WEBVTT - Smart Talks with IBM - AI for Business: Multiplying the impact of AI 0:00:04.440 --> 0:00:12.639 Welcome to Tech Stuff, a production from iHeartRadio. Today, we 0:00:12.680 --> 0:00:15.640 are witnessed to one of those rare moments in history, 0:00:16.000 --> 0:00:19.239 the rise of an innovative technology with the potential to 0:00:19.360 --> 0:00:24.080 radically transform business in society forever. That technology, of course, 0:00:24.560 --> 0:00:28.120 is artificial intelligence, and it's the central focus for this 0:00:28.320 --> 0:00:32.279 new season of Smart Talks with IBM. Join hosts from 0:00:32.360 --> 0:00:36.080 your favorite Pushkin podcasts as they talk with industry experts 0:00:36.080 --> 0:00:39.680 and leaders to explore how businesses can integrate AI into 0:00:39.720 --> 0:00:43.040 their workflows and help drive real change in this new 0:00:43.159 --> 0:00:46.800 era of AI, and of course, host Malcolm Gladwell will 0:00:46.840 --> 0:00:49.160 be there to guide you through the season and throw 0:00:49.240 --> 0:00:52.120 in his two cents as well. Look out for new 0:00:52.159 --> 0:00:55.040 episodes of Smart Talks with IBM every other week on 0:00:55.080 --> 0:00:59.320 the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts, 0:00:59.520 --> 0:01:03.760 and learn more at IBM dot com slash smart Talks. 0:01:05.720 --> 0:01:09.160 Hello, Hello, Welcome to Smart Talks with IBM, a podcast 0:01:09.160 --> 0:01:14.959 from Pushkin Industries, iHeartRadio and IBM. I'm Malcolm Glabo. This season, 0:01:15.120 --> 0:01:19.360 we're continuing our conversation with new creators visionaries who are 0:01:19.400 --> 0:01:23.360 creatively applying technology in business to drive change, but with 0:01:23.400 --> 0:01:27.920 a focus on the transformative power of artificial intelligence and 0:01:27.959 --> 0:01:31.040 what it means to leverage AI as a game changing 0:01:31.160 --> 0:01:35.360 multiplier for your business. Our guest today is Kareem Yousef, 0:01:35.640 --> 0:01:40.200 Senior Vice President of Product Management and Growth for IBM Software. 0:01:40.720 --> 0:01:44.679 Kareem's focus at IBM is on product strategy, thinking about 0:01:44.680 --> 0:01:47.680 the roadmap for IBM Software products and how they can 0:01:47.720 --> 0:01:53.120 deliver effective and compelling customer experiences with the current boom 0:01:53.160 --> 0:01:56.279 and generative AI. Kareem's job is to help businesses figure 0:01:56.280 --> 0:02:00.520 out how they can apply artificial intelligence at scale to 0:02:00.560 --> 0:02:04.520 help solve big problems and boost productivity at new orders 0:02:04.520 --> 0:02:08.640 of magnitude. In today's episode, you'll hear Kareem explain how 0:02:08.720 --> 0:02:12.919 AI powered by foundation models can make AI adoption by 0:02:13.080 --> 0:02:18.040 enterprise businesses even easier, how generative AI will change the 0:02:18.040 --> 0:02:21.680 way businesses process data and make decisions, and how these 0:02:21.720 --> 0:02:27.639 considerations influence the design of Watson x, IBM's next generation 0:02:28.200 --> 0:02:32.960 AI and data platform. Kareem spoke with Jacob Goldstein, host 0:02:32.960 --> 0:02:37.040 of the Pushkin podcast What's Your Problem. A veteran business journalist, 0:02:37.240 --> 0:02:40.760 Jacob has reported for The Wall Street Journal, the Miami Herald, 0:02:41.120 --> 0:02:46.160 and was a longtime host of the NPR program Planet Money. Okay, 0:02:46.639 --> 0:02:47.880 let's get to the interview. 0:02:49.440 --> 0:02:52.560 I'm Jacob Goldstein. I'm one of the hosts at Pushkin 0:02:52.720 --> 0:02:55.600 and a correspondent on this show, and I'm delighted to 0:02:55.600 --> 0:02:57.200 have you here. Can you introduce yourself? 0:02:57.639 --> 0:03:00.520 Ah? Hi, I'm Kareem Yusuf. I'm this and your vice 0:03:00.520 --> 0:03:03.920 president of Product management and Growth for IBM Software. You 0:03:03.919 --> 0:03:05.960 can think of me as the chief product officer for 0:03:06.000 --> 0:03:06.880 IBM Software. 