speaker 1: Are there are we they're responsible people were the planet.

834
00:42:10,040 --> 00:42:12,640
Speaker 1: But do you think, but what if someone in China

835
00:42:12,680 --> 00:42:14,600
Speaker 1: doesn't agree and they just go and do something completely

836
00:42:14,600 --> 00:42:17,640
Speaker 1: different to what you where you decided, for example, they're

837
00:42:17,680 --> 00:42:18,280
Speaker 1: not nations.

838
00:42:18,400 --> 00:42:21,080
Speaker 2: Yeah, I mean these are accessible technologies, right, so that

839
00:42:21,400 --> 00:42:24,759
Speaker 2: there's nothing that can stop people who really seriously wanted

840
00:42:24,800 --> 00:42:27,239
Speaker 2: to do that. But there's a whole raft of technologies,

841
00:42:27,360 --> 00:42:31,439
Speaker 2: you know, chemical weapons, biological weapons, cluster dimmunitions, blinding lasers.

842
00:42:31,480 --> 00:42:34,120
Speaker 2: There's actually quite a bit of stuff that we've we've

843
00:42:34,120 --> 00:42:40,640
Speaker 2: actually got UN conventions on and we don't use even

844
00:42:40,640 --> 00:42:42,759
Speaker 2: though it would. You know, we can't stop people using

845
00:42:42,800 --> 00:42:47,480
Speaker 2: chemical weapons. Yes, said for example, Star uses them, Fusions

846
00:42:47,480 --> 00:42:50,160
Speaker 2: probably used them. You know, they occasionally get used. But

847
00:42:50,200 --> 00:42:53,560
Speaker 2: when they do, the world is outraged. There's resolutions on

848
00:42:53,600 --> 00:42:55,560
Speaker 2: the floor of the u N there's headlines in the

849
00:42:55,560 --> 00:42:56,160
Speaker 2: New York Times.

850
00:42:56,160 --> 00:42:57,640
Speaker 1: But do you think it means any.

851
00:42:59,000 --> 00:43:02,840
Speaker 2: I think the world will be a much darker, dangerous

852
00:43:02,880 --> 00:43:07,440
Speaker 2: place if we didn't, if we morally hadn't decided that

853
00:43:07,440 --> 00:43:09,319
Speaker 2: that's not the way that we should need to fight war.

854
00:43:10,560 --> 00:43:13,080
Speaker 2: What troubles me is that, you know, I think we'll

855
00:43:13,120 --> 00:43:16,200
Speaker 2: get there with the darker uses of AI in a

856
00:43:16,280 --> 00:43:19,439
Speaker 2: military setting. What worries me is that most of those

857
00:43:19,480 --> 00:43:22,319
Speaker 2: other examples, like chemical weapons, it was only after we

858
00:43:22,320 --> 00:43:24,359
Speaker 2: saw the horrors, for example, the horrors of the First

859
00:43:24,360 --> 00:43:27,320
Speaker 2: World War, you know what the First World World poets

860
00:43:27,320 --> 00:43:31,120
Speaker 2: and people told us about. When we saw that happening,

861
00:43:31,560 --> 00:43:34,960
Speaker 2: did we have the courage to decide, actually, we should,

862
00:43:35,000 --> 00:43:39,319
Speaker 2: you know, try together collectively globally, we should do something

863
00:43:39,320 --> 00:43:41,960
Speaker 2: about this. And we've done something about this and limited

864
00:43:42,000 --> 00:43:45,200
Speaker 2: the use to which chemical weapons we go. What I

865
00:43:45,320 --> 00:43:48,480
Speaker 2: worry that we're going to have to see atrocities of

866
00:43:48,560 --> 00:43:50,800
Speaker 2: AI being used on the battlefield and so on to

867
00:43:51,480 --> 00:43:55,160
Speaker 2: murder civilians before we have the courage to actually take action.

868
00:43:56,200 --> 00:43:58,759
Speaker 2: I say to people, it's easy to imagine what that

869
00:43:58,800 --> 00:44:00,759
Speaker 2: world will be if we don't do any thing. It

870
00:44:00,800 --> 00:44:03,239
Speaker 2: will look like one of those bad Hollywood movies. Will

871
00:44:03,280 --> 00:44:04,799
Speaker 2: it look like what happened Hiroshima?

872
00:44:04,960 --> 00:44:07,360
Speaker 1: Then they dropped the atomic bondah and you had those

873
00:44:07,760 --> 00:44:09,880
Speaker 1: that are the photo that's in my mind or the

874
00:44:10,160 --> 00:44:13,120
Speaker 1: images in my mind, is that little girl running down

875
00:44:13,160 --> 00:44:16,600
Speaker 1: the street naked with burns or over a body like

876
00:44:17,480 --> 00:44:18,640
Speaker 1: her arms outstretched.

877
00:44:19,480 --> 00:44:22,720
Speaker 2: But these will be used to target women and children

878
00:44:22,800 --> 00:44:24,360
Speaker 2: and civilians and innocent people.

879
00:44:25,280 --> 00:44:27,319
Speaker 1: You're going to speaking of the United Nations and you

880
00:44:27,320 --> 00:44:29,839
Speaker 1: said next week, yes, So what are you going to tell.

881
00:44:29,760 --> 00:44:32,200
Speaker 2: Them, I mean, say, we can have to make these choices,

882
00:44:33,000 --> 00:44:37,000
Speaker 2: that it's wrong to think that, you know, the US

883
00:44:37,000 --> 00:44:38,719
Speaker 2: has been pushing back against this. The China has been

884
00:44:38,719 --> 00:44:43,080
Speaker 2: pushing against this because because I think they wrongly feel

885
00:44:43,160 --> 00:44:46,359
Speaker 2: that they've they've got some strategic advantage. They've got, you know,

886
00:44:46,360 --> 00:44:49,160
Speaker 2: a technical advantage over the rest of the world. But

887
00:44:49,239 --> 00:44:52,040
Speaker 2: to think that they're going to keep that is misguided.

888
00:44:52,680 --> 00:44:56,040
Speaker 1: Do you think they're just pushing underground then middle of

889
00:44:56,120 --> 00:44:59,239
Speaker 1: the Manhattan Project and other projects. I mean, I had

890
00:44:59,280 --> 00:45:02,480
Speaker 1: I had someone's in the chair recently talking about how

891
00:45:03,800 --> 00:45:07,880
Speaker 1: this is the theory, how most of the major countries

892
00:45:07,920 --> 00:45:12,200
Speaker 1: around the world have had experiences, have actually UFO information,

893
00:45:12,760 --> 00:45:15,560
Speaker 1: but they won't reveal it to anybody because they're keeping

894
00:45:15,560 --> 00:45:17,880
Speaker 1: it for their own advantage. They're trying to examine the

895
00:45:18,080 --> 00:45:21,520
Speaker 1: materials and everything around it as to how they can

896
00:45:21,520 --> 00:45:23,440
Speaker 1: get some military advantage out of it, which is why

897
00:45:23,480 --> 00:45:24,640
Speaker 1: they're not sharing it globally.

898
00:45:25,320 --> 00:45:27,960
Speaker 2: You know. And it's a nice conspiracy theory, but it's

899
00:45:28,000 --> 00:45:30,879
Speaker 2: a good one. But I wonder how you can keep

900
00:45:30,880 --> 00:45:32,439
Speaker 2: so many people quiet for so long.

901
00:45:32,920 --> 00:45:34,680
Speaker 1: Yeah, well they're out a lot of them are out

902
00:45:34,680 --> 00:45:36,279
Speaker 1: there now talking about it. But do you think that

903
00:45:37,120 --> 00:45:39,520
Speaker 1: and I'm not suggesting I agree with that, but do

904
00:45:39,560 --> 00:45:42,160
Speaker 1: you think that all we're going to do is pushed underground?

905
00:45:44,160 --> 00:45:46,520
Speaker 2: Well, these technologies are going to exist, you know, whether

906
00:45:46,600 --> 00:45:50,440
Speaker 2: they exist already? Yes, And that's the fundamental challenge we

907
00:45:50,480 --> 00:45:54,480
Speaker 2: have with a technology like arstic intelligence. It's completely due use.

908
00:45:55,440 --> 00:45:58,879
Speaker 2: So for example, you know, take your AI drone that's

909
00:45:58,880 --> 00:46:01,520
Speaker 2: going to autonom target people. It needs to be able

910
00:46:01,560 --> 00:46:05,440
Speaker 2: to track people. Will you know, it needs to be

911
00:46:05,480 --> 00:46:08,600
Speaker 2: able to track soldiers on the ground so it can

912
00:46:08,600 --> 00:46:12,680
Speaker 2: target them. Well, the same algorithms again into our autonomous

913
00:46:12,680 --> 00:46:16,120
Speaker 2: cars to tract pedestrians so it can avoid them. We

914
00:46:16,160 --> 00:46:19,680
Speaker 2: want autonomous cars. A thousand people will die next year

915
00:46:20,080 --> 00:46:24,520
Speaker 2: in Australia in road traffic accidents, mostly caused by human error.

916
00:46:25,000 --> 00:46:27,600
Speaker 2: Once we have autonomous cars, we won't have those deaths.

917
00:46:27,920 --> 00:46:30,480
Speaker 2: It will go to close to zero. So that's going

918
00:46:30,520 --> 00:46:33,600
Speaker 2: to be a fantastic you know, bonus to our lives.

919
00:46:35,080 --> 00:46:38,680
Speaker 2: You know, if you survive birth, the most likely cause

920
00:46:38,719 --> 00:46:41,280
Speaker 2: of death to you're in your forties is car accident.

