1 00:00:00,040 --> 00:00:06,040 Are AI agents an enterprise savior, workforce apocalypse, or 2 00:00:06,040 --> 00:00:09,080 just another tech bubble waiting to burst? 3 00:00:09,480 --> 00:00:15,840 Today on CXO Talk episode 876, we cut through the noise with 4 00:00:15,880 --> 00:00:21,320 Bill First, CEO of HFS Research and one of the most respected 5 00:00:21,320 --> 00:00:26,640 industry analysts in the world. When we talk about AI agents and 6 00:00:26,640 --> 00:00:29,440 agentic AI, what what are we actually referring to? 7 00:00:29,440 --> 00:00:32,720 What do we mean by that? It's really the ability to 8 00:00:32,720 --> 00:00:37,040 replicate human behaviour in software. 9 00:00:37,040 --> 00:00:39,400 It's as simple as that. Whether it's mimicking our 10 00:00:39,400 --> 00:00:44,360 voices or supporting us in doing our day-to-day work, It's, it's 11 00:00:44,360 --> 00:00:47,560 really like the augmentation of humanity and software. 12 00:00:47,560 --> 00:00:50,560 And we we talk about the blending of, you know, humans 13 00:00:50,560 --> 00:00:54,200 and technology. This is really where it's at and 14 00:00:54,200 --> 00:00:58,200 it's it's it's something that, you know, we dreamt about for a 15 00:00:58,200 --> 00:01:01,480 long time, but there's only really starting to come into 16 00:01:01,480 --> 00:01:03,960 reality, but at an alarming pace. 17 00:01:03,960 --> 00:01:08,240 I don't know, you know, we see, you know, a lot of fun things 18 00:01:08,240 --> 00:01:11,960 flying around on X and LinkedIn and all these types of things. 19 00:01:11,960 --> 00:01:15,440 It's just incredible, you know, how much development there's 20 00:01:15,440 --> 00:01:19,880 been in voice and video in in just the last six to nine 21 00:01:19,880 --> 00:01:21,760 months. So we're we're going through a 22 00:01:21,760 --> 00:01:25,160 complete revolution and Agentic is right at the front of that 23 00:01:25,160 --> 00:01:30,400 from a technology perspective. Why is agentic AI so important 24 00:01:30,400 --> 00:01:35,680 and at this particular time? I'd like to rewind back to the 25 00:01:35,680 --> 00:01:39,160 early 20 tens when you might have heard of a technology 26 00:01:39,160 --> 00:01:43,160 called RPA, Robotic Process Automation, which got very big 27 00:01:43,160 --> 00:01:46,400 and very height within the technology world. 28 00:01:46,400 --> 00:01:49,160 We actually coined the phrase alongside a company called 29 00:01:49,160 --> 00:01:52,520 Blueprism when we launched it in 2012 and we did the first 30 00:01:52,760 --> 00:01:55,960 analyst papers on it. And at the time we were talking 31 00:01:55,960 --> 00:02:02,520 about RPA replicating human, human behaviour in software, 32 00:02:02,520 --> 00:02:07,000 which would allow us to scale more effectively, threatened 33 00:02:07,200 --> 00:02:10,240 elements like offshore outsourcing, because companies 34 00:02:10,240 --> 00:02:14,080 could technically consider having less offshore resources 35 00:02:14,080 --> 00:02:15,880 when you can automate a lot of this stuff. 36 00:02:16,480 --> 00:02:20,320 But the problem with RPA was the technology didn't scale well. 37 00:02:20,320 --> 00:02:23,520 It was very brittle. But the the concept was there. 38 00:02:23,640 --> 00:02:26,920 But that was really all about following instructions, easy, 39 00:02:27,080 --> 00:02:30,040 easily. It was about eliminating manual 40 00:02:30,040 --> 00:02:33,080 effort waste, you know, which was wasted on repetitive tasks. 41 00:02:33,560 --> 00:02:37,840 Then everyone remembers the influx of gem AI nearly nearly 42 00:02:37,840 --> 00:02:42,640 it's going to be its third year with ChatGPT really came public 43 00:02:43,280 --> 00:02:47,160 nearly three years ago and that really changed the game in terms 44 00:02:47,160 --> 00:02:51,880 of it became the productivity amplifier that accelerates 45 00:02:51,880 --> 00:02:55,400 creative and analytical work that really bottlenecks humans. 46 00:02:55,400 --> 00:02:58,560 It's the ability to create content and this is like one of 47 00:02:58,560 --> 00:03:02,560 the first times we've had non-technical people have that 48 00:03:02,560 --> 00:03:07,200 ability to start to create content, create data, augment 49 00:03:07,200 --> 00:03:09,360 their work, create code. Even. 50 00:03:09,360 --> 00:03:12,240 You know, there's a lot, lot of discussions going on around how 51 00:03:12,240 --> 00:03:14,600 much code can be eliminated now because of Gen. 52 00:03:14,600 --> 00:03:16,880 AI. And that was all about creating 53 00:03:16,880 --> 00:03:20,000 based on prompts. Now we're into the Gen. 54 00:03:20,000 --> 00:03:24,320 AI phase, which is about understanding goals and figuring 55 00:03:24,320 --> 00:03:28,640 out how to achieve them. So Gentic AI is a collaborative 56 00:03:28,640 --> 00:03:33,240 actor that removes the need for constant human oversight of 57 00:03:33,240 --> 00:03:37,080 complex processes. It's self directing in many 58 00:03:37,080 --> 00:03:39,480 respects. It coordinates multiple tasks. 59 00:03:39,480 --> 00:03:44,360 It transforms entire workforces, it creates new organizational 60 00:03:44,360 --> 00:03:47,880 paradigms. But it's not about the fact that 61 00:03:47,880 --> 00:03:50,880 it sounds great. What's exciting about a genetic 62 00:03:50,880 --> 00:03:55,600 is it really does work and it's and it's working at an alarming 63 00:03:55,600 --> 00:03:59,840 pace that is making in, in reality, many people are 64 00:03:59,840 --> 00:04:02,000 comfortable. Some people are loving it and 65 00:04:02,000 --> 00:04:04,600 they're embracing it and they're realizing, wow, I can do my job 66 00:04:04,600 --> 00:04:06,840 so much better. And I'm I'm an analyst. 67 00:04:06,840 --> 00:04:10,200 I can tell you how Jane Turk and Jenna are helping me do my job. 68 00:04:10,840 --> 00:04:14,960 But this is the most, I think, impactful wave in this AI 69 00:04:14,960 --> 00:04:19,240 continuum that takes us to the next phase, which we're terming 70 00:04:19,240 --> 00:04:23,280 artificial general intelligence, which is much more self-directed 71 00:04:23,280 --> 00:04:26,360 intelligence that overcomes human cognitive limitations 72 00:04:26,360 --> 00:04:29,720 across all domains. And eventually, you know, 73 00:04:29,720 --> 00:04:33,280 artificial super intelligence, which is about computers 74 00:04:33,280 --> 00:04:37,000 outperforming humans. We're not there yet, obviously. 75 00:04:37,000 --> 00:04:40,520 But you know, I watched Terminator, Terminator One with 76 00:04:40,520 --> 00:04:43,120 my son the other day. I hadn't seen that in about 30 77 00:04:43,120 --> 00:04:44,440 years. And that brought me back. 78 00:04:45,200 --> 00:04:49,760 They actually predicted in 2029 was when humans, computers 79 00:04:49,760 --> 00:04:52,480 become self aware. So that if anyone's got nothing 80 00:04:52,480 --> 00:04:55,600 better to do this weekend, watch the rerun of Terminator One. 81 00:04:55,600 --> 00:04:59,120 It's uncanny. Oh, they got the timings right 82 00:04:59,120 --> 00:05:01,720 on this thing. So I want to latch onto a 83 00:05:01,720 --> 00:05:06,960 particular point that you made. You said it really works, right? 