1
00:00:00,120 --> 00:00:02,840
Speaker 1: Hey everyone, it's Robert and Joe here. Today we've got

2
00:00:02,840 --> 00:00:04,720
Speaker 1: something a little bit different to share with you. It

3
00:00:04,840 --> 00:00:08,000
Speaker 1: is a new season of the Smart Talks with IBM

4
00:00:08,119 --> 00:00:09,119
Speaker 1: podcast series.

5
00:00:09,600 --> 00:00:11,680
Speaker 2: Today we are witnessed to one of those rare moments

6
00:00:11,680 --> 00:00:14,360
Speaker 2: in history, the rise of an innovative technology with the

7
00:00:14,360 --> 00:00:18,680
Speaker 2: potential to radically transform business and society forever. The technology,

8
00:00:18,760 --> 00:00:22,200
Speaker 2: of course, is artificial intelligence, and it's the central focus

9
00:00:22,239 --> 00:00:24,800
Speaker 2: for this new season of Smart Talks with IBM.

10
00:00:25,320 --> 00:00:28,400
Speaker 1: Join hosts from your favorite Pushkin podcasts as they talk

11
00:00:28,480 --> 00:00:31,640
Speaker 1: with industry experts and leaders to explore how businesses can

12
00:00:31,680 --> 00:00:35,360
Speaker 1: integrate AI into their workflows and help drive real change

13
00:00:35,400 --> 00:00:38,160
Speaker 1: in this new era of AI. And of course, host

14
00:00:38,280 --> 00:00:40,440
Speaker 1: Malcolm Gladwell will be there to guide you through the

15
00:00:40,479 --> 00:00:42,640
Speaker 1: season and throw in his two cents as well.

16
00:00:43,120 --> 00:00:46,120
Speaker 2: Look out for new episodes of Smart Talks with IBM

17
00:00:46,400 --> 00:00:49,519
Speaker 2: every other week on the iHeartRadio app, Apple Podcasts, or

18
00:00:49,560 --> 00:00:53,360
Speaker 2: wherever you get your podcasts, and learn more at IBM

19
00:00:53,479 --> 00:00:55,480
Speaker 2: dot com slash smart talks.

20
00:00:57,560 --> 00:01:01,000
Speaker 3: Hello, Hello, Welcome to Smart Talks with IBM, a podcast

21
00:01:01,000 --> 00:01:06,800
Speaker 3: from Pushkin Industries, iHeartRadio and IBM. I'm Malcolm Glabo. This season,

22
00:01:06,959 --> 00:01:11,200
Speaker 3: we're continuing our conversation with new creators visionaries who are

23
00:01:11,240 --> 00:01:15,200
Speaker 3: creatively applying technology in business to drive change, but with

24
00:01:15,240 --> 00:01:19,760
Speaker 3: a focus on the transformative power of artificial intelligence and

25
00:01:19,800 --> 00:01:22,880
Speaker 3: what it means to leverage AI as a game changing

26
00:01:23,040 --> 00:01:27,200
Speaker 3: multiplier for your business. Our guest today is Kareem Yousef,

27
00:01:27,480 --> 00:01:32,039
Speaker 3: Senior Vice President of Product Management and Growth for IBM Software.

28
00:01:32,560 --> 00:01:36,520
Speaker 3: Kareem's focus at IBM is on product strategy, thinking about

29
00:01:36,520 --> 00:01:39,560
Speaker 3: the roadmap for IBM Software products and how they can

30
00:01:39,560 --> 00:01:44,959
Speaker 3: deliver effective and compelling customer experiences with the current boom

31
00:01:45,000 --> 00:01:48,120
Speaker 3: and generative AI. Kareem's job is to help businesses figure

32
00:01:48,160 --> 00:01:52,320
Speaker 3: out how they can apply artificial intelligence at scale to

33
00:01:52,400 --> 00:01:56,360
Speaker 3: help solve big problems and boost productivity at new orders

34
00:01:56,360 --> 00:02:00,520
Speaker 3: of magnitude. In today's episode, you'll hear Kareem explaining how

35
00:02:00,560 --> 00:02:04,760
Speaker 3: AI powered by foundation models can make AI adoption by

36
00:02:04,920 --> 00:02:09,880
Speaker 3: enterprise businesses even easier, how generative AI will change the

37
00:02:09,919 --> 00:02:13,520
Speaker 3: way businesses process data and make decisions, and how these

38
00:02:13,560 --> 00:02:19,519
Speaker 3: considerations influenced the design of Watson x, IBM's next generation

39
00:02:20,040 --> 00:02:24,800
Speaker 3: AI and data platform. Kareem spoke with Jacob Goldstein, host

40
00:02:24,800 --> 00:02:28,880
Speaker 3: of the Pushkin podcast What's Your problem. A veteran business journalist,

41
00:02:29,080 --> 00:02:32,600
Speaker 3: Jacob has reported for The Wall Street Journal, the Miami Herald,

42
00:02:32,960 --> 00:02:38,000
Speaker 3: and was a longtime host of the NPR program Planet Money. Okay,

43
00:02:38,480 --> 00:02:39,720
Speaker 3: let's get to the interview.

44
00:02:41,280 --> 00:02:44,400
Speaker 4: I'm Jacob Goldstein. I'm one of the hosts at Pushkin

45
00:02:44,560 --> 00:02:47,440
Speaker 4: and a correspondent on this show, and I'm delighted to

46
00:02:47,440 --> 00:02:48,960
Speaker 4: have you here. Can you introduce yourself?

47
00:02:49,480 --> 00:02:49,639
Speaker 5: Ah?

48
00:02:49,760 --> 00:02:51,120
Speaker 6: Hi, I'm Kareem Yusuf.

49
00:02:51,320 --> 00:02:54,200
Speaker 5: I'm the senior vice president of Product Management and Growth

50
00:02:54,240 --> 00:02:56,600
Speaker 5: for IBM Software. You can think of me as the

51
00:02:56,680 --> 00:02:58,800
Speaker 5: chief product officer for IBM Software.

52
00:02:59,160 --> 00:03:03,880
Speaker 4: Okay, sounds like a big job. We're here today to

53
00:03:03,919 --> 00:03:09,840
Speaker 4: talk about AI. We've heard really an extraordinary amount in

54
00:03:09,880 --> 00:03:13,840
Speaker 4: the last few months about chat GPT and you know,

55
00:03:13,919 --> 00:03:16,480
Speaker 4: particularly in how it's used in the very kind of

56
00:03:16,480 --> 00:03:19,760
Speaker 4: consumer facing way. But I'm curious what is the rise

57
00:03:19,800 --> 00:03:22,600
Speaker 4: of chat GPT and you know, AI more generally, what

58
00:03:22,639 --> 00:03:24,000
Speaker 4: does it mean for business?