921
00:46:42,239 --> 00:46:45,799
Speaker 2: I mean, it's terrible, and most of us sadly have

922
00:46:46,040 --> 00:46:47,960
Speaker 2: you know, in our families or in our close friends

923
00:46:48,000 --> 00:46:50,759
Speaker 2: know someone's life which was touched by one of those

924
00:46:50,760 --> 00:46:53,239
Speaker 2: car accidents. And we're going to look back in twenty

925
00:46:53,320 --> 00:46:55,720
Speaker 2: or thirty years time and say, you know, it's funny

926
00:46:55,760 --> 00:46:57,160
Speaker 2: that we don't have any of that anymore.

927
00:46:57,640 --> 00:46:59,120
Speaker 1: Well, it's funny. When I was four and a half

928
00:46:59,200 --> 00:47:01,920
Speaker 1: years of age, I first started school on four and

929
00:47:01,960 --> 00:47:06,880
Speaker 1: a half and I got off the bus at the

930
00:47:06,920 --> 00:47:10,400
Speaker 1: bus stop where I live and because the bus was

931
00:47:10,440 --> 00:47:12,520
Speaker 1: higher than it wasn't only little, but are you running

932
00:47:12,520 --> 00:47:14,520
Speaker 1: for it a half yeah, and he didn't see me,

933
00:47:14,760 --> 00:47:18,560
Speaker 1: and he will hit me, ran over me the bus

934
00:47:18,560 --> 00:47:21,120
Speaker 1: a bus yeah, oh my god. Yeah, and lucky to

935
00:47:21,160 --> 00:47:23,000
Speaker 1: have you a land him in hospital for quite a

936
00:47:23,040 --> 00:47:26,160
Speaker 1: long time. But point being is probably in the future

937
00:47:26,360 --> 00:47:29,880
Speaker 1: buses will have senses all around them which are probably

938
00:47:30,320 --> 00:47:32,160
Speaker 1: you know, even if they're not an autonomous if still

939
00:47:32,160 --> 00:47:34,800
Speaker 1: have sensors around it which would automatically stop the bus

940
00:47:35,000 --> 00:47:37,600
Speaker 1: from proceeding because a little sense that there is a

941
00:47:37,640 --> 00:47:39,759
Speaker 1: little person or there's something a little in front of it,

942
00:47:39,880 --> 00:47:42,279
Speaker 1: something in front of it.

943
00:47:42,080 --> 00:47:43,799
Speaker 2: As an example of that. You know, one of the

944
00:47:43,840 --> 00:47:47,120
Speaker 2: newest terrorist weapons is to go rent a truck or

945
00:47:47,160 --> 00:47:49,880
Speaker 2: a bus and drive us into a crowd. Yeah, that

946
00:47:49,920 --> 00:47:52,240
Speaker 2: won't be possible in five or ten years time because

947
00:47:52,480 --> 00:47:54,880
Speaker 2: every truck will have cameras on the front and if

948
00:47:54,920 --> 00:47:57,840
Speaker 2: there's a pedestrian in front of it, it will break automatically.

949
00:47:57,880 --> 00:47:58,719
Speaker 2: You won't be able to drive it.

950
00:47:58,840 --> 00:48:00,960
Speaker 1: And that's our AI, that's working.

951
00:48:02,360 --> 00:48:04,319
Speaker 2: People to be able to recognize the roads and to

952
00:48:04,360 --> 00:48:07,600
Speaker 2: see the pedestrians and then to make the decision to stop,

953
00:48:07,680 --> 00:48:10,480
Speaker 2: so that there's a great positive terrorist won't be able

954
00:48:10,520 --> 00:48:12,160
Speaker 2: to do that in the future. They can hire as

955
00:48:12,160 --> 00:48:14,399
Speaker 2: many trucks as they like, they can try and drive

956
00:48:14,440 --> 00:48:15,800
Speaker 2: them to crowds, but they were to stop.

957
00:48:16,040 --> 00:48:19,759
Speaker 1: And also just someone like a young kid not knowing

958
00:48:19,760 --> 00:48:22,160
Speaker 1: any better, they won't what might get run over?

959
00:48:22,560 --> 00:48:23,520
Speaker 2: Run over, young kid.

960
00:48:23,760 --> 00:48:26,560
Speaker 1: I remember for six weeks, I remember it. Remember I

961
00:48:26,560 --> 00:48:28,400
Speaker 1: don't remember being I still remember being a hospital.

962
00:48:28,440 --> 00:48:31,520
Speaker 2: I don't remember that many more entrepreneurs that get to

963
00:48:31,520 --> 00:48:33,000
Speaker 2: grow up become success.

964
00:48:33,000 --> 00:48:35,919
Speaker 1: Maybe there's going to be too many. But just on

965
00:48:35,960 --> 00:48:38,800
Speaker 1: this is you're probably well aware of the big argument

966
00:48:38,840 --> 00:48:43,280
Speaker 1: in Australia right now that Australia in the developed nations

967
00:48:43,360 --> 00:48:47,480
Speaker 1: is probably not probably is has i should say, the

968
00:48:47,560 --> 00:48:53,480
Speaker 1: lowest productivity growth of any nation in the world developed

969
00:48:53,520 --> 00:48:57,120
Speaker 1: nations right now, What do you think AI is going

970
00:48:57,160 --> 00:49:01,160
Speaker 1: to do to help productivity in a place like Australia, Well, if.

971
00:49:01,040 --> 00:49:04,600
Speaker 2: You believe the economists, all the economists come out with numbers,

972
00:49:04,600 --> 00:49:07,640
Speaker 2: which is it can lift our productivity by fifteen twenty.

973
00:49:07,440 --> 00:49:09,360
Speaker 1: Percent, which is significant.

974
00:49:09,440 --> 00:49:11,680
Speaker 2: Yes, and I see a few other things coming along

975
00:49:11,719 --> 00:49:14,960
Speaker 2: that can offer fifteen twenty percent lifts of our productivity.

976
00:49:15,080 --> 00:49:15,560
Speaker 1: That's good.

977
00:49:15,719 --> 00:49:16,319
Speaker 2: Yeah, that's good.

978
00:49:16,360 --> 00:49:16,920
Speaker 1: That's a good thing.

979
00:49:17,239 --> 00:49:17,839
Speaker 2: That's a good thing.

980
00:49:18,400 --> 00:49:20,799
Speaker 1: Should we embrace it? Because the flip side of that is,

981
00:49:21,840 --> 00:49:24,360
Speaker 1: you know, no I read in or hear about this often,

982
00:49:24,800 --> 00:49:26,600
Speaker 1: is that there's going to be some stage where no

983
00:49:26,600 --> 00:49:28,279
Speaker 1: one actually goes to work anymore. No one, no one

984
00:49:28,320 --> 00:49:30,880
Speaker 1: actually works anymore, No, no, no one, but lead the

985
00:49:30,880 --> 00:49:34,680
Speaker 1: majority of people, their jobs are going to become redundant.

986
00:49:35,520 --> 00:49:38,080
Speaker 1: But the same productivity, the same man A taxs paid,

987
00:49:38,239 --> 00:49:39,840
Speaker 1: which means the govern's got the money and we just

988
00:49:40,000 --> 00:49:40,960
Speaker 1: pay you to stay home.

989
00:49:41,120 --> 00:49:41,359
Speaker 2: Yep.

990
00:49:42,520 --> 00:49:43,160
Speaker 1: Well, I think.

991
00:49:43,400 --> 00:49:46,319
Speaker 2: People forget this has already happened, right, This happened with

992
00:49:46,320 --> 00:49:50,080
Speaker 2: the Industrial Revolution, before the people forget the weekend was

993
00:49:50,120 --> 00:49:52,840
Speaker 2: an invention of the dust Revolution. Before the indust revolution,

994
00:49:52,880 --> 00:49:55,720
Speaker 2: you work seminary week some went up, some went down.

995
00:49:55,880 --> 00:49:58,279
Speaker 2: You were working all that time. And then because of

996
00:49:58,320 --> 00:50:01,560
Speaker 2: the productivity benefits of you know, the automation machinery that

997
00:50:01,600 --> 00:50:05,080
Speaker 2: we brought in in the Dust Revolution, we got we

998
00:50:05,160 --> 00:50:07,279
Speaker 2: got to say, you know, we want to share some

999
00:50:07,320 --> 00:50:08,919
Speaker 2: of that. We want Sunday off to go to church,

1000
00:50:09,440 --> 00:50:12,879
Speaker 2: then Saturday afternoon off, then all of Saturday. And then

1001
00:50:13,040 --> 00:50:15,400
Speaker 2: I don't know why we stopped asking why did we

1002
00:50:15,400 --> 00:50:17,960
Speaker 2: stop asking for more? I mean, what was There's nothing

1003
00:50:17,960 --> 00:50:20,799
Speaker 2: magic about two days off but every seven. There's nothing

1004
00:50:20,840 --> 00:50:22,840
Speaker 2: about you know, the earth going around the sun that

1005
00:50:22,840 --> 00:50:26,320
Speaker 2: says it's only two days off seven. It's a human construction,

1006
00:50:27,360 --> 00:50:32,080
Speaker 2: which was we could have carried on asking for more.

1007
00:50:32,680 --> 00:50:34,960
Speaker 2: And you know, interestingly, there are experiments on a four

1008
00:50:35,040 --> 00:50:38,040
Speaker 2: day week in New Zealand and Europe. Hundreds of thousand

1009
00:50:38,040 --> 00:50:40,719
Speaker 2: people are doing these experiments now and the experiments all

1010
00:50:40,760 --> 00:50:44,040
Speaker 2: have two conclusions. One is that people do as much

1011
00:50:44,040 --> 00:50:45,680
Speaker 2: work in four days as they used to do in five.