84 00:05:07,240 --> 00:05:11,280 Can you elaborate on that? Because that's kind of the magic 85 00:05:11,280 --> 00:05:16,920 point, right? This is not a lab based theory, 86 00:05:16,920 --> 00:05:18,400 it's something that actually works. 87 00:05:18,480 --> 00:05:21,560 So tell us about that. Most companies right now have 88 00:05:21,560 --> 00:05:24,000 created some stand alone single agents. 89 00:05:24,000 --> 00:05:28,200 So that's but that's an agent that can handle maybe one 90 00:05:28,200 --> 00:05:31,080 specific task or function. So that could be like an e-mail 91 00:05:31,080 --> 00:05:34,240 writer or even a meeting scheduler, things like that. 92 00:05:34,240 --> 00:05:37,800 Even like copilot, you can use copilot right now to summarize 93 00:05:37,800 --> 00:05:41,120 your emails and remind you to do things. 94 00:05:41,120 --> 00:05:44,080 Or you can use Fireflies, which is a really popular tool, or 95 00:05:44,080 --> 00:05:47,840 LinkedIn, not so much LinkedIn, Zoom AI needed to summarize 96 00:05:47,840 --> 00:05:49,920 meetings. Those are single agents, believe 97 00:05:49,920 --> 00:05:52,400 it or not, and they're already working. 98 00:05:52,400 --> 00:05:55,360 People are already, you know, pretty excited about or did you 99 00:05:55,360 --> 00:05:57,600 turn your fireflies on? So we got a good summary of this 100 00:05:57,600 --> 00:06:00,200 meeting. Where this starts to get really 101 00:06:00,200 --> 00:06:03,760 exciting is when we start to build functional multi agents, 102 00:06:03,760 --> 00:06:07,160 which is multiple agents work together within a single 103 00:06:07,200 --> 00:06:10,960 business function. So that could be a sales team of 104 00:06:10,960 --> 00:06:14,400 agents handling prospecting or qualification and follow-ups, 105 00:06:14,680 --> 00:06:17,520 that sort of thing. And eventually we get to 106 00:06:17,520 --> 00:06:20,520 something we're calling horizontal multi agents, which 107 00:06:20,520 --> 00:06:23,920 is where you get different agents collaborating across 108 00:06:23,920 --> 00:06:27,800 various business functions and and even other supply chain 109 00:06:27,800 --> 00:06:30,640 partners. So that could be sales agents 110 00:06:30,640 --> 00:06:33,760 working with marketing and customer service agents. 111 00:06:33,920 --> 00:06:37,600 So you're actually building out capabilities and business 112 00:06:37,600 --> 00:06:40,120 functions beyond one single function. 113 00:06:40,880 --> 00:06:46,160 It works because you just got to try it like I, I, I'd love to. 114 00:06:46,160 --> 00:06:49,360 There's a demo, it's called Super Film, I think it was where 115 00:06:49,600 --> 00:06:53,520 you could actually put an avatar of me in our research website 116 00:06:53,520 --> 00:06:58,960 and ask me questions and I would literally in my voice dig into 117 00:06:58,960 --> 00:07:02,120 our research and communicate them back to you using using 118 00:07:02,120 --> 00:07:04,360 voice. You just got to see it to see 119 00:07:04,360 --> 00:07:07,320 how effective this is. Now, is it perfect? 120 00:07:07,320 --> 00:07:10,400 No. Is it as accurate as talking to 121 00:07:10,400 --> 00:07:12,640 a human being? Not yet. 122 00:07:13,040 --> 00:07:17,280 But in many respects, we're creating agents that are 123 00:07:17,280 --> 00:07:20,200 becoming very, very supportive in our jobs. 124 00:07:20,200 --> 00:07:23,960 I mean, I'll, I'll tell you, for example, I wrote a piece on 125 00:07:23,960 --> 00:07:27,480 tariffs the other day and I put together here's like big 126 00:07:27,480 --> 00:07:29,720 tournament and things about terrorists and people might have 127 00:07:29,720 --> 00:07:31,920 even read it. And for a bit of fun, I pumped 128 00:07:31,920 --> 00:07:36,360 it through Chatchy, BT pro and and outcomes. 129 00:07:37,680 --> 00:07:39,880 I said, can you pump? Can you can you replicate this 130 00:07:39,880 --> 00:07:42,320 using Phil first voice just for a bit of fun. 131 00:07:42,320 --> 00:07:44,560 And it came out sounding like the sort of thing that I would 132 00:07:44,560 --> 00:07:46,720 have written. And then I asked it to turn it 133 00:07:46,720 --> 00:07:49,160 down a bit, that sort of thing. And then I produced another 134 00:07:49,160 --> 00:07:51,440 piece. Well, I said, can you produce me 135 00:07:51,440 --> 00:07:54,720 a chart that shows life expectancy in the US and versus 136 00:07:54,720 --> 00:07:56,520 other countries and health issues? 137 00:07:56,880 --> 00:07:58,960 And it starts putting information all over the place. 138 00:07:58,960 --> 00:08:01,400 And then you start to train, train the model. 139 00:08:01,400 --> 00:08:03,840 So start starting to become your own personal agent. 140 00:08:03,840 --> 00:08:05,400 And it's getting to know what I need. 141 00:08:05,720 --> 00:08:08,120 And then it's like, can you produce this in the HFS? 142 00:08:08,240 --> 00:08:11,240 Find some colours and you're programming in the colours to 143 00:08:11,240 --> 00:08:13,760 use and everything. So you're really building out 144 00:08:14,360 --> 00:08:18,440 something that can literally become your go to at work, you 145 00:08:18,440 --> 00:08:20,240 know, so there's so many different uses. 146 00:08:20,240 --> 00:08:22,640 And, you know, I don't even know if we're going to call these 147 00:08:22,640 --> 00:08:25,960 agents in another 6 or 12 months, but this is just how 148 00:08:25,960 --> 00:08:29,760 we're partnering with technology now where we don't have to go to 149 00:08:29,760 --> 00:08:31,600 people to get things done all the time now. 150 00:08:31,600 --> 00:08:35,240 We can get so much done ourselves and then we as human 151 00:08:35,240 --> 00:08:38,600 beings become the creators of that content. 152 00:08:39,080 --> 00:08:41,280 Like we say, I've got a big business meeting to go to 153 00:08:41,280 --> 00:08:43,520 tomorrow. You're going to get so much of 154 00:08:43,520 --> 00:08:46,320 content that you need you. You set your agenda to the 155 00:08:46,320 --> 00:08:50,440 technology, it gathers you to what you need and it allows you 156 00:08:50,640 --> 00:08:54,280 to then curate that to make you effective as a human being. 157 00:08:54,720 --> 00:09:00,640 How is this different from having an interactive chat with 158 00:09:00,640 --> 00:09:03,680 Chat GT? So what's unique about agents? 159 00:09:03,920 --> 00:09:07,560 But before you answer, I just want to remind everybody, our 160 00:09:07,560 --> 00:09:11,440 regular listeners know this, that you can ask your questions. 161 00:09:11,440 --> 00:09:15,280 So right now, if you're watching on Twitter, pop your question 162 00:09:15,280 --> 00:09:20,800 using the hashtag into the into Twitter using cxotalk #cxotalk. 163 00:09:21,240 --> 00:09:25,520 If you're watching on LinkedIn, pop your question into the chat. 164 00:09:25,760 --> 00:09:30,240 This is your opportunity to ask one of the top analysts in the 165 00:09:30,240 --> 00:09:32,360 world pretty much whatever you want. 166 00:09:32,360 --> 00:09:35,520 So take advantage of it. And we have some questions that 167 00:09:35,520 --> 00:09:39,320 are coming in now. But first again, so you're 168 00:09:39,320 --> 00:09:42,240 describing an interactive process. 169 00:09:42,240 --> 00:09:46,880 You go to a meeting, you give it notes, you then say modify this 170 00:09:46,880 --> 00:09:49,080 or modify that sounds like ChatGPT. 