1 00:00:11,970 --> 00:00:14,930 Hello and welcome to the SAP Learning Insights podcast. 2 00:00:14,930 --> 00:00:18,010 I am David Chaviano today with Thomas Yenovine to talk about 3 00:00:18,010 --> 00:00:21,970 everybody's favorite topic, Artificial intelligence type of 4 00:00:21,970 --> 00:00:25,090 recording is 17th of September 2024. 5 00:00:25,090 --> 00:00:27,930 I think a little bit of the hype has worn down, but I think it's 6 00:00:27,930 --> 00:00:32,369 quite a solid topic to talk about in a more maybe rational 7 00:00:32,369 --> 00:00:35,080 fashion. So let's talk about AI adoption, 8 00:00:35,080 --> 00:00:37,520 specifically within organizations and business. 9 00:00:37,520 --> 00:00:40,280 So the technologies out there, how do we get people to actually 10 00:00:40,280 --> 00:00:42,720 start using it? Before we jump into it, we 11 00:00:42,840 --> 00:00:44,880 always want to get our hero's origin story. 12 00:00:45,440 --> 00:00:48,080 Thomas, take a moment. Please introduce yourself. 13 00:00:48,080 --> 00:00:50,040 Where are you from? What have you studied and how 14 00:00:50,040 --> 00:00:52,920 did you find yourself at SAP? Hey, David. 15 00:00:53,040 --> 00:00:55,680 Hi, everyone. Thanks so much for having me as 16 00:00:55,680 --> 00:00:58,920 a guest. Yeah, my hero's story. 17 00:01:00,520 --> 00:01:03,280 So, OK, currently I'm living near Khalshwa. 18 00:01:03,280 --> 00:01:08,440 Originally I'm from the South of Germany, Hybron Weinsburg, near 19 00:01:08,440 --> 00:01:10,800 Stuttgart. So I didn't make it that far. 20 00:01:10,800 --> 00:01:14,400 However, due my studies, I was in San Diego studying a little 21 00:01:14,400 --> 00:01:18,760 bit in California, and I studied organizational psychology. 22 00:01:19,040 --> 00:01:21,440 Yeah, so I think that's super fascinating. 23 00:01:22,000 --> 00:01:24,480 And this actually even brought me to SAP. 24 00:01:24,560 --> 00:01:28,320 So I started as a student very long ago, 1998. 25 00:01:29,080 --> 00:01:33,160 I saw in the internal learning organizational development team 26 00:01:33,720 --> 00:01:35,320 and yeah, I had a good time there. 27 00:01:35,320 --> 00:01:39,240 But then I moved on to product management and now since eight 28 00:01:39,240 --> 00:01:41,600 years I'm doing business development in Europe. 29 00:01:41,720 --> 00:01:45,840 We have 4 topics around learning, change management and 30 00:01:45,840 --> 00:01:49,320 also some cloudified services. All right, very cool. 31 00:01:49,320 --> 00:01:52,320 So often I think most of our guests, they end up studying 32 00:01:52,320 --> 00:01:54,160 something and then doing something completely different. 33 00:01:54,160 --> 00:01:57,560 SAP but it seems like you're quite consistent in what you 34 00:01:57,560 --> 00:01:59,800 study, organizational psychology, business 35 00:01:59,800 --> 00:02:03,640 development, and in this topic in particular, adoption of AI. 36 00:02:04,280 --> 00:02:06,680 So that involves a lot of psychology within the large 37 00:02:06,680 --> 00:02:08,360 organizations and the change of management. 38 00:02:08,360 --> 00:02:11,720 So I think you're the perfect person to be talking about this 39 00:02:11,720 --> 00:02:15,280 topic today. So, starting from scratch, maybe 40 00:02:15,280 --> 00:02:18,560 you can give us a a general introduction on the topic of AI 41 00:02:18,560 --> 00:02:23,040 adoption and why it's important. So one of the challenges of 42 00:02:23,040 --> 00:02:26,520 course, if the, if you look at the term of adoption, it's like 43 00:02:26,520 --> 00:02:29,080 a buzzword, yeah. So people have different 44 00:02:29,080 --> 00:02:32,200 meanings or project different meanings into it. 45 00:02:32,560 --> 00:02:37,160 So for some it's just the consumption of whatever digital 46 00:02:37,160 --> 00:02:40,080 tool. For some it's acceptance and 47 00:02:40,080 --> 00:02:41,680 actually use it. Yeah. 48 00:02:41,680 --> 00:02:46,960 So whatever I I think it's a mix and even the implementation or 49 00:02:46,960 --> 00:02:53,440 or the transformation. So what I find interesting also 50 00:02:53,440 --> 00:02:55,400 look at the different stakeholder interests. 51 00:02:55,640 --> 00:02:57,920 Yeah. So like if management looks at 52 00:02:57,920 --> 00:03:01,960 adoption, they want ROI, they want cost reduction, they want 53 00:03:01,960 --> 00:03:06,520 growth, they want profit because they buy software and the goal 54 00:03:06,520 --> 00:03:11,160 is to pick a management goal. Yeah, IT, but if you talk to 55 00:03:11,160 --> 00:03:15,040 someone from IT, they perhaps they won low integration costs 56 00:03:15,040 --> 00:03:17,920 or security. So I really other things or 57 00:03:17,920 --> 00:03:20,480 perhaps even automation of their own processes. 58 00:03:22,080 --> 00:03:25,600 And if you are someone in the line of business, like in HR, of 59 00:03:25,600 --> 00:03:29,000 course they have their own goals like perhaps want to process 60 00:03:29,000 --> 00:03:34,000 more recruits or whatever. And I think where you actually 61 00:03:34,160 --> 00:03:38,640 should start is the actual user. What they usually want is less 62 00:03:38,640 --> 00:03:41,080 frustration. If they have a new tool, I would 63 00:03:41,080 --> 00:03:45,320 say they want to have some individual benefits that their 64 00:03:45,320 --> 00:03:48,120 work is per simpler, fast, better. 65 00:03:48,160 --> 00:03:50,040 Yeah. And I think that's that's 66 00:03:50,040 --> 00:03:53,800 important to know. And, and also there are 67 00:03:53,800 --> 00:03:59,600 different other, I would say words or narratives around 68 00:03:59,600 --> 00:04:02,520 adoption. So one is adoption, which is the 69 00:04:02,520 --> 00:04:05,880 actual usage, but you also have the adaption. 