0:03:07.320 --> 0:03:12.040 Okay, sounds like a big job. We're here today to 0:03:12.080 --> 0:03:18.000 talk about AI. We've heard really an extraordinary amount in 0:03:18.040 --> 0:03:22.000 the last few months about chat GPT and you know, 0:03:22.080 --> 0:03:24.560 particularly in how it's used in the very kind of 0:03:24.639 --> 0:03:27.919 consumer facing way. But I'm curious what is the rise 0:03:27.960 --> 0:03:30.760 of chat GPT and you know, AI more generally, what 0:03:30.800 --> 0:03:32.160 does it mean for business? 0:03:32.720 --> 0:03:35.080 Well, you know, it's if you kind of step back 0:03:35.120 --> 0:03:39.760 and think about what really happens. You know, in a business, 0:03:40.240 --> 0:03:43.160 you're really talking about a set of processes, right, you know, 0:03:43.240 --> 0:03:47.120 activities that represent what a business needs to get done, 0:03:47.280 --> 0:03:50.680 whether it's product they produce and then sell or service 0:03:50.760 --> 0:03:54.640 that they provide. And inherent to operating the business, I 0:03:54.640 --> 0:03:58.440 would say are two very key factors. Data and then 0:03:58.480 --> 0:04:02.320 the decisions you make around data and then actually lastly 0:04:02.840 --> 0:04:06.760 the processes or activities you do in accordance with that decision. 0:04:07.520 --> 0:04:10.600 So if you then think about AI as applied to 0:04:10.680 --> 0:04:13.400 business right in that context, well, the first place it 0:04:13.440 --> 0:04:16.480 often starts is how do you make sense of a 0:04:16.520 --> 0:04:20.560 lot of the data associated with driving the business? And 0:04:20.640 --> 0:04:23.320 so AI has always been, in my mind at its 0:04:23.360 --> 0:04:29.560 foremost about gaining insights then lead in to supporting decisions, 0:04:30.040 --> 0:04:34.360 and ultimately ending at helping to automate the activities that 0:04:34.560 --> 0:04:38.200 then are executed as a result of those decisions. So 0:04:38.240 --> 0:04:40.480 that's kind of my simple way of thinking of AI, 0:04:40.800 --> 0:04:43.560 and we can obviously coloring with examples, but that's my 0:04:43.640 --> 0:04:46.000 simplest way of thinking about AI. When you think about 0:04:46.080 --> 0:04:49.400 in the business context, gain insights from masses of data 0:04:49.720 --> 0:04:52.240 to support decisions and then drive. 0:04:52.040 --> 0:04:55.320 Actions, that's a really helpful framework. And then if we 0:04:55.360 --> 0:04:57.880 think about sort of what's happening in the world now 0:04:57.920 --> 0:05:02.080 with you know, enterprise businesses NAI, what are you seeing 0:05:02.120 --> 0:05:05.360 with enterprise adoption of AI at this moment? 0:05:06.200 --> 0:05:10.000 So we're really talking about almost a tale of two periods. 0:05:10.000 --> 0:05:12.400 So let me first of all kind of take you 0:05:12.480 --> 0:05:16.479 back before the advent of what I would call generative AI, 0:05:16.560 --> 0:05:19.919 and the whole chat gpt to what has been going 0:05:20.000 --> 0:05:22.480 on in what I would term the realm of more 0:05:22.600 --> 0:05:26.640 standardized machine learning models. A lot of what has been 0:05:26.640 --> 0:05:29.040 going on has been very much in the realms of 0:05:29.160 --> 0:05:34.440 certain things like anomaly detection or optimization, right, using machine 0:05:34.480 --> 0:05:36.920 learning models to do that kind of work, and where 0:05:36.960 --> 0:05:41.800 might it apply well, think of anomaly detection in security 0:05:41.920 --> 0:05:46.440 software right detecting threats based upon different events flowing through 0:05:47.120 --> 0:05:52.760 or in enterprise asset management software monitoring equipment and detecting 0:05:52.839 --> 0:05:58.080 anomalies within their behavior, or even in IT automation software 0:05:58.400 --> 0:06:01.680 once again detecting anomalies based upon what's going on with 0:06:01.760 --> 0:06:06.360 various IT events and then tasks that should occur. Optimizations 0:06:06.480 --> 0:06:09.640 often play around in the realm as you might imagine 0:06:09.800 --> 0:06:13.279 to solve problems of resource optimization, whether you think of 0:06:13.320 --> 0:06:16.919 that in the context of application resource management for IT 0:06:17.680 --> 0:06:20.160 