171 00:09:49,120 --> 00:09:53,440 How are agents different from LLMS and and our usage as we 172 00:09:53,440 --> 00:09:58,120 know and love it today? ChatGPT is useful because you 173 00:09:58,120 --> 00:09:59,960 can use it for a specific prompt. 174 00:09:59,960 --> 00:10:02,960 So I need some information on this or that get some 175 00:10:02,960 --> 00:10:06,760 information quickly produce this, do that, do this, do that. 176 00:10:06,760 --> 00:10:11,400 An actual agent is the virtual Co worker who is completing end 177 00:10:11,400 --> 00:10:14,240 to end processes for you. So it's self directs and 178 00:10:14,240 --> 00:10:18,320 coordinates multiple tasks. So once you've so say I'm using 179 00:10:18,320 --> 00:10:20,960 the example of I could use an agent to help me with my 180 00:10:20,960 --> 00:10:25,480 research. I would develop train this as a 181 00:10:25,480 --> 00:10:28,240 virtual Co worker to be like my research assistant. 182 00:10:28,520 --> 00:10:32,480 So it would start to overtime learn what I do, what I need, 183 00:10:32,480 --> 00:10:34,720 how I do it. So you can start to interact 184 00:10:34,720 --> 00:10:37,160 with this like a virtual Co worker, like a research 185 00:10:37,160 --> 00:10:40,200 assistant, for example. And you can leverage this to, 186 00:10:40,320 --> 00:10:43,040 you know, create whole new organizational paradigms. 187 00:10:43,400 --> 00:10:46,720 I'm not joking. In 12 months time you can say, 188 00:10:46,720 --> 00:10:49,600 hey, they feel I need to have a meeting with you next week to 189 00:10:49,600 --> 00:10:54,160 talk about XYZI can literally have you talk to my agent like 190 00:10:54,160 --> 00:10:57,240 who will look at my diary and coordinate what I need in time 191 00:10:57,240 --> 00:10:58,760 and maybe ask particular questions. 192 00:10:58,960 --> 00:11:04,840 So we are training virtual Co workers to do the jobs that we 193 00:11:04,840 --> 00:11:07,040 either used to do ourselves or someone else did. 194 00:11:07,400 --> 00:11:10,040 And we can start to get into real examples of this. 195 00:11:10,040 --> 00:11:14,680 But the challenge is, you know, going to your staff in another 196 00:11:14,680 --> 00:11:20,000 company and asking them to almost recreate their jobs into 197 00:11:20,000 --> 00:11:23,680 software, which is very different than saying train up 198 00:11:23,680 --> 00:11:27,600 another human being and transfer tasks from yourself to another 199 00:11:27,600 --> 00:11:29,840 human. We're now expecting people, 200 00:11:29,840 --> 00:11:36,200 including ourselves, to transfer human work tasks into software 201 00:11:36,360 --> 00:11:38,680 That technically frees us up to do other things. 202 00:11:38,920 --> 00:11:41,600 Oh, let's be honest, could make us redundant, right? 203 00:11:41,600 --> 00:11:45,040 We're not needed anymore. We can actually leverage agents 204 00:11:45,040 --> 00:11:49,320 to do the jobs of humans. And we're now seeing enterprises 205 00:11:49,760 --> 00:11:52,760 who are really trying to have an AI first mindset. 206 00:11:53,000 --> 00:11:57,920 They're now insisting before you hire any new staff, you need to 207 00:11:57,920 --> 00:12:00,160 show that this work can't be done by AI. 208 00:12:00,720 --> 00:12:02,600 We've reaching that point quite quickly. 209 00:12:03,000 --> 00:12:04,800 This wasn't like maybe 20 years ago. 210 00:12:04,800 --> 00:12:08,720 People used to say, hey, if you hire new staff, can we see if 211 00:12:08,720 --> 00:12:11,400 that work can be done offshore in the Philippines or India or 212 00:12:11,400 --> 00:12:14,000 something? Now C-Suite directives are, can 213 00:12:14,000 --> 00:12:17,840 we not do this with AI? So the whole point of agents a 214 00:12:17,840 --> 00:12:22,240 really this ability for companies to grow and scale in a 215 00:12:22,240 --> 00:12:24,640 way that you don't need to keep adding more and more people. 216 00:12:24,880 --> 00:12:27,880 You do a lot more with the people you have. 217 00:12:28,080 --> 00:12:30,320 And I think that's the positive way to think about this. 218 00:12:30,320 --> 00:12:33,000 It's, you know, I want to run a marketing campaign. 219 00:12:33,000 --> 00:12:38,120 Can I develop a planning agent who can coordinate and breakdown 220 00:12:38,360 --> 00:12:40,800 the campaign requests into specific tasks? 221 00:12:41,120 --> 00:12:44,400 Can I create a research agent that gathers market 222 00:12:44,400 --> 00:12:47,960 intelligence? Can I create a creative agent 223 00:12:48,120 --> 00:12:52,440 that develops, you know, creative assets and messaging? 224 00:12:52,800 --> 00:12:57,160 Can I develop a strategy agent that optimizes my campaign and 225 00:12:57,160 --> 00:12:59,200 engagement across marketing channels? 226 00:12:59,280 --> 00:13:00,720 Right. You can go on and on about 227 00:13:01,480 --> 00:13:04,080 almost every new staff member you need. 228 00:13:04,080 --> 00:13:07,080 You can create an agent for like it could be, hey, I need 229 00:13:07,400 --> 00:13:11,600 somebody to manage social media and I need to do automated 230 00:13:11,600 --> 00:13:13,360 LinkedIn updates, that sort of thing. 231 00:13:13,680 --> 00:13:18,200 Or it could be, you know, even a campaign coordination agent to 232 00:13:18,200 --> 00:13:22,000 synthesise inputs from all agents into a cohesive campaign. 233 00:13:22,000 --> 00:13:26,560 So it's it's creating people into software. 234 00:13:26,560 --> 00:13:29,120 So, you know, we call this thing, you probably heard of 235 00:13:29,120 --> 00:13:32,000 services as software, but this is what's happening in the 236 00:13:32,000 --> 00:13:37,160 services industry right now is companies are starting to think 237 00:13:37,160 --> 00:13:41,480 about how can I replicate the services I'm receiving from an 238 00:13:41,480 --> 00:13:45,320 IBM or an Infosys or one of these companies and receive this 239 00:13:45,760 --> 00:13:50,200 using agent agentic software versus why do I keep having to 240 00:13:50,200 --> 00:13:53,160 add more people all the time. So that's the real nub of 241 00:13:53,520 --> 00:13:57,280 Agentic and I think why it's causing, you know, excitement 242 00:13:57,280 --> 00:14:01,320 and friction at the same time. Go to cxotalk.com, subscribe to 243 00:14:01,320 --> 00:14:05,560 our newsletter and join us. Join our community and join our 244 00:14:05,560 --> 00:14:09,320 live shows. So we have a a very interesting 245 00:14:09,320 --> 00:14:14,760 question coming from Twitter from Anthony Scrifignano, who is 246 00:14:14,760 --> 00:14:19,880 the former Chief data scientist of Dun and Bradstreet and has 247 00:14:19,880 --> 00:14:22,320 been a guest on CXO Talk a number of times. 248 00:14:23,040 --> 00:14:30,120 And he's asking about the unintended risks or unintended 249 00:14:30,280 --> 00:14:36,120 harms that can emerge. Can you talk about that aspect 250 00:14:36,120 --> 00:14:38,320 of it? I think lots of people are are 251 00:14:38,320 --> 00:14:43,880 concerned about the impact of AI agents on the workforce. 