70 00:04:06,240 --> 00:04:11,480 And I think that's, I would say almost equally important because 71 00:04:11,480 --> 00:04:15,360 if you have a new software and with AI, it's really disruptive. 72 00:04:15,360 --> 00:04:18,880 You can do very new things. You can automate knowledge work, 73 00:04:18,880 --> 00:04:24,360 you can really improve tasks, like in text processing, for 74 00:04:24,360 --> 00:04:29,120 example, up to 80% or so, for example, regards of timing. 75 00:04:30,120 --> 00:04:33,160 But what, what does this mean? How do we adapt as an 76 00:04:33,160 --> 00:04:35,920 organization? Adoption is one side of the 77 00:04:35,920 --> 00:04:39,640 metal, but on the other side we need to adapt our jobs, our 78 00:04:39,640 --> 00:04:42,800 tasks, our processes. Perhaps some tasks are not 79 00:04:43,200 --> 00:04:47,040 needed anymore, but there are new tasks possible, for example. 80 00:04:48,040 --> 00:04:49,320 Yeah. And of course, if you as an 81 00:04:49,320 --> 00:04:53,040 organization do not adopt, the competition is certainly going 82 00:04:53,040 --> 00:04:55,560 to be doing it. So it's quite an important topic 83 00:04:55,560 --> 00:04:59,520 to stay ahead of the curve here at least, and if you're really 84 00:04:59,520 --> 00:05:04,160 good, be a trendsetter. So adoption, I think in my mind 85 00:05:04,160 --> 00:05:06,840 at least, it's focusing a lot on the human component. 86 00:05:06,840 --> 00:05:09,600 So the technology is there, but there's a lot of, I guess, 87 00:05:09,600 --> 00:05:13,960 psychology involved in getting humans to actually start 88 00:05:13,960 --> 00:05:16,040 adopting new technology within an organization. 89 00:05:16,040 --> 00:05:18,000 So you have people that do things a certain way, and 90 00:05:18,000 --> 00:05:19,760 they're obviously going to have to change the way they do things 91 00:05:19,760 --> 00:05:21,080 if you're going to adopt new technology. 92 00:05:21,560 --> 00:05:26,600 So what are the key benefits to humans and an AI adoption 93 00:05:26,600 --> 00:05:28,160 situation within the organization? 94 00:05:28,640 --> 00:05:31,200 Of course, it depends on the use case because there are different 95 00:05:31,200 --> 00:05:34,840 use cases around AI. Now, if you talk about AII think 96 00:05:34,840 --> 00:05:38,080 recently, we often talk about generative AI, right, which 97 00:05:38,320 --> 00:05:42,000 really is powered by large language models and so on. 98 00:05:42,000 --> 00:05:45,720 And there, of course, it's working with text, for example, 99 00:05:46,000 --> 00:05:48,160 like if we're a sales professional, marketing 100 00:05:48,160 --> 00:05:54,120 professional or HR professional. And often you really can be much 101 00:05:54,120 --> 00:05:57,480 more productive. So you can do much more things 102 00:05:57,560 --> 00:06:01,120 shorter time or you can do additional things like just once 103 00:06:01,120 --> 00:06:04,600 more example for our podcast, we also have a German speaking 104 00:06:04,600 --> 00:06:07,240 podcast around education, the education Newscast. 105 00:06:07,520 --> 00:06:12,480 We now do transcripts automated, which is voice to text. 106 00:06:12,800 --> 00:06:17,560 And from that whole transcripts, so the pages we do summary also 107 00:06:17,560 --> 00:06:20,920 with AI tool. I think we tried it once or 108 00:06:20,920 --> 00:06:25,040 twice, but usually it takes at least two hours to transcribe 109 00:06:25,040 --> 00:06:29,280 and summarise a podcast. Now if we have this AI chain, 110 00:06:29,720 --> 00:06:33,320 let's say the transcription and the summary, perhaps it's down 111 00:06:33,320 --> 00:06:37,680 to 20 minutes or so. So I think we're quite safe some 112 00:06:37,680 --> 00:06:42,680 time for example, but also have an extra service because earlier 113 00:06:42,680 --> 00:06:46,640 times we perhaps often didn't have to have the time to to to 114 00:06:46,640 --> 00:06:49,280 do that to offer that additional service. 115 00:06:50,280 --> 00:06:52,480 OK. So, so the the promise of of 116 00:06:52,480 --> 00:06:54,320 increased productivity is certainly there. 117 00:06:55,240 --> 00:06:57,600 What would be some change in management considerations that 118 00:06:57,600 --> 00:07:00,520 an organization to take with this people when implementing 119 00:07:00,520 --> 00:07:03,040 this technology? The productivity of course, 120 00:07:03,040 --> 00:07:06,640 especially in, in our German area, that's the number one 121 00:07:06,640 --> 00:07:08,680 thing. But I think you also can be more 122 00:07:08,680 --> 00:07:13,000 creative, you know, because you have the word knowledge now in 123 00:07:13,000 --> 00:07:18,640 let's say TCPT or in atrophic cloud or whatever you use and 124 00:07:18,640 --> 00:07:22,840 you can research, you can brainstorm, you have a, let's 125 00:07:22,840 --> 00:07:26,600 say, own body. And that's also something what 126 00:07:26,600 --> 00:07:29,960 really makes you more creative. So often people look at the 127 00:07:29,960 --> 00:07:34,240 productivity, but I've think it also can make your, for example, 128 00:07:34,240 --> 00:07:36,440 your knowledge work also more creative. 129 00:07:36,880 --> 00:07:38,960 Yeah, yeah. I, I can report from personal 130 00:07:38,960 --> 00:07:40,400 experience. It's, it's always interesting to 131 00:07:40,400 --> 00:07:43,400 give various prompts when you want to present and a large 132 00:07:43,400 --> 00:07:45,280 language model a problem that you're facing. 133 00:07:45,720 --> 00:07:48,760 And you can ask it like how would Jeff Bezos address it, 134 00:07:48,760 --> 00:07:52,240 such as this situation, what would be Bill Gates strategy and 135 00:07:52,240 --> 00:07:54,080 on solving this problem, stuff like that. 136 00:07:54,080 --> 00:07:56,400 And it's interesting to sort of answers you get that you may not 137 00:07:56,400 --> 00:08:00,280 have considered otherwise. Absolutely. 