59
00:03:24,560 --> 00:03:26,920
Speaker 5: Well, you know, it's if you kind of step back

60
00:03:26,960 --> 00:03:31,600
Speaker 5: and think about what really happens. You know, in a business,

61
00:03:32,080 --> 00:03:35,000
Speaker 5: you're really talking about a set of processes, right, you know,

62
00:03:35,080 --> 00:03:38,960
Speaker 5: activities that represent what a business needs to get done,

63
00:03:39,120 --> 00:03:42,520
Speaker 5: whether it's product, they produce and then sell or service

64
00:03:42,600 --> 00:03:46,480
Speaker 5: that they provide. And inherent to operating the business, I

65
00:03:46,480 --> 00:03:50,280
Speaker 5: would say, are two very key factors. Data and then

66
00:03:50,320 --> 00:03:53,680
Speaker 5: the decisions you make around that data, and then actually

67
00:03:53,760 --> 00:03:57,840
Speaker 5: lastly the processes or activities you do in accordance with

68
00:03:57,920 --> 00:04:01,440
Speaker 5: that decision. So if if you then think about AI

69
00:04:01,760 --> 00:04:04,520
Speaker 5: as applied to business right in that context, well, the

70
00:04:04,560 --> 00:04:07,360
Speaker 5: first place it often starts is how do you make

71
00:04:07,400 --> 00:04:11,520
Speaker 5: sense of a lot of the data associated with driving

72
00:04:11,560 --> 00:04:14,320
Speaker 5: the business? And so AI has always been, in my

73
00:04:14,440 --> 00:04:20,000
Speaker 5: mind at its foremost about gaining insights, then leading to

74
00:04:20,080 --> 00:04:25,240
Speaker 5: supporting decisions, and ultimately ending at helping to automate the

75
00:04:25,279 --> 00:04:29,720
Speaker 5: activities that then are executed as a result of those decisions.

76
00:04:29,920 --> 00:04:32,320
Speaker 5: So that's kind of my simple way of thinking of AI,

77
00:04:32,680 --> 00:04:35,400
Speaker 5: and we can obviously coloring with examples, but that's my

78
00:04:35,480 --> 00:04:37,839
Speaker 5: simplest way of thinking about AI. When you think about

79
00:04:37,920 --> 00:04:41,240
Speaker 5: in the business context, gain insights from masses of data

80
00:04:41,560 --> 00:04:44,880
Speaker 5: to support decisions and then drive actions.

81
00:04:44,760 --> 00:04:47,360
Speaker 4: That's a really helpful framework. And then if we think

82
00:04:47,360 --> 00:04:49,800
Speaker 4: about sort of what's happening in the world now with

83
00:04:50,440 --> 00:04:54,039
Speaker 4: you know, enterprise businesses NAI, what are you seeing with

84
00:04:54,440 --> 00:04:56,000
Speaker 4: enterprise adoption of AI?

85
00:04:56,440 --> 00:05:00,440
Speaker 5: At this moment, so we're really talking about most a

86
00:05:00,600 --> 00:05:02,839
Speaker 5: tale of two periods. So let me first of all

87
00:05:03,080 --> 00:05:06,800
Speaker 5: kind of take you back before the advent of what

88
00:05:06,839 --> 00:05:09,640
Speaker 5: I would call generative AI and the whole chat gpt

89
00:05:10,680 --> 00:05:13,039
Speaker 5: to what has been going on in what I would

90
00:05:13,120 --> 00:05:17,440
Speaker 5: term the realm of more standardized machine learning models. A

91
00:05:17,480 --> 00:05:19,720
Speaker 5: lot of what has been going on has been very

92
00:05:19,800 --> 00:05:23,320
Speaker 5: much in the realms of certain things like anomaly detection

93
00:05:23,839 --> 00:05:27,800
Speaker 5: or optimization, right, using machine learning models to do that

94
00:05:27,880 --> 00:05:30,520
Speaker 5: kind of work, and where might it apply well, think

95
00:05:30,560 --> 00:05:36,440
Speaker 5: of anomaly detection in security software right detecting threats based

96
00:05:36,480 --> 00:05:41,640
Speaker 5: upon different events flowing through or in enterprise asset management

97
00:05:41,720 --> 00:05:47,239
Speaker 5: software monitoring equipment and detecting anomalies within their behavior, or

98
00:05:47,279 --> 00:05:52,200
Speaker 5: even in IT automation software once again detecting anomalies based

99
00:05:52,240 --> 00:05:55,120
Speaker 5: upon what's going on with various IT events, and then

100
00:05:55,240 --> 00:06:00,000
Speaker 5: tasks that should occur. Optimizations often play around in the realm,

101
00:06:00,240 --> 00:06:04,320
Speaker 5: as you might imagine to solve problems of resource optimization,

102
00:06:04,400 --> 00:06:07,200
Speaker 5: whether you think of that in the context of application

103
00:06:07,360 --> 00:06:11,039
Speaker 5: resource management for IT or in the context of supply chain.

104
00:06:11,440 --> 00:06:16,040
Speaker 5: These have been very classical applications of machine learning AI

105
00:06:16,520 --> 00:06:19,160
Speaker 5: to really make sense of the data and provide a

106
00:06:19,279 --> 00:06:24,880
Speaker 5: basis to drive decisions. Now, what is characterized by all

107
00:06:24,920 --> 00:06:27,960
Speaker 5: those examples of given and the state of the art

108
00:06:28,040 --> 00:06:31,680
Speaker 5: of that kind of technology has always been it's very

109
00:06:31,920 --> 00:06:37,320
Speaker 5: task specific. So there was a air quotes, if I may,

110
00:06:37,800 --> 00:06:41,880
Speaker 5: kind of limitation in the sense that the tar it

111
00:06:41,960 --> 00:06:45,000
Speaker 5: had to be very task specific. And so we've seen

112
00:06:45,040 --> 00:06:48,039
Speaker 5: a lot of broad based adoption within the enterprise, right,

113
00:06:48,080 --> 00:06:52,280
Speaker 5: but it's very very task specific, as you might imagine. Now,

114
00:06:52,480 --> 00:06:55,880
Speaker 5: what has happened recently and has been brought to the

115
00:06:56,080 --> 00:07:00,760
Speaker 5: four has been this discussion of generative AI, which is

116
00:07:00,880 --> 00:07:05,560
Speaker 5: powered by a very specific innovation, this notion of foundation models.

117
00:07:06,240 --> 00:07:09,480
Speaker 5: And in the simplest way to think about it, it's

118
00:07:09,560 --> 00:07:14,760
Speaker 5: about training this large model that can then be refined

119
00:07:16,240 --> 00:07:21,840
Speaker 5: to various tasks. And the easiest one that everybody recognized

120
00:07:21,880 --> 00:07:24,720
Speaker 5: at the moment is the notion of a large language model,

121
00:07:25,080 --> 00:07:28,240
Speaker 5: a model that has an understanding of a lot of

122
00:07:28,280 --> 00:07:31,440
Speaker 5: the elements of a language such that it can be

123
00:07:31,560 --> 00:07:35,600
Speaker 5: refined to a variety of tasks. Write an essay, answer

124
00:07:35,640 --> 00:07:40,080
Speaker 5: a question, singer songs, so on, answers so forth. I

125
00:07:40,240 --> 00:07:44,080
Speaker 5: like to liken the power if you like, and this

126
00:07:44,120 --> 00:07:47,000
Speaker 5: will speak to the why everybody is so excited about it?

127
00:07:47,040 --> 00:07:49,160
Speaker 6: Why would argue at an inflection point?

128
00:07:49,480 --> 00:07:53,160
Speaker 5: I like to liken it to teaching a child the alphabet.