252 00:14:43,880 --> 00:14:46,080 You spoke about the positive aspects, but what about the 253 00:14:46,080 --> 00:14:48,920 unintended risks? First thing is you've got the 254 00:14:49,000 --> 00:14:53,440 more general risks of AI. So when you read something sent 255 00:14:53,440 --> 00:14:57,520 to you now a lot of people are thinking was this written by AI 256 00:14:57,520 --> 00:15:01,040 or human being right. That's that's a big problem and 257 00:15:01,040 --> 00:15:03,680 I see that as an opportunity for a research firm like us, because 258 00:15:03,680 --> 00:15:06,360 it's people need real more than ever. 259 00:15:06,520 --> 00:15:12,120 And do you trust information that's all produced using agents 260 00:15:12,120 --> 00:15:15,520 and genetic software? Is it is it truly trustable? 261 00:15:15,520 --> 00:15:20,200 Is it reliable where the source is coming from and that sort of 262 00:15:20,200 --> 00:15:22,640 thing. And I think a bigger risk right 263 00:15:22,640 --> 00:15:27,000 now is can you trust information from people? 264 00:15:28,240 --> 00:15:32,360 The other issues, obviously with interactions and hallucinations 265 00:15:32,360 --> 00:15:34,840 and these types of things, all the types of teething problems 266 00:15:34,840 --> 00:15:37,760 you'll have with honestly any type of technology. 267 00:15:38,000 --> 00:15:40,880 You know, you, I remember when we were getting into more 268 00:15:41,480 --> 00:15:44,160 sophisticated accounting applications 20 years ago when 269 00:15:44,160 --> 00:15:47,720 people worried then about software malfunctioning and 270 00:15:47,720 --> 00:15:50,600 producing, you know, incorrect calculations and stuff like 271 00:15:50,600 --> 00:15:52,400 that. So a lot of it is trusting the 272 00:15:52,400 --> 00:15:56,920 software, trusting the security of that software as well, and 273 00:15:56,920 --> 00:16:01,200 understanding how to navigate your way around this climate 274 00:16:01,200 --> 00:16:03,840 because it's only going to get more confusing and more 275 00:16:03,840 --> 00:16:06,320 worrying. And, you know, three times, you 276 00:16:06,320 --> 00:16:11,360 know, we're getting more sophisticated spamming phishing 277 00:16:11,520 --> 00:16:13,840 stuff or, you know, we're getting them everyday and texts 278 00:16:13,840 --> 00:16:15,640 and emails and quite convincing ones. 279 00:16:15,640 --> 00:16:20,040 Sometimes it's like, you know, I get I get my own staff coming to 280 00:16:20,040 --> 00:16:23,680 me saying, hey, Phil, did you send me this text about getting 281 00:16:23,680 --> 00:16:25,480 Amazon vouchers, things like that. 282 00:16:25,480 --> 00:16:28,920 So we can talk about this for a very long time. 283 00:16:29,120 --> 00:16:31,600 All the different risks, all the different worries, all the 284 00:16:31,600 --> 00:16:36,480 different concerns. And the other thing is, you 285 00:16:36,480 --> 00:16:38,560 know, where do you want it to impact? 286 00:16:38,560 --> 00:16:45,040 So you know, I sat in a room with 10 senior level AI decision 287 00:16:45,040 --> 00:16:50,080 makers in banking just a couple of weeks ago and they had a lot 288 00:16:50,080 --> 00:16:54,160 of common issues, which was we really want to leverage Agantech 289 00:16:54,160 --> 00:16:56,200 to improve our customer experience function. 290 00:16:56,320 --> 00:16:59,080 But you know, it's all about the customer experience with using 291 00:16:59,120 --> 00:17:01,320 banking apps and technology and that sort of thing. 292 00:17:01,760 --> 00:17:04,640 And a lot of their customers, they still want to talk to a 293 00:17:04,640 --> 00:17:06,720 human being, right? Especially when you're getting 294 00:17:06,720 --> 00:17:10,400 into your finances and, and, and, and loans and borrowing and 295 00:17:10,400 --> 00:17:14,560 that sort of thing is, you know, how far do you go before you can 296 00:17:14,560 --> 00:17:18,119 truly trust the technology versus versus the people? 297 00:17:18,119 --> 00:17:20,920 And and I think, I don't think we have a full answer for that 298 00:17:20,920 --> 00:17:23,960 just yet. We have a question from Arsalan 299 00:17:23,960 --> 00:17:29,200 Khan, who says agentic AI requires the correct data at the 300 00:17:29,200 --> 00:17:33,600 right time with the right human and systems integration, 301 00:17:33,840 --> 00:17:37,400 eventually these agents become autonomous. 302 00:17:37,720 --> 00:17:40,920 What happens to humans then? So he's asking about this 303 00:17:40,920 --> 00:17:46,560 boundary between human work and autonomous agent work. 304 00:17:46,960 --> 00:17:50,320 This isn't just about enterprises trying to cut costs 305 00:17:50,320 --> 00:17:54,280 from the place people with bots. This is about us as human 306 00:17:54,280 --> 00:17:56,840 beings. We're all, we're all threatened 307 00:17:56,840 --> 00:17:59,120 by this and we all have opportunities with this. 308 00:17:59,200 --> 00:18:04,120 And if you're in a job where you can be effectively replicated 309 00:18:04,120 --> 00:18:08,080 and replaced, you kind of you kind of know that and you need 310 00:18:08,080 --> 00:18:12,640 to figure out how do I continue to add value in an enterprise? 311 00:18:12,960 --> 00:18:15,320 And I think the value comes from collaboration. 312 00:18:15,320 --> 00:18:17,640 It comes from people skills, it comes from empathy. 313 00:18:18,200 --> 00:18:22,480 And if you can become a great person everybody likes to work 314 00:18:22,480 --> 00:18:25,880 with and you become very thoughtful about what you do and 315 00:18:25,880 --> 00:18:28,960 you start to collaborate beyond your existing area, you become 316 00:18:28,960 --> 00:18:32,040 very valuable to your company. And you know, I can, I can go 317 00:18:32,040 --> 00:18:36,960 through many examples of this. You know, I have a, a guy 318 00:18:36,960 --> 00:18:39,720 running my IT systems who actually was a procurement guy 319 00:18:39,840 --> 00:18:43,880 just a couple, about 3 or 4 years ago, But he, he broadened 320 00:18:43,880 --> 00:18:47,840 his knowledge into understanding how to manage HubSpot and 321 00:18:49,480 --> 00:18:52,560 accounting software. He manages our stack of social 322 00:18:52,560 --> 00:18:53,920 and Grammarly and all this sort of stuff. 323 00:18:54,160 --> 00:18:57,680 And as part of his job, he started to get to know all his 324 00:18:57,680 --> 00:19:00,280 colleagues in different departments in the company, like 325 00:19:00,280 --> 00:19:03,240 analysts and, and, and finance and HR and all this sort of 326 00:19:03,240 --> 00:19:04,680 stuff. And then you start to develop 327 00:19:04,680 --> 00:19:08,040 real value to deliver across your organization. 