1012
00:50:47,280 --> 00:50:49,279
Speaker 2: You you know, stop having all the bullshit meetings. You

1013
00:50:49,320 --> 00:50:51,480
Speaker 2: get down to it and focus on what you need

1014
00:50:51,520 --> 00:50:54,000
Speaker 2: to do so you can pay them as much because

1015
00:50:54,000 --> 00:50:56,160
Speaker 2: they're doing just as much. Right, so business is just

1016
00:50:56,200 --> 00:50:58,640
Speaker 2: as productive as they were when you were working five

1017
00:50:58,840 --> 00:51:02,120
Speaker 2: And second, people are happy. Who would have imagined.

1018
00:51:01,800 --> 00:51:07,239
Speaker 1: Well that I from wondering about though the happier I

1019
00:51:07,239 --> 00:51:09,399
Speaker 1: guess abby people said they're going to work. They're still

1020
00:51:09,440 --> 00:51:12,239
Speaker 1: enjoying their colleagues Monday to Thursday, for example, find I

1021
00:51:12,280 --> 00:51:14,560
Speaker 1: said it's something that they have an opportunity to enjoy

1022
00:51:14,600 --> 00:51:16,520
Speaker 1: their family. They're still getting paid the same amount of

1023
00:51:16,560 --> 00:51:18,720
Speaker 1: money because the company should be making the same amount

1024
00:51:18,719 --> 00:51:20,120
Speaker 1: of money, so the company can afford to pay them

1025
00:51:20,320 --> 00:51:22,160
Speaker 1: the same amount of money. That will be people who

1026
00:51:22,200 --> 00:51:24,040
Speaker 1: will be seduced into paying you less, but it doesn't

1027
00:51:24,080 --> 00:51:26,960
Speaker 1: matter trying to pay you less. Do you see that

1028
00:51:27,040 --> 00:51:29,399
Speaker 1: in the is this some? Is this a consequence of AO?

1029
00:51:29,520 --> 00:51:31,920
Speaker 1: Do you see this coming out in the future that

1030
00:51:32,040 --> 00:51:35,919
Speaker 1: will we definitely will be working less days, because isn't

1031
00:51:35,960 --> 00:51:38,240
Speaker 1: that one way fixing up the fact that some people

1032
00:51:38,320 --> 00:51:41,080
Speaker 1: will have no job? Yes, so instead of saying you

1033
00:51:41,120 --> 00:51:44,759
Speaker 1: don't work any days, I'll work one less day and

1034
00:51:44,840 --> 00:51:47,440
Speaker 1: you can you will work? Yeah, yeah, whatever, some along

1035
00:51:47,480 --> 00:51:48,759
Speaker 1: those lines. You know what I mean, like, you know,

1036
00:51:48,760 --> 00:51:53,839
Speaker 1: because there might be for every one person loses their job,

1037
00:51:53,880 --> 00:51:56,000
Speaker 1: there might be five people keep the job. So five people,

1038
00:51:56,680 --> 00:51:59,480
Speaker 1: I'll do it. If six people keep the job, five

1039
00:51:59,480 --> 00:52:01,560
Speaker 1: people keep the job, we should say so fire people

1040
00:52:01,600 --> 00:52:04,400
Speaker 1: going to all say we'll work one day less. You

1041
00:52:04,520 --> 00:52:06,080
Speaker 1: the one person have lost your job, you can now

1042
00:52:06,120 --> 00:52:08,640
Speaker 1: work five days or one of those lines, you know, mathematically.

1043
00:52:09,160 --> 00:52:12,040
Speaker 1: So that is that something people are talking about now.

1044
00:52:12,760 --> 00:52:14,920
Speaker 2: I think we should be having these conversations. I mean,

1045
00:52:14,920 --> 00:52:18,359
Speaker 2: I think it's an example of how we went through

1046
00:52:18,400 --> 00:52:22,160
Speaker 2: some big structural changes through the Industrial Revolution. We introduced,

1047
00:52:22,280 --> 00:52:26,400
Speaker 2: you know, the weekend unions, labor laws to protect the

1048
00:52:26,480 --> 00:52:29,360
Speaker 2: rights of workers, to spread the benefits around, universal education

1049
00:52:29,400 --> 00:52:32,440
Speaker 2: so we're educated for those new jobs, Universal welfare so

1050
00:52:32,480 --> 00:52:34,680
Speaker 2: that when we were unemployed we weren't in the poorhouse,

1051
00:52:35,840 --> 00:52:37,839
Speaker 2: Universal pensions so that we could retire at the end

1052
00:52:37,840 --> 00:52:40,279
Speaker 2: of our working lives. We did a lot to you know,

1053
00:52:40,320 --> 00:52:42,239
Speaker 2: spread the benefits around and make sure it was good.

1054
00:52:42,280 --> 00:52:43,640
Speaker 1: So the structure sort of exists.

1055
00:52:44,320 --> 00:52:46,360
Speaker 2: Well, now we're going through another big change, and I

1056
00:52:46,400 --> 00:52:50,839
Speaker 2: think we should be thinking equally radically about this. And

1057
00:52:50,920 --> 00:52:53,640
Speaker 2: you know, there are deep fundamental questions here, you know,

1058
00:52:53,840 --> 00:52:56,879
Speaker 2: many people derive their value through what they do.

1059
00:52:57,040 --> 00:52:58,960
Speaker 1: They identify with that, and then we.

1060
00:52:59,040 --> 00:53:01,000
Speaker 2: Meet someone, you say, what do you do? What work

1061
00:53:01,040 --> 00:53:03,120
Speaker 2: do you do? Right? Device, Well, if we're in a

1062
00:53:03,200 --> 00:53:06,160
Speaker 2: world where work is less important, now, how are we

1063
00:53:06,239 --> 00:53:08,440
Speaker 2: going to define our meaning? How? What does our purpose?

1064
00:53:08,840 --> 00:53:12,680
Speaker 1: So therefore a consequence of AI and a consequence of

1065
00:53:13,239 --> 00:53:16,040
Speaker 1: more efficiency and better productivity as a result of AI,

1066
00:53:16,120 --> 00:53:19,000
Speaker 1: which is one of the things I should do. Maybe

1067
00:53:19,040 --> 00:53:22,520
Speaker 1: government or somebody has now gone an obligation to actually

1068
00:53:22,560 --> 00:53:26,560
Speaker 1: help us redefine what we do. So in other words,

1069
00:53:26,560 --> 00:53:28,640
Speaker 1: I've only worked three days a week. That's the time

1070
00:53:28,680 --> 00:53:31,479
Speaker 1: I was working five days a week. That was my life.

1071
00:53:31,480 --> 00:53:35,160
Speaker 1: Now I'm working less than one half of my ev seven.

1072
00:53:35,320 --> 00:53:37,080
Speaker 1: By the way, seven days is fiction as well, But

1073
00:53:37,360 --> 00:53:41,080
Speaker 1: don't worry. We just assume seven days a week is correct.

1074
00:53:41,600 --> 00:53:43,759
Speaker 1: But four of those seven I'm not working. So all

1075
00:53:43,800 --> 00:53:46,080
Speaker 1: of a sudden, I'm at home. I'm not doing anything

1076
00:53:46,840 --> 00:53:49,080
Speaker 1: I don't think to do. It's government's got to start

1077
00:53:49,080 --> 00:53:53,520
Speaker 1: to re educate us to or to encourage us to,

1078
00:53:54,200 --> 00:53:56,879
Speaker 1: because we don't really sitting at home doing nothing. Maybe

1079
00:53:56,880 --> 00:53:59,919
Speaker 1: we do, but most of we'll get jacked that pretty quick.

1080
00:54:00,640 --> 00:54:04,200
Speaker 2: Yes, you're right. I mean I always say to remind people,

1081
00:54:04,200 --> 00:54:06,879
Speaker 2: you know, the purpose of education. When we educate people,

1082
00:54:06,880 --> 00:54:08,759
Speaker 2: it's not just to prepare them for work. It's to

1083
00:54:08,760 --> 00:54:12,040
Speaker 2: prepare them to take advantage of their lives, to be

1084
00:54:12,440 --> 00:54:15,080
Speaker 2: good citizens, to be able to do stuff and you know,

1085
00:54:15,600 --> 00:54:16,560
Speaker 2: make the world a better play.

1086
00:54:16,600 --> 00:54:18,239
Speaker 1: Do you think we're how to enjoy ourselves though, Like

1087
00:54:19,719 --> 00:54:21,399
Speaker 1: if we're only going to work three days a week,

1088
00:54:21,440 --> 00:54:24,160
Speaker 1: because we identify with ourselves working five days a week.

1089
00:54:24,239 --> 00:54:25,600
Speaker 1: And as you say, this is going to happen quickly,

1090
00:54:26,120 --> 00:54:32,560
Speaker 1: so quicker than we think. We've got to learn relearn

1091
00:54:32,600 --> 00:54:35,520
Speaker 1: how to fill in those extra couple of days a week.

1092
00:54:35,640 --> 00:54:38,520
Speaker 2: We are, but I mean, that's the great strength of humans, right,

1093
00:54:38,560 --> 00:54:41,239
Speaker 2: we're so quick to there we go back, cast your

1094
00:54:41,280 --> 00:54:45,120
Speaker 2: mind back to COVID, and we immediately learned the value

1095
00:54:45,280 --> 00:54:47,680
Speaker 2: of being able to come together a bit, being able

1096
00:54:47,719 --> 00:54:49,640
Speaker 2: to go out and you know, have a walk in

1097
00:54:49,680 --> 00:54:53,080
Speaker 2: the park and so on, those things that we remember.

1098
00:54:53,280 --> 00:54:56,279
Speaker 2: Reminded so quickly about the value of those things, and

1099
00:54:56,320 --> 00:54:59,040
Speaker 2: we appreciated those things so much more.

1100
00:54:59,360 --> 00:55:01,600
Speaker 1: Of course, a lot of people during the COVID period

1101
00:55:01,960 --> 00:55:07,919
Speaker 1: became addicted to social Yeah. Do you think is there.