138 00:08:00,880 --> 00:08:02,800 And going in into like the change of management. 139 00:08:02,800 --> 00:08:07,120 So I imagine not everybody would be convinced in an organization 140 00:08:07,120 --> 00:08:10,520 based on what we just said. So I managers have to go and put 141 00:08:10,560 --> 00:08:15,240 a little bit more effort into implementing this technology 142 00:08:15,240 --> 00:08:18,000 into an organization. So what considerations should 143 00:08:18,080 --> 00:08:22,480 leadership take regarding its people and specifically when 144 00:08:22,480 --> 00:08:26,600 during such an implementation? So I think first of all you and 145 00:08:26,600 --> 00:08:31,520 I think that's really with every new technology pretty similar. 146 00:08:31,760 --> 00:08:35,799 I think you should think about how do people react to something 147 00:08:35,799 --> 00:08:39,200 new, to new innovation. And there you have the diffusion 148 00:08:39,200 --> 00:08:42,440 model, which is the old model, I think many of us know it, which 149 00:08:42,440 --> 00:08:48,600 says that just perhaps to to, I think 12% of people are 150 00:08:48,600 --> 00:08:52,240 innovators or early adopters. They like every new offering, 151 00:08:52,240 --> 00:08:55,960 every new tool, they like to change, they like to try out new 152 00:08:55,960 --> 00:08:57,920 things. But then you have perhaps the 153 00:08:57,920 --> 00:09:03,280 early maturity, yeah, you know, this typical curve, this normal 154 00:09:03,560 --> 00:09:07,040 distribution, and then you have a late maturity and then you 155 00:09:07,040 --> 00:09:12,040 have laggards. So perhaps just 12% of the whole 156 00:09:12,280 --> 00:09:15,880 base of people of your target group, they will jump on it 157 00:09:16,040 --> 00:09:20,000 absolutely motivated and then the first will question it and 158 00:09:20,000 --> 00:09:23,600 look at the individual benefit, not just for the new or 159 00:09:23,600 --> 00:09:26,720 technology sake. And then perhaps there's even a 160 00:09:26,720 --> 00:09:30,920 larger portion also they are more let's say hesitant or 161 00:09:30,920 --> 00:09:34,800 perhaps even this is critic. So I think this is one point to 162 00:09:34,800 --> 00:09:38,240 look on that and perhaps in in the beginning work with the 163 00:09:38,240 --> 00:09:42,520 innovators and earlier adopters to understand the use cases, how 164 00:09:42,520 --> 00:09:48,560 they use it, how it works. And then, yeah, regarding change 165 00:09:48,560 --> 00:09:53,080 management, I think, I think with every IT software 166 00:09:53,080 --> 00:09:57,240 implementation also if it's a large as Farhana implementation, 167 00:09:57,240 --> 00:09:59,360 there's certain things to consider. 168 00:09:59,800 --> 00:10:05,120 So of course people may need to be aware, yeah, like why the 169 00:10:05,120 --> 00:10:07,920 change is coming. So you need to have a strategy 170 00:10:07,920 --> 00:10:12,120 and narrative, whole story. They need to have somehow a 171 00:10:12,120 --> 00:10:16,000 desire, so something emotional. I think that's important to also 172 00:10:16,000 --> 00:10:17,960 understand your own individual benefit. 173 00:10:17,960 --> 00:10:22,360 Often this is something which is missed because managers perhaps 174 00:10:22,360 --> 00:10:26,640 select the new software due to manager KPIs, but they not 175 00:10:26,640 --> 00:10:31,040 certainly correlate to to, you know, less restoration 176 00:10:31,160 --> 00:10:35,240 expectation, for example. So yeah, employees need to 177 00:10:35,400 --> 00:10:39,600 understand what's in it for them and also let's say, get that 178 00:10:39,600 --> 00:10:43,200 desire, that individual desire, then they need to build up 179 00:10:43,200 --> 00:10:45,720 knowledge here. Is it like how to change or what 180 00:10:45,720 --> 00:10:51,400 is changing in detail, like which of my tasks at my top are 181 00:10:51,400 --> 00:10:55,280 now impacted of AI? And I think of course what's 182 00:10:55,280 --> 00:10:56,920 similar, but it's still different. 183 00:10:56,920 --> 00:10:59,560 They need to the possibility to develop abilities or 184 00:10:59,560 --> 00:11:02,560 competencies. Yeah, like really in the daily 185 00:11:02,560 --> 00:11:07,920 top really now do something different like not write a 186 00:11:07,920 --> 00:11:13,040 resume in HR or write and own let's say block or so in 187 00:11:13,040 --> 00:11:17,520 marketing or e-mail, but use AI tool to support you. 188 00:11:17,720 --> 00:11:20,520 And I think this ability, of course, it's different than the 189 00:11:20,520 --> 00:11:22,920 awareness knowledge or the desire. 190 00:11:23,520 --> 00:11:26,200 And then I think this is very simplified. 191 00:11:26,200 --> 00:11:28,920 Yeah. But I think then if people are 192 00:11:28,920 --> 00:11:32,960 able and they know what's changing and so on, but also 193 00:11:32,960 --> 00:11:36,840 it's always important this this reinforcement that that you 194 00:11:36,840 --> 00:11:41,200 reinforce new behavior because something like what we have 195 00:11:41,200 --> 00:11:44,040 experienced perhaps in in the corona pandemic. 196 00:11:44,040 --> 00:11:47,880 So we tried out new things, but then suddenly if everything is 197 00:11:47,880 --> 00:11:50,560 over, people want to go back to physical events. 198 00:11:50,560 --> 00:11:55,080 They do some managers want 100% people back in the office. 