129
00:07:54,280 --> 00:07:57,239
Speaker 5: When you teach a child an alphabet, it's a set

130
00:07:57,280 --> 00:08:01,280
Speaker 5: of letters, right, Let's call that our foundation model. But

131
00:08:01,520 --> 00:08:06,280
Speaker 5: over time that knowledge of the alphabet is tuned to

132
00:08:06,320 --> 00:08:08,920
Speaker 5: read a book, write an essay, do a composition, create

133
00:08:08,960 --> 00:08:12,160
Speaker 5: a song, write a poem, write an invoice. You understand

134
00:08:12,160 --> 00:08:16,120
Speaker 5: what I mean, right, And so from one foundation model

135
00:08:16,560 --> 00:08:21,840
Speaker 5: you can support multiple targeted tasks as opposed sticking with

136
00:08:21,880 --> 00:08:26,520
Speaker 5: the analogy to having a model for reading, writing, thinking

137
00:08:26,520 --> 00:08:29,080
Speaker 5: of doing a poem, doing an essay, so on and

138
00:08:29,120 --> 00:08:32,959
Speaker 5: so forth, and so in the enterprise context, that means

139
00:08:33,000 --> 00:08:36,240
Speaker 5: that we're now talking about being able to unlock even

140
00:08:36,240 --> 00:08:41,600
Speaker 5: additional value at scale because of the nation of nature

141
00:08:41,679 --> 00:08:47,000
Speaker 5: foundation models and their appeal to generative use cases. Generative

142
00:08:47,040 --> 00:08:49,679
Speaker 5: in this case meaning creation of new content.

143
00:08:50,440 --> 00:08:55,040
Speaker 4: So let's talk about Watson x. IBM recently announced what's

144
00:08:55,080 --> 00:08:57,120
Speaker 4: an X Just first of all, what is that? What

145
00:08:57,200 --> 00:08:57,880
Speaker 4: is what's an X?

146
00:08:58,720 --> 00:09:02,439
Speaker 5: Well, what's an extra to our is our brand for

147
00:09:02,559 --> 00:09:07,000
Speaker 5: our platform, the WHATSAP platform for really taking advantage of

148
00:09:07,240 --> 00:09:11,800
Speaker 5: generative AI within the enterprise, within business, and so when

149
00:09:11,840 --> 00:09:14,800
Speaker 5: you begin to think about what does that mean while,

150
00:09:14,840 --> 00:09:17,160
Speaker 5: it leads you to the components of what's next and

151
00:09:17,520 --> 00:09:19,439
Speaker 5: to a set of use cases. So let me paint

152
00:09:19,440 --> 00:09:23,920
Speaker 5: a few quick pictures for you here. What's anex first

153
00:09:23,920 --> 00:09:28,680
Speaker 5: of all, is about enabling our customers to manipulate models

154
00:09:29,000 --> 00:09:33,240
Speaker 5: against their task, manipulate these foundation models against their task.

155
00:09:33,760 --> 00:09:37,720
Speaker 5: Our belief is that the world is a multi model world,

156
00:09:38,320 --> 00:09:40,720
Speaker 5: right and especially when you think about it in the

157
00:09:40,760 --> 00:09:44,679
Speaker 5: context of business. Models are going to come from various sources,

158
00:09:45,000 --> 00:09:48,000
Speaker 5: the ones we supply, the ones out there in open source,

159
00:09:48,040 --> 00:09:50,520
Speaker 5: and so of you. But there are activities you need

160
00:09:50,559 --> 00:09:54,120
Speaker 5: to do around these models to as I said, apply

161
00:09:54,280 --> 00:09:55,440
Speaker 5: them to your use case.

162
00:09:55,559 --> 00:09:57,400
Speaker 6: And we'll talk about use cases in a bit. So

163
00:09:57,920 --> 00:09:58,439
Speaker 6: what's next.

164
00:09:58,480 --> 00:10:01,880
Speaker 5: The AI is that vironment that build a tool if

165
00:10:01,920 --> 00:10:04,680
Speaker 5: you like, for being able to do those manipulation of

166
00:10:04,800 --> 00:10:08,840
Speaker 5: models to meet your specific use case. Thinks that people

167
00:10:08,920 --> 00:10:13,040
Speaker 5: will recognize in the field prompt engineering, prompt tuning, fine tuning,

168
00:10:13,480 --> 00:10:16,480
Speaker 5: those kind of activities which are all the manipulation of

169
00:10:16,559 --> 00:10:20,040
Speaker 5: models to meet your use case. Yeah, the second component

170
00:10:20,480 --> 00:10:24,040
Speaker 5: is dot data. So what's the next dot data is Essentially,

171
00:10:24,559 --> 00:10:27,800
Speaker 5: a next generation data store is based upon something referred

172
00:10:27,800 --> 00:10:31,760
Speaker 5: to as an open data lakehouse architecture that helps to

173
00:10:31,920 --> 00:10:36,000
Speaker 5: bring together the data that's needed to actually do the AI.

174
00:10:36,200 --> 00:10:38,960
Speaker 5: In this case, when you think about manipulating a model,

175
00:10:39,160 --> 00:10:42,600
Speaker 5: a foundation model, you're generally using some data to prompt it,

176
00:10:42,800 --> 00:10:45,480
Speaker 5: tune it, to train it to your use cases. And

177
00:10:45,559 --> 00:10:49,400
Speaker 5: so we provide a very open data store that allows

178
00:10:49,440 --> 00:10:51,680
Speaker 5: all manner of data and formats to be brought through

179
00:10:51,720 --> 00:10:55,800
Speaker 5: to do that. And the third component is what's next

180
00:10:55,880 --> 00:10:59,080
Speaker 5: dot governance, And this is all about the framework and

181
00:10:59,120 --> 00:11:04,200
Speaker 5: the toolkit rep required to apply the right governance principles

182
00:11:04,240 --> 00:11:07,520
Speaker 5: across doing this kind of work, because when you're deploying

183
00:11:08,080 --> 00:11:12,560
Speaker 5: AI within the enterprise, governance is actually important, right, It's

184
00:11:12,640 --> 00:11:15,920
Speaker 5: critical to understand why is your data coming from, what

185
00:11:16,080 --> 00:11:19,000
Speaker 5: data did you add in, How is your model performing?

186
00:11:19,440 --> 00:11:21,720
Speaker 5: Are you able to keep an appropriate audit trail of

187
00:11:21,720 --> 00:11:26,160
Speaker 5: your activities for your own internal policy and compliance needs

188
00:11:26,280 --> 00:11:27,760
Speaker 5: or for regulatory needs as well.

189
00:11:28,360 --> 00:11:32,560
Speaker 4: So this platform, the system that you're describing, I'm curious,

190
00:11:33,120 --> 00:11:36,439
Speaker 4: how is it different from the you know, the generitive

191
00:11:36,480 --> 00:11:38,920
Speaker 4: AI options that you know we've all been hearing about

192
00:11:38,960 --> 00:11:40,200
Speaker 4: sort of in the press.

193
00:11:40,520 --> 00:11:43,840
Speaker 5: Well, I think it really comes down to the ethos

194
00:11:43,960 --> 00:11:47,040
Speaker 5: or the principles that first of all drive the work

195
00:11:47,040 --> 00:11:50,160
Speaker 5: that we're doing. The first I would fixate on is

196
00:11:50,400 --> 00:11:51,240
Speaker 5: being open.

197
00:11:52,200 --> 00:11:52,400
Speaker 6: Right.