or in the context of supply chain. These have been 0:06:20.279 --> 0:06:25.479 very classical applications of machine learning AI to really make 0:06:25.560 --> 0:06:30.360 sense of the data and provide a basis to drive decisions. Now, 0:06:30.760 --> 0:06:34.920 what is characterized by all those examples have given and 0:06:35.000 --> 0:06:37.680 the state of the art of that kind of technology 0:06:38.240 --> 0:06:43.039 has always been it's very task specific. So there was 0:06:43.120 --> 0:06:48.040 a air quotes, if I may, kind of limitation in 0:06:48.120 --> 0:06:50.720 the sense that the tak it had to be very 0:06:50.880 --> 0:06:53.800 task specific. And so we've seen a lot of broad 0:06:53.839 --> 0:06:57.560 based adoption within the enterprise, right, but it's very very 0:06:57.680 --> 0:07:01.760 task specific. As you might imagine. Now, what has happened 0:07:02.040 --> 0:07:05.960 recently and has been brought to the four has been 0:07:06.040 --> 0:07:09.640 this discussion of generative AI, which is powered by a 0:07:09.760 --> 0:07:14.880 very specific innovation, this notion of foundation models. And in 0:07:14.920 --> 0:07:18.440 the simplest way to think about it, it's about training 0:07:18.480 --> 0:07:25.480 this large model that can then be refined to various tasks. 0:07:26.480 --> 0:07:30.600 And the easiest one that everybody recognized at the moment 0:07:30.680 --> 0:07:33.840 is the notion of a large language model, a model 0:07:33.880 --> 0:07:37.080 that has an understanding of a lot of the elements 0:07:37.080 --> 0:07:40.480 of a language such that it can be refined to 0:07:40.720 --> 0:07:44.320 a variety of tasks. Write an essay, answer a question, 0:07:44.560 --> 0:07:48.720 singer songs, so on, answers so forth. I like to 0:07:49.080 --> 0:07:52.720 liken the power if you like, and this will speak 0:07:52.760 --> 0:07:55.360 to the why everybody is so excited about it. Why 0:07:55.400 --> 0:07:58.440 would argue at an inflection point? I like to liken 0:07:58.480 --> 0:08:02.960 it to teaching a child the alphabet. When you teach 0:08:02.960 --> 0:08:06.480 a child and alphabet, it's a set of letters. Right, 0:08:06.800 --> 0:08:11.280 Let's call that our foundation model. But over time that 0:08:11.480 --> 0:08:15.080 knowledge of the alphabet is tuned to read a book, 0:08:15.120 --> 0:08:17.720 write an essay, do a composition, create a song, write 0:08:17.760 --> 0:08:20.880 a poem, write an invoice. You understand what I mean, right, 0:08:21.120 --> 0:08:26.920 And so from one foundation model you can support multiple 0:08:27.040 --> 0:08:31.360 targeted tasks as opposed sticking with the analogy to having 0:08:31.360 --> 0:08:35.640 a model for reading, writing, thinking of doing a poem, 0:08:35.920 --> 0:08:38.560 doing an essay, so on and so forth. And so 0:08:38.679 --> 0:08:42.200 in the enterprise context, that means that we're now talking 0:08:42.240 --> 0:08:46.680 about being able to unlock even additional value at scale 0:08:47.600 --> 0:08:52.120 because of the nation of nature foundation models and their 0:08:52.160 --> 0:08:56.120 appeal to generative use cases. Generative in this case meaning 0:08:56.559 --> 0:08:57.840 creation of new content. 0:08:58.600 --> 0:09:02.920 So let's talk about what's in X. IBM recently announced 0:09:02.920 --> 0:09:05.120 what's an X. Just first of all, what is that? 0:09:05.160 --> 0:09:06.000 What is what's an X? 0:09:06.880 --> 0:09:10.480 Well, what's an X refers to our is our brand 0:09:10.520 --> 0:09:14.560 for our platform, the WhatsApp platform for really taking advantage 0:09:15.080 --> 0:09:19.560 of generative AI within the enterprise, within business. And so 0:09:19.760 --> 0:09:22.720 when you begin to think about what does that mean, 0:09:22.800 --> 0:09:24.840 while it leads you to the components of what's an 0:09:25.040 --> 0:09:27.199 X and to a set of use cases. So let 0:09:27.200 --> 0:09:31.400 me paint a few quick pictures for you here. What's 0:09:31.440 --> 0:09:35.000 an X first of all, is about enabling our customers 0:09:35.080 --> 0:09:40.080 to manipulate models against their task, manipulate these