199 00:11:55,280 --> 00:11:59,200 But yeah, yeah, you don't you don't reinforce the good things 200 00:11:59,360 --> 00:12:01,720 or the changes. I think that that's equally 201 00:12:01,920 --> 00:12:03,120 important. Important, Yeah. 202 00:12:03,960 --> 00:12:07,320 So that's very simple and there are different models of 203 00:12:07,320 --> 00:12:09,880 frameworks. Perhaps we can put something in 204 00:12:09,880 --> 00:12:14,800 the show notes, the different categories and and so on, but I 205 00:12:14,800 --> 00:12:17,520 think pretty simplistic. These are some of the change 206 00:12:17,520 --> 00:12:20,080 management considerations. OK. 207 00:12:20,120 --> 00:12:22,280 Yeah, that, that'll sound great if we can get those into the 208 00:12:22,280 --> 00:12:26,560 notes. And I think it first step, the 209 00:12:26,560 --> 00:12:28,720 very first step was, is maybe a willingness to actually 210 00:12:28,720 --> 00:12:30,200 communicate with one's organization. 211 00:12:30,600 --> 00:12:34,000 I think that's often a complaint from many organizations that 212 00:12:34,400 --> 00:12:37,920 management up there in the ivory tower make a decision and they 213 00:12:37,920 --> 00:12:41,320 get the news very late and feel like they didn't really have any 214 00:12:41,320 --> 00:12:44,320 sort of say in the whole thing as this is just happening to 215 00:12:44,320 --> 00:12:46,480 them and it's done to them by the managers. 216 00:12:46,480 --> 00:12:50,560 So best to avoid such situations with effective communication and 217 00:12:50,560 --> 00:12:52,880 identifying all those people along the bell curve that you 218 00:12:52,880 --> 00:12:56,080 mentioned and that communicate with them with a message that 219 00:12:56,160 --> 00:12:58,000 resonates with them and their concerns. 220 00:12:58,960 --> 00:13:02,640 So Speaking of concerns, at the top of the list, I would say for 221 00:13:02,640 --> 00:13:08,080 most I would say anti AI people in the workforce are concerns 222 00:13:08,080 --> 00:13:13,320 about job security. What would be your best way to 223 00:13:13,320 --> 00:13:15,320 address concerns about job security? 224 00:13:15,880 --> 00:13:18,600 I think that's not already going into the implementation 225 00:13:18,600 --> 00:13:22,440 considerations. So I think this concerns of 226 00:13:22,440 --> 00:13:26,960 fears, I think it's fear and resistance, they are often more 227 00:13:26,960 --> 00:13:31,400 diffuse or generic, like even AI will kill us all. 228 00:13:31,400 --> 00:13:35,280 That's super concern that this topic future. 229 00:13:35,920 --> 00:13:40,080 But now some other thing perhaps depending on my job, there will 230 00:13:40,080 --> 00:13:43,920 be some negative changes that I perhaps even will lose my job, 231 00:13:44,120 --> 00:13:47,440 but others perhaps even will think and my job won't be as 232 00:13:47,440 --> 00:13:51,280 important anymore. So they lose status, which is or 233 00:13:51,280 --> 00:13:54,520 can be equally, let's say, you know, fearful. 234 00:13:55,280 --> 00:13:58,320 So I think the best thing of course is, and that's already 235 00:13:58,320 --> 00:14:02,040 key for the implementation that they really look on the use 236 00:14:02,040 --> 00:14:04,400 cases of AI in the end. Yeah. 237 00:14:04,400 --> 00:14:07,200 So it's not this black box strange thing. 238 00:14:07,400 --> 00:14:11,040 But in the end, you need to look what the use cases are affected 239 00:14:11,040 --> 00:14:14,760 or tasks in the end are affected by AI in my daily work. 240 00:14:15,120 --> 00:14:19,280 And also how does this change and we know from earlier, let's 241 00:14:19,280 --> 00:14:25,480 say automation waves, I would say we had I think 567 years 242 00:14:25,480 --> 00:14:30,040 ago, we had already a first AI wave with more manual chat bots 243 00:14:30,040 --> 00:14:34,600 with machine learning coming up. Also there people at fears and 244 00:14:34,600 --> 00:14:38,120 already there what you saw it's not that full top which is 245 00:14:38,120 --> 00:14:43,520 eliminated it's tasks which are perhaps not some I think some 246 00:14:43,520 --> 00:14:48,040 are fully eliminated like typing text from somewhere else. 247 00:14:48,040 --> 00:14:52,640 You now have text recognition and transcription and so on or 248 00:14:52,640 --> 00:14:56,960 translation, but even translation in this example, 249 00:14:56,960 --> 00:15:02,000 it's not fully, Yeah, now done automated you always. 250 00:15:02,000 --> 00:15:06,320 And that's also inside how the AI tools work. 251 00:15:06,440 --> 00:15:11,240 You need to yeah, review it. Yeah, so, but perhaps it's 8090% 252 00:15:11,480 --> 00:15:15,240 some tasks, some tasks are perhaps just 10% impacted. 253 00:15:15,240 --> 00:15:18,840 And I think this task and the lists and also depending on use 254 00:15:18,840 --> 00:15:23,440 case to your daily top, I think this really needs to be done on 255 00:15:23,440 --> 00:15:26,800 the top level or team level. OK. 256 00:15:26,800 --> 00:15:30,480 And just basically letting people know that this is freeing 257 00:15:30,480 --> 00:15:33,200 them up to do other, probably more interesting things 258 00:15:33,200 --> 00:15:36,640 throughout the day and not necessarily replacing them 259 00:15:36,640 --> 00:15:38,600 completely. Absolutely. 260 00:15:38,600 --> 00:15:43,240 And what I said you can't replace many people anyhow 261 00:15:43,280 --> 00:15:46,400 because it's individual tasks which are more or less 262 00:15:46,400 --> 00:15:49,280 automated. And perhaps in some very special 263 00:15:49,280 --> 00:15:52,520 like like what I mentioned, of course, a translator or show on 264 00:15:52,520 --> 00:15:57,480 the list, depending if it's very standardised data, you can to a 265 00:15:57,480 --> 00:16:00,400 high degree automate, but still you need to have the quality 266 00:16:00,400 --> 00:16:04,760 control, then you perhaps can, yeah, translate more or create 267 00:16:04,760 --> 00:16:08,440 more output. And would this also be 268 00:16:08,440 --> 00:16:12,120 convincing to sorts of folks that would say, we've always 269 00:16:12,120 --> 00:16:14,800 done it this way and it works just fine as it is? 270 00:16:14,800 --> 00:16:17,280 So that that's one of the favorite things I've heard from 271 00:16:17,680 --> 00:16:20,840 various parts of my professional experience thus far. 272 00:16:21,840 --> 00:16:24,360 People that just figure why change, why improve? 