328 00:19:08,640 --> 00:19:11,240 And I think no matter what role you're in, if you're on a sales 329 00:19:11,240 --> 00:19:14,200 role, you're on a delivery role, you're on a tech role, you need 330 00:19:14,200 --> 00:19:19,160 to become broader and more aligned to your business to add 331 00:19:19,160 --> 00:19:21,840 value there. Because if you if you become 332 00:19:23,000 --> 00:19:29,040 just a replicatable solo task driven professional, you do run 333 00:19:29,040 --> 00:19:32,720 a risk. So you know, you wouldn't 334 00:19:32,720 --> 00:19:35,120 believe some of the conversations I have with CIOs 335 00:19:35,120 --> 00:19:39,000 right now who are under immense pressure to wipe out costs 336 00:19:39,000 --> 00:19:43,040 because of code. One major organization I spoke 337 00:19:43,040 --> 00:19:46,960 to produces half a billion lines of code a year to keep that 338 00:19:46,960 --> 00:19:50,440 organization functioning. And they've been tasked with 339 00:19:50,440 --> 00:19:54,760 eradicating 90% of the effort because you don't need to have 340 00:19:54,760 --> 00:19:59,160 armies of legacy coders anymore. A lot of these code code can be 341 00:19:59,560 --> 00:20:04,320 rewritten using Gen. AI and other types of AI 342 00:20:04,320 --> 00:20:07,560 software now. So, you know, we're, we're just 343 00:20:07,560 --> 00:20:10,280 all facing the challenge of how relevant are we now? 344 00:20:10,280 --> 00:20:15,760 I think you can't replace the humanity and the human ability 345 00:20:15,760 --> 00:20:19,520 to be empathetic. To collaborate, to energize 346 00:20:19,520 --> 00:20:22,120 people and to curate content that is real. 347 00:20:22,120 --> 00:20:26,040 I still believe and I think more than ever, we're going to be hit 348 00:20:26,040 --> 00:20:30,680 with so much AI. Fake information, or could be 349 00:20:30,680 --> 00:20:32,800 real information, but it's written by AI. 350 00:20:32,880 --> 00:20:35,080 We want to read stuff written by people. 351 00:20:35,600 --> 00:20:39,400 AI can help us as humans get much better what we do. 352 00:20:39,400 --> 00:20:42,160 It can help us become better communicators, maybe want more 353 00:20:42,160 --> 00:20:45,280 productive, get more done, like I told you earlier. 354 00:20:45,960 --> 00:20:49,600 And I find I'm becoming way more productive as an analyst because 355 00:20:49,600 --> 00:20:55,880 I've now got, you know, some AI tools which can develop charts, 356 00:20:55,880 --> 00:21:01,240 synthesize data, get me some bits I want so I can, I can 357 00:21:01,240 --> 00:21:06,280 answer my questions. Be specific on how these agents 358 00:21:06,280 --> 00:21:08,440 help you and your job. Tell us the tools you're using, 359 00:21:08,440 --> 00:21:10,680 and then we're going to go back and get some more questions. 360 00:21:10,680 --> 00:21:13,360 Questions are coming in. If you want to develop real 361 00:21:13,360 --> 00:21:15,800 value within your own organization, you have to run 362 00:21:15,800 --> 00:21:19,960 boot camps with your own colleagues to present to each 363 00:21:19,960 --> 00:21:22,360 other how you're using these tools to be more effective at 364 00:21:22,360 --> 00:21:24,760 your job. We've even, we've even hired an 365 00:21:24,760 --> 00:21:28,440 AI expert who's a full time employee within our company. 366 00:21:28,920 --> 00:21:32,120 She's probably listening to this that who's actually working 367 00:21:32,120 --> 00:21:35,960 across our operations people, our analysts, she's working with 368 00:21:36,320 --> 00:21:39,840 Amazon and, and, and a company called Lizier, for example, to 369 00:21:39,840 --> 00:21:43,760 identify how we deliver our research to our clients. 370 00:21:43,760 --> 00:21:46,920 So while yes, I, I can go on about the personal Productivity 371 00:21:46,920 --> 00:21:50,960 Tools I use, we're using agentic to transform our whole business 372 00:21:50,960 --> 00:21:56,000 because we're in the information business and we have set up a 373 00:21:56,000 --> 00:21:59,360 fairly complex system. We're using an agentic solution 374 00:21:59,360 --> 00:22:03,240 called Lizier, which is a, it's a start up, but it's, it's in a 375 00:22:03,240 --> 00:22:06,600 pretty mature phase. They're very popular and they're 376 00:22:06,600 --> 00:22:12,600 powered by AWS to produce at scale the ability for we have 377 00:22:12,600 --> 00:22:18,280 like 150,000 subscribers to go in and create their own research 378 00:22:18,280 --> 00:22:22,440 support agents to help them leverage, get the most out of 379 00:22:22,440 --> 00:22:24,600 HFS. So that's how we're using it 380 00:22:24,600 --> 00:22:26,880 from a corporate standpoint, from a personal standpoint, 381 00:22:27,120 --> 00:22:31,080 right now I use, I'm using chat CPT Pro. 382 00:22:31,560 --> 00:22:35,200 So I paid the extra money. I'm not sure I need the $200 a 383 00:22:35,200 --> 00:22:38,280 month package, but I'm loving it right now because it gives me a 384 00:22:38,280 --> 00:22:42,240 lot of query time. It, the computing power is a 385 00:22:42,240 --> 00:22:44,440 little challenging. Sometimes it takes a bit of time 386 00:22:44,440 --> 00:22:48,480 to produce everything I need. I'm finding that effective. 387 00:22:48,800 --> 00:22:53,240 I'm using deep research from Perplexity, which is pretty good 388 00:22:53,240 --> 00:22:55,120 as well. And I've also been experimenting 389 00:22:55,120 --> 00:23:01,920 rather tools like Claude, which is the anthropic tool, And I've 390 00:23:01,920 --> 00:23:05,640 also looked at some other tools that can be fairly effective, 391 00:23:05,640 --> 00:23:08,200 like Gemini, I'm still not completely convinced by, but 392 00:23:08,200 --> 00:23:11,000 other people love it. So a lot of this is, you know, 393 00:23:11,000 --> 00:23:13,680 people finding technologies that they think are better than 394 00:23:13,680 --> 00:23:16,760 others and they like the way they're interacting with these 395 00:23:16,760 --> 00:23:18,600 tools. But the new, the new suite from 396 00:23:18,600 --> 00:23:21,880 ChatGPT Pro is excellent. You've got the image creation, 397 00:23:21,880 --> 00:23:24,400 you've got the operations piece, you've got the deep research 398 00:23:24,400 --> 00:23:27,200 piece. What I'm seeing right now, this 399 00:23:27,200 --> 00:23:30,400 thing is pretty good and we're going to get to a stage fairly 400 00:23:30,400 --> 00:23:33,760 quickly where we're going to be whittled down to maybe 3 or 4 401 00:23:33,760 --> 00:23:37,120 powerhouses in this space who are going to be dominating the 402 00:23:37,120 --> 00:23:41,200 progression here. I use so many different LLMS, 403 00:23:41,200 --> 00:23:43,720 I'm always experimenting to see which one is better. 404 00:23:43,960 --> 00:23:51,280 Here is a question from Wes Andrews who says you jokingly 405 00:23:51,280 --> 00:23:55,880 referenced Terminator earlier, but given the struggles that 406 00:23:56,560 --> 00:24:01,360 that AI and other sectors are having with establishing 407 00:24:01,360 --> 00:24:07,280 frameworks, guardrail standards such as NIST and GDPR, what do 408 00:24:07,280 --> 00:24:10,640 you suggest? And I'll just mention also to 409 00:24:10,640 --> 00:24:14,760 folks that last week we had two members of the House of Lords 410 00:24:14,760 --> 00:24:16,840 from the UK discussing these issues. 