foundation models 0:09:40.320 --> 0:09:43.800 against their task. Our belief is that the world is 0:09:43.880 --> 0:09:48.160 a multi model world, right and especially when you think 0:09:48.200 --> 0:09:51.320 about it in the context of business. Models are going 0:09:51.400 --> 0:09:54.560 to come from various sources, the ones we supply, the 0:09:54.559 --> 0:09:56.680 ones out there in open source, and so of you. 0:09:57.000 --> 0:09:59.600 But there are activities you need to do around these 0:09:59.640 --> 0:10:03.600 models to as I said, apply them to your use case. 0:10:03.679 --> 0:10:05.560 And we'll talk about use cases in a bit. So 0:10:06.080 --> 0:10:09.480 what's next. Dot AI is that environment that build a 0:10:09.559 --> 0:10:11.960 tool if you like, for being able to do those 0:10:12.000 --> 0:10:16.480 manipulation of models to meet your specific use case. Thinks 0:10:16.520 --> 0:10:20.400 that people will recognize in the field prompt engineering, prompt tuning, 0:10:20.559 --> 0:10:23.840 fine tuning, those kinds of activities which are all the 0:10:23.920 --> 0:10:27.160 manipulation of models to meet your use case. Yeah. The 0:10:27.200 --> 0:10:30.679 second component is dot data, So what's the next? Dot 0:10:30.800 --> 0:10:34.800 data is essentially a next generation data store is based 0:10:34.880 --> 0:10:38.840 upon something referred to as an open data lakehouse architecture 0:10:39.160 --> 0:10:42.960 that helps to bring together the data that's needed to 0:10:43.120 --> 0:10:45.520 actually do the AI. In this case, when you think 0:10:45.520 --> 0:10:49.160 about manipulating a model, a foundation model, you're generally using 0:10:49.200 --> 0:10:51.920 some data to prompt it, tune it, to train it 0:10:51.960 --> 0:10:55.800 to your use cases. And so we provide a very 0:10:55.880 --> 0:10:58.600 open data store that allows all manner of data and 0:10:58.640 --> 0:11:01.319 formats to be brought through. Today you that and the 0:11:01.360 --> 0:11:05.720 third component is what's next up governance And this is 0:11:05.760 --> 0:11:09.680 all about the framework and the toolkit required to apply 0:11:09.920 --> 0:11:14.080 the right governance principles across doing this kind of work, 0:11:14.320 --> 0:11:19.000 because when you're deploying AI within the enterprise, governance is 0:11:19.040 --> 0:11:23.000 actually important, right, It's critical to understand why is your 0:11:23.080 --> 0:11:25.800 data coming from? What data did you add in? How 0:11:25.920 --> 0:11:28.520 is your model performing? Are you able to keep an 0:11:28.559 --> 0:11:32.280 appropriate audit trail of your activities for your own internal 0:11:32.640 --> 0:11:35.920 policy and compliance needs or for regulatory needs as well. 0:11:36.520 --> 0:11:40.720 So this platform, the system that you're describing, I'm curious, 0:11:41.280 --> 0:11:44.520 how is it different from the you know, the generative 0:11:44.640 --> 0:11:47.040 AI options that you know we've all been hearing about 0:11:47.080 --> 0:11:48.000 sort of in the press. 0:11:48.679 --> 0:11:51.960 Well, I think it really comes down to the ethos 0:11:52.120 --> 0:11:55.160 or the principles that first of all drive the work 0:11:55.200 --> 0:11:58.280 that we're doing. The first I would fixate on is 0:11:58.559 --> 0:12:02.880 being open. Right. We fundamentally believe that to do this 0:12:02.960 --> 0:12:05.680 kind of work within the enterprise, you need an open 0:12:05.760 --> 0:12:09.360 platform that, as I said, is open to all manner 0:12:09.360 --> 0:12:11.880 of models from all sources. It's one of the reasons 0:12:11.880 --> 0:12:15.079 why we announced our partnership with hugging Face to make 0:12:15.120 --> 0:12:18.960 sure that our clients can gain access to open source 0:12:19.040 --> 0:12:22.320 innovation within the platform to do their work. 0:12:22.800 --> 0:12:24.440 And hugging Face, to be clear, is sort of the 0:12:25.320 --> 0:12:27.280 open source AI kind of hub. 0:12:27.760 --> 0:12:31.320 That's right, that's correct. Yes, it's a marketplace hub for 0:12:31.520 --> 0:12:35.880 all kind of ecosystem coordinator for open source