273 00:16:24,840 --> 00:16:27,360 Would the procedure of convincing them be similar to 274 00:16:27,360 --> 00:16:30,680 the job security folks? I think in the end you need to. 275 00:16:30,680 --> 00:16:33,080 So there are different aspects to that. 276 00:16:33,240 --> 00:16:34,840 Of course. One thing is the individual 277 00:16:34,840 --> 00:16:37,960 benefit that people understand that what's in it for them, what 278 00:16:37,960 --> 00:16:41,840 is changing and how can they cope, how to, let's say, build 279 00:16:41,880 --> 00:16:44,640 up the new tasks, the new skills and so on. 280 00:16:45,120 --> 00:16:47,640 That's one thing. The other thing is, of course, a 281 00:16:47,640 --> 00:16:49,680 change story for the company. Yeah. 282 00:16:49,920 --> 00:16:55,360 What's the overall benefit? So not just perhaps some very, 283 00:16:55,360 --> 00:16:58,560 let's say, rational stuff. I think important is that also 284 00:16:58,560 --> 00:17:01,920 this is connected to somehow emotions person to the company 285 00:17:01,920 --> 00:17:05,400 culture in the best way. For example, what I really 286 00:17:05,400 --> 00:17:08,839 liked, it's not AI, but similar Hugo Boss. 287 00:17:09,240 --> 00:17:13,440 Their, their slogan for the S4 implementation was it's for 288 00:17:13,480 --> 00:17:15,800 hookah boss, like S for hookah boss. 289 00:17:15,920 --> 00:17:19,960 So already it's connected to their, let's say company logo 290 00:17:20,040 --> 00:17:22,680 and claim. And they had their own claim. 291 00:17:22,680 --> 00:17:26,760 They created merch like T-shirts, like own, let's say 292 00:17:26,800 --> 00:17:31,520 marketing again to build up that desire and this emotional thing. 293 00:17:32,440 --> 00:17:37,080 Of course, perhaps I would say 10, 12% of people you anyhow 294 00:17:37,080 --> 00:17:38,960 can't reach because they are negative. 295 00:17:38,960 --> 00:17:43,080 And you need to think in the end, is it the skills or is it 296 00:17:43,080 --> 00:17:45,400 the motivation? If it's the skills, you can help 297 00:17:45,400 --> 00:17:50,120 them or yeah, somehow, if it's a motivation or if they even are 298 00:17:50,120 --> 00:17:53,680 opponent, of course, then it's again a cultural topic. 299 00:17:53,680 --> 00:17:58,120 You do, you just ignore them or you give them other jobs or you 300 00:17:58,120 --> 00:18:00,840 fire them. I think that then, of course, in 301 00:18:00,840 --> 00:18:03,040 every transformation, more cultural topic. 302 00:18:03,800 --> 00:18:06,120 OK. And for new people coming into a 303 00:18:06,120 --> 00:18:09,040 situation like this, so like digital natives, people who 304 00:18:09,040 --> 00:18:12,760 maybe are not so institutionalized and the way 305 00:18:12,760 --> 00:18:15,360 the organization used to be coming in with a fresh mind that 306 00:18:15,360 --> 00:18:19,560 may be more open to an AI adoption situation, but 307 00:18:19,560 --> 00:18:21,560 obviously more junior, maybe they don't have as much 308 00:18:21,560 --> 00:18:24,240 influence. What can those new people do to 309 00:18:24,240 --> 00:18:27,920 help an organization and their and the adoption process and how 310 00:18:27,920 --> 00:18:31,200 can leadership identify those people and and get them to help? 311 00:18:31,800 --> 00:18:34,160 I think in the in the end it's similar, yeah. 312 00:18:34,160 --> 00:18:37,200 So I think they need to be aware, they need to build the 313 00:18:37,200 --> 00:18:41,240 desire, they need to build the knowledge and abilities 314 00:18:42,000 --> 00:18:45,640 regarding their tasks. And I wouldn't say that younger 315 00:18:45,640 --> 00:18:47,840 people, Percy, are more digital fluent. 316 00:18:47,840 --> 00:18:49,840 So that's also, I think a disbelief. 317 00:18:50,160 --> 00:18:54,480 I think they're older people who are also, let's say pretty much 318 00:18:54,680 --> 00:18:57,440 digital fluent or open to new technologies. 319 00:18:58,160 --> 00:19:02,720 But yeah, I think in the end it's also, let's say 320 00:19:02,720 --> 00:19:06,320 motivational topic. If someone you like, let's say a 321 00:19:06,320 --> 00:19:09,600 new coworker is extremely passionate about that. 322 00:19:09,600 --> 00:19:14,640 And even, yeah, builds own apps or whatever, tries regularly new 323 00:19:14,640 --> 00:19:19,040 things then here see that it's perhaps one of the innovators 324 00:19:19,360 --> 00:19:23,120 and you can think about giving them a special job like a change 325 00:19:23,120 --> 00:19:25,840 agent or multiplier and training others. 326 00:19:26,000 --> 00:19:29,720 And I think managers need to of course, recognise this. 327 00:19:29,720 --> 00:19:33,960 I think Speaking of managers, I think for them it's very 328 00:19:33,960 --> 00:19:38,960 important that they not just have some executive, let's say 329 00:19:38,960 --> 00:19:44,280 presentations and high level, let's say cognitive info on what 330 00:19:44,280 --> 00:19:46,080 AI is and what the business case is. 331 00:19:46,080 --> 00:19:48,680 Of course, that's important. They need to understand the use 332 00:19:48,680 --> 00:19:52,960 cases and need to perhaps also prioritize, let's say, which 333 00:19:53,120 --> 00:19:57,320 have the highest value, which doable, let's say this triple 334 00:19:57,320 --> 00:20:00,640 prioritisation criteria. But I think they also should 335 00:20:00,640 --> 00:20:05,440 somehow hands on understand what AI does because otherwise you 336 00:20:05,440 --> 00:20:10,840 know, they they try to talk to experts or to everyone and don't 337 00:20:10,840 --> 00:20:14,840 understand the subjects. And AI, such a disruptive 338 00:20:15,080 --> 00:20:18,760 technology I would say. I think you also should train 339 00:20:18,760 --> 00:20:20,920 managers and they should be enabled. 