1102
00:55:07,800 --> 00:55:09,960
Speaker 2: A very seductive, very addictive.

1103
00:55:10,160 --> 00:55:12,360
Speaker 1: Is there a potential problem because that that is AI

1104
00:55:12,560 --> 00:55:14,360
Speaker 1: working You either be time why.

1105
00:55:14,200 --> 00:55:15,840
Speaker 2: It's going to be super charged by it to be

1106
00:55:16,960 --> 00:55:19,040
Speaker 2: because the content is going to be our generated content,

1107
00:55:19,120 --> 00:55:21,480
Speaker 2: that it's going to be personalized for you. It's the

1108
00:55:21,640 --> 00:55:25,040
Speaker 2: AI algorithm's course that are finding your sweet spots and things.

1109
00:55:24,960 --> 00:55:28,680
Speaker 1: And keep seeing you the same stuff and then testing

1110
00:55:28,719 --> 00:55:31,279
Speaker 1: where you want to go from there? Do you think

1111
00:55:31,320 --> 00:55:34,920
Speaker 1: that therefore, which is why we should.

1112
00:55:34,640 --> 00:55:38,040
Speaker 2: Be concerned about the power of the people in Silicon

1113
00:55:38,120 --> 00:55:40,759
Speaker 2: Valley who are making decisions to yeah, let's put the

1114
00:55:40,760 --> 00:55:43,480
Speaker 2: beauty filters back in yeah, despite the fact that they

1115
00:55:43,600 --> 00:55:45,920
Speaker 2: know that we're not going to be able to resist

1116
00:55:46,000 --> 00:55:47,920
Speaker 2: using them, that we're all going to be Oh, I

1117
00:55:47,960 --> 00:55:50,279
Speaker 2: can't compare myself to these other people have got the

1118
00:55:50,320 --> 00:55:53,440
Speaker 2: beauty filters. Are I have to use the beauty filters myself?

1119
00:55:53,640 --> 00:55:57,160
Speaker 1: Because for me, the way I see myself advancing in

1120
00:55:57,200 --> 00:56:01,080
Speaker 1: life is to learn more things. But to learn things

1121
00:56:01,440 --> 00:56:03,160
Speaker 1: they've got nothing to do with what I already know.

1122
00:56:04,680 --> 00:56:08,120
Speaker 1: But the whole of the time for that well, at

1123
00:56:08,120 --> 00:56:10,319
Speaker 1: the essence of AIS, it doesn't work that way. It's

1124
00:56:10,320 --> 00:56:11,759
Speaker 1: going to say no, I'm going to keep giving you

1125
00:56:11,800 --> 00:56:13,440
Speaker 1: what I know you like, especially if it's on a

1126
00:56:13,480 --> 00:56:16,440
Speaker 1: social environment. Do you think the conversation needs to be

1127
00:56:16,520 --> 00:56:20,319
Speaker 1: had as to how do I encourage Mark when he's

1128
00:56:20,360 --> 00:56:22,200
Speaker 1: only working three days a week he said a four days,

1129
00:56:22,440 --> 00:56:26,040
Speaker 1: seven days, five days a week. How do I encourage

1130
00:56:26,080 --> 00:56:30,319
Speaker 1: Mark this conversation to or how to I encourage Mark

1131
00:56:30,400 --> 00:56:34,880
Speaker 1: Zuckerberg or those people control these environments to start to

1132
00:56:34,920 --> 00:56:38,440
Speaker 1: feed Mark something different, to encourage him to start to

1133
00:56:38,440 --> 00:56:42,640
Speaker 1: think more broadly or laterally about other things. What do

1134
00:56:42,680 --> 00:56:43,840
Speaker 1: you think about that conversation?

1135
00:56:44,560 --> 00:56:46,640
Speaker 2: As I've seen the reason why we couldn't think of

1136
00:56:46,680 --> 00:56:49,759
Speaker 2: that algorithmically, how we couldn't actually say to the algorithms, no,

1137
00:56:49,840 --> 00:56:51,960
Speaker 2: actually we don't want more of the same. We want

1138
00:56:52,480 --> 00:56:54,799
Speaker 2: If you play my favorite song to me too many times,

1139
00:56:54,800 --> 00:56:56,680
Speaker 2: it's going to stop being my favorite song. I'm going

1140
00:56:56,680 --> 00:56:59,120
Speaker 2: to get tired of it. So play me other things.

1141
00:56:59,320 --> 00:57:01,440
Speaker 2: You know, I remember, you know, I used to go

1142
00:57:01,480 --> 00:57:03,719
Speaker 2: to the music shop and browse through the music and

1143
00:57:04,800 --> 00:57:09,719
Speaker 2: find things by chance serendipity. It's fantastic for finding musicians

1144
00:57:09,719 --> 00:57:11,960
Speaker 2: that I would never have otherwise found, just because they

1145
00:57:11,960 --> 00:57:13,960
Speaker 2: happen to be next to where I wanted on the shelf.

1146
00:57:15,200 --> 00:57:19,280
Speaker 2: So you know, maybe we howgroythmically can encourage serendipity. But

1147
00:57:19,320 --> 00:57:23,880
Speaker 2: the problem here is the perverse incentives that the incentives

1148
00:57:23,880 --> 00:57:27,000
Speaker 2: that the media companies the tech companies have to you know,

1149
00:57:27,120 --> 00:57:29,560
Speaker 2: keep you there in the short term to keep you

1150
00:57:29,840 --> 00:57:34,600
Speaker 2: scrolling infinite scrolling through what their content for the now,

1151
00:57:35,240 --> 00:57:37,479
Speaker 2: because you know, obviously they're sending your time, they're selling

1152
00:57:37,520 --> 00:57:40,680
Speaker 2: your adverts and everything in and now, which is good

1153
00:57:40,680 --> 00:57:42,680
Speaker 2: for their revenues in the short term, but bad for

1154
00:57:42,760 --> 00:57:45,680
Speaker 2: you and indeed probably bad for them actually in the

1155
00:57:45,720 --> 00:57:48,760
Speaker 2: longer term, which is that you know, maybe they should

1156
00:57:48,800 --> 00:57:52,640
Speaker 2: just be saying, you know, chatter gbt'd ever says, you know, tybe,

1157
00:57:52,680 --> 00:57:54,720
Speaker 2: you've been asking me questions for three hours. It's three

1158
00:57:54,760 --> 00:57:57,480
Speaker 2: o'clock in the morning. Time to go to bed. You know,

1159
00:57:57,640 --> 00:58:00,720
Speaker 2: it always ends with an open question because they're trying

1160
00:58:00,720 --> 00:58:01,920
Speaker 2: to sell you more token.

1161
00:58:01,800 --> 00:58:04,040
Speaker 1: Or three more questions. Yes, yeah, it's funny. You know.

1162
00:58:04,880 --> 00:58:08,200
Speaker 1: I remember in two thousand and one, excuse me. In

1163
00:58:08,240 --> 00:58:11,240
Speaker 1: two thousand and I got asked by the then Minister

1164
00:58:12,200 --> 00:58:16,880
Speaker 1: for Finance Jaie Hockey to share a task force as

1165
00:58:16,920 --> 00:58:28,320
Speaker 1: part of Australia's OCD obligations to build a code of

1166
00:58:28,360 --> 00:58:34,919
Speaker 1: Practice for consumer protection for Internet vendors. So we're twenty

1167
00:58:34,960 --> 00:58:38,640
Speaker 1: two thous and one. It's simple things like spam you

1168
00:58:38,720 --> 00:58:40,560
Speaker 1: should you you know, what's the relationship you got to

1169
00:58:40,600 --> 00:58:43,120
Speaker 1: have opped in opted out? Just a code of practice,

1170
00:58:43,160 --> 00:58:46,680
Speaker 1: not legislation, but a code of practice. Simple things like

1171
00:58:46,760 --> 00:58:51,040
Speaker 1: if you know, if you go to a vendors website

1172
00:58:51,880 --> 00:58:54,280
Speaker 1: and if you want to buy pajamas for your kid,

1173
00:58:54,760 --> 00:58:57,600
Speaker 1: there should be the code of practice has said that

1174
00:58:57,680 --> 00:59:02,479
Speaker 1: you should disclose if the pajamas a fly noble or

1175
00:59:02,520 --> 00:59:06,080
Speaker 1: returns by something doesn't hit. What's my returns policies? You

1176
00:59:06,120 --> 00:59:08,280
Speaker 1: have to have it laid out. This is in two thousand, Well,

1177
00:59:08,320 --> 00:59:11,160
Speaker 1: it's not that long ago. It's not very long ago.

1178
00:59:11,640 --> 00:59:14,560
Speaker 1: Australia was the first country, so we produced a paper

1179
00:59:14,600 --> 00:59:17,280
Speaker 1: in a document that was the first country to first

1180
00:59:17,360 --> 00:59:21,360
Speaker 1: over CD country that actually put out this code of practice,

1181
00:59:21,920 --> 00:59:24,960
Speaker 1: and we're all quite proud of ourselves. But of course

1182
00:59:24,960 --> 00:59:28,720
Speaker 1: a lot of that stuff's now embedded in legislation. Do

1183
00:59:28,760 --> 00:59:33,200
Speaker 1: you think that the world is now at a point

1184
00:59:33,560 --> 00:59:37,760
Speaker 1: where we either need a code of practice or or

1185
00:59:38,040 --> 00:59:42,240
Speaker 1: legislation because then it was directed at millions and millions

1186
00:59:42,240 --> 00:59:44,920
Speaker 1: of indors. Do you think we will have the same

1187
00:59:44,960 --> 00:59:48,480
Speaker 1: equivalent for the twelve or fifteen powerful people who run

1188
00:59:48,520 --> 00:59:54,120
Speaker 1: all these AI platforms to say, dudes, this is a

1189
00:59:54,120 --> 00:59:57,760
Speaker 1: code of practice, or even stronger, this is the legislation,

1190
00:59:58,040 --> 00:59:59,920
Speaker 1: because code practice always ends up getting code of fi

1191
01:00:01,280 --> 01:00:04,680
Speaker 1: into law, which is what happened. Do you think this

1192
01:00:04,800 --> 01:00:06,280
Speaker 1: sort of thing should be done? And is there any

1193
01:00:06,280 --> 01:00:08,680
Speaker 1: inquiries into this stuff going on at the moment anywhere

1194
01:00:08,680 --> 01:00:09,120
Speaker 1: in the world.