models. And 0:12:35.960 --> 0:12:38.720 I believe there's a lot of innovation going on out there. So, 0:12:38.800 --> 0:12:44.840 first of all, open becomes important. The second targeted So 0:12:45.040 --> 0:12:50.960 our focus is very much on enabling these business use cases, right, 0:12:51.200 --> 0:12:53.080 And you might say what kind of use cases are 0:12:53.120 --> 0:12:55.319 we talking about? I give you three very quick ones 0:12:55.360 --> 0:12:58.600 that with our customers are focused on a lot of 0:12:58.640 --> 0:13:03.320 focus around and enhancing customer service use cases. Think of 0:13:03.360 --> 0:13:07.640 this as chatbots or digital assistance that are further trained 0:13:08.200 --> 0:13:11.319 in more and more information about what the company has 0:13:11.360 --> 0:13:14.760 to offer, or could be internal policies, external policy, so 0:13:14.800 --> 0:13:17.560 on and so forth. This means a platform that makes 0:13:17.600 --> 0:13:22.160 it really easy to bring your own data to train 0:13:22.280 --> 0:13:26.640 and tune the model, while protecting your own data as 0:13:26.720 --> 0:13:31.839 extremely important for the enterprise right. Another important use case 0:13:32.160 --> 0:13:34.160 seeing a lot of focused on what i'd call AI 0:13:34.200 --> 0:13:38.600 based orchestration or automation of task whereby think about like 0:13:38.960 --> 0:13:42.959 an HR professional as an example, going through a job 0:13:43.000 --> 0:13:47.040 requisition is able to interact with multiple systems via a 0:13:47.120 --> 0:13:52.359 very simple chat interface and have work dynamically sequenced to 0:13:52.400 --> 0:13:56.719 support them in doing their task. That once again requires 0:13:56.800 --> 0:14:00.360 a notion of working with models and technology in a 0:14:00.440 --> 0:14:03.439 way that in many ways can be unique to how 0:14:03.440 --> 0:14:06.280 a business wishes to work and indeed, in various cases 0:14:06.320 --> 0:14:09.640 can embody what they consider their their secret source or 0:14:09.640 --> 0:14:13.719 their differentiated advantage. So once again, a platform that recognizes 0:14:13.760 --> 0:14:16.640 that and designed for business that's not the same scope 0:14:16.720 --> 0:14:21.280 or frame of reference for a consumer platform. And then 0:14:21.400 --> 0:14:23.360 you know, we're also seeing a lot of work around 0:14:23.720 --> 0:14:28.720 cod generation, application modernization, you know, and people enhancing their skills. 0:14:28.920 --> 0:14:32.960 So targeted becomes really important. Mentioned open and I mentioned 0:14:33.040 --> 0:14:36.400 targeted targeted to the business to the use cases that 0:14:36.440 --> 0:14:40.960 they need to do. Underpinning that then it's trusted. So 0:14:41.040 --> 0:14:44.120 everything I gave you in those targeted use cases talk 0:14:44.200 --> 0:14:51.320 about handling enterprise proprietary and specific data. We are trusted 0:14:51.360 --> 0:14:53.840 in this regard right. We have been serving the business 0:14:54.280 --> 0:14:57.720 for many, many a year, and we are designing our 0:14:57.800 --> 0:15:00.920 platform and even our principles and way of creating to 0:15:01.120 --> 0:15:04.280 recognize and enable that. Both in terms of the work 0:15:04.320 --> 0:15:07.880 we do around the governance framework and transparency that you're 0:15:07.920 --> 0:15:10.560 able to gain and apply, but even in the way 0:15:10.600 --> 0:15:15.040 we allow our platform to be deployed in multiple kind 0:15:15.080 --> 0:15:18.800 of locations, of footprints consumed as a service on a hyperscaler, 0:15:19.160 --> 0:15:22.720 running your own private footprint on prem or your cloud footprint. 