411 00:24:16,840 --> 00:24:20,000 So if you care about these issues, listen to our last show 412 00:24:20,000 --> 00:24:21,640 and you can get the transcript on our site. 413 00:24:21,880 --> 00:24:25,680 But Phil, what what about this this framework and guardrails 414 00:24:25,680 --> 00:24:28,200 set of issues? We look deeply into this because 415 00:24:28,200 --> 00:24:33,680 we cover Global Services a lot with an HFS and every different 416 00:24:33,680 --> 00:24:37,280 region has slightly different attitudes towards AI. 417 00:24:37,960 --> 00:24:41,880 So obviously you mentioned GDPR is, is huge in the UK and 418 00:24:41,880 --> 00:24:45,320 Europe. India is a little bit more of a 419 00:24:45,320 --> 00:24:49,360 free fall right now with how they're accepting AI based 420 00:24:49,360 --> 00:24:54,360 solutions and US, you know, this could be the second coming with 421 00:24:54,360 --> 00:24:58,320 the tech Bros driving a lot of policy here. 422 00:24:58,320 --> 00:25:00,920 So I think we're still waiting to see how a lot of this shapes 423 00:25:00,920 --> 00:25:03,280 up. EU has typically been the most 424 00:25:04,320 --> 00:25:08,640 closed from a framework perspective and demanding in 425 00:25:08,640 --> 00:25:11,440 terms of compliance. And anyone running a business 426 00:25:11,440 --> 00:25:16,760 knows how challenging running GDPR practices has been in 427 00:25:16,760 --> 00:25:19,240 recent years to get to, to the other side. 428 00:25:21,680 --> 00:25:26,520 But I, I do think that as this continues to evolve, the need 429 00:25:26,520 --> 00:25:29,320 for common frameworks is going to become more and more 430 00:25:29,320 --> 00:25:34,840 paramount and the need for cooperation is, is going to 431 00:25:34,840 --> 00:25:37,800 continue to proliferate. They're really doing, look 432 00:25:37,800 --> 00:25:42,200 what's going on politically across the world right now in, 433 00:25:42,240 --> 00:25:45,320 in many ways, this is going to actually bring I think a lot of 434 00:25:45,640 --> 00:25:48,360 regions close to the governments and regions close together and 435 00:25:48,360 --> 00:25:51,920 which may actually drive better cooperation with AI. 436 00:25:52,040 --> 00:25:58,440 So for example, I was hearing today about a strong movement to 437 00:25:58,440 --> 00:26:01,480 create the China less supply chain, right? 438 00:26:01,800 --> 00:26:05,800 So how can countries start to group together to manufacture 439 00:26:05,800 --> 00:26:09,680 goods outside of China to avoid these potential tariffs, right? 440 00:26:10,240 --> 00:26:13,480 And in that case, you need to sort of build a supply chain 441 00:26:13,480 --> 00:26:17,000 competency that sensors and responds, that manages 442 00:26:17,000 --> 00:26:20,640 inventory, that brings cooperation together and these 443 00:26:20,640 --> 00:26:23,560 types of things. So I, I think the need to build 444 00:26:24,200 --> 00:26:27,480 and supply chain standards, trading standards, you know, 445 00:26:27,480 --> 00:26:32,600 around AI, I, I, I think this is just going to, it's only just 446 00:26:32,600 --> 00:26:35,200 beginning and we're going to see a lot more of this emerge in the 447 00:26:35,200 --> 00:26:39,400 next couple of years. What about enterprise adoption? 448 00:26:39,400 --> 00:26:45,080 Where are we today? AI agents are still relatively 449 00:26:45,080 --> 00:26:48,680 new. There's lots of promise, but in 450 00:26:48,680 --> 00:26:52,280 terms of actual usage and enterprise adoption? 451 00:26:52,600 --> 00:26:55,720 I can share the latest and greatest that we've been working 452 00:26:55,720 --> 00:26:59,440 with. We spoke to over 1000 major 453 00:26:59,440 --> 00:27:04,760 enterprises looking at the adoption of of Gen. 454 00:27:04,760 --> 00:27:13,160 AI and the Gen. tech and 45% of them are either worried about 455 00:27:13,160 --> 00:27:18,360 job loss or they're resistant to change and adoption is I'd say 456 00:27:18,360 --> 00:27:21,360 fairly diminished. The other at the end of the 457 00:27:21,360 --> 00:27:29,120 spectrum, only 15% of AI leaders are generally positive about AI 458 00:27:29,120 --> 00:27:34,840 adoption and they have fairly integrated views of where 459 00:27:34,840 --> 00:27:36,560 they're going. They have a strong culture of 460 00:27:36,560 --> 00:27:39,240 support and they're they're embracing this. 461 00:27:39,520 --> 00:27:43,440 And then in the middle, you've got about 40% of enterprises 462 00:27:43,440 --> 00:27:47,840 where they're still in that sort of pilot purgatory phase. 463 00:27:48,640 --> 00:27:50,600 Their culture is becoming more adaptive. 464 00:27:50,600 --> 00:27:53,240 They're recognizing the benefits of AI, but they're not there 465 00:27:53,240 --> 00:27:56,720 yet. So in terms of actual adoption, 466 00:27:56,760 --> 00:28:01,600 you've only got about 15%, maybe a little more, who are getting 467 00:28:01,600 --> 00:28:04,880 to the point where they have a real clear vision and 468 00:28:04,880 --> 00:28:06,160 understanding of where they're going. 469 00:28:07,080 --> 00:28:10,560 One thing that is crystal clear is we're seeing immense pressure 470 00:28:10,560 --> 00:28:16,600 coming from the board level people and also C-Suite leaders 471 00:28:16,600 --> 00:28:20,880 in organizations to drive AI adoption a lot faster. 472 00:28:20,880 --> 00:28:24,720 There's real pressure coming right from the top to really 473 00:28:24,720 --> 00:28:28,280 embrace and become more effective as you know, AI first 474 00:28:28,280 --> 00:28:30,720 cultures. So, but the reality is we're 475 00:28:30,720 --> 00:28:34,640 still at early days, You know, we, we, we've been talking about 476 00:28:34,640 --> 00:28:37,120 this. It's, you know, for a long time, 477 00:28:37,120 --> 00:28:42,720 but the reality is ChatGPT 35 only came in not even 2 1/2 to 478 00:28:42,720 --> 00:28:47,200 three years ago. So we're playing catch up, but 479 00:28:47,200 --> 00:28:49,600 what's happening is the technology is staring it on our 480 00:28:49,600 --> 00:28:54,880 face. It is really here we've got big 481 00:28:54,880 --> 00:28:56,640 firms really trying to get on top of it. 482 00:28:56,640 --> 00:28:59,360 You've got the big software companies like Salesforce in 483 00:28:59,360 --> 00:29:02,680 particular with their Agentforce roll out and service. 484 00:29:02,680 --> 00:29:06,080 Now somebody's business is really trying to muscle in on a 485 00:29:06,080 --> 00:29:09,560 gentic because they see that as an opportunity to take market 486 00:29:09,560 --> 00:29:12,640 share away from services firms. And at the other flip side, 487 00:29:12,640 --> 00:29:16,680 you've got services companies like Accenture really trying to 488 00:29:17,280 --> 00:29:20,160 become more dominant in the services of software realm as 489 00:29:20,160 --> 00:29:23,320 well. So adoption is alone is the is 490 00:29:23,320 --> 00:29:26,880 the real answer to this, but the pressure is there and and it's 491 00:29:26,880 --> 00:29:30,840 on like never before. I just want to invite everybody 492 00:29:30,840 --> 00:29:34,400 listening to join the CXO Talk community. 493 00:29:35,200 --> 00:29:40,240 Go to cxotalk.com and subscribe to our newsletter so we can 494 00:29:40,240 --> 00:29:42,800 notify you of upcoming conversations. 495 00:29:42,800 --> 00:29:47,360 We do this every week and you guys who are listening, you are 496 00:29:47,640 --> 00:29:51,000 the cream of the crop. So join, subscribe to our 497 00:29:51,000 --> 00:29:55,240 newsletter and join these conversations and add your 498 00:29:55,240 --> 00:29:57,360 points of view and your questions. 499 00:29:57,640 --> 00:30:02,440 So we have an important question from Arsalan Khan who says how 500 00:30:02,440 --> 00:30:08,160 do you convince the C-Suite that agentic AI is not just a fancy 501 00:30:08,160 --> 00:30:12,240 chatbot before they move on to the next shiny object? 