1195
01:00:09,520 --> 01:00:12,320
Speaker 2: Yeah, I mean, I think policy personal politicians are working

1196
01:00:12,360 --> 01:00:15,040
Speaker 2: up to that they could do something. There was certainly,

1197
01:00:15,040 --> 01:00:17,160
Speaker 2: I think a feeling of impotence that, you know, how

1198
01:00:17,240 --> 01:00:20,560
Speaker 2: can we push back against these powerful companies. They're wealthier

1199
01:00:20,560 --> 01:00:25,120
Speaker 2: than nations, they operate outs outside our national borders and

1200
01:00:25,280 --> 01:00:28,200
Speaker 2: pay tax. They're pay tax. You know, the rules don't

1201
01:00:28,200 --> 01:00:29,120
Speaker 2: seem to apply to them.

1202
01:00:29,240 --> 01:00:30,320
Speaker 1: Yeah, well they're not people.

1203
01:00:31,000 --> 01:00:33,400
Speaker 2: Why they're not people? So you know, we're just going

1204
01:00:33,440 --> 01:00:35,240
Speaker 2: to have to accept it. And I think now there's

1205
01:00:35,240 --> 01:00:38,600
Speaker 2: a growing realization and a huge pressure from the public

1206
01:00:38,640 --> 01:00:41,160
Speaker 2: as well to do something. And we saw the success

1207
01:00:41,200 --> 01:00:45,640
Speaker 2: of the social media you know, age delay laws demonstrating

1208
01:00:45,680 --> 01:00:47,640
Speaker 2: that we can we can protect young people from this,

1209
01:00:47,720 --> 01:00:50,640
Speaker 2: and there's being in the press other legislation around, you know,

1210
01:00:51,200 --> 01:00:54,800
Speaker 2: making it criminal to distribute new defied pictures of people,

1211
01:00:54,880 --> 01:00:57,280
Speaker 2: because that was becoming a problem in our schools.

1212
01:00:58,480 --> 01:01:01,280
Speaker 1: As in fake fake images of faking a friend your

1213
01:01:01,360 --> 01:01:03,600
Speaker 1: friend and then spread around the whole school.

1214
01:01:05,520 --> 01:01:09,080
Speaker 2: So yeah, I think there's we've demonstrated we can do things,

1215
01:01:09,320 --> 01:01:11,520
Speaker 2: and there's a real appetite now to do things. I

1216
01:01:11,560 --> 01:01:14,600
Speaker 2: really do feel that with the new things that we're

1217
01:01:14,600 --> 01:01:16,040
Speaker 2: going to be able to do with AI that werey're

1218
01:01:16,040 --> 01:01:18,960
Speaker 2: going to be new harms. We talked about how people

1219
01:01:19,000 --> 01:01:21,479
Speaker 2: are having relationships with AIS, how people are using AI

1220
01:01:21,600 --> 01:01:26,600
Speaker 2: the therapists, people are committing suicides, and the chat pot

1221
01:01:26,600 --> 01:01:30,400
Speaker 2: is encouraging that behavior. Like, I think there's we've also

1222
01:01:30,440 --> 01:01:35,800
Speaker 2: seen that the tech companies themselves, Oh you know, there's

1223
01:01:35,840 --> 01:01:37,920
Speaker 2: perverse incentives been played for them not to do the

1224
01:01:38,000 --> 01:01:41,720
Speaker 2: right thing, not to worry necessary so much about our safety.

1225
01:01:41,840 --> 01:01:44,920
Speaker 1: Do you think we might end up having an AI policeman?

1226
01:01:45,000 --> 01:01:48,400
Speaker 1: So you know, like I'm on one of the platforms

1227
01:01:48,400 --> 01:01:54,680
Speaker 1: and I'm doing what I'm doing, but somehow I allow

1228
01:01:55,400 --> 01:01:58,840
Speaker 1: or permit an II policeman to always enter into my

1229
01:01:58,920 --> 01:02:02,920
Speaker 1: world and say, Mark, you're going too far. You need

1230
01:02:02,960 --> 01:02:06,080
Speaker 1: to go to sleep outside of outside of, for example,

1231
01:02:06,080 --> 01:02:11,800
Speaker 1: whatever I'm using. So it's actually an artificial intelligence policeman

1232
01:02:12,160 --> 01:02:17,680
Speaker 1: police person that actually not policing but effectively monitors what

1233
01:02:17,760 --> 01:02:20,760
Speaker 1: I'm doing. But across the board on everything I would I.

1234
01:02:20,760 --> 01:02:22,480
Speaker 2: Wouldn't call it a policeman. I call it a guardian.

1235
01:02:22,840 --> 01:02:24,000
Speaker 1: Okay, that's a better word.

1236
01:02:24,040 --> 01:02:27,439
Speaker 2: Yeah, AI guardian. And I'm pretty convinced on your phone

1237
01:02:27,440 --> 01:02:29,520
Speaker 2: you're going to have an AI guardian that sits there

1238
01:02:29,560 --> 01:02:33,040
Speaker 2: watching what's going on and says, Mark, be careful. This

1239
01:02:33,080 --> 01:02:37,080
Speaker 2: looks like it's a scam advert lock. This company's trying

1240
01:02:37,080 --> 01:02:39,280
Speaker 2: to steal some data out of you here, and.

1241
01:02:39,440 --> 01:02:41,439
Speaker 1: Well, Mark, you might be spent too much time. Yeah,

1242
01:02:41,520 --> 01:02:44,960
Speaker 1: and also just tells me howmuch, says hey, like actually

1243
01:02:45,160 --> 01:02:48,880
Speaker 1: prompts me with a notification or a nudge, because the

1244
01:02:48,920 --> 01:02:51,160
Speaker 1: nudging is still very powerful. Yes, I mean it, which,

1245
01:02:51,240 --> 01:02:52,760
Speaker 1: by the way, is what most of the platform is

1246
01:02:52,760 --> 01:02:55,360
Speaker 1: doing to you all the time. But it's so subtle

1247
01:02:56,160 --> 01:02:58,560
Speaker 1: you don't even feel the nudge. But something is a

1248
01:02:58,600 --> 01:02:59,760
Speaker 1: little less subtle, and.

1249
01:03:00,320 --> 01:03:02,760
Speaker 2: The nudges that they're giving are noticary for your benefit.

1250
01:03:02,880 --> 01:03:06,200
Speaker 1: No, no, no, they're pushing to get.

1251
01:03:06,000 --> 01:03:08,040
Speaker 2: You to spend more time on the platform. Or to

1252
01:03:08,040 --> 01:03:10,960
Speaker 2: buy more stuff totally. Whereas you know, we want to

1253
01:03:10,960 --> 01:03:13,920
Speaker 2: be nudged in ways to help for us to say, like, okay,

1254
01:03:14,800 --> 01:03:18,600
Speaker 2: you know, make you a happier, more content rounded person.

1255
01:03:18,800 --> 01:03:21,640
Speaker 1: You know, it could be nearly an education program for

1256
01:03:21,760 --> 01:03:23,840
Speaker 1: you too, that could sort of saying, well, the reason

1257
01:03:23,920 --> 01:03:26,600
Speaker 1: you need this or the reason this might be a

1258
01:03:26,600 --> 01:03:31,080
Speaker 1: good idea this guardian ship for adults as well as kids,

1259
01:03:31,680 --> 01:03:33,360
Speaker 1: is because we want to make you a happy person.

1260
01:03:33,400 --> 01:03:36,440
Speaker 1: We don't want you to become sort of living in

1261
01:03:36,440 --> 01:03:41,080
Speaker 1: this weird world where it's not real because the definition

1262
01:03:41,200 --> 01:03:44,600
Speaker 1: real will become brought into question. We won't know, But

1263
01:03:45,200 --> 01:03:47,040
Speaker 1: I think that's look for me, I reckon, that's a

1264
01:03:47,040 --> 01:03:50,440
Speaker 1: great idea. I would actually I would actually allow that

1265
01:03:50,440 --> 01:03:54,720
Speaker 1: app to live in my world. But not not not

1266
01:03:54,720 --> 01:03:57,280
Speaker 1: not to wrap me out of the knuckles. But maybe

1267
01:03:57,760 --> 01:04:00,000
Speaker 1: if it's you know, maybe you can you can enable

1268
01:04:00,000 --> 01:04:02,000
Speaker 1: a lot of empowering. You say, listen. If I don't

1269
01:04:02,040 --> 01:04:03,800
Speaker 1: listen to you, and I continue to straighten no matter

1270
01:04:03,840 --> 01:04:06,440
Speaker 1: what you tell me, please advise my partner, or please

1271
01:04:06,480 --> 01:04:10,360
Speaker 1: advise my oldest son or daughter, or my mom or dad,

1272
01:04:10,440 --> 01:04:12,880
Speaker 1: or whatever the case may be, you opt into that

1273
01:04:13,600 --> 01:04:15,680
Speaker 1: it's not automatic you can opt in. In other words,

1274
01:04:15,720 --> 01:04:18,720
Speaker 1: you're still in control. Is that it? But is that

1275
01:04:18,880 --> 01:04:21,400
Speaker 1: being done anywhere? Is anyone doing anything like that? Sounds

1276
01:04:21,440 --> 01:04:23,600
Speaker 1: like a fantastic I was just saying, we were just thinking.