198
00:11:52,920 --> 00:11:55,400
Speaker 5: We fundamentally believe that to do this kind of work

199
00:11:55,440 --> 00:11:58,720
Speaker 5: within the enterprise, you need an open platform that, as

200
00:11:58,720 --> 00:12:02,000
Speaker 5: I said, is open to all manner of models from

201
00:12:02,040 --> 00:12:04,520
Speaker 5: all sources. It's one of the reasons why we announced

202
00:12:04,520 --> 00:12:07,760
Speaker 5: our partnership with hugging Face to make sure that our

203
00:12:07,840 --> 00:12:12,800
Speaker 5: clients can gain access to open source innovation within the

204
00:12:12,880 --> 00:12:14,160
Speaker 5: platform to do their work.

205
00:12:14,640 --> 00:12:16,320
Speaker 4: And hugging Face, to be clear, is sort of the

206
00:12:17,160 --> 00:12:19,120
Speaker 4: open source AI kind of hub.

207
00:12:19,640 --> 00:12:20,680
Speaker 6: That's right, that's correct.

208
00:12:20,720 --> 00:12:25,080
Speaker 5: Yes, it's a marketplace hub for all kind of ecosystem

209
00:12:25,160 --> 00:12:28,560
Speaker 5: coordinator for open source models. And I believe there's a

210
00:12:28,559 --> 00:12:31,319
Speaker 5: lot of innovation going on out there. So first of all,

211
00:12:31,880 --> 00:12:37,840
Speaker 5: open becomes important. The second targeted, So our focus is

212
00:12:38,040 --> 00:12:43,160
Speaker 5: very much on enabling these business use cases, right, And

213
00:12:43,200 --> 00:12:45,000
Speaker 5: you might say what kind of use cases are we

214
00:12:45,080 --> 00:12:47,360
Speaker 5: talking about? I give you three very quick ones that

215
00:12:47,960 --> 00:12:50,880
Speaker 5: with our customers are focused on a lot of focus

216
00:12:50,960 --> 00:12:55,600
Speaker 5: around enhancing customer service use cases. Think of this as

217
00:12:55,720 --> 00:13:00,560
Speaker 5: chatbots or digital assistance that are further trained in more

218
00:13:00,600 --> 00:13:03,880
Speaker 5: and more information about what the company has to offer

219
00:13:04,120 --> 00:13:06,760
Speaker 5: or could be internal policies, external policies, and so on

220
00:13:06,760 --> 00:13:09,560
Speaker 5: and so forth. This means a platform that makes it

221
00:13:09,640 --> 00:13:14,240
Speaker 5: really easy to bring your own data to train and

222
00:13:14,360 --> 00:13:19,199
Speaker 5: tune the model, while protecting your own data as extremely

223
00:13:19,240 --> 00:13:24,280
Speaker 5: important for the enterprise right. Another important use case seeing

224
00:13:24,280 --> 00:13:26,360
Speaker 5: a lot of focused on what i'd call AI based

225
00:13:26,480 --> 00:13:30,920
Speaker 5: orchestration or automation of task whereby think about like an

226
00:13:31,160 --> 00:13:35,440
Speaker 5: HR professional as an example, going through a job requisition

227
00:13:35,880 --> 00:13:39,160
Speaker 5: is able to interact with multiple systems via a very

228
00:13:39,160 --> 00:13:44,680
Speaker 5: simple chat interface and have work dynamically sequenced to support

229
00:13:44,720 --> 00:13:48,720
Speaker 5: them in doing their tasks. That once again requires a

230
00:13:48,840 --> 00:13:52,559
Speaker 5: notion of working with models and technology in a way

231
00:13:52,600 --> 00:13:55,360
Speaker 5: that in many ways can be unique to how a

232
00:13:55,440 --> 00:13:58,319
Speaker 5: business wishes to work and indeed, in various cases can

233
00:13:58,360 --> 00:14:01,640
Speaker 5: embody what they can do, their secret source or their

234
00:14:01,640 --> 00:14:05,680
Speaker 5: differentiated advantage. So once again, a platform that recognizes that

235
00:14:05,760 --> 00:14:08,640
Speaker 5: and designed for business that's not the same scope or

236
00:14:08,679 --> 00:14:13,440
Speaker 5: frame of reference for a consumer platform. And then you know,

237
00:14:13,480 --> 00:14:16,319
Speaker 5: we're also seeing a lot of work around cod generation,

238
00:14:16,480 --> 00:14:20,920
Speaker 5: application modernization, you know, and people enhancing their skills, so

239
00:14:21,080 --> 00:14:22,960
Speaker 5: targeted becomes really important.

240
00:14:23,160 --> 00:14:25,080
Speaker 6: Mentioned open, and I mentioned.

241
00:14:24,880 --> 00:14:28,240
Speaker 5: Targeted, targeted to the business to the use cases that

242
00:14:28,280 --> 00:14:32,800
Speaker 5: they need to do underpinning that, then it's trusted. So

243
00:14:32,880 --> 00:14:35,960
Speaker 5: everything I gave you in those targeted use cases talk

244
00:14:36,040 --> 00:14:43,160
Speaker 5: about handling enterprise proprietary and specific data. We are trusted

245
00:14:43,200 --> 00:14:45,680
Speaker 5: in this regard right. We have been serving the business

246
00:14:46,120 --> 00:14:49,600
Speaker 5: for many, many a year, and we are designing our

247
00:14:49,640 --> 00:14:52,760
Speaker 5: platform and even our principles and way of operating to

248
00:14:53,000 --> 00:14:56,160
Speaker 5: recognize and enable that. Both in terms of the work

249
00:14:56,160 --> 00:14:59,680
Speaker 5: we do around the governance framework and transparency that you're

250
00:14:59,720 --> 00:15:02,160
Speaker 5: able to to gain and apply, but even in the

251
00:15:02,200 --> 00:15:06,280
Speaker 5: way we allow our platform to be deployed in multiple

252
00:15:07,160 --> 00:15:10,640
Speaker 5: locations of footprints, consumed as a service on a hyperscaler,

253
00:15:11,000 --> 00:15:14,560
Speaker 5: running your own private footprint on prem or your cloud footprint.

254
00:15:14,880 --> 00:15:17,120
Speaker 5: All of these need to be brought together to meet

255
00:15:17,160 --> 00:15:22,160
Speaker 5: the needs of an actual enterprise business. My last comment

256
00:15:22,240 --> 00:15:26,200
Speaker 5: is where I think we're fundamentally differentiated is we're really

257
00:15:26,320 --> 00:15:33,520
Speaker 5: about empowering our customers to take advantage of AI to

258
00:15:33,640 --> 00:15:39,160
Speaker 5: unleash the intelligence capabilities productivity of their own business. This

259
00:15:39,200 --> 00:15:43,040
Speaker 5: isn't about, oh, we've established a bunch of APIs that

260
00:15:43,080 --> 00:15:46,800
Speaker 5: you can ask questions. This is about how do you

261
00:15:47,360 --> 00:15:52,320
Speaker 5: craft what you need for your business to deliver differentiated

262
00:15:52,440 --> 00:15:58,840
Speaker 5: value to your customers, shareholders, employees with all the appropriate

263
00:15:58,880 --> 00:16:01,240
Speaker 5: protections as well well. And so there's a lot of

264
00:16:01,280 --> 00:16:03,280
Speaker 5: focus on what we've done with the platform and the

265
00:16:03,320 --> 00:16:05,920
Speaker 5: tool set to enable that, to enable what we like

266
00:16:05,960 --> 00:16:11,560
Speaker 5: to call AI value creators, not just consumers of AI value.

