1
00:00:00,160 --> 00:00:03,160
Speaker 1: Talk to Shane Cooper in a short while. He's the

2
00:00:03,200 --> 00:00:07,840
Speaker 1: head of Digital Advisory at fourvist Mazas in South Africa,

3
00:00:08,520 --> 00:00:11,880
Speaker 1: corporate governance and consulting. Are we going to be asking

4
00:00:11,920 --> 00:00:16,000
Speaker 1: if your organization is truly ready for AI? In twenty

5
00:00:16,120 --> 00:00:20,639
Speaker 1: twenty six and based on Shane's recent work, he focuses

6
00:00:20,680 --> 00:00:24,520
Speaker 1: on institutional nor portfolio management. He've uws AI as a

7
00:00:24,600 --> 00:00:28,560
Speaker 1: layer that stops the rot early by catching things human's

8
00:00:28,640 --> 00:00:31,880
Speaker 1: miss in massive data sets, and that we can understand

9
00:00:32,360 --> 00:00:38,440
Speaker 1: it can kind of take the lift. The heavy stuff

10
00:00:39,080 --> 00:00:43,519
Speaker 1: is philosophy. AI readiness is about reallocating human efforts. You

11
00:00:43,560 --> 00:00:46,919
Speaker 1: aren't ready if you just automated a task. You are

12
00:00:46,960 --> 00:00:50,360
Speaker 1: ready when your human experts are freed up to perform

13
00:00:50,479 --> 00:00:55,240
Speaker 1: high level oversight and strategy while they trace trust. Of course,

14
00:00:55,280 --> 00:01:01,720
Speaker 1: the AI is intelligent ingestion and heavy heavy lifting. So yeah,

15
00:01:02,000 --> 00:01:09,520
Speaker 1: what is operational organizational readiness in twenty sixteen? Welcome Shane Cooper.

16
00:01:09,560 --> 00:01:11,120
Speaker 1: It's great to have you with us.

17
00:01:11,560 --> 00:01:13,200
Speaker 2: Thank you very much. It's great to be with.

18
00:01:13,160 --> 00:01:18,959
Speaker 1: You, Shane. I would imagine that the heavy lifting is

19
00:01:19,080 --> 00:01:22,360
Speaker 1: important you get a greater sense of your business, of

20
00:01:22,400 --> 00:01:25,440
Speaker 1: your business. But you've also mentioned a shift from mechanical

21
00:01:25,440 --> 00:01:29,080
Speaker 1: to judgment work. Let's just understand that what does that mean.

22
00:01:29,160 --> 00:01:32,920
Speaker 1: What's the mechanical work you're referring to, and what is

23
00:01:32,959 --> 00:01:36,120
Speaker 1: the judgment work that you want to see achieved.

24
00:01:37,640 --> 00:01:42,319
Speaker 2: So from a go back to first principle of mechanical work,

25
00:01:42,840 --> 00:01:47,680
Speaker 2: one could argue is typically quite deterministic. You have a

26
00:01:47,920 --> 00:01:52,320
Speaker 2: very clear understanding of the activities that you need to undertake,

27
00:01:53,440 --> 00:01:57,720
Speaker 2: and I would call i'll probably raughly categorize that work

28
00:01:57,760 --> 00:02:02,240
Speaker 2: as mundane tasks. Where judgment work, you know, it's in

29
00:02:02,240 --> 00:02:06,320
Speaker 2: the word, it requires of us to be very considered

30
00:02:06,400 --> 00:02:09,880
Speaker 2: with the tasks that we do and to provide a

31
00:02:09,960 --> 00:02:13,280
Speaker 2: judgment as we perform our task on a daily basis.

32
00:02:13,800 --> 00:02:17,919
Speaker 2: And the argument that we are putting forward is that

33
00:02:18,520 --> 00:02:22,720
Speaker 2: with the advent of AI as we know it, there

34
00:02:22,800 --> 00:02:26,640
Speaker 2: is a risk that as knowledge work begins to be

35
00:02:26,960 --> 00:02:32,120
Speaker 2: undertaken by artificial intelligence, organizations are not setting themselves up

36
00:02:32,120 --> 00:02:33,080
Speaker 2: for success.

37
00:02:34,480 --> 00:02:36,200
Speaker 1: I get that, I get that, but we need to

38
00:02:36,200 --> 00:02:38,920
Speaker 1: move some paradigms. Yeah, so let's just take a step back.

39
00:02:39,320 --> 00:02:43,160
Speaker 1: How do we measure readiness in a seem that historically

40
00:02:43,200 --> 00:02:47,359
Speaker 1: has been maybe rewarded for high volume mechanical output.

41
00:02:49,360 --> 00:02:51,640
Speaker 2: Yeah, and I think, I mean, it's a very good question.

42
00:02:51,680 --> 00:02:56,080
Speaker 2: And if I were to position that question with an

43
00:02:56,160 --> 00:02:59,960
Speaker 2: answer slightly differently, I think what we're what we're witnessing

44
00:03:00,440 --> 00:03:03,280
Speaker 2: in a number of industries around the world is that

45
00:03:03,639 --> 00:03:06,320
Speaker 2: the mechanical tasks and even if you were to, you know,

46
00:03:06,400 --> 00:03:13,680
Speaker 2: to consider something like coding, there is a significant disruption underway.