1277
01:04:23,600 --> 01:04:26,480
Speaker 1: I think we might have just invented something. Yes, just

1278
01:04:26,480 --> 01:04:30,000
Speaker 1: talk about your book. So I have the These books

1279
01:04:30,000 --> 01:04:32,440
Speaker 1: are both the same. But this book's called the Shortest

1280
01:04:32,480 --> 01:04:34,280
Speaker 1: History of AI but it's obviously by you. And this

1281
01:04:34,320 --> 01:04:36,680
Speaker 1: is the English version of this strain version. I like

1282
01:04:36,680 --> 01:04:39,600
Speaker 1: this one because you signed this one. What and you've

1283
01:04:39,600 --> 01:04:42,680
Speaker 1: done a number of books. You've done produced number of books,

1284
01:04:42,680 --> 01:04:44,080
Speaker 1: so as well as you know you speak at lots

1285
01:04:44,080 --> 01:04:47,040
Speaker 1: of events. You know it's particularly you know you're not

1286
01:04:47,040 --> 01:04:50,120
Speaker 1: a nations, which is pretty cool. But what is the shortest?

1287
01:04:50,800 --> 01:04:54,320
Speaker 2: Was never in my bucket list. I never expected, never wanted.

1288
01:04:55,120 --> 01:04:58,400
Speaker 2: But you know, if you know, you see the harms,

1289
01:04:58,440 --> 01:05:00,000
Speaker 2: you think, well, if I could do something about it

1290
01:05:00,120 --> 01:05:03,280
Speaker 2: to I will you know. I know what my daughter's

1291
01:05:03,440 --> 01:05:05,560
Speaker 2: term me when she's grown up and say, Dad, you

1292
01:05:05,640 --> 01:05:06,840
Speaker 2: had a chance to do something.

1293
01:05:06,920 --> 01:05:07,600
Speaker 1: Yeah, why didn't you?

1294
01:05:07,680 --> 01:05:08,200
Speaker 2: Why didn't you?

1295
01:05:08,480 --> 01:05:13,160
Speaker 1: She's your guardian. So this says six the six essential

1296
01:05:13,320 --> 01:05:18,160
Speaker 1: ideas that animated obviously that's AI. What do you mean

1297
01:05:18,200 --> 01:05:20,240
Speaker 1: by that? The six essential ideas.

1298
01:05:20,160 --> 01:05:23,040
Speaker 2: So it's a short history of AI six chapters. Each

1299
01:05:23,160 --> 01:05:26,040
Speaker 2: chapter is is one idea that some of the magic

1300
01:05:26,120 --> 01:05:28,480
Speaker 2: behind art of intelligence. So one of the chapters is

1301
01:05:28,480 --> 01:05:31,520
Speaker 2: about neural networks and chat gipt and how how they work.

1302
01:05:31,600 --> 01:05:33,640
Speaker 2: Tries to explain and explains the history, where did they

1303
01:05:33,680 --> 01:05:35,800
Speaker 2: come from? Where did that how did that technology come?

1304
01:05:35,840 --> 01:05:37,480
Speaker 2: And some of the colorful characters.

1305
01:05:37,280 --> 01:05:39,920
Speaker 1: Because there's a lot of let's call it jargon in

1306
01:05:40,040 --> 01:05:45,000
Speaker 1: AI world, neural networks, large language models, and all these

1307
01:05:45,080 --> 01:05:47,320
Speaker 1: sorts of things, and it's nearly like it's to me,

1308
01:05:47,440 --> 01:05:52,680
Speaker 1: it looks like it's designed to the average person to say, look,

1309
01:05:54,040 --> 01:05:57,280
Speaker 1: just accept it's too complicated for you. Just go along

1310
01:05:57,320 --> 01:06:01,680
Speaker 1: with it. That this nearly it is nearly a new language.

1311
01:06:02,280 --> 01:06:05,880
Speaker 2: It is. But let's forget we're trying to make machines

1312
01:06:05,920 --> 01:06:09,200
Speaker 2: that are intelligent, and you know intelligence is you know,

1313
01:06:09,360 --> 01:06:11,680
Speaker 2: our brain is the most complex thing in the universe. Yep,

1314
01:06:11,920 --> 01:06:14,919
Speaker 2: by far. Yeah, nothing approaches the complexity, so we're trying

1315
01:06:14,920 --> 01:06:16,800
Speaker 2: to match that. So it's not surprising that it gets

1316
01:06:16,800 --> 01:06:20,520
Speaker 2: a bit technical and it does seem a bit magical

1317
01:06:20,640 --> 01:06:24,040
Speaker 2: to us. And we're easily taken in when the machines

1318
01:06:24,080 --> 01:06:26,760
Speaker 2: start talking back to us, because no one else has

1319
01:06:26,800 --> 01:06:29,760
Speaker 2: ever only how the humans, but the only thing that

1320
01:06:29,800 --> 01:06:32,320
Speaker 2: ever talked back to us were other humans, other intelligent beings.

1321
01:06:32,360 --> 01:06:36,000
Speaker 2: So it's easy to be fooled and seduced by the

1322
01:06:36,080 --> 01:06:38,760
Speaker 2: machines when they started talking to us because our only

1323
01:06:38,840 --> 01:06:41,600
Speaker 2: experience of things that talk back to us are other humans,

1324
01:06:42,800 --> 01:06:44,720
Speaker 2: all of their intelligence and everything else that goes with

1325
01:06:44,800 --> 01:06:45,360
Speaker 2: being humans.

1326
01:06:45,520 --> 01:06:47,800
Speaker 1: You own a book called twenty sixty two The World

1327
01:06:47,840 --> 01:06:49,600
Speaker 1: that AOI made. This is looking into the future.

1328
01:06:49,960 --> 01:06:51,760
Speaker 2: Yes, the year twenty sixty two.

1329
01:06:51,840 --> 01:06:54,680
Speaker 1: Yeah, the year twenty sixty two. So why they're year?

1330
01:06:55,000 --> 01:06:57,960
Speaker 2: Well I gave a talk in North Sydney and this

1331
01:06:58,280 --> 01:07:00,560
Speaker 2: old lady turned up at the talk and she said, oh,

1332
01:07:00,680 --> 01:07:02,720
Speaker 2: so it's nothing to do with my post code. Then

1333
01:07:06,040 --> 01:07:08,600
Speaker 2: I'm very sorry. It's got you here on the forced protector.

1334
01:07:08,600 --> 01:07:09,520
Speaker 2: It's all about the year.

1335
01:07:09,480 --> 01:07:11,120
Speaker 1: Twenty sixty or why twenty sixty two?

1336
01:07:11,560 --> 01:07:14,520
Speaker 2: Well? I surveyed three hundred colleagues other experts around the

1337
01:07:14,520 --> 01:07:17,280
Speaker 2: world in AI. I asked them the simple question, when

1338
01:07:17,320 --> 01:07:21,040
Speaker 2: would computers match humans in their intelligence? And the average

1339
01:07:21,080 --> 01:07:23,680
Speaker 2: answer they said this was ten years ago. It was

1340
01:07:23,720 --> 01:07:26,200
Speaker 2: twenty sixty two. I think I suspect if I did

1341
01:07:26,240 --> 01:07:29,240
Speaker 2: the survey today, they might say twenty forty two. Oh really, yes,

1342
01:07:29,320 --> 01:07:32,080
Speaker 2: I think the date has moved forwards quite significantly.

1343
01:07:32,280 --> 01:07:34,720
Speaker 1: And what do you mean by match our intelligence?

1344
01:07:34,800 --> 01:07:37,400
Speaker 2: To do everything that we can do? That requires intelligence?

1345
01:07:38,720 --> 01:07:43,000
Speaker 2: And what do you mean by intelligence to be able

1346
01:07:43,040 --> 01:07:48,120
Speaker 2: to write a poem, do a tax return, answer a

1347
01:07:48,200 --> 01:07:50,440
Speaker 2: maths question? They can do this, plan a holiday.

1348
01:07:50,640 --> 01:07:54,720
Speaker 1: He can do that now though, yes, why twenty forty two?

1349
01:07:55,200 --> 01:07:57,959
Speaker 1: What is it that do you think that a computer? Sorry,

1350
01:07:58,200 --> 01:08:00,720
Speaker 1: I I will that we can do that. I I

1351
01:08:00,880 --> 01:08:01,760
Speaker 1: can't do anything.

1352
01:08:01,840 --> 01:08:05,640
Speaker 2: Well, So they don't have our creativity, they don't have

1353
01:08:05,720 --> 01:08:08,919
Speaker 2: our emotional intelligence, they don't have our social intelligence. There's

1354
01:08:08,960 --> 01:08:12,640
Speaker 2: still still a very what's called jagged intelligence. You know,

1355
01:08:12,720 --> 01:08:15,640
Speaker 2: there's really good at something so that they can answer

1356
01:08:15,800 --> 01:08:18,960
Speaker 2: international maths olympiad questions. But you know better than almost

1357
01:08:19,000 --> 01:08:21,439
Speaker 2: anyone on the planet. And yet you can ask them,

1358
01:08:21,439 --> 01:08:23,439
Speaker 2: as you know, a simple puzzle, you know, like, here's

1359
01:08:23,439 --> 01:08:28,599
Speaker 2: a simple one that defeats chattypt You know, I've got

1360
01:08:28,640 --> 01:08:31,479
Speaker 2: a two liter jug and a five liter jug. How

1361
01:08:31,520 --> 01:08:35,680
Speaker 2: do I measure out three liters of liquid? And you know,

1362
01:08:35,760 --> 01:08:38,200
Speaker 2: the answer right, you feel the five liter jug to

1363
01:08:38,320 --> 01:08:40,479
Speaker 2: the top, you pour it into the two liter jug,

1364
01:08:40,520 --> 01:08:43,400
Speaker 2: and you've got three liters left to it. But you

1365
01:08:43,520 --> 01:08:46,559
Speaker 2: ask CHATTYPT and it says, it gives you this twelve

1366
01:08:46,600 --> 01:08:50,160
Speaker 2: step plan of pouring water backwards and fords between the jugs,

1367
01:08:51,520 --> 01:08:54,639
Speaker 2: and eventually it says, and now in the two liter jug,

1368
01:08:54,720 --> 01:08:58,200
Speaker 2: you've got three liters of liquid. Batshit crazy. It's not

1369
01:08:59,320 --> 01:09:00,960
Speaker 2: deeply under earning the problem that you.