340 00:20:20,920 --> 00:20:26,080 And what I actually haven't seen yet often is use cases how 341 00:20:26,080 --> 00:20:29,160 managers can improve their own knowledge work. 342 00:20:29,520 --> 00:20:33,160 And honestly, I suspect many managers are still working 343 00:20:33,160 --> 00:20:37,440 mainly on excellent PowerPoint, which is in an AI world, perhaps 344 00:20:37,440 --> 00:20:41,160 not the go to tools anymore. Oh yeah, actually I've 345 00:20:41,160 --> 00:20:45,560 experienced 2 very exciting, let's say fresher new new 346 00:20:45,560 --> 00:20:48,840 formats. So one was a prompt hackathon 347 00:20:49,440 --> 00:20:54,800 where you actually just get minimum input regarding let's 348 00:20:54,800 --> 00:20:59,560 say theory and you have maximum hands on let's say work, for 349 00:20:59,560 --> 00:21:01,840 example. There we formulated different 350 00:21:01,960 --> 00:21:05,120 tasks and especially challenges. We did it for learning 351 00:21:05,120 --> 00:21:10,160 professionals and they worked in five hours in small groups on 352 00:21:10,160 --> 00:21:13,720 the different challenges and then also talk to each other and 353 00:21:13,720 --> 00:21:16,600 share their their learnings. I think that's extremely 354 00:21:16,600 --> 00:21:18,840 helpful. I know for example, SAP 355 00:21:18,840 --> 00:21:23,160 internally, but other companies like Deutsche Telecom in BV, 356 00:21:23,160 --> 00:21:26,440 whoever they do that in different formats, perhaps two 357 00:21:26,480 --> 00:21:29,800 hours or only. But the, the key is really work 358 00:21:29,800 --> 00:21:34,080 on challenges from your daily job and really maximum, let's 359 00:21:34,080 --> 00:21:35,920 say hands on. I think that's, that's a very 360 00:21:35,920 --> 00:21:40,360 cool format because it then let's experience, let's say the 361 00:21:40,360 --> 00:21:42,280 new technology on how to handle it. 362 00:21:42,320 --> 00:21:46,400 I think that's important, but I also did, but I also like over a 363 00:21:46,400 --> 00:21:49,720 longer period of time to work on in a learning group. 364 00:21:49,760 --> 00:21:53,760 I did this with my team and I used that their workbook based 365 00:21:53,760 --> 00:21:57,800 approach like Germany. There's a whole movement of Leno 366 00:21:57,800 --> 00:21:59,920 S guides, a lot of open source guides. 367 00:21:59,920 --> 00:22:01,640 We can put that also in the show notes. 368 00:22:02,120 --> 00:22:05,680 And there you have again, a little bit of input and quite 369 00:22:06,000 --> 00:22:10,160 different kind of tasks or talent just again, and you do 370 00:22:10,160 --> 00:22:14,200 that with a team and I always have some you do meet once a 371 00:22:14,200 --> 00:22:18,040 week, then you do some homework and reflect and then you meet 372 00:22:18,040 --> 00:22:20,400 again. So it's, let's say, over a 373 00:22:20,400 --> 00:22:23,200 longer period of time. And this is, I would say 374 00:22:23,200 --> 00:22:27,800 different iterations. You also very, very much build 375 00:22:27,800 --> 00:22:31,720 up new competencies and yeah, really relate and reflect that 376 00:22:31,720 --> 00:22:34,520 to your daily work. So I think those two elements 377 00:22:34,520 --> 00:22:39,360 are also pretty helpful for hands on more fun learning 378 00:22:39,360 --> 00:22:40,720 approach. OK. 379 00:22:40,800 --> 00:22:43,480 And making it real for for the job and making it real for 380 00:22:43,480 --> 00:22:45,000 everybody. Involved and relevant they're. 381 00:22:45,840 --> 00:22:47,120 Very cool. And they're called Hack. 382 00:22:47,160 --> 00:22:50,360 They're like a hackathon. Right, a hackathon or name, but 383 00:22:50,360 --> 00:22:52,760 you can call it for the whole like OK. 384 00:22:54,960 --> 00:22:56,240 Got it. Cool. 385 00:22:56,560 --> 00:23:00,040 And you mentioned understanding managers and leadership, 386 00:23:00,040 --> 00:23:02,400 understanding the technology a bit better. 387 00:23:03,520 --> 00:23:08,040 Can we maybe go into some specific examples of how this 388 00:23:08,040 --> 00:23:10,440 technology can be applied in an organization? 389 00:23:10,440 --> 00:23:12,920 So we covered sort of like the marketing and communications 390 00:23:13,320 --> 00:23:15,800 part, generating text and stuff like that. 391 00:23:16,160 --> 00:23:21,480 But is there anything else maybe SAP process related that Jewel 392 00:23:21,480 --> 00:23:25,880 is doing in the future that we could use as an example of how 393 00:23:26,560 --> 00:23:28,680 I'm assuming a large language model can automate? 