0:15:23.040 --> 0:15:25.280 All of these need to be brought together to meet 0:15:25.320 --> 0:15:30.320 the needs of an actual enterprise business. My last comment 0:15:30.400 --> 0:15:34.360 is where I think we're fundamentally differentiated is we're really 0:15:34.480 --> 0:15:41.680 about empowering our customers to take advantage of AI to 0:15:41.800 --> 0:15:47.320 unleash the intelligence, capabilities productivity of their own business. This 0:15:47.360 --> 0:15:51.200 isn't about, oh, we've established a bunch of APIs that 0:15:51.240 --> 0:15:54.960 you can ask questions. This is about how do you 0:15:55.480 --> 0:15:59.920 craft what you need for your business to deliver different 0:16:00.080 --> 0:16:06.280 shaped value to your customers, shareholders, employees with all the 0:16:06.400 --> 0:16:09.400 appropriate protections as well. And so there's a lot of 0:16:09.440 --> 0:16:11.440 focus on what we've done with the platform and the 0:16:11.480 --> 0:16:14.080 tool set to enable that, to enable what we like 0:16:14.120 --> 0:16:19.720 to call AI value creators, not just consumers of AI value. 0:16:20.480 --> 0:16:26.480 When you were talking about basically enterprise adoption of AI, 0:16:27.320 --> 0:16:30.880 you use the word trust, and I'm curious, you know, 0:16:31.360 --> 0:16:35.720 what does trust mean in the context of AI and 0:16:35.800 --> 0:16:36.400 the enterprise. 0:16:37.800 --> 0:16:43.120 I would kind of deconstruct trust along these k avenues. 0:16:45.240 --> 0:16:49.520 If AI is about giving you insights to help you 0:16:49.560 --> 0:16:55.800 support decisions, how do you trust what insights it's provided? 0:16:56.680 --> 0:17:01.760 What data did it use? What did it consider based 0:17:01.840 --> 0:17:08.120 upon that data that therefore led to the insight provided? 0:17:10.440 --> 0:17:14.280 Why is this important? Why this notion of trust? Well, One, 0:17:14.680 --> 0:17:17.520 you're about to make a decision, so you want to 0:17:17.600 --> 0:17:22.360 understand the basis for a decision. It's no different than 0:17:22.400 --> 0:17:25.520 me asking you something and then saying, okay, can you 0:17:25.560 --> 0:17:27.879 explain you're working? Right, that would be a notion of 0:17:27.920 --> 0:17:32.080 trust that we establish and a very natural interaction as humans, right, 0:17:32.080 --> 0:17:35.399 we do it all the time, right, So there is 0:17:35.440 --> 0:17:39.000 that element. The other reason why it becomes important if 0:17:39.040 --> 0:17:43.680 you're applying AI into business processes and therefore how your 0:17:43.720 --> 0:17:48.239 business works. You want to make sure that you know 0:17:48.320 --> 0:17:52.600 what biases are built in to any decision or not 0:17:53.040 --> 0:17:56.880 or if the AI the model in effect is drifting 0:17:57.920 --> 0:18:01.840 away from kind of the parameters that you would want 0:18:01.880 --> 0:18:06.320 it to remain within, right or go trust and so 0:18:07.440 --> 0:18:11.320 in many ways, that's one big aspect of trusting the 0:18:11.400 --> 0:18:14.879 technology because you're applying it into decisions you need to 0:18:14.880 --> 0:18:17.240 make every day, and you need to know in very 0:18:17.240 --> 0:18:20.840 simple terms how it works and how it is working. 0:18:22.280 --> 0:18:25.760 The element of trust that I think is important in 0:18:25.760 --> 0:18:32.080 this discussion. Who are you getting your AI from? That's 0:18:32.320 --> 0:18:35.840 very important to us as a company here at IBM. Right, 0:18:36.440 --> 0:18:42.320 given we serve business, that trust becomes extremely important and 0:18:42.359 --> 0:18:44.080 what are the elements of that trust? What are the 0:18:44.119 --> 0:18:49.000 customers trying to understand? Well, first and foremost, what's your 0:18:49.000 --> 0:18:52.920 ethos around AI? We're very clear on the customer's data 0:18:53.080 --> 0:18:56.880 is their data when they tune or refine those models 0:18:56.920 --> 0:19:00.359 to meet their use cases. That is all THEIRS actually 0:19:00.359 --> 0:19:02.880 provide the ability for them to do that in very 0:19:02.960 --> 0:19:07.800 isolated and protected ways as they choose, and we never 0:19:07.960 --> 0:19:13.200 use that data without explicit opting and permissions. Right, customers 0:19:13.240 --> 0:19:15.440 might say Oh yeah, use this so that you can 0:19:15.440 --> 0:19:18.600 make a generally overall better model. But it's full awareness, 0:19:18.800 --> 0:19:23.160 full transparency that is important. That's a trust of who 0:19:23.200 --> 0:19:26.520 you're doing business with. So that's how I think about trust. 