502 00:30:12,240 --> 00:30:17,320 What are the challenges and the opportunities associated with 503 00:30:17,320 --> 00:30:21,000 this we're. Past that point where C-Suite 504 00:30:21,000 --> 00:30:24,440 can keep denying that this is just another fancy chat bot. 505 00:30:24,440 --> 00:30:32,320 I think if, if you're leading ACX function in particular, you 506 00:30:32,320 --> 00:30:36,200 know, you, you, if you, if you're not familiar with how 507 00:30:36,200 --> 00:30:41,720 easily replicable call centre Asians are with Jane, you know, 508 00:30:41,720 --> 00:30:44,360 with, with smart agents, you know, right now you shouldn't be 509 00:30:44,360 --> 00:30:49,000 in a job anymore. To be quite harsh about it. 510 00:30:49,480 --> 00:30:53,040 I can tell you, you know, just an example of an organization I 511 00:30:53,040 --> 00:30:57,600 spoke to with 50,000 onshore, onshore staff responding to 512 00:30:57,600 --> 00:31:00,640 healthcare inquiry calls and the, the leader basically said, 513 00:31:00,640 --> 00:31:04,360 look, the bottom line is, is there's the same 6 questions 514 00:31:04,360 --> 00:31:06,080 being asked over and over and over again. 515 00:31:06,280 --> 00:31:08,920 We've already run the analysis. We can literally replace half 516 00:31:08,920 --> 00:31:12,680 these people with, with intelligent bots and they call 517 00:31:12,680 --> 00:31:15,920 them, they call them empathy bots very, very quickly. 518 00:31:16,080 --> 00:31:18,440 We're not going to do it straight away, but we know the 519 00:31:18,440 --> 00:31:22,160 possibility is there. And I think this is this is a 520 00:31:22,160 --> 00:31:25,120 typical case across a lot of companies is they're very aware 521 00:31:25,120 --> 00:31:28,160 of what they can do, but they they're still yet to have that 522 00:31:28,160 --> 00:31:30,720 burning trigger platform to go do it. 523 00:31:31,040 --> 00:31:35,840 My concern is if we plunge into, you know, a, a deep recession, 524 00:31:37,080 --> 00:31:39,760 you're going to see some organizations literally come out 525 00:31:39,760 --> 00:31:42,600 and say, we're just going to start relying a lot more on AI 526 00:31:42,600 --> 00:31:43,960 and we're going to let people go. 527 00:31:44,080 --> 00:31:49,000 So my vain hope is we don't fall into recession so we can have a 528 00:31:49,000 --> 00:31:52,000 more positive view of people and technology. 529 00:31:52,000 --> 00:31:56,880 But there is that risk there that a negative economy can 530 00:31:56,880 --> 00:32:01,480 drive a lot more weaponized AI where companies would say, look, 531 00:32:01,480 --> 00:32:03,800 well, let's just replace these people, we don't need them 532 00:32:03,800 --> 00:32:07,360 anymore. So I, I don't think companies I 533 00:32:07,360 --> 00:32:11,520 speak to are not aware of this. It's more how advanced they are 534 00:32:11,520 --> 00:32:13,920 with embracing this. Are they prepared to do 535 00:32:13,920 --> 00:32:16,120 anything? And my concern is I do talk to a 536 00:32:16,120 --> 00:32:18,320 lot of enterprises. We have a lot of summits and 537 00:32:18,320 --> 00:32:21,960 round tables on top of our research where, you know, people 538 00:32:21,960 --> 00:32:23,720 want to talk. But when it comes down to what 539 00:32:23,720 --> 00:32:26,080 are you actually doing, They're not doing, They're not doing a 540 00:32:26,080 --> 00:32:27,240 lot. And I think what I just said to 541 00:32:27,240 --> 00:32:32,800 about the 15%, that's not a big #15% are kind of on the path. 542 00:32:33,280 --> 00:32:36,760 The rest are either still figuring it out or they're not 543 00:32:36,760 --> 00:32:39,080 on the path. And and this is just going to 544 00:32:39,080 --> 00:32:42,880 become more pronounced as we go through the next few months of 545 00:32:43,120 --> 00:32:47,880 macroeconomic turbulence. You just made wild comment, 546 00:32:48,160 --> 00:32:51,960 which is, and I don't want to put words in your mouth, but it 547 00:32:51,960 --> 00:32:56,920 seems like you just said that the technology is becoming so 548 00:32:56,920 --> 00:33:06,280 good at sufficient number of use cases that an economic downturn 549 00:33:06,280 --> 00:33:12,400 can push many companies to replace many workers because 550 00:33:12,400 --> 00:33:17,040 those use cases and the effectiveness are so broadly 551 00:33:17,040 --> 00:33:21,120 dispersed even even today or if not today soon enough. 552 00:33:21,400 --> 00:33:24,040 The technology is available. It's there. 553 00:33:24,720 --> 00:33:26,200 I think companies are aware of it. 554 00:33:26,880 --> 00:33:30,160 I do believe as well most enterprises don't tread lightly 555 00:33:30,160 --> 00:33:35,640 on the fact that, hey, let's go replace 5000 people with, you 556 00:33:35,640 --> 00:33:44,240 know, 1000 or 500, you know, or get augmented consultants who 557 00:33:44,240 --> 00:33:48,360 can manage a team of bots. But one of the things that has 558 00:33:48,360 --> 00:33:54,520 been looked at in industry is why do you need 500 people in 559 00:33:54,520 --> 00:33:59,200 India, for example, running a bunch of coding or app support, 560 00:33:59,200 --> 00:34:02,800 that sort of thing, when you can potentially replace them with a 561 00:34:02,800 --> 00:34:07,760 team of maybe 25 people who are local and onshore who are 562 00:34:08,120 --> 00:34:10,520 supported and augmented by Gentic technology. 563 00:34:10,840 --> 00:34:16,040 So it's this ability to reduce the scale of people numbers that 564 00:34:16,040 --> 00:34:20,639 you have and they'll augment higher value, you know, people 565 00:34:21,239 --> 00:34:24,040 with, with the Gentic to support them. 566 00:34:24,040 --> 00:34:27,719 And you know, we, we put out some research recently around, 567 00:34:28,880 --> 00:34:32,719 you know, the impact of tariffs, for example, that could have a 568 00:34:32,719 --> 00:34:39,360 real impact on what we call it could drive the whole services 569 00:34:39,360 --> 00:34:42,719 and software adoption curve, right? 570 00:34:42,719 --> 00:34:46,120 Because suddenly it's like if it becomes really difficult to 571 00:34:46,120 --> 00:34:49,920 manage a disparate global workforce, manufacturing goods 572 00:34:50,080 --> 00:34:52,880 and all over the world, you need to bring stuff back home. 573 00:34:54,639 --> 00:34:59,040 You know, suddenly, hey, I can actually do what I need in the 574 00:34:59,320 --> 00:35:02,160 US with a smaller number of staff. 575 00:35:02,240 --> 00:35:05,320 They might be more expensive, but I don't need as many. 576 00:35:05,320 --> 00:35:07,240 And they're supported by this technology. 577 00:35:07,560 --> 00:35:10,480 So we are at a point where companies are starting to make 578 00:35:10,480 --> 00:35:14,880 much more radical assumptions on what they can do. 579 00:35:15,160 --> 00:35:18,440 You may have seen a recent announcement from the bank Citi, 580 00:35:18,440 --> 00:35:24,240 Citibank, who have decided to reduce their 144 service 581 00:35:24,240 --> 00:35:26,000 provider relationships down to 50. 582 00:35:26,320 --> 00:35:29,760 And they've actually increased the numbers of staff that they 583 00:35:29,760 --> 00:35:34,000 have on shore in the United States and some other regions 584 00:35:34,040 --> 00:35:36,120 who are directly within the company. 