267
00:16:12,320 --> 00:16:18,320
Speaker 4: When you were talking about basically enterprise adoption of AI,

268
00:16:19,160 --> 00:16:22,720
Speaker 4: you use the word trust, and I'm curious, you know,

269
00:16:23,200 --> 00:16:27,600
Speaker 4: what does trust mean in the context of AI and

270
00:16:27,640 --> 00:16:28,240
Speaker 4: the enterprise.

271
00:16:29,640 --> 00:16:34,960
Speaker 5: I would kind of deconstruct trust along these k avenues.

272
00:16:37,080 --> 00:16:41,360
Speaker 5: If AI is about giving you insights to help you

273
00:16:41,400 --> 00:16:47,680
Speaker 5: support decisions, how do you trust what insights it's provided?

274
00:16:48,520 --> 00:16:53,600
Speaker 5: What data did it use? What did it consider based

275
00:16:53,680 --> 00:16:59,920
Speaker 5: upon that data that therefore led to the insight provided?

276
00:17:02,280 --> 00:17:06,120
Speaker 5: Why is this important? Why this notion of trust? Well, one,

277
00:17:06,560 --> 00:17:09,359
Speaker 5: you're about to make a decision, so you want to

278
00:17:09,440 --> 00:17:14,200
Speaker 5: understand the basis for a decision. It's no different than

279
00:17:14,280 --> 00:17:17,359
Speaker 5: me asking you something and then saying, okay, can you

280
00:17:17,400 --> 00:17:19,720
Speaker 5: explain you're working? Right, that would be a notion of

281
00:17:19,760 --> 00:17:23,920
Speaker 5: trust that we establish and a very natural interaction as humans, Right,

282
00:17:23,920 --> 00:17:27,240
Speaker 5: we do it all the time, right, So there is

283
00:17:27,280 --> 00:17:30,840
Speaker 5: that element. The other reason why it becomes important if

284
00:17:30,880 --> 00:17:35,520
Speaker 5: you're applying AI into business processes and therefore how your

285
00:17:35,560 --> 00:17:40,080
Speaker 5: business works, you want to make sure that you know

286
00:17:40,160 --> 00:17:44,439
Speaker 5: what biases are built in to any decision or not

287
00:17:44,880 --> 00:17:48,720
Speaker 5: or if the AI the model in effect is drifting

288
00:17:49,800 --> 00:17:53,680
Speaker 5: away from kind of the parameters that you would want

289
00:17:53,720 --> 00:17:58,159
Speaker 5: it to remain within, right or go trust and so

290
00:17:59,280 --> 00:18:03,040
Speaker 5: in many a ways, that's one big aspect of trusting

291
00:18:03,080 --> 00:18:06,600
Speaker 5: the technology because you're applying it into decisions you need

292
00:18:06,640 --> 00:18:08,720
Speaker 5: to make every day, and you need to know in

293
00:18:08,880 --> 00:18:12,679
Speaker 5: very simple terms how it works and how it is working.

294
00:18:14,119 --> 00:18:17,600
Speaker 5: The element of trust that I think is important in

295
00:18:17,600 --> 00:18:23,920
Speaker 5: this discussion. Who are you getting your AI from? That's

296
00:18:24,160 --> 00:18:27,680
Speaker 5: very important to us as a company here at IBM. Right,

297
00:18:28,280 --> 00:18:34,159
Speaker 5: given we serve business, that trust becomes extremely important. And

298
00:18:34,200 --> 00:18:35,919
Speaker 5: what are the elements of that trust? What are the

299
00:18:35,960 --> 00:18:37,720
Speaker 5: customers trying to understand?

300
00:18:38,240 --> 00:18:38,560
Speaker 6: Well?

301
00:18:38,960 --> 00:18:43,080
Speaker 5: First and foremost, what's your ethos around AI? We're very

302
00:18:43,160 --> 00:18:46,440
Speaker 5: clear on the customer's data is their data. When they

303
00:18:46,600 --> 00:18:49,879
Speaker 5: tune or refine those models to meet their use cases,

304
00:18:49,920 --> 00:18:53,080
Speaker 5: that is all theirs, and we actually provide the ability

305
00:18:53,119 --> 00:18:56,080
Speaker 5: for them to do that in very isolated and protected

306
00:18:56,119 --> 00:19:00,560
Speaker 5: ways as they choose, and we never use their data

307
00:19:00,680 --> 00:19:06,080
Speaker 5: without explicit opting and permissions. Right customers might say, oh yeah,

308
00:19:06,320 --> 00:19:08,359
Speaker 5: use this so that you can make a generally overall

309
00:19:08,400 --> 00:19:13,560
Speaker 5: better model. But it's full awareness, full transparency that is important.

310
00:19:13,680 --> 00:19:16,720
Speaker 5: That's a trust of who you're doing business with. So

311
00:19:16,880 --> 00:19:19,760
Speaker 5: that's how I think about trust. How do you build

312
00:19:19,800 --> 00:19:26,360
Speaker 5: systems you trust? And are you working with people you trust?

313
00:19:27,119 --> 00:19:29,480
Speaker 3: I find Kareem's point about trust when it comes to

314
00:19:29,560 --> 00:19:32,960
Speaker 3: data to be so important because as powerful as AI

315
00:19:33,040 --> 00:19:37,080
Speaker 3: tools can be, their helpfulness is dependent on how trustworthy

316
00:19:37,119 --> 00:19:40,960
Speaker 3: the data is. Humans will have to decide if our data,

317
00:19:41,040 --> 00:19:44,439
Speaker 3: our decision making, and our AI insights live up to

318
00:19:44,480 --> 00:19:47,560
Speaker 3: the vision we hope to achieve in business. As Green

319
00:19:47,640 --> 00:19:51,200
Speaker 3: and Jacob continue the conversation, Jacob asks some more practical

320
00:19:51,280 --> 00:19:56,640
Speaker 3: questions about how businesses can adopt AI into their own processes.

321
00:19:57,280 --> 00:20:02,920
Speaker 4: Let's listen, how can businesses move toward integrating AI as

322
00:20:02,960 --> 00:20:06,320
Speaker 4: part of their core business model instead of, you know,

323
00:20:06,359 --> 00:20:08,320
Speaker 4: sort of as an add on on the periphery.

324
00:20:09,119 --> 00:20:10,320
Speaker 6: It's funny, you know.