394 00:23:29,120 --> 00:23:33,480 I think that the number one even above HR and the marketing 395 00:23:33,840 --> 00:23:37,200 that's a top functions is everything with software, of 396 00:23:37,200 --> 00:23:40,560 course, yeah, because they're also language models are very 397 00:23:40,560 --> 00:23:44,320 capable. So if you have any task related 398 00:23:44,320 --> 00:23:49,280 software, software development, software support and so on, of 399 00:23:49,280 --> 00:23:52,480 course, there you you will if I have a high value because 400 00:23:52,480 --> 00:23:57,240 there's a high portion or high degree of, let's say, possible 401 00:23:57,240 --> 00:24:00,280 automation. And so like with the GitHub 402 00:24:00,280 --> 00:24:03,720 Copilot or with other tools. And I think that's just the 403 00:24:03,720 --> 00:24:05,080 beginning. Yeah. 404 00:24:05,520 --> 00:24:09,800 So certainly that and then, well, I think there are certain 405 00:24:09,800 --> 00:24:13,720 studies, I don't know it out of my head, of course, everything 406 00:24:13,720 --> 00:24:17,640 around education. Yeah, I think they're special 407 00:24:17,640 --> 00:24:22,120 generative AI is very helpful. Also everything around support, 408 00:24:22,160 --> 00:24:24,880 let's say more the call centre work. 409 00:24:24,920 --> 00:24:31,280 I think they're so, so so the the last generation of let's say 410 00:24:31,280 --> 00:24:35,760 support ticketing and similar tools was very annoying, I would 411 00:24:35,760 --> 00:24:39,400 say if it was AI based. But I think that's now really 412 00:24:39,400 --> 00:24:42,600 getting better. And also the technology on the 413 00:24:42,600 --> 00:24:44,920 one hand, you have the world knowledge in the large language 414 00:24:44,920 --> 00:24:46,800 models. And then you can extend that, 415 00:24:46,800 --> 00:24:50,440 yeah, with grounding and different kind of tools like 416 00:24:50,440 --> 00:24:54,880 vector databases to feed or to. I don't say train because it's 417 00:24:54,880 --> 00:24:59,040 not the training, but to ground your own knowledge into it. 418 00:24:59,040 --> 00:25:03,760 Like, I think this is also very beneficial, yeah. 419 00:25:03,960 --> 00:25:09,000 And can you comment on what SAP specifically is doing in terms 420 00:25:09,000 --> 00:25:12,920 of learning and enablement for AI adoption? 421 00:25:13,680 --> 00:25:15,560 Yeah. So we have different offerings. 422 00:25:15,800 --> 00:25:19,760 We have these free offerings. So especially for SAP experts, 423 00:25:19,760 --> 00:25:23,720 different learning journeys on more on the basics, perhaps we 424 00:25:23,720 --> 00:25:26,240 can research that on some business AI. 425 00:25:26,240 --> 00:25:29,400 They're very worthwhile. They're very good courses on 426 00:25:29,400 --> 00:25:31,560 Essex. I think that's something also 427 00:25:31,560 --> 00:25:35,920 what you need to reflect. So like also see that ties back 428 00:25:35,920 --> 00:25:39,760 to your earlier questions that you have some Essex frameworks 429 00:25:39,760 --> 00:25:43,880 which, for example, describe use cases, what you don't do or 430 00:25:43,880 --> 00:25:48,320 which you usually expect always, for example, if AI is touching 431 00:25:48,320 --> 00:25:52,000 somehow personal data or customer data and so on. 432 00:25:52,280 --> 00:25:54,640 So I think, yeah, so there's this free offering. 433 00:25:54,960 --> 00:25:57,760 Then we have the SAP Learning Hub, yeah, which is our, let's 434 00:25:57,760 --> 00:26:01,320 say, premium learning platform that you have always live 435 00:26:01,320 --> 00:26:04,960 sessions by experts. Also several ones are touching 436 00:26:05,000 --> 00:26:07,840 AI topics. We have the learning systems 437 00:26:07,840 --> 00:26:09,680 where you can try it out hands on. 438 00:26:09,680 --> 00:26:12,720 I think that's important like what I mentioned, so not just 439 00:26:12,720 --> 00:26:17,600 read or hear or see, but also try out, let's say if you and 440 00:26:18,280 --> 00:26:22,440 the training systems, they have real data in in them to 441 00:26:22,440 --> 00:26:24,480 practice. So you don't need to build up 442 00:26:24,480 --> 00:26:28,280 everything by your own. So I think that's certainly 443 00:26:28,320 --> 00:26:30,200 something to to upskill yourself. 444 00:26:30,800 --> 00:26:35,160 I think the end users, I don't know, they perhaps don't need 445 00:26:35,160 --> 00:26:37,880 that much upscaling. They need more, let's say 446 00:26:38,000 --> 00:26:42,040 generic, yeah, skills like prompt engineering, for example, 447 00:26:42,360 --> 00:26:44,800 like the actual tool handling there. 448 00:26:44,800 --> 00:26:48,280 We also have different offerings like SAP Enable now a support 449 00:26:48,280 --> 00:26:51,600 set, which also has no generative AI embedded, for 450 00:26:51,600 --> 00:26:55,240 example, writing a system or you can summarize content and so on. 451 00:26:56,160 --> 00:26:58,040 So that's on the enablement side. 452 00:26:58,240 --> 00:27:00,840 What we also perhaps can put in the show notes. 453 00:27:01,040 --> 00:27:05,520 We have also big toolbox in the road map navigator, which is, 454 00:27:05,520 --> 00:27:08,880 let's say our central tool for everything around our 455 00:27:08,880 --> 00:27:12,200 implementation methodology sub activate. 456 00:27:12,320 --> 00:27:15,600 And that's their own cross topic page on organizational change 457 00:27:15,600 --> 00:27:19,200 management with many different, for example, templates. 458 00:27:19,320 --> 00:27:22,920 And the good thing is that's really aligned to a whole SAP. 459 00:27:23,480 --> 00:27:26,440 Yeah, implementation project like for discover, prepare, 460 00:27:26,440 --> 00:27:28,320 explore to realize, deploy and run. 461 00:27:28,720 --> 00:27:31,760 And they have templates, let's say how to create the change 462 00:27:31,760 --> 00:27:36,360 strategy and vision, communication, enablement and so 463 00:27:36,360 --> 00:27:37,480 on. And monitoring. 464 00:27:37,480 --> 00:27:41,400 I think that's also important. And of course we have next to 465 00:27:41,400 --> 00:27:45,920 this free offerings, we have also services what you can order 466 00:27:46,080 --> 00:27:48,960 if you like. OK, can you repeat the name of 467 00:27:48,960 --> 00:27:51,320 that toolbox again? That'd be definitely should get 468 00:27:51,320 --> 00:27:54,320 down to the show notes. Yeah, it's a road map viewer and 469 00:27:54,320 --> 00:27:57,680 there you have a cost topic page on organizational change 470 00:27:57,680 --> 00:28:00,960 management with mandate template. 