47
00:03:13,919 --> 00:03:17,840
Speaker 2: I think there is a broad acceptance that the lower

48
00:03:17,960 --> 00:03:21,919
Speaker 2: run of employment in the knowledge workspace around the world

49
00:03:22,120 --> 00:03:28,360
Speaker 2: is severely under threat. Early indications at the moment in

50
00:03:28,400 --> 00:03:32,840
Speaker 2: the US is suggesting that youngsters that are emerging from

51
00:03:33,240 --> 00:03:38,000
Speaker 2: from universities are finding it difficult to find jobs. Now,

52
00:03:38,240 --> 00:03:41,280
Speaker 2: coming back to the question around how you measure a

53
00:03:41,320 --> 00:03:45,920
Speaker 2: performance for people that are traditionally in that space, those

54
00:03:46,080 --> 00:03:49,800
Speaker 2: that you know, those performance metrics are relatively simple. You

55
00:03:49,840 --> 00:03:53,200
Speaker 2: have a very clear job description, and you're very clear

56
00:03:53,240 --> 00:03:57,800
Speaker 2: on what what success looks like. The paradigm shift that

57
00:03:57,840 --> 00:04:02,400
Speaker 2: we're looking at is when technology like artificial intelligence arrives,

58
00:04:02,600 --> 00:04:05,040
Speaker 2: and I can assure you it will arrive whether you

59
00:04:05,160 --> 00:04:07,680
Speaker 2: like it or not, it is going to change the

60
00:04:07,720 --> 00:04:11,840
Speaker 2: way organizations look at how those tasks are performed. So

61
00:04:12,120 --> 00:04:16,720
Speaker 2: wholesale disruption of traditional processes and indeed wholesale disruptions of

62
00:04:16,760 --> 00:04:19,240
Speaker 2: what people actually do on a day to day basis.

63
00:04:20,920 --> 00:04:26,880
Speaker 1: Okay, so let's talk about readiness then, and just what

64
00:04:26,920 --> 00:04:30,960
Speaker 1: are we in for the members of a board that

65
00:04:31,000 --> 00:04:34,760
Speaker 1: are listening rates now, the owners of companies, Well, what's

66
00:04:34,760 --> 00:04:38,920
Speaker 1: that hidden cost of AA readiness that we are overlooking

67
00:04:38,920 --> 00:04:41,640
Speaker 1: at this moment in time or maybe resisting at this

68
00:04:41,720 --> 00:04:42,320
Speaker 1: moment in time.

69
00:04:43,800 --> 00:04:46,560
Speaker 2: I think the general spine of the argument is that

70
00:04:47,279 --> 00:04:52,320
Speaker 2: AI interest is high at both board and executive level,

71
00:04:52,360 --> 00:04:56,160
Speaker 2: but operational readiness is often weak. And if there are

72
00:04:56,200 --> 00:05:00,600
Speaker 2: board members listening, I would caution that one shouldn't confuse

73
00:05:01,680 --> 00:05:06,760
Speaker 2: executive interest in AI for readiness, and they typically five

74
00:05:06,960 --> 00:05:10,279
Speaker 2: things that they should look out for. So one, of

75
00:05:10,320 --> 00:05:14,000
Speaker 2: course is in a classically strategy, do we have a

76
00:05:14,160 --> 00:05:17,600
Speaker 2: board back roadmap linked to proper business outcomes? Now, I

77
00:05:17,600 --> 00:05:19,680
Speaker 2: would argue that if you if I were to call

78
00:05:19,720 --> 00:05:21,360
Speaker 2: a board member at three o'clock in the morning and

79
00:05:21,640 --> 00:05:25,760
Speaker 2: ask them, could they list the outcomes that they've agreed

80
00:05:25,760 --> 00:05:27,719
Speaker 2: that AI should be addressing, they should be able to

81
00:05:27,760 --> 00:05:31,520
Speaker 2: answer those. So strategy is critical. The other one is

82
00:05:32,000 --> 00:05:35,160
Speaker 2: use case value. So those use cases that you've identified,

83
00:05:35,560 --> 00:05:39,000
Speaker 2: are they tied to measurable value? If not, then then

84
00:05:39,040 --> 00:05:41,240
Speaker 2: I would argue that you're just dealing with AI theater,

85
00:05:41,320 --> 00:05:44,039
Speaker 2: which is which is problematic. The third one is data

86
00:05:44,160 --> 00:05:48,520
Speaker 2: and infrastructure, and this is so classically any digital transformation

87
00:05:49,360 --> 00:05:51,680
Speaker 2: process that you'd undertake has to consider this. So is

88
00:05:51,720 --> 00:05:55,279
Speaker 2: the data that you have in your organization reliable? Is

89
00:05:55,279 --> 00:05:58,560
Speaker 2: it accessible and good enough? And can the infra infrastructure

90
00:05:58,560 --> 00:06:03,120
Speaker 2: scale safely and cost effectively. The fourth one is governance,

91
00:06:03,240 --> 00:06:06,839
Speaker 2: risk and security, And I mean we are all and