325
00:20:10,400 --> 00:20:13,480
Speaker 5: My simple answer to that is it's actually the simplest

326
00:20:13,480 --> 00:20:16,920
Speaker 5: thing in the world. To do by thinking about your business,

327
00:20:18,280 --> 00:20:24,560
Speaker 5: thinking about your elements of differentiation, and then thinking about

328
00:20:25,520 --> 00:20:30,720
Speaker 5: how AI can help you extend expand those Right, what

329
00:20:31,040 --> 00:20:32,960
Speaker 5: do you want to be known for? I picked a

330
00:20:33,160 --> 00:20:36,920
Speaker 5: very simple use case of customer service interaction. Almost every

331
00:20:36,920 --> 00:20:39,480
Speaker 5: business needs to do that and wants to do it better,

332
00:20:39,880 --> 00:20:41,680
Speaker 5: and so it becomes a way to start. But then

333
00:20:41,680 --> 00:20:43,800
Speaker 5: as you begin to work your way through, you think

334
00:20:43,800 --> 00:20:47,959
Speaker 5: about various automation of business processes. You think about decisions

335
00:20:48,000 --> 00:20:50,760
Speaker 5: that need to be made right, or how can individuals

336
00:20:50,800 --> 00:20:53,520
Speaker 5: be helped? How can they be made more productive? I

337
00:20:53,560 --> 00:20:56,959
Speaker 5: think always becomes a very important one. Right, So, and

338
00:20:57,040 --> 00:20:59,959
Speaker 5: you can apply this in many context a financial analyst

339
00:21:00,160 --> 00:21:03,359
Speaker 5: looking at reams of data and trying to derive insights.

340
00:21:03,880 --> 00:21:06,560
Speaker 5: How do you leverage AI to make that financial analyst

341
00:21:06,840 --> 00:21:10,080
Speaker 5: even more powerful? And so that's how I advise you, know, people,

342
00:21:10,080 --> 00:21:12,640
Speaker 5: to always look at it. Think about your task, think

343
00:21:12,640 --> 00:21:16,280
Speaker 5: about your business processes, think about where help is needed

344
00:21:16,359 --> 00:21:19,400
Speaker 5: or where new value could be unlocked, and then you're

345
00:21:19,400 --> 00:21:22,520
Speaker 5: applying AI as a tool to achieve that end.

346
00:21:23,440 --> 00:21:26,800
Speaker 4: One of the themes we return to on this show a

347
00:21:26,840 --> 00:21:32,680
Speaker 4: lot is creativity and the relationship between technology and creativity

348
00:21:33,520 --> 00:21:38,440
Speaker 4: and I'm curious how you think that AI can help

349
00:21:38,480 --> 00:21:40,280
Speaker 4: people be more creative at work.

350
00:21:42,160 --> 00:21:45,040
Speaker 5: I think AI can help people be more creative at

351
00:21:45,080 --> 00:21:49,080
Speaker 5: work by automating the mundane to unlock your mind to

352
00:21:49,080 --> 00:21:52,119
Speaker 5: be able to focus on higher value. You know, I've

353
00:21:52,280 --> 00:21:55,840
Speaker 5: used a couple of times I've talked about deriving insights

354
00:21:55,880 --> 00:21:59,520
Speaker 5: from data right to drive informed decisions.

355
00:22:00,440 --> 00:22:01,240
Speaker 6: If you can.

356
00:22:01,240 --> 00:22:05,600
Speaker 5: Use AI to gather a lot more insights into one

357
00:22:05,680 --> 00:22:08,640
Speaker 5: place than you could typically do yourself, or more manually

358
00:22:08,640 --> 00:22:11,440
Speaker 5: you'd have to like write it down, look at six spreadsheets,

359
00:22:11,480 --> 00:22:14,560
Speaker 5: copy from here to there, then you actually have more

360
00:22:14,640 --> 00:22:18,600
Speaker 5: time to look at that data, digest those insights, and

361
00:22:18,720 --> 00:22:21,160
Speaker 5: think about what do I need to do with these

362
00:22:21,160 --> 00:22:23,960
Speaker 5: as a business, which direction do I want to go?

363
00:22:24,480 --> 00:22:27,560
Speaker 5: I think of its freeing us up to do more

364
00:22:27,640 --> 00:22:30,800
Speaker 5: of what we actually as humans do extremely well.

365
00:22:30,760 --> 00:22:32,439
Speaker 6: Which is actually that creative thinking.

366
00:22:33,240 --> 00:22:36,840
Speaker 5: Exactly simple terms, why do we use a calculator to

367
00:22:36,920 --> 00:22:41,639
Speaker 5: do arithmetic? It's not that we cannot necessarily knock it

368
00:22:41,720 --> 00:22:44,400
Speaker 5: out ourselves. But if you're trying to balance your checkbook,

369
00:22:44,560 --> 00:22:46,720
Speaker 5: to use an old phrase or dare I.

370
00:22:46,680 --> 00:22:52,919
Speaker 4: Say, what's a check But so let us modernize that.

371
00:22:54,119 --> 00:22:58,080
Speaker 5: If you're trying to check your expenses for the month

372
00:22:58,359 --> 00:23:03,040
Speaker 5: and your performance against budget. Yes, you could print out

373
00:23:03,080 --> 00:23:08,160
Speaker 5: all your statements, circle everything and add it all up,

374
00:23:09,119 --> 00:23:13,359
Speaker 5: or you could begin to use technology to improve that

375
00:23:13,480 --> 00:23:15,719
Speaker 5: experience so you can get more time to think about

376
00:23:16,160 --> 00:23:19,000
Speaker 5: what really am I learning from my spending patterns and

377
00:23:19,040 --> 00:23:20,800
Speaker 5: what do I want to do about it. It's a

378
00:23:20,920 --> 00:23:24,560
Speaker 5: very simple personal example, but I think it's fundamentally what

379
00:23:24,560 --> 00:23:28,200
Speaker 5: we're talking about here, and that's always been in my mind,

380
00:23:28,240 --> 00:23:32,760
Speaker 5: the promise of technology freeing us up to actually apply

381
00:23:32,840 --> 00:23:37,360
Speaker 5: ourselves to higher value thought and higher value problems.

382
00:23:37,920 --> 00:23:42,240
Speaker 4: So we've been talking basically about the present so far,

383
00:23:42,320 --> 00:23:44,960
Speaker 4: and I'm curious if you think about the future and

384
00:23:45,000 --> 00:23:48,520
Speaker 4: you think, you know, medium to long term, how do

385
00:23:48,560 --> 00:23:52,040
Speaker 4: you think AI is going to transform business? And you know,

386
00:23:52,080 --> 00:23:56,399
Speaker 4: how can people now business leaders now prepare for what's coming.

387
00:23:57,520 --> 00:24:01,879
Speaker 5: So to an earlier common I made, I do really

388
00:24:01,960 --> 00:24:06,639
Speaker 5: think that we are at an inflection point with the

389
00:24:06,720 --> 00:24:12,359
Speaker 5: advancement of the technologies of AI. I talked about foundation models.

390
00:24:13,200 --> 00:24:17,440
Speaker 5: We definitely at the cusp of being able to address

391
00:24:17,560 --> 00:24:21,840
Speaker 5: use cases at scale that were more challenging before.

392
00:24:22,320 --> 00:24:24,440
Speaker 6: And so I do think.

393
00:24:24,280 --> 00:24:30,760
Speaker 5: The future looks like a lot more generative AI surfacing

394
00:24:30,880 --> 00:24:35,720
Speaker 5: within the enterprise and within business processes and manifesting in

395
00:24:35,720 --> 00:24:41,480
Speaker 5: interesting ways. I think it's almost a given that any

396
00:24:42,880 --> 00:24:45,840
Speaker 5: piece of software, right, whether you think of it in

397
00:24:45,880 --> 00:24:48,879
Speaker 5: terms of an application or you think about it in

398
00:24:48,960 --> 00:24:51,880
Speaker 5: terms of you know, the interact with the website will

399
00:24:51,960 --> 00:24:57,960
Speaker 5: have conversational enabled interfaces from the analyst saying give me

400
00:24:58,000 --> 00:25:00,560
Speaker 5: the latest reports for the last three months, you know,

401
00:25:00,680 --> 00:25:04,240
Speaker 5: typing that or saying it versus the right click file

402
00:25:04,480 --> 00:25:07,120
Speaker 5: blah blah. I think you're going to see that change

403
00:25:07,160 --> 00:25:10,959
Speaker 5: in interaction to more conversational interaction.