0:19:27.160 --> 0:19:30.720 How do you build systems you trust? And are you 0:19:31.760 --> 0:19:34.479 working with people you trust? 0:19:35.280 --> 0:19:37.600 I find Kareem's point about trust when it comes to 0:19:37.720 --> 0:19:41.080 data to be so important because as powerful as AI 0:19:41.200 --> 0:19:45.240 tools can be, their helpfulness is dependent on how trustworthy 0:19:45.280 --> 0:19:49.120 the data is. Humans will have to decide if our data, 0:19:49.200 --> 0:19:52.600 our decision making, and our AI insights live up to 0:19:52.640 --> 0:19:55.720 the vision we hope to achieve in business. As Green 0:19:55.800 --> 0:19:59.359 and Jacob continue the conversation, Jacob asks some more practical 0:19:59.440 --> 0:20:04.720 questions about how businesses can adopt AI into their own processes. 0:20:05.440 --> 0:20:11.040 Let's listen, how can businesses move toward integrating AI as 0:20:11.119 --> 0:20:14.479 part of their core business model instead of, you know, 0:20:14.520 --> 0:20:16.480 sort of as an add on on the periphery. 0:20:17.280 --> 0:20:20.119 It's funny, you know. My simple answer to that is 0:20:20.320 --> 0:20:22.720 it's actually the simplest thing in the world to do. 0:20:23.280 --> 0:20:30.080 By thinking about your business, thinking about your elements of differentiation, 0:20:31.359 --> 0:20:37.640 and then thinking about how AI can help you extend 0:20:37.800 --> 0:20:40.160 expand those Right, what do you want to be known for. 0:20:40.480 --> 0:20:44.040 I picked a very simple use case of customer service interaction. 0:20:44.400 --> 0:20:46.919 Almost every business needs to do that and wants to 0:20:46.960 --> 0:20:49.560 do it better, and so it becomes a way to stop. 0:20:49.560 --> 0:20:51.359 But then as you begin to work your way through, 0:20:51.600 --> 0:20:55.040 you think about various automation of business processes. You think 0:20:55.040 --> 0:20:58.080 about decisions that need to be made right or how 0:20:58.119 --> 0:21:01.119 can individuals be helped, how can they made more productive? 0:21:01.560 --> 0:21:04.879 I think always becomes a very important one. Right, So, 0:21:05.000 --> 0:21:07.760 and you can apply this in many context a financial 0:21:07.800 --> 0:21:11.520 analyst looking at reams of data and trying to derive insights. 0:21:12.040 --> 0:21:14.720 How do you leverage AI to make that financial analyst 0:21:14.960 --> 0:21:17.879 even more powerful? And so that's how I advise you, know, 0:21:17.920 --> 0:21:20.440 people to always look at it. Think about your task, 0:21:20.600 --> 0:21:24.119 think about your business processes, think about where help is 0:21:24.160 --> 0:21:27.360 needed or where new value could be unlocked, and then 0:21:27.400 --> 0:21:30.680 you're applying AI as a tool to achieve that end. 0:21:31.600 --> 0:21:34.680 One of the themes we return to on this show 0:21:34.840 --> 0:21:40.840 a lot is creativity and the relationship between technology and creativity. 0:21:41.640 --> 0:21:46.600 And I'm curious how you think that AI can help 0:21:46.640 --> 0:21:48.439 people be more creative at work. 0:21:50.320 --> 0:21:53.200 I think AI can help people be more creative at 0:21:53.240 --> 0:21:57.240 work by automating the mundane to unlock your mind to 0:21:57.240 --> 0:22:00.359 be able to focus on higher value. You know, I've 0:22:00.440 --> 0:22:04.000 used a couple of times I've talked about deriving insights 0:22:04.040 --> 0:22:09.200 from data right to drive informed decisions. If you can 0:22:09.400 --> 0:22:14.119 use AI to gather a lot more insights into one place, 0:22:14.160 --> 0:22:17.000 then you could typically do yourself or more manually, you'd 0:22:17.040 --> 0:22:19.600 have to like write it down, look at six spreadsheets, 0:22:19.600 --> 0:22:22.720 copy from here to there. Then you actually have more 0:22:22.800 --> 0:22:26.760 time to look at that data, digest those insights, and 0:22:26.880 --> 0:22:29.280 think about what do I need to do with these 0:22:29.320 --> 0:22:32.120 as a business, which direction do I want to go? 0:22:32.640 --> 0:22:35.720 I think of its freeing us