92
00:06:06,960 --> 00:06:09,919
Speaker 2: all of your listeners, I'm sure are attuned to the

93
00:06:09,960 --> 00:06:14,480
Speaker 2: fact that cyber risk is a real challenge for organizations

94
00:06:14,520 --> 00:06:16,000
Speaker 2: and countries around the world, So you need to make

95
00:06:16,000 --> 00:06:18,160
Speaker 2: sure that the controls that you have in place are

96
00:06:18,160 --> 00:06:21,719
Speaker 2: proportionate to the risk. And then, lastly and probably most importantly,

97
00:06:21,760 --> 00:06:24,200
Speaker 2: which talks a little bit about what we were talking earlier,

98
00:06:24,480 --> 00:06:27,400
Speaker 2: is around people and adoption. Do we have the talent,

99
00:06:28,279 --> 00:06:32,520
Speaker 2: is the training in place? And is there broad organizational

100
00:06:32,560 --> 00:06:36,359
Speaker 2: willingness to embed AI into operations? So these are the

101
00:06:36,400 --> 00:06:39,600
Speaker 2: practical questions that board members need to ask if their

102
00:06:39,640 --> 00:06:42,479
Speaker 2: executive teams, and if there are gaps in any of

103
00:06:42,480 --> 00:06:44,960
Speaker 2: those five, then you probably have issues.

104
00:06:46,640 --> 00:06:49,200
Speaker 1: Wow, okay. As Shane Cooper is our guest head of

105
00:06:49,240 --> 00:06:53,360
Speaker 1: Digital Advisory at fours Mazaars. The question that we are

106
00:06:53,400 --> 00:06:56,279
Speaker 1: asking is your organization truly ready for AI in twenty

107
00:06:56,400 --> 00:06:58,400
Speaker 1: twenty sex do you also just stand the risk if

108
00:06:58,440 --> 00:07:01,600
Speaker 1: you are not maybe going to put more effort then

109
00:07:02,160 --> 00:07:06,720
Speaker 1: on this particular front of lagging behind and and how

110
00:07:06,760 --> 00:07:11,680
Speaker 1: could lagging behind impact your business? Shame well.

111
00:07:11,920 --> 00:07:16,720
Speaker 2: I think the the the consensus across the world is

112
00:07:16,760 --> 00:07:20,160
Speaker 2: that that AI and we probably see AIS as a tsunami.

113
00:07:20,200 --> 00:07:22,480
Speaker 2: You have to be ready for it. And there are

114
00:07:22,600 --> 00:07:25,640
Speaker 2: arguments around whether you whether you want to to to

115
00:07:26,120 --> 00:07:29,280
Speaker 2: delve into the space or not. I think that if

116
00:07:29,280 --> 00:07:31,920
Speaker 2: you don't, there will be a bifurcation of society. There

117
00:07:31,920 --> 00:07:34,080
Speaker 2: will be those that adopt AI and those who don't.

118
00:07:34,840 --> 00:07:36,480
Speaker 2: And I think the risk that you have if you

119
00:07:36,600 --> 00:07:39,520
Speaker 2: if you're not focusing on this intensively, is that you're

120
00:07:39,520 --> 00:07:42,400
Speaker 2: going to be wasting your investment. You have to focus,

121
00:07:42,440 --> 00:07:44,840
Speaker 2: even if it is from a risk perspective that you

122
00:07:44,920 --> 00:07:47,800
Speaker 2: are aware of how your business is going to be

123
00:07:47,840 --> 00:07:50,240
Speaker 2: disrupted by competitors who arrive in your space. I think

124
00:07:50,280 --> 00:07:52,520
Speaker 2: there's a there's a very real risk that you develop

125
00:07:52,560 --> 00:07:55,160
Speaker 2: a blind spot around who is going to compete with

126
00:07:55,240 --> 00:07:57,320
Speaker 2: you in your world if you do not if you

127
00:07:57,320 --> 00:07:59,480
Speaker 2: do not pay attention to to how AI is going

128
00:07:59,480 --> 00:08:00,400
Speaker 2: to impact your business.

129
00:08:01,480 --> 00:08:03,840
Speaker 1: So let's talk about and you say, A handles the

130
00:08:03,880 --> 00:08:08,800
Speaker 1: mechanical tasks, leaving humans to focus on judgment. Let's look

131
00:08:08,840 --> 00:08:14,240
Speaker 1: at middle management. So readiness involves redesigning roles and if

132
00:08:14,280 --> 00:08:17,520
Speaker 1: AI is doing what percentage of the routine work do

133
00:08:17,560 --> 00:08:19,080
Speaker 1: you imagine it does one hundred percent?

134
00:08:21,680 --> 00:08:25,240
Speaker 2: I think you look at this stage, it's it's way

135
00:08:25,320 --> 00:08:29,600
Speaker 2: too early to expect anything more than fifty or sixty percent.

136
00:08:29,640 --> 00:08:31,560
Speaker 2: So this is this term that we referred to as

137
00:08:31,680 --> 00:08:34,800
Speaker 2: as human in the loop. So once you deploy AI

138
00:08:35,400 --> 00:08:38,480
Speaker 2: to take over the mundane tasks, you do have to

139
00:08:38,559 --> 00:08:41,040
Speaker 2: ensure that there is a human in the loop to

140
00:08:41,040 --> 00:08:44,720
Speaker 2: to verify whether whether the output is as you expect it.