404
00:25:11,560 --> 00:25:13,960
Speaker 6: I think, particularly chat based.

405
00:25:13,880 --> 00:25:17,560
Speaker 4: We forget that the graphical user interface is just a metaphor, right,

406
00:25:17,600 --> 00:25:20,960
Speaker 4: It's not like the way computers work. It's just an interface.

407
00:25:21,000 --> 00:25:23,840
Speaker 4: And if chat is a better interface, people will use chat.

408
00:25:24,560 --> 00:25:26,600
Speaker 6: And I think we're going to see that rarely explode.

409
00:25:26,640 --> 00:25:30,080
Speaker 5: And that's powered by a lot of this generative AI work,

410
00:25:30,160 --> 00:25:33,040
Speaker 5: because it becomes for it to feel natural, for it

411
00:25:33,080 --> 00:25:36,200
Speaker 5: to be as informed to readily, as I said, link

412
00:25:36,280 --> 00:25:37,520
Speaker 5: things to get and orchestrate.

413
00:25:37,600 --> 00:25:38,280
Speaker 6: That's a big part.

414
00:25:38,320 --> 00:25:41,959
Speaker 5: So I think I see that happening and the appropriate

415
00:25:42,040 --> 00:25:45,480
Speaker 5: or associated productivity on locks. You begin to see with

416
00:25:45,600 --> 00:25:50,080
Speaker 5: that will just change what kind of decisions, the ease

417
00:25:50,200 --> 00:25:53,600
Speaker 5: with which we can make more and more informed business decisions.

418
00:25:53,920 --> 00:25:58,440
Speaker 5: And so for me, it's that rolling out at scale,

419
00:25:58,840 --> 00:26:04,159
Speaker 5: touching everything procurement. HR think about the advent of the

420
00:26:04,200 --> 00:26:09,560
Speaker 5: spreadsheet and how many different roles.

421
00:26:09,400 --> 00:26:10,760
Speaker 6: It just ended up touching.

422
00:26:11,040 --> 00:26:14,840
Speaker 5: And everybody can use or does use a spreadsheeting business

423
00:26:14,920 --> 00:26:17,560
Speaker 5: in some shape, size or form. So I think of

424
00:26:17,640 --> 00:26:20,560
Speaker 5: this as AI at scale. And so what it therefore

425
00:26:20,680 --> 00:26:25,160
Speaker 5: means from as you said, getting prepared, Well, it's all

426
00:26:25,200 --> 00:26:28,919
Speaker 5: about gaining first of all, the right understanding of the

427
00:26:29,000 --> 00:26:31,480
Speaker 5: technologies and part of what a lot we'll be talking

428
00:26:31,520 --> 00:26:36,000
Speaker 5: about necessary ingredients begin to be well, where do I

429
00:26:36,040 --> 00:26:38,760
Speaker 5: want to apply it first? What data do I need

430
00:26:38,760 --> 00:26:42,679
Speaker 5: to bring together to readily support that? What unlocks what

431
00:26:42,840 --> 00:26:44,840
Speaker 5: new value? And I think it's going to be like

432
00:26:44,960 --> 00:26:47,119
Speaker 5: this rollout, right, you got to start with this project

433
00:26:47,119 --> 00:26:49,679
Speaker 5: and then there's another project, and very soon it will

434
00:26:49,720 --> 00:26:53,680
Speaker 5: be so much it will be ubiquitous in the way

435
00:26:53,800 --> 00:26:56,679
Speaker 5: it supports the work we need to do. That it

436
00:26:56,680 --> 00:26:58,760
Speaker 5: will just speak to a new way of us working

437
00:26:59,160 --> 00:27:02,400
Speaker 5: that is, when you now look back, will be pretty.

438
00:27:02,080 --> 00:27:03,639
Speaker 6: Different from how we work today.

439
00:27:04,359 --> 00:27:09,040
Speaker 5: You see the seeds today but I would argue, think

440
00:27:09,080 --> 00:27:12,439
Speaker 5: of that now, like fully bloomed, it's a forest, not

441
00:27:12,520 --> 00:27:14,800
Speaker 5: a not a flowerbed, you know, yeah, yeah.

442
00:27:14,760 --> 00:27:20,000
Speaker 4: Yeah, great, one other one other sort of loose thread

443
00:27:20,040 --> 00:27:23,399
Speaker 4: I wanna I want to return to uh. And that's

444
00:27:23,520 --> 00:27:28,720
Speaker 4: that's governance, right, you talked about governance and maybe just

445
00:27:28,720 --> 00:27:31,000
Speaker 4: just to help sort of set the table, like you

446
00:27:31,400 --> 00:27:34,920
Speaker 4: mentioned it in a broadway but narrowly, what does governance

447
00:27:35,000 --> 00:27:38,720
Speaker 4: mean in the context of IBM's work on enterprise A high.

448
00:27:38,720 --> 00:27:45,320
Speaker 5: I think, as the Wood tries to suggest, it is

449
00:27:45,480 --> 00:27:53,240
Speaker 5: about having the way to govern one's activities in this realm,

450
00:27:53,520 --> 00:28:01,879
Speaker 5: which really speaks to policies, rules, and frameworks within which

451
00:28:02,000 --> 00:28:06,439
Speaker 5: to understand all of that. Now, before we dive in

452
00:28:06,480 --> 00:28:10,480
Speaker 5: the direction of regulation, which is where people often go,

453
00:28:11,520 --> 00:28:18,320
Speaker 5: policies can be all internal. So think about it this way.

454
00:28:18,560 --> 00:28:22,199
Speaker 5: If I say to you, when I build AI, I

455
00:28:22,320 --> 00:28:26,320
Speaker 5: do not use my customer's data. Is their customer's data?

456
00:28:26,680 --> 00:28:31,440
Speaker 5: Then from a governance perspective, I need processes that ensure

457
00:28:31,600 --> 00:28:35,760
Speaker 5: I know what data I'm using and I can prove

458
00:28:36,600 --> 00:28:39,640
Speaker 5: to myself just first of all internally, forget about anybody

459
00:28:39,680 --> 00:28:42,160
Speaker 5: else that I'm actually adhering.

460
00:28:41,880 --> 00:28:42,680
Speaker 6: To the policies.

461
00:28:42,720 --> 00:28:47,120
Speaker 5: I've laid out that, in my mind, is a lot

462
00:28:47,160 --> 00:28:49,880
Speaker 5: of what governance is about. And in the context of AI,

463
00:28:50,560 --> 00:28:53,680
Speaker 5: it always tends to I think structure around three key

464
00:28:53,720 --> 00:28:57,040
Speaker 5: areas data where did it come from? And what did

465
00:28:57,080 --> 00:28:58,600
Speaker 5: I do with it? And how did I apply it?