1370
01:09:01,000 --> 01:09:05,960
Speaker 1: Are is that because it's taken that answer from somewhere else,

1371
01:09:06,080 --> 01:09:07,280
Speaker 1: or it actually deducted and.

1372
01:09:07,600 --> 01:09:10,040
Speaker 2: It's been trained. There's lots of example where it puzzles

1373
01:09:10,080 --> 01:09:13,080
Speaker 2: on the internet where people have talked about jug questions

1374
01:09:13,080 --> 01:09:15,519
Speaker 2: where you pourp but it and it's sort of just

1375
01:09:15,600 --> 01:09:19,160
Speaker 2: giving you a blurred average of those problems where it's

1376
01:09:19,360 --> 01:09:21,479
Speaker 2: lots of pouring liquids backwards. And for us, it's not

1377
01:09:21,600 --> 01:09:25,400
Speaker 2: really sitting there thinking through logically what the steps are.

1378
01:09:25,520 --> 01:09:29,200
Speaker 1: So would like as someone who's like you're a mathematicism,

1379
01:09:29,240 --> 01:09:32,519
Speaker 1: but as someone like you think, would you say, therefore,

1380
01:09:33,280 --> 01:09:35,879
Speaker 1: artificial intelligence is not really making a deduction.

1381
01:09:37,040 --> 01:09:39,160
Speaker 2: Yes, it's not reasoning the way that we're reasing.

1382
01:09:39,240 --> 01:09:41,120
Speaker 1: Yeah, the way we reason I should say the way

1383
01:09:41,200 --> 01:09:42,200
Speaker 1: we would make it deduction.

1384
01:09:42,520 --> 01:09:45,519
Speaker 2: Yes, and yes, I mean we're working on how we

1385
01:09:45,560 --> 01:09:49,120
Speaker 2: can actually add more deductive capabilities to these language models.

1386
01:09:49,160 --> 01:09:52,240
Speaker 2: But they're they're you know, retrieving the answers from you know,

1387
01:09:52,320 --> 01:09:54,559
Speaker 2: what they were trained on more than they are reasoning

1388
01:09:54,600 --> 01:09:55,200
Speaker 2: about the answer.

1389
01:09:55,280 --> 01:09:57,720
Speaker 1: What do you expect for twenty twenty six? That's the

1390
01:09:57,760 --> 01:09:59,960
Speaker 1: opposite of twenty sixty two? But like, what do you

1391
01:10:00,080 --> 01:10:02,400
Speaker 1: what are you looking forward to in twenty twenty six

1392
01:10:02,680 --> 01:10:04,559
Speaker 1: this year when it comes to artificial intelligence?

1393
01:10:04,760 --> 01:10:06,960
Speaker 2: Well, I mean, the exciting thing happening now is is

1394
01:10:07,000 --> 01:10:10,920
Speaker 2: that we have these AI agents that instead of having

1395
01:10:11,040 --> 01:10:12,880
Speaker 2: these chatbots which sit there and wait for you to

1396
01:10:12,920 --> 01:10:16,360
Speaker 2: ask a question, they actually have some agencies, some ability

1397
01:10:16,400 --> 01:10:19,080
Speaker 2: to go and do stuff. So, as an example, when

1398
01:10:19,120 --> 01:10:21,479
Speaker 2: you've got a new employee, there's a bunch of stuff

1399
01:10:21,479 --> 01:10:23,000
Speaker 2: you've got to do. You've got to give you know,

1400
01:10:23,080 --> 01:10:24,840
Speaker 2: you've got to give them a log in, You've got

1401
01:10:24,960 --> 01:10:26,800
Speaker 2: to they've got to do the health and safety on

1402
01:10:26,920 --> 01:10:31,280
Speaker 2: board training. They've got to get a path for the building,

1403
01:10:31,880 --> 01:10:33,640
Speaker 2: they've you know, they've got to you've got to get

1404
01:10:33,640 --> 01:10:37,519
Speaker 2: their financial data so you can pay their pay their

1405
01:10:37,560 --> 01:10:39,519
Speaker 2: wages at the end of the first week, there's a

1406
01:10:39,600 --> 01:10:42,040
Speaker 2: bunch of you know, tasks you've got to do. Well. Now,

1407
01:10:42,160 --> 01:10:44,400
Speaker 2: instead of you having to go through and do all

1408
01:10:44,479 --> 01:10:46,599
Speaker 2: those things yourself, you can just tay to the agent, okay,

1409
01:10:47,360 --> 01:10:49,200
Speaker 2: can you on board this new person, and it breaks

1410
01:10:49,240 --> 01:10:51,160
Speaker 2: the task down into the different things, and then it

1411
01:10:51,240 --> 01:10:53,080
Speaker 2: takes each of those tasks and break those down into

1412
01:10:53,080 --> 01:10:54,759
Speaker 2: the things you have to do. To get the people's

1413
01:10:54,800 --> 01:10:57,599
Speaker 2: financial information, you've got to get some you know, identity documents,

1414
01:10:57,640 --> 01:10:59,800
Speaker 2: you've got to get their bank co ordinance. You've got

1415
01:10:59,840 --> 01:11:01,720
Speaker 2: to just to them in the accountancy system. You've got

1416
01:11:01,760 --> 01:11:03,960
Speaker 2: to set up the payroll. And then it said, you know,

1417
01:11:04,080 --> 01:11:05,680
Speaker 2: to register the payroll. It's got to get you know,

1418
01:11:05,720 --> 01:11:08,360
Speaker 2: all the information their date of birth, the tax number

1419
01:11:09,120 --> 01:11:10,519
Speaker 2: and so on, and it goes through all those things.

1420
01:11:10,640 --> 01:11:11,760
Speaker 2: It does all that stuff for you.

1421
01:11:12,960 --> 01:11:15,640
Speaker 1: And do you see for the twenty twenty six so

1422
01:11:15,960 --> 01:11:18,320
Speaker 1: I mean you just see just more efficiency games or

1423
01:11:18,360 --> 01:11:20,920
Speaker 1: do you see something large coming out of all it's

1424
01:11:20,960 --> 01:11:22,840
Speaker 1: something really disruptive.

1425
01:11:24,160 --> 01:11:27,120
Speaker 2: Well, the disruptive thing that you have there is suddenly

1426
01:11:27,280 --> 01:11:32,400
Speaker 2: that the computers are doing stuff. They're changing your files there,

1427
01:11:32,800 --> 01:11:36,760
Speaker 2: maybe spending money, maybe the ordering stuff and that's going

1428
01:11:36,800 --> 01:11:37,360
Speaker 2: to go wrong.

1429
01:11:37,439 --> 01:11:39,879
Speaker 1: Sometimes that will be disruptive.

1430
01:11:40,000 --> 01:11:41,760
Speaker 2: That's going to be disrupted, and it's like, who are

1431
01:11:41,760 --> 01:11:43,839
Speaker 2: we going to blame because we can't blame the computer,

1432
01:11:44,439 --> 01:11:47,479
Speaker 2: but what will do? Also, things will go wrong at

1433
01:11:47,520 --> 01:11:49,760
Speaker 2: speed because this is a computer speed, not human speed.

1434
01:11:49,840 --> 01:11:52,519
Speaker 1: So if it's trying to reverse, it could be hard

1435
01:11:52,560 --> 01:11:53,040
Speaker 1: to reverse.

1436
01:11:53,240 --> 01:11:55,559
Speaker 2: I mean, there was a story in the paper yesterday

1437
01:11:55,600 --> 01:11:58,680
Speaker 2: where the head of safety at open AI had all

1438
01:11:59,000 --> 01:12:02,840
Speaker 2: email was deleted by one of these agents. She's the

1439
01:12:02,920 --> 01:12:06,040
Speaker 2: head of safety, like a rookie mistake. She gave it

1440
01:12:06,120 --> 01:12:09,559
Speaker 2: the power to clean up her mailbox. You know, I'd

1441
01:12:09,640 --> 01:12:12,120
Speaker 2: love to have someone who could go through my mailbox

1442
01:12:12,160 --> 01:12:14,720
Speaker 2: and clean out all the junk, but also include out

1443
01:12:14,760 --> 01:12:16,120
Speaker 2: all the stuff that she wanted to keep.

1444
01:12:16,320 --> 01:12:20,320
Speaker 1: Because the AI agent had agency and made the decisions.

1445
01:12:19,960 --> 01:12:21,080
Speaker 2: Yes, made the decisions.

1446
01:12:21,280 --> 01:12:24,400
Speaker 1: So that that means the instructions and the rules that

1447
01:12:24,520 --> 01:12:30,320
Speaker 1: have been given to the AAI agent were not not

1448
01:12:30,439 --> 01:12:32,280
Speaker 1: complete or maybe not accurately.