141
00:08:44,720 --> 00:08:47,559
Speaker 2: Because we need to remember that AI is by and

142
00:08:47,679 --> 00:08:51,280
Speaker 2: large probabilistic. It's you know, it's not a deterministic system.

143
00:08:51,360 --> 00:08:55,080
Speaker 2: So there is the risk that AI does something that

144
00:08:55,120 --> 00:08:57,800
Speaker 2: you would not like. But I think that over the

145
00:08:58,000 --> 00:09:01,920
Speaker 2: over the next year or two, in particular, as agentic

146
00:09:02,040 --> 00:09:05,559
Speaker 2: AI becomes more mature, I think up to eighty or

147
00:09:05,640 --> 00:09:09,520
Speaker 2: ninety percent of mechanical knowledge work activity can be taken

148
00:09:09,600 --> 00:09:11,920
Speaker 2: up by AI. And the argument here is that, you know,

149
00:09:11,960 --> 00:09:17,080
Speaker 2: you move the resource availability in an organization away from

150
00:09:17,120 --> 00:09:20,720
Speaker 2: those mundane tasks to be sure, focused on the value added.

151
00:09:20,600 --> 00:09:26,000
Speaker 1: Judgmental hejudge and judgment as well. Okay, right, let's let's

152
00:09:26,040 --> 00:09:28,280
Speaker 1: just take that that line of thoughts a little further

153
00:09:28,440 --> 00:09:32,360
Speaker 1: so that the AI ready organization then as upskilled, it's

154
00:09:32,400 --> 00:09:36,200
Speaker 1: its managers to interpret AI reasoning. And you say that

155
00:09:36,320 --> 00:09:41,680
Speaker 1: is a probable probabilistic rather than deterministic, I would become

156
00:09:41,720 --> 00:09:43,480
Speaker 1: increasingly deterministic as well.

157
00:09:44,679 --> 00:09:48,640
Speaker 2: Yes, absolutely, Now you know, there's this this term in

158
00:09:48,920 --> 00:09:52,280
Speaker 2: the AI space called hallucination. There are some some more

159
00:09:52,960 --> 00:09:57,240
Speaker 2: colorful terminology, aren't there, But there is there is the

160
00:09:57,360 --> 00:10:01,400
Speaker 2: risk that that AI systems, even to they produce output

161
00:10:01,480 --> 00:10:04,920
Speaker 2: that is not connected to reality, reality or truth. But

162
00:10:05,000 --> 00:10:09,959
Speaker 2: the the the it is now known why these large

163
00:10:10,040 --> 00:10:15,000
Speaker 2: language models hallucinate and and the training and the retraining

164
00:10:15,000 --> 00:10:17,920
Speaker 2: that these models are undergoing is reducing that risk. And

165
00:10:17,960 --> 00:10:22,280
Speaker 2: there are also tools and guardrails that you can put

166
00:10:22,320 --> 00:10:24,880
Speaker 2: in place in organizations to reduce that risk.

167
00:10:26,800 --> 00:10:30,800
Speaker 1: Okay, right, this is a this is a lot to

168
00:10:30,920 --> 00:10:35,120
Speaker 1: wrap the mind around. I would imagine that many firms

169
00:10:35,200 --> 00:10:40,920
Speaker 1: have very significant let's call them data lakes lakes nearly

170
00:10:41,360 --> 00:10:45,080
Speaker 1: or they swamps. Yeah, so what is that that first

171
00:10:45,080 --> 00:10:50,360
Speaker 1: step in making that unstructured data truly AI ready, what

172
00:10:50,360 --> 00:10:51,160
Speaker 1: what do they do with that?

173
00:10:52,559 --> 00:10:56,040
Speaker 2: Well? This is this is actually a fairly fairly common

174
00:10:56,080 --> 00:10:59,800
Speaker 2: problem that many organizations have, especially older organizations that have

175
00:11:00,040 --> 00:11:04,680
Speaker 2: ultiple systems. The idea is that if you would to imagine,

176
00:11:04,840 --> 00:11:08,680
Speaker 2: I mean the lake references is there because it's a

177
00:11:08,679 --> 00:11:12,400
Speaker 2: good analogy for what we're solving for is that all

178
00:11:12,440 --> 00:11:15,800
Speaker 2: of the data in your organization that you use to

179
00:11:15,880 --> 00:11:19,560
Speaker 2: drive decisions needs to be in a single place, and

180
00:11:19,600 --> 00:11:21,440
Speaker 2: it needs to be clean. In other words, I mean

181
00:11:21,480 --> 00:11:24,120
Speaker 2: it's quite common for data sets to be to be

182
00:11:24,240 --> 00:11:27,760
Speaker 2: dirty or unstructured. You have to as your first step

183
00:11:27,800 --> 00:11:29,920
Speaker 2: on a journeys to ensure that your data is clean,

184
00:11:30,640 --> 00:11:33,760
Speaker 2: and then you overlay that with very interesting technologies out

185
00:11:33,760 --> 00:11:35,840
Speaker 2: there that you can You know, you're in an now

186
00:11:35,880 --> 00:11:39,160
Speaker 2: in a position as an executive to truly make data

187
00:11:39,200 --> 00:11:41,640
Speaker 2: driven decisions and decisions that are made on the fly.