466
00:28:58,680 --> 00:29:03,960
Speaker 5: And where did I use it? And then usage what

467
00:29:04,000 --> 00:29:07,080
Speaker 5: do I expect this model to do? Is this model

468
00:29:07,120 --> 00:29:10,080
Speaker 5: still performing the way I think it should be performing.

469
00:29:11,360 --> 00:29:14,800
Speaker 5: What are my processes to address whether they answered that

470
00:29:14,920 --> 00:29:18,280
Speaker 5: question is yes or no? And manage that through? And

471
00:29:18,320 --> 00:29:21,880
Speaker 5: then importantly so this is then to bridge to regulation.

472
00:29:22,440 --> 00:29:24,320
Speaker 5: If you take a look at what's going on in

473
00:29:24,760 --> 00:29:29,400
Speaker 5: the world of AI regulation and our point of view

474
00:29:29,480 --> 00:29:32,200
Speaker 5: on this, by the way, is that you actually regulate

475
00:29:32,280 --> 00:29:36,960
Speaker 5: the use cases, not the technology. Then from a governance perspective,

476
00:29:37,520 --> 00:29:42,280
Speaker 5: how are you able to clearly understand, track and account

477
00:29:42,560 --> 00:29:46,120
Speaker 5: for what use cases you are leveraging AI for? And

478
00:29:46,160 --> 00:29:50,280
Speaker 5: then back to my earlier comments, how that AI is performing.

479
00:29:50,320 --> 00:29:53,000
Speaker 4: And when you talk about governance, how do you make

480
00:29:53,040 --> 00:29:57,920
Speaker 4: sure that you have the governance you need without inhibiting innovation?

481
00:29:58,840 --> 00:29:59,800
Speaker 6: I think what is.

482
00:30:01,600 --> 00:30:04,800
Speaker 5: And this is key a key design point for what

483
00:30:04,800 --> 00:30:07,880
Speaker 5: we're doing with what's the next is how you make

484
00:30:08,120 --> 00:30:15,440
Speaker 5: governance seamless institute versus another activity that you do right,

485
00:30:15,840 --> 00:30:20,160
Speaker 5: And so our goal is to try and drive that

486
00:30:20,280 --> 00:30:25,520
Speaker 5: kind of seamless interactions or value add in terms of governance,

487
00:30:25,920 --> 00:30:30,400
Speaker 5: so that when oh, let's pull through the history right

488
00:30:30,640 --> 00:30:33,280
Speaker 5: of everything we've done here, or what prompts we've created,

489
00:30:33,640 --> 00:30:38,480
Speaker 5: or what data we've used, it's kind of already there, right,

490
00:30:38,720 --> 00:30:41,800
Speaker 5: and so you can feel free to be innovating and

491
00:30:41,840 --> 00:30:45,080
Speaker 5: testing out your different prompts and all that stuff, or

492
00:30:45,080 --> 00:30:47,720
Speaker 5: bringing in your data sets without saying, oh, before I

493
00:30:47,840 --> 00:30:50,240
Speaker 5: do that, I need to make sure I run this checker.

494
00:30:50,280 --> 00:30:53,320
Speaker 5: And now you can kind of bring it in systems

495
00:30:53,400 --> 00:30:56,120
Speaker 5: kind of automatically categorizing it, and then you can go

496
00:30:56,160 --> 00:30:59,240
Speaker 5: in and later verified, validate, or explore say I'm no

497
00:30:59,320 --> 00:31:01,880
Speaker 5: longer going to take this path based upon these facts.

498
00:31:02,240 --> 00:31:02,520
Speaker 6: I think.

499
00:31:02,560 --> 00:31:04,400
Speaker 5: The more we can make it more of a natural

500
00:31:04,440 --> 00:31:09,080
Speaker 5: extension of the activities that need to be done, the

501
00:31:09,120 --> 00:31:11,360
Speaker 5: more we can make it then just a part of

502
00:31:11,400 --> 00:31:13,440
Speaker 5: what needs to be done. And as you're to your point,

503
00:31:14,120 --> 00:31:17,800
Speaker 5: gain our governance needs or supports the governance needs of

504
00:31:17,800 --> 00:31:22,560
Speaker 5: our customers without stifling the innovation of the individuals at

505
00:31:22,560 --> 00:31:27,280
Speaker 5: the glass trying to think through, iteratively, think through new

506
00:31:27,480 --> 00:31:28,760
Speaker 5: value ways to do work.

507
00:31:30,000 --> 00:31:33,320
Speaker 4: Excellent. Let me ask you. Are there things I didn't

508
00:31:33,360 --> 00:31:35,320
Speaker 4: ask you that I should. Are there things you want

509
00:31:35,360 --> 00:31:36,760
Speaker 4: to talk about that we didn't talk about.

510
00:31:37,800 --> 00:31:39,840
Speaker 6: I think we covered quite a lot of true.

511
00:31:40,520 --> 00:31:43,600
Speaker 5: No, I think we we covered the bases there.

512
00:31:45,960 --> 00:31:49,160
Speaker 3: Earlier, Green mentioned that we are at an inflection point

513
00:31:49,200 --> 00:31:54,000
Speaker 3: in AI technology. Implementing AI in business will get easier,

514
00:31:54,400 --> 00:31:58,600
Speaker 3: and AI platforms like watsonex can empower even the largest

515
00:31:58,720 --> 00:32:03,280
Speaker 3: enterprise businesses to reinvent the way they run. As Greem said,

516
00:32:03,640 --> 00:32:06,680
Speaker 3: in the same way the spreadsheet took over business operations,

517
00:32:07,120 --> 00:32:11,000
Speaker 3: the adoption of AI at enterprise scale could be just

518
00:32:11,200 --> 00:32:15,480
Speaker 3: as ubiquitous. It's not an overstatement to say that a

519
00:32:15,560 --> 00:32:21,520
Speaker 3: new era of work may be upon us. I'm Malcolm Gladwell.

520
00:32:21,800 --> 00:32:26,480
Speaker 3: This is a paid advertisement from IBM. Smart Talks with

521
00:32:26,520 --> 00:32:30,440
Speaker 3: IBM is produced by Matt Romano, David jaw Nisha Venkat

522
00:32:30,640 --> 00:32:35,000
Speaker 3: and Royston Deserve with Jacob Goldstein. We're edited by Lydia

523
00:32:35,040 --> 00:32:39,360
Speaker 3: gene Kott. Our engineers are Jason Gambrel, Sarah Bruguier and

524
00:32:39,440 --> 00:32:44,800
Speaker 3: Ben Tolliday. Theme song by Gramoscope Special thanks to Carli Migliori,

525
00:32:45,040 --> 00:32:48,480
Speaker 3: Andy Kelly, Kathy Callahan and eight Bar and the eight

526
00:32:48,520 --> 00:32:52,360
Speaker 3: Bar and IBM teams, as well as the Pushkin marketing team.

527
00:32:52,760 --> 00:32:55,720
Speaker 3: Smart Talks with IBM is a production of Pushkin Industries

528
00:32:56,040 --> 00:33:00,200
Speaker 3: and Ruby's studio at iHeartMedia. To find more push can

529
00:33:00,280 --> 00:33:05,120
Speaker 3: podcast listen on the iHeartRadio app, Apple Podcasts, or wherever

530
00:33:05,560 --> 00:33:07,240
Speaker 3: you listen to podcasts.