1449
01:12:32,520 --> 01:12:34,080
Speaker 2: It wasn't just that the rules were in complete, but

1450
01:12:34,120 --> 01:12:38,040
Speaker 2: also the AI is not perfect. You know, we talked

1451
01:12:38,040 --> 01:12:41,600
Speaker 2: about how they hallucinate, They make mistakes, you know, just

1452
01:12:41,680 --> 01:12:43,839
Speaker 2: like humans are not perfect. These are not perfect either.

1453
01:12:44,040 --> 01:12:46,479
Speaker 1: So we be, and I guess that's sort of part

1454
01:12:46,479 --> 01:12:49,880
Speaker 1: of it too. I think we should learn to expect imperfection.

1455
01:12:50,200 --> 01:12:52,920
Speaker 2: They are, we should, But I mean, the interesting thing

1456
01:12:53,040 --> 01:12:56,160
Speaker 2: is people have higher standards for machines than they'd have

1457
01:12:56,439 --> 01:12:56,800
Speaker 2: to have.

1458
01:12:56,960 --> 01:12:57,559
Speaker 1: That's weird.

1459
01:12:57,880 --> 01:13:00,840
Speaker 2: So you say to people, you know, we've got self

1460
01:13:00,920 --> 01:13:04,479
Speaker 2: driving cars now, and they're roughly the level of humans

1461
01:13:04,680 --> 01:13:07,000
Speaker 2: in terms of, you know, the number of mistakes they make,

1462
01:13:07,080 --> 01:13:10,200
Speaker 2: and some ways they're better than humans they make fewer mistakes,

1463
01:13:10,200 --> 01:13:14,000
Speaker 2: but people are still very reluctant to give them responsibility

1464
01:13:14,080 --> 01:13:16,880
Speaker 2: to drive. So you say, well, okay, well how much

1465
01:13:16,960 --> 01:13:19,280
Speaker 2: better than humans? If they do, they have to be

1466
01:13:19,360 --> 01:13:23,000
Speaker 2: ten times better than humans before we can entrust them

1467
01:13:23,040 --> 01:13:27,360
Speaker 2: with the responsibility of driving without human oversight. And then

1468
01:13:27,400 --> 01:13:29,160
Speaker 2: at some point, you know, okay, so maybe we say

1469
01:13:29,200 --> 01:13:32,960
Speaker 2: it's ten Well, what happens when they're a hundred times

1470
01:13:33,120 --> 01:13:37,240
Speaker 2: better than humans? Do we let humans still drive? Because

1471
01:13:37,439 --> 01:13:40,360
Speaker 2: humans are going to be killing people. Machines are provably

1472
01:13:40,439 --> 01:13:42,080
Speaker 2: not going to be killed. They can't get drunk. They

1473
01:13:42,080 --> 01:13:44,320
Speaker 2: can't get drunk, they're going to be tired, they overtired. Yeah,

1474
01:13:44,320 --> 01:13:45,960
Speaker 2: they're not going to be speeding, They're not going to

1475
01:13:46,000 --> 01:13:50,160
Speaker 2: be breaking the road rules. At what point do we say, humans,

1476
01:13:50,280 --> 01:13:53,439
Speaker 2: you know you're killing you know, your fellow citizens, maybe

1477
01:13:53,439 --> 01:13:54,200
Speaker 2: we shouldn't allow this.

1478
01:13:54,439 --> 01:13:56,040
Speaker 1: Well, a friend of mine the other day told me

1479
01:13:56,120 --> 01:14:01,080
Speaker 1: that Tom she was in San Francisco and the incinet

1480
01:14:01,160 --> 01:14:05,360
Speaker 1: went down, stopped and they all stopped, which is the

1481
01:14:05,400 --> 01:14:07,200
Speaker 1: stop stop there and then, which is the safe thing

1482
01:14:07,240 --> 01:14:10,280
Speaker 1: to do totally. But there were other non waymos on

1483
01:14:10,400 --> 01:14:14,599
Speaker 1: the road. Yes, humans people driving cars. I'm now called

1484
01:14:14,640 --> 01:14:18,560
Speaker 1: to non non and you're driving and and like I

1485
01:14:18,640 --> 01:14:21,200
Speaker 1: all of a sudden the things because literally they're just yeah,

1486
01:14:22,040 --> 01:14:23,880
Speaker 1: they're not used, they're not going very fast because they're

1487
01:14:23,880 --> 01:14:27,719
Speaker 1: all governed. But nonetheless they just stopped because something happened

1488
01:14:27,720 --> 01:14:32,320
Speaker 1: in obviously and yeah, and we keep forgetting that. The

1489
01:14:33,280 --> 01:14:40,559
Speaker 1: the Achilles heel of all artificial intelligence is that is power. Yes,

1490
01:14:41,640 --> 01:14:43,920
Speaker 1: I mean that's what drives everything. If you turn off

1491
01:14:43,960 --> 01:14:49,200
Speaker 1: the switch, not literally, if you know what I mean metaphorically,

1492
01:14:49,200 --> 01:14:52,080
Speaker 1: if you just turn the power off this thing that

1493
01:14:52,120 --> 01:14:55,519
Speaker 1: we're really worried about, or somehow you can if the

1494
01:14:55,600 --> 01:14:59,599
Speaker 1: thing is battery based, like it's some other way it's powered,

1495
01:15:00,080 --> 01:15:02,880
Speaker 1: not not by electricity with something, not by put money

1496
01:15:02,880 --> 01:15:05,360
Speaker 1: in a wall. If somebody you can disconnect the power

1497
01:15:06,000 --> 01:15:07,880
Speaker 1: thing stops, it does.

1498
01:15:07,960 --> 01:15:10,800
Speaker 2: But that's the problem as well. I am forced to

1499
01:15:10,840 --> 01:15:14,040
Speaker 2: write this fantastic science fiction story one hundred years ago

1500
01:15:14,680 --> 01:15:19,160
Speaker 2: titled When the Machine Stops, about how wig becoming ever

1501
01:15:19,280 --> 01:15:22,439
Speaker 2: more dependent upon the machines and the machines are becoming

1502
01:15:22,720 --> 01:15:29,400
Speaker 2: ever more when we're losing touch with actually how they work.

1503
01:15:31,600 --> 01:15:33,519
Speaker 2: And the short story is about what happens, you know,

1504
01:15:33,600 --> 01:15:37,160
Speaker 2: when the machines eventually stop break, no one knows how

1505
01:15:37,160 --> 01:15:39,080
Speaker 2: to fix them, and the world grinds to a halt.

1506
01:15:40,120 --> 01:15:40,360
Speaker 1: Yeah.

1507
01:15:40,520 --> 01:15:42,120
Speaker 2: So it's not that you know that we turn the

1508
01:15:42,200 --> 01:15:44,960
Speaker 2: machines off. It's problems. But when the machine stops, what happens?

1509
01:15:45,520 --> 01:15:48,080
Speaker 2: Could we still do anything? And we're already completely dependent.

1510
01:15:48,240 --> 01:15:50,840
Speaker 2: If the Internet goes down, the stock market stops, the

1511
01:15:51,240 --> 01:15:56,840
Speaker 2: air traffic you know, you know in the university, you know,

1512
01:15:56,880 --> 01:15:59,400
Speaker 2: the internet goes down, people say, well, time to go home.

1513
01:16:00,000 --> 01:16:02,519
Speaker 1: I think we can do yeah go home? Yeah, Well

1514
01:16:03,040 --> 01:16:03,519
Speaker 1: I'm wishing.

1515
01:16:03,720 --> 01:16:05,720
Speaker 2: And now even that is problematic because lots of people

1516
01:16:05,760 --> 01:16:08,920
Speaker 2: have got smart doorbells and so on, and they can't

1517
01:16:08,960 --> 01:16:10,000
Speaker 2: get into their homes anymore.

1518
01:16:10,640 --> 01:16:13,559
Speaker 1: Yeah, because it's either thing of printer or some sort

1519
01:16:13,560 --> 01:16:17,240
Speaker 1: of biometric you can't get out. We were in an

1520
01:16:17,280 --> 01:16:20,800
Speaker 1: old fashion key hidden somewhere exactly somewhere, and that's your

1521
01:16:20,840 --> 01:16:21,920
Speaker 1: redundancy process.

1522
01:16:23,080 --> 01:16:24,840
Speaker 2: My wife was saying, should we get some of those

1523
01:16:24,840 --> 01:16:29,240
Speaker 2: smart doorbells and things. I said, no, keys are fantastic.

1524
01:16:29,320 --> 01:16:33,120
Speaker 1: They work totally the redundancy and you can have and

1525
01:16:33,240 --> 01:16:35,519
Speaker 1: you can have a second key. If you lose that key,

1526
01:16:35,560 --> 01:16:37,479
Speaker 1: you can have another key hidden somewhere, which is what

1527
01:16:37,560 --> 01:16:42,760
Speaker 1: most people do. I really appreciate this conversation. I've had

1528
01:16:42,760 --> 01:16:44,240
Speaker 1: a lot of fun talking to you about it, but

1529
01:16:44,720 --> 01:16:46,560
Speaker 1: equally on the funds off on the other side of

1530
01:16:46,640 --> 01:16:50,320
Speaker 1: fund is a serious topic. Yeah, a very serious topic topic,

1531
01:16:50,360 --> 01:16:51,960
Speaker 1: and I wish you the best when you go to

1532
01:16:51,960 --> 01:16:55,320
Speaker 1: the United Nations next week and hopefully, you know, we

1533
01:16:55,439 --> 01:16:59,560
Speaker 1: can start to get some better road rules around or

1534
01:17:00,160 --> 01:17:02,400
Speaker 1: get the conversation going about what the road rules are

1535
01:17:02,840 --> 01:17:05,639
Speaker 1: for the future. Indeed, thank you very much.

1536
01:17:05,760 --> 01:17:06,400
Speaker 2: Been a pleasure by