188
00:11:42,960 --> 00:11:48,760
Speaker 1: I see, Dean, No, it's not related. But so how

189
00:11:48,800 --> 00:11:54,840
Speaker 1: do we balance AI autonomy or ugentic AI with what

190
00:11:54,920 --> 00:11:58,720
Speaker 1: you suggest strict requirements for what are human in the

191
00:11:58,800 --> 00:12:00,480
Speaker 1: loop traceability.

192
00:12:00,000 --> 00:12:05,760
Speaker 2: Early, Yeah, and again these are these are our challenges

193
00:12:05,800 --> 00:12:08,400
Speaker 2: that organizations are around the world and in the technologies

194
00:12:08,800 --> 00:12:12,000
Speaker 2: technologists are grappling with. Is the extent to which you

195
00:12:12,120 --> 00:12:16,640
Speaker 2: release full agentic systems into an organization, you know, in

196
00:12:17,040 --> 00:12:20,400
Speaker 2: high risk environments in the medical space as an example,

197
00:12:21,160 --> 00:12:23,440
Speaker 2: Clearly you're not going to take that risk. Now you

198
00:12:23,559 --> 00:12:27,840
Speaker 2: potentially look to to deploy agentic tool sets in areas

199
00:12:27,880 --> 00:12:30,839
Speaker 2: where it is okay if you know, if it if

200
00:12:30,840 --> 00:12:33,040
Speaker 2: it were to hallucinate or to perform an activity that

201
00:12:33,080 --> 00:12:35,720
Speaker 2: you wouldn't. But it is important that we begin to

202
00:12:35,800 --> 00:12:41,400
Speaker 2: explore this because the whole agenic AI concept will transform

203
00:12:41,440 --> 00:12:45,239
Speaker 2: the way organizations set up processes and systems with Craccastically,

204
00:12:45,320 --> 00:12:48,840
Speaker 2: businesses are set up for humans to interact with systems.

205
00:12:49,320 --> 00:12:53,200
Speaker 2: Agentic AI is going to transform that and and but

206
00:12:53,240 --> 00:12:55,040
Speaker 2: you do need to be aware that that for the

207
00:12:55,200 --> 00:12:58,800
Speaker 2: for the time being, there are some concerns around how

208
00:12:58,960 --> 00:13:01,319
Speaker 2: these systems will behave, But that doesn't mean that you

209
00:13:01,320 --> 00:13:02,280
Speaker 2: shouldn't be exploring.

210
00:13:02,920 --> 00:13:05,360
Speaker 1: Does that talk to the missing piece in the puzzle?

211
00:13:05,400 --> 00:13:10,240
Speaker 1: Does that talk to the lack of regulation in this environment.

212
00:13:11,280 --> 00:13:14,800
Speaker 2: Yeah, absolutely, there is. There are some standards beginning to

213
00:13:15,040 --> 00:13:18,480
Speaker 2: emerge around the world. The problem of course with policies

214
00:13:18,559 --> 00:13:22,480
Speaker 2: that it tends to move as fast as treacle and

215
00:13:22,520 --> 00:13:26,440
Speaker 2: with technology moving quickly and keeping pace with technology advancements

216
00:13:26,480 --> 00:13:30,240
Speaker 2: is a challenge. We are beginning to see some strong

217
00:13:30,320 --> 00:13:34,760
Speaker 2: regulations emerging from Europe South Africa. As a positioning paper,

218
00:13:35,200 --> 00:13:38,800
Speaker 2: I suspect that probably around twenty twenty seven will begin

219
00:13:38,880 --> 00:13:42,120
Speaker 2: to see good frameworks come in together. But it is

220
00:13:42,240 --> 00:13:46,040
Speaker 2: important that we do have the policy guard rails in

221
00:13:46,120 --> 00:13:50,079
Speaker 2: place to manage how these things are implemented, especially in

222
00:13:50,400 --> 00:13:51,439
Speaker 2: sensitive environments.

223
00:13:52,200 --> 00:13:55,800
Speaker 1: Okay, just coming back to the roles that AI are

224
00:13:55,840 --> 00:14:01,080
Speaker 1: playing and well playing in the very future. My research

225
00:14:01,120 --> 00:14:04,040
Speaker 1: suggests what attend to twenty percent reduction in traditional middle

226
00:14:04,480 --> 00:14:07,600
Speaker 1: management roles? We're seeing that at present? Can you confirm that?