up to do more 0:22:35.760 --> 0:22:39.120 of what we actually as humans do extremely well, which 0:22:39.160 --> 0:22:43.320 is actually that creative thinking exactly simple terms. Why do 0:22:43.400 --> 0:22:47.760 we use a calculator to do arithmetic? It's not that 0:22:47.800 --> 0:22:51.120 we cannot necessarily knock it out ourselves. But if you're 0:22:51.119 --> 0:22:53.960 trying to balance your checkbook, to use an old phrase 0:22:54.320 --> 0:23:01.080 or dare I say, what's a check but so modernize that. 0:23:02.280 --> 0:23:06.240 If you're trying to check your expenses for the month 0:23:06.520 --> 0:23:11.200 and your performance against budget, yes you could print out 0:23:11.240 --> 0:23:16.320 all your statements, circle everything and add it all up. 0:23:17.240 --> 0:23:22.120 Or you could begin to use technology to improve that experience, 0:23:22.160 --> 0:23:24.560 so you can get more time to think about what 0:23:24.640 --> 0:23:27.280 really am I learning from my spending patterns and what 0:23:27.320 --> 0:23:29.240 do I want to do about it. It's a very 0:23:29.280 --> 0:23:32.920 simple personal example, but I think it's fundamentally what we're 0:23:32.960 --> 0:23:36.359 talking about here, and that's always been in my mind, 0:23:36.400 --> 0:23:40.919 the promise of technology freeing us up to actually apply 0:23:41.000 --> 0:23:45.520 ourselves to higher value thought and higher value problems. 0:23:46.080 --> 0:23:50.320 So we've been talking basically about the present so far, 0:23:50.440 --> 0:23:53.040 and I'm curious if if you think about the future 0:23:53.040 --> 0:23:56.560 and you think, you know, medium to long term, how 0:23:56.560 --> 0:23:59.479 do you think AI is going to transform business? And 0:24:00.080 --> 0:24:03.720 you know, how can people now, business leaders now prepare 0:24:03.760 --> 0:24:04.560 for what's coming. 0:24:05.640 --> 0:24:10.040 So to an earlier comment I made, I do really 0:24:10.119 --> 0:24:14.800 think that we are at an inflection point with the 0:24:14.880 --> 0:24:20.480 advancement of the technologies of AI. I talked about foundation models. 0:24:21.359 --> 0:24:25.600 We definitely at the cusp of being able to address 0:24:25.720 --> 0:24:30.640 use cases at scale that were more challenging before, and 0:24:30.720 --> 0:24:35.440 so I do think the future looks like a lot 0:24:35.560 --> 0:24:40.919 more generative AI surfacing within the enterprise and within business 0:24:40.960 --> 0:24:47.600 processes and manifesting in interesting ways. I think it's almost 0:24:47.640 --> 0:24:53.400 a given that any piece of software right think, whether 0:24:53.440 --> 0:24:55.960 you think of it in terms of an application or 0:24:56.000 --> 0:24:58.080 you think about it in terms of you know, the 0:24:58.240 --> 0:25:04.480 interact with the website will have conversational enabled interfaces from 0:25:04.520 --> 0:25:07.159 the analyst saying give me the latest reports for the 0:25:07.240 --> 0:25:10.200 last three months, you know, typing that or saying it 0:25:10.440 --> 0:25:14.040 versus the right click file blah blah. I think you're 0:25:14.080 --> 0:25:19.119 going to see that change in interaction to more conversational interaction. 0:25:19.720 --> 0:25:22.040 I think, particularly chat based. 0:25:22.040 --> 0:25:25.679 We forget that the graphical user interface is just a metaphor, right, 0:25:25.760 --> 0:25:29.119 It's not like the way computers work. It's just an interface. 0:25:29.160 --> 0:25:32.000 And if chat is a better interface, people will use chat. 0:25:32.680 --> 0:25:34.760 And I think we're going to see that rarely explode. 0:25:34.800 --> 0:25:37.920 And that's powered by a lot of this generative AI 0:25:38.000 --> 0:25:41.000 work because it becomes for it to feel natural, for 0:25:41.080 --> 0:25:43.879 it to be as informed, to readily, as I said, 0:25:44.119 --> 0:25:46.440 link things to get and orchestrate. That's a big part. 0:25:46.440 --> 0:25:50.080 So I think I see that happening and the appropriate 0:25:50.200 --> 0:25:53.639 or associated productivity on locks you begin to see with 0:25:53.760 --> 0:25:58.240