585 00:35:36,120 --> 00:35:39,960 Because they what they want to do is they want to spend less on 586 00:35:39,960 --> 00:35:43,880 the legacy and more on the new. So I'm not trying to say 587 00:35:43,880 --> 00:35:45,960 companies are just going to fire everybody or replace them with 588 00:35:45,960 --> 00:35:49,640 bots, but I think a lot of smart businesses are thinking, how do 589 00:35:49,640 --> 00:35:54,640 we stop spending billions of dollars on maintaining legacy 590 00:35:54,640 --> 00:35:59,320 applications and legacy systems when we really want to reinvest 591 00:35:59,320 --> 00:36:03,000 that money in modernized thinking, modernized agentic 592 00:36:03,000 --> 00:36:06,200 technology, that sort of thing. So what some companies are 593 00:36:06,200 --> 00:36:09,600 doing, and I use the example of Citibank is they're trying to 594 00:36:09,600 --> 00:36:14,840 stop the cost of the old so then they can bring you back in house 595 00:36:14,840 --> 00:36:18,680 and then start to think about how do we reinvest in and the 596 00:36:19,000 --> 00:36:21,120 technology they need to take them to a different place. 597 00:36:21,280 --> 00:36:23,440 So I don't think companies are thinking right now about how do 598 00:36:23,440 --> 00:36:25,800 we just get rid of people. They're actually thinking about 599 00:36:25,800 --> 00:36:30,880 how do we break from the past. I did a great podcast with Jason 600 00:36:30,880 --> 00:36:35,080 Aberbrook, who's one of the leading minds in HR technology. 601 00:36:35,120 --> 00:36:37,880 You know, he, he's a Mercer these days and he talks there's 602 00:36:37,880 --> 00:36:41,640 like this CHROS across all the big global 50 companies. 603 00:36:41,640 --> 00:36:46,440 And he, he actually came out and said these companies have so 604 00:36:46,440 --> 00:36:48,200 much data. They, they don't want to do with 605 00:36:48,200 --> 00:36:49,400 it. They can't join it up. 606 00:36:49,400 --> 00:36:53,120 They can't make decisions on it. It's got to the point where he's 607 00:36:53,120 --> 00:36:56,440 got clients who are literally thinking, oh, just just get, 608 00:36:56,440 --> 00:37:01,120 let's just trash this old systems and rebuild, rebuild 609 00:37:01,120 --> 00:37:03,160 with the new. And I think this is where some 610 00:37:03,160 --> 00:37:05,560 of these conversations are happening with a Gentech, which 611 00:37:05,560 --> 00:37:09,040 is how do we start to really build out the new and and make a 612 00:37:09,040 --> 00:37:11,960 break from from the legacy that's been holding us back for 613 00:37:11,960 --> 00:37:15,360 so long. Anthony Scrifignano makes a 614 00:37:15,360 --> 00:37:19,480 comment on Twitter directly addressing this point that you 615 00:37:19,480 --> 00:37:23,200 were just discussing. He says it's equally likely that 616 00:37:23,200 --> 00:37:29,560 the C-Suite is being taken to task for not adopting more to 617 00:37:29,720 --> 00:37:33,480 drive down cost. He says the KPIs need to be more 618 00:37:33,480 --> 00:37:37,680 than just cost savings. What new problems are being 619 00:37:37,680 --> 00:37:43,200 addressed that were unaddressed before being enabled by this 620 00:37:43,200 --> 00:37:45,960 technology? And it sounds like you're saying 621 00:37:46,200 --> 00:37:50,120 the same thing, that cost savings is a part of it, but 622 00:37:50,120 --> 00:37:52,720 there's also a whole set of new opportunities. 623 00:37:53,080 --> 00:37:58,280 I would agree that the same fundamental issues have remained 624 00:37:58,280 --> 00:38:05,360 for a very long time in terms of changing we, we call it, you 625 00:38:05,360 --> 00:38:08,400 know, paying off your debts, your technical debts, your 626 00:38:08,400 --> 00:38:12,000 people debt, your process debt, your data debt within companies 627 00:38:12,000 --> 00:38:15,240 and, and, and, and this inertia of companies refusing to change. 628 00:38:15,240 --> 00:38:19,680 And there's so many managers and leaders within enterprise who, 629 00:38:19,920 --> 00:38:22,640 let's be honest, have got away with not having to do much 630 00:38:22,640 --> 00:38:24,400 different for the last 20-30 years. 631 00:38:24,400 --> 00:38:27,400 I mean, we still have companies operating the processes that 632 00:38:27,840 --> 00:38:30,080 were designed before the Second World War, some even the 633 00:38:30,080 --> 00:38:33,640 industrial revolution. So what is different this time? 634 00:38:33,640 --> 00:38:39,480 I think what's different is the technology is much more 635 00:38:39,480 --> 00:38:42,680 pronounced. It's much more ready, it's much 636 00:38:42,680 --> 00:38:47,520 more scalable. And there's a final exhaustion 637 00:38:47,520 --> 00:38:51,360 where you talk to CIOs off record, they'll all tell you one 638 00:38:51,360 --> 00:38:54,000 thing. They are fed up spending 10% a 639 00:38:54,000 --> 00:38:57,240 year on their services firms and then 10% increase is every year 640 00:38:57,240 --> 00:39:01,480 on their software license hikes. SAS is becoming a legacy 641 00:39:01,480 --> 00:39:04,840 paradigm and services people just don't want to keep paying 642 00:39:04,840 --> 00:39:07,280 more and more and more. You can't keep going up this 643 00:39:07,280 --> 00:39:10,960 exponential cost curve. Eventually the chickens come 644 00:39:10,960 --> 00:39:14,800 home to roost and, and, and I think C-Suite executives are 645 00:39:14,800 --> 00:39:19,080 really being held to task now. Can you deliver an AI first 646 00:39:19,080 --> 00:39:23,680 organization where a culture has to change within the company? 647 00:39:23,680 --> 00:39:25,640 And I think that's the problem we've got with a lot of these 648 00:39:25,640 --> 00:39:29,360 businesses is they haven't got the right cultures to shift, 649 00:39:30,040 --> 00:39:31,840 shift forward and really embrace. 650 00:39:31,840 --> 00:39:34,880 And you know, while I would agree, I don't think the 651 00:39:34,880 --> 00:39:37,080 fundamental issues have changed all that much. 652 00:39:37,720 --> 00:39:41,920 What is changing is the onus on AI that's coming right from the 653 00:39:41,920 --> 00:39:44,880 top. Because when RPA came in, in 654 00:39:44,880 --> 00:39:48,920 2012, the reason why one of the reasons it failed was the CIO. 655 00:39:48,960 --> 00:39:53,000 It would get dumped on the CI OS docket and it would eventually 656 00:39:53,000 --> 00:39:55,920 get dumbed down two or three layers into the what we call the 657 00:39:55,920 --> 00:39:57,840 frozen middle within the organization. 658 00:39:57,840 --> 00:40:01,280 And that's when technology solutions go to die. 659 00:40:01,920 --> 00:40:03,840 That's not happening so much with Agentic. 660 00:40:04,160 --> 00:40:10,080 But Phil, I remember those RPA days and I remember software 661 00:40:10,080 --> 00:40:14,920 companies describing RPA just like you're talking about 662 00:40:14,920 --> 00:40:19,800 agentic AI right now, which is it's going to save us money. 663 00:40:19,800 --> 00:40:22,720 We're not going to need as many employees. 664 00:40:23,160 --> 00:40:25,760 This is going to, it's going to be great. 665 00:40:27,320 --> 00:40:29,280 But the promise was never fulfilled. 666 00:40:29,280 --> 00:40:34,360 So what's different and how do we see our way through the hype? 667 00:40:34,480 --> 00:40:39,480 I think what is different is that 15% of high performers and 668 00:40:39,640 --> 00:40:44,160 I think the following 15% behind are organizations where the 669 00:40:44,160 --> 00:40:47,400 leadership have realized they can no longer keep painting lip 670 00:40:