227
00:14:08,720 --> 00:14:15,000
Speaker 2: Yeah, absolutely, and perhaps less in less so in South Africa,

228
00:14:15,800 --> 00:14:18,920
Speaker 2: but certainly in areas that have that have aggressively adopted

229
00:14:19,320 --> 00:14:22,520
Speaker 2: these technologies and US is the US and China are

230
00:14:22,560 --> 00:14:27,560
Speaker 2: often good reference points to see what's coming management. Absolutely,

231
00:14:27,680 --> 00:14:31,360
Speaker 2: anywhere between ten and twenty five percent of middle management

232
00:14:31,440 --> 00:14:35,360
Speaker 2: is being impacted, and as I mentioned earlier, the bottom

233
00:14:35,480 --> 00:14:39,280
Speaker 2: run of knowledge work is being significantly impacted.

234
00:14:39,760 --> 00:14:43,040
Speaker 1: How do you avoid the loss of institutional knowledge? How

235
00:14:43,080 --> 00:14:46,200
Speaker 1: do you avoid let's call it cultural collapse. Do you

236
00:14:46,400 --> 00:14:50,240
Speaker 1: restructure the roles of that middle management that are no

237
00:14:50,280 --> 00:14:51,080
Speaker 1: longer required.

238
00:14:52,400 --> 00:14:56,760
Speaker 2: That's a that's a tough question to answer, and I

239
00:14:56,760 --> 00:15:01,960
Speaker 2: think the easy answers to say, let's let's repossession and

240
00:15:02,040 --> 00:15:06,440
Speaker 2: create new roles. But of course, you know, when executives

241
00:15:06,520 --> 00:15:11,080
Speaker 2: and boards and shareholders see an opportunity to increase your

242
00:15:11,160 --> 00:15:14,400
Speaker 2: net margin, you know, normally those decisions are taken. So

243
00:15:14,440 --> 00:15:18,240
Speaker 2: that's a that's a very very tricky situation to be in.

244
00:15:18,960 --> 00:15:22,600
Speaker 2: I think there is a real risk of fragmentation in

245
00:15:22,680 --> 00:15:27,720
Speaker 2: terms of institutional knowledge and culture disruption from from AAR.

246
00:15:28,160 --> 00:15:30,440
Speaker 2: I don't think there is a debate about that. And

247
00:15:30,480 --> 00:15:32,600
Speaker 2: then you know, the question now is going to be

248
00:15:32,640 --> 00:15:34,920
Speaker 2: is how do you how do you prevent that from

249
00:15:35,040 --> 00:15:38,360
Speaker 2: you know, from from properly disturbing how organizations maintain their

250
00:15:38,480 --> 00:15:39,440
Speaker 2: their cultures.

251
00:15:39,600 --> 00:15:44,320
Speaker 1: So in terms of the territory of being future ready, yeah,

252
00:15:44,880 --> 00:15:46,920
Speaker 1: let's also just talk about that ten to twenty percent,

253
00:15:47,000 --> 00:15:49,680
Speaker 1: and that would accrue directly to the bottom line. Or

254
00:15:49,720 --> 00:15:53,440
Speaker 1: what can and that may be a very let's just say,

255
00:15:54,880 --> 00:15:58,520
Speaker 1: you know something that that your shoulders would would would

256
00:15:58,560 --> 00:16:01,359
Speaker 1: want you to do, but you're risking something in the process.

257
00:16:01,040 --> 00:16:05,560
Speaker 2: Right, No, absolutely, And I did mention you know zero

258
00:16:05,600 --> 00:16:09,640
Speaker 2: point five on those tests that boarded boards need to

259
00:16:10,120 --> 00:16:15,360
Speaker 2: need to consider change management and an organizational readiness is

260
00:16:15,440 --> 00:16:18,720
Speaker 2: absolutely something to consider. And I don't know if you've

261
00:16:18,800 --> 00:16:21,640
Speaker 2: seen this, but over the last four or five weeks

262
00:16:21,640 --> 00:16:25,560
Speaker 2: there have been some significant announcements by large tech companies

263
00:16:25,560 --> 00:16:29,280
Speaker 2: around wholesale removal of people in organizations. You know, upwards

264
00:16:29,280 --> 00:16:32,720
Speaker 2: of twenty thirty thousand people's jobs are being affected. There's

265
00:16:32,760 --> 00:16:34,600
Speaker 2: a little bit of a debate around whether that's AI

266
00:16:34,800 --> 00:16:38,440
Speaker 2: washing where they're using AI as an excuse, but for

267
00:16:38,480 --> 00:16:42,000
Speaker 2: sure the risk is there, and I think boards need

268
00:16:42,040 --> 00:16:45,200
Speaker 2: to need to take that into consideration, and I think

269
00:16:45,280 --> 00:16:49,960
Speaker 2: broader policy discussion around what role does does government need

270
00:16:50,000 --> 00:16:53,920
Speaker 2: to play to ensure that AI does not significantly impact

271
00:16:54,240 --> 00:16:56,760
Speaker 2: the working class because South Africa, you know, that's the

272
00:16:56,800 --> 00:16:59,960
Speaker 2: last thing we need now is for it to have

273
00:17:00,040 --> 00:17:02,800
Speaker 2: a significant impact on youngster's looking for jobs.

274
00:17:02,960 --> 00:17:04,880
Speaker 1: Well, you're looking at the middle class, when you're looking

275
00:17:04,880 --> 00:17:07,120
Speaker 1: at middle management roles that are impacted.

276
00:17:07,320 --> 00:17:10,400
Speaker 2: Absolutely absolutely so, I mean, and there's there's there's no

277
00:17:10,560 --> 00:17:13,879
Speaker 2: debates about it. One of the one of the discussions

278
00:17:13,880 --> 00:17:17,480
Speaker 2: also under ways the extent to which people in organizations

279
00:17:17,560 --> 00:17:20,879
Speaker 2: need to become very familiar with the use of AI

280
00:17:21,680 --> 00:17:24,359
Speaker 2: to be a manager of agents. You know, you can

281
00:17:24,400 --> 00:17:28,360
Speaker 2: imagine a middle manager who used to look after people

282
00:17:28,480 --> 00:17:31,600
Speaker 2: now need to be comfortable in both dealing with people

283
00:17:31,680 --> 00:17:36,200
Speaker 2: and AI agents. So so there is also a position

284
00:17:36,280 --> 00:17:38,680
Speaker 2: that is put forward to say that if you are

285
00:17:38,720 --> 00:17:42,600
Speaker 2: in an environment that is intensively a knowledge work environment,

286
00:17:42,960 --> 00:17:45,920
Speaker 2: you it is incumbent upon you to become familiar with

287
00:17:45,920 --> 00:17:48,399
Speaker 2: with how AI works and you know to the to

288
00:17:48,480 --> 00:17:50,800
Speaker 2: the extent that you're able to manage agents that are

289
00:17:50,800 --> 00:17:53,200
Speaker 2: deployed in the environment, and that is to really to

290
00:17:53,400 --> 00:17:57,679
Speaker 2: secure your let's call it your your future in the

291
00:17:57,680 --> 00:18:02,640
Speaker 2: employed space. I think middle management in organizations are unfortunately

292
00:18:03,000 --> 00:18:03,840
Speaker 2: going to be targeted.

293
00:18:04,480 --> 00:18:06,000
Speaker 1: Let's let's talk about it. And we see that a

294
00:18:06,000 --> 00:18:11,440
Speaker 1: lot on social media companies organizations running really expensive kind

295
00:18:11,480 --> 00:18:16,800
Speaker 1: of pilots. Is that signaling that they are AI ready now?

296
00:18:16,880 --> 00:18:20,600
Speaker 2: Absolutely not so. So It's often often the case where

297
00:18:21,000 --> 00:18:26,680
Speaker 2: where you know boards mistake these pilots for for AI readiness.

298
00:18:26,680 --> 00:18:30,280
Speaker 2: I mean, doing a pilot doesn't necessarily translate into a capability.

299
00:18:31,080 --> 00:18:33,760
Speaker 2: And part of the challenge as well is that because

300
00:18:33,840 --> 00:18:36,800
Speaker 2: these technologies are so new and perhaps in many instances

301
00:18:36,880 --> 00:18:41,320
Speaker 2: quite complicated, boards are uncomfortable in asking the questions around

302
00:18:41,400 --> 00:18:43,080
Speaker 2: you know what, what what is this and how is

303
00:18:43,080 --> 00:18:45,840
Speaker 2: it going to impact my business? And I think it's

304
00:18:45,920 --> 00:18:51,119
Speaker 2: super critical for South African organizations in particular to be

305
00:18:51,240 --> 00:18:54,600
Speaker 2: okay with asking questions run what this tech is and

306
00:18:54,680 --> 00:18:59,080
Speaker 2: to really challenge the executives teams on giving them feedback

307
00:18:59,119 --> 00:19:01,520
Speaker 2: on how this is going to impact their business, because

308
00:19:01,560 --> 00:19:04,240
Speaker 2: if they don't, the risk is that somebody else, some

309
00:19:04,359 --> 00:19:06,200
Speaker 2: other organization is going to take their lunch.

310
00:19:06,440 --> 00:19:09,520
Speaker 1: Yeah, we're talking AI readiness in twenty twenty six, and

311
00:19:09,560 --> 00:19:13,199
Speaker 1: it's it evolved very significantly in the past year and

312
00:19:13,320 --> 00:19:16,520
Speaker 1: in the past months. While some people are still trying

313
00:19:16,520 --> 00:19:20,399
Speaker 1: to understand how AI fits into the big picture, it

314
00:19:20,560 --> 00:19:23,440
Speaker 1: is the big picture. And maybe just I like your

315
00:19:23,480 --> 00:19:26,480
Speaker 1: analogy about the child's plates and the adults plates. I

316
00:19:26,480 --> 00:19:28,600
Speaker 1: don't know if you can help us just to make

317
00:19:28,720 --> 00:19:33,119
Speaker 1: sense of in a very simple way that thing that

318
00:19:33,200 --> 00:19:35,360
Speaker 1: AI is doing at this moment. In time, and we'll

319
00:19:35,400 --> 00:19:38,520
Speaker 1: be doing a little more so in the future.

320
00:19:39,520 --> 00:19:44,320
Speaker 2: Yeah, I'm sure all of us had that moment three

321
00:19:44,560 --> 00:19:47,800
Speaker 2: so that two three years ago when chachipt first began

322
00:19:47,880 --> 00:19:51,399
Speaker 2: a consumer product. You know, that was where that was

323
00:19:51,440 --> 00:19:53,760
Speaker 2: a way were you moment where you understood the power

324
00:19:53,800 --> 00:19:58,800
Speaker 2: of this technology, and you know, organizations now had to

325
00:19:59,160 --> 00:20:02,560
Speaker 2: for the first time catch up in terms of, you know,

326
00:20:02,600 --> 00:20:06,160
Speaker 2: how do we deploy this kind of technology into environments.

327
00:20:05,920 --> 00:20:09,040
Speaker 2: As human beings, we use this on a daily basis.

328
00:20:09,600 --> 00:20:12,960
Speaker 2: Our children use this technology, and I would you know,

329
00:20:13,640 --> 00:20:15,840
Speaker 2: the one thing I like to talk about in terms

330
00:20:15,880 --> 00:20:19,359
Speaker 2: of the promise that this has for our very young

331
00:20:19,680 --> 00:20:23,600
Speaker 2: continents of Africa is that for the first time, our

332
00:20:23,640 --> 00:20:27,480
Speaker 2: youngsters will now have tools available to them to build

333
00:20:27,520 --> 00:20:30,560
Speaker 2: their own futures. And I do think that whilst there

334
00:20:30,680 --> 00:20:33,120
Speaker 2: is some concern around how AI is going to take

335
00:20:33,200 --> 00:20:35,280
Speaker 2: jobs down the line, I do think that the flip

336
00:20:35,320 --> 00:20:38,199
Speaker 2: side of that is going to provide significant opportunities for

337
00:20:38,280 --> 00:20:41,679
Speaker 2: youngsters to embrace this technology to build their own futures.

338
00:20:41,760 --> 00:20:44,400
Speaker 2: So this is not just an adult thing. Youngsters can

339
00:20:44,520 --> 00:20:47,160
Speaker 2: use it and at tack even as we have it today,

340
00:20:47,160 --> 00:20:49,640
Speaker 2: whether it's for free or for the twenty twenty dollars

341
00:20:49,680 --> 00:20:51,520
Speaker 2: a month subscription is extremely powerful.

342
00:20:52,240 --> 00:20:54,600
Speaker 1: Knowledge is becoming a utility.

343
00:20:57,000 --> 00:20:57,760
Speaker 2: What does that exactly?

344
00:20:57,880 --> 00:20:59,440
Speaker 1: What does that mean for the man in the street

345
00:21:00,080 --> 00:21:00,960
Speaker 1: can know anything?

346
00:21:01,760 --> 00:21:04,160
Speaker 2: Yes, And I think that aligns with the point I've

347
00:21:04,200 --> 00:21:06,800
Speaker 2: just made that you know, in late terms, you can

348
00:21:06,840 --> 00:21:10,560
Speaker 2: ask a question in any language and you can get

349
00:21:10,560 --> 00:21:13,560
Speaker 2: a you know, let's call it a PhD equivalent or

350
00:21:13,560 --> 00:21:17,280
Speaker 2: at least a graduate equivalent answer for for knowledge. So

351
00:21:17,640 --> 00:21:21,400
Speaker 2: you know, the days are gone where you know, you'd

352
00:21:21,440 --> 00:21:24,199
Speaker 2: have to lean on an expert that you have to

353
00:21:24,280 --> 00:21:29,160
Speaker 2: pay for services from. Now you just have in your

354
00:21:29,200 --> 00:21:34,320
Speaker 2: pocket a PhD across just about any any knowledge there

355
00:21:34,560 --> 00:21:37,119
Speaker 2: to answer a question for you. It's it's incredible, and

356
00:21:37,160 --> 00:21:38,800
Speaker 2: I mean to the extent you can then go further

357
00:21:38,880 --> 00:21:40,720
Speaker 2: to say, well none of that that I have this

358
00:21:40,800 --> 00:21:44,440
Speaker 2: information help me develop an application to to take to market.

359
00:21:44,520 --> 00:21:45,360
Speaker 2: So very exciting.

360
00:21:45,440 --> 00:21:50,679
Speaker 1: So it informs strategy. Yeah, and it's the way to go.

361
00:21:50,760 --> 00:21:53,720
Speaker 1: It's it's sus I think a whole lot for us

362
00:21:53,800 --> 00:21:56,840
Speaker 1: to still come to understand about it, but yeah, I

363
00:21:56,880 --> 00:21:59,440
Speaker 1: kind of do it quickly because it is also evolving

364
00:21:59,520 --> 00:22:02,199
Speaker 1: very quickly. And I think that's the message from Shane Cooper,

365
00:22:02,440 --> 00:22:05,000
Speaker 1: head of Digital Advisory at fourvis Missas. Thank you very

366
00:22:05,040 --> 00:22:08,320
Speaker 1: much for your time, Shane. It's been so informative, or

367
00:22:08,359 --> 00:22:09,919
Speaker 1: at least I hope it was. For me. It was.

368
00:22:10,480 --> 00:22:12,840
Speaker 1: It's appreciated and we'll lean on you for some more

369
00:22:12,880 --> 00:22:17,000
Speaker 1: related information in the future, Shane. Shane brings the time

370
00:22:17,080 --> 00:22:18,600
Speaker 1: to three minutes to eleven