471 00:28:01,560 --> 00:28:05,200 What's also very good tool, which you really should think 472 00:28:05,200 --> 00:28:08,800 about using in the very beginning is from the SAP App 473 00:28:08,800 --> 00:28:11,160 house. It's our organization which 474 00:28:11,160 --> 00:28:16,400 supports customers in let's say we are user research and 475 00:28:16,680 --> 00:28:19,480 everything around design, syncing and so on. 476 00:28:19,920 --> 00:28:23,920 And they have toolkit for generative AI exploration 477 00:28:23,920 --> 00:28:29,360 workshop to really find out the use cases, priorise them and 478 00:28:29,360 --> 00:28:31,600 this I think that's even free so. 479 00:28:32,040 --> 00:28:36,000 And I know they are currently working on other offerings 480 00:28:36,000 --> 00:28:38,320 actually already. OK. 481 00:28:38,360 --> 00:28:40,440 Yeah, that's a lot of information that SAP is putting 482 00:28:40,440 --> 00:28:43,600 out. And hopefully we can make the 483 00:28:43,600 --> 00:28:47,400 show notes of this podcast a a central repository of all that 484 00:28:47,400 --> 00:28:49,920 information if if one doesn't already exist out there. 485 00:28:50,760 --> 00:28:52,560 All right, that that covers a lot. 486 00:28:52,560 --> 00:28:55,880 And since we're approaching 30 minutes, I've taken plenty of 487 00:28:55,880 --> 00:28:59,760 your valuable time. So I always want to close on 488 00:28:59,760 --> 00:29:03,400 some final words of wisdom from you, Sir, if you have any words 489 00:29:03,400 --> 00:29:06,480 of wisdom for new people who are entering the workforce in this 490 00:29:06,480 --> 00:29:09,640 time of change and also for those who are maybe already 491 00:29:10,240 --> 00:29:13,040 established and we'll be facing these these changes as they 492 00:29:13,040 --> 00:29:14,240 come. All right. 493 00:29:14,240 --> 00:29:17,960 So, so I think like what we already mentioned, I think it's 494 00:29:17,960 --> 00:29:22,720 important to yeah, leverage change management and yeah, 495 00:29:23,400 --> 00:29:27,480 organizational development, let's say, as part of AI 496 00:29:27,480 --> 00:29:31,640 adoption, because let's say trust is so important there and 497 00:29:31,640 --> 00:29:35,040 perhaps even fear and real assistance are also high to 498 00:29:35,040 --> 00:29:37,320 think about ethics and trust and so on. 499 00:29:37,720 --> 00:29:40,560 I think it's important to be used and human centered. 500 00:29:40,840 --> 00:29:42,640 Yeah. So think about the end users 501 00:29:42,640 --> 00:29:47,080 like leverage, for example, this different powers design thinking 502 00:29:47,080 --> 00:29:50,080 tools. And perhaps to end with some 503 00:29:50,080 --> 00:29:52,360 final words. So I'm a big fan of ESA 504 00:29:52,360 --> 00:29:56,760 Molechia, who is professor at MIT and he just wrote the Co 505 00:29:56,760 --> 00:29:59,760 intelligence and book living and working with AI. 506 00:30:00,280 --> 00:30:05,320 And he has four key rules. Let's say 1 is always invite AI 507 00:30:05,320 --> 00:30:08,000 to the table. I think that's important, yeah. 508 00:30:08,000 --> 00:30:12,080 Like always, whatever you do, like new project, new task, 509 00:30:12,440 --> 00:30:17,080 think about how can an AI help me if it's with pictures, video, 510 00:30:17,240 --> 00:30:21,160 text, whatever or code. So always invite AI to the 511 00:30:21,160 --> 00:30:23,400 table, but be the human in the loop. 512 00:30:23,720 --> 00:30:25,440 I think that's also important. Yeah. 513 00:30:25,440 --> 00:30:28,080 To use empathy, your own knowledge. 514 00:30:28,360 --> 00:30:33,000 We know AI can hallucinate, so we need to double check on Fact 515 00:30:33,000 --> 00:30:36,680 Check content. Then treat AI like a person but 516 00:30:36,680 --> 00:30:38,720 tell it what kind of person it is. 517 00:30:39,200 --> 00:30:42,800 That relates to the whole prompt engineering ways. 518 00:30:42,800 --> 00:30:46,400 So it's important that you breathe their eye with a lot of 519 00:30:46,400 --> 00:30:48,840 context knowledge. And in prompt engineering with 520 00:30:49,120 --> 00:30:52,240 defining prompts you already can do that if you don't want to go 521 00:30:52,240 --> 00:30:55,080 further. So like telling what Pezona you 522 00:30:55,080 --> 00:30:58,000 are, what are the tasks, what does this output and so on. 523 00:30:58,480 --> 00:31:02,040 So treat AI like a person, but tell it what kind of person it 524 00:31:02,040 --> 00:31:05,360 is and assume that's the worst AI you will ever use. 525 00:31:05,640 --> 00:31:09,720 So I think that's, we know now that we have the strawberry 526 00:31:09,720 --> 00:31:13,840 model from open eye. So it's so much fast progressing 527 00:31:13,840 --> 00:31:17,920 then even the entropics models suddenly have become better, at 528 00:31:17,920 --> 00:31:21,520 least I I think then the ones from open my eyes. 529 00:31:21,520 --> 00:31:25,880 So I think it's a real race always gets better and faster. 530 00:31:26,240 --> 00:31:31,240 So I think everyone should start now to deal that and put it on 531 00:31:31,240 --> 00:31:35,480 the table all. Right, some great guidance and 532 00:31:35,480 --> 00:31:37,440 great tips and lots of good information. 533 00:31:37,960 --> 00:31:40,000 Thank you so much for being on the podcast, Thomas. 534 00:31:40,640 --> 00:31:42,640 Thank you. Thanks for the invitation. 535 00:31:42,640 --> 00:31:44,800 Yeah, All the best, everyone. Bye. 536 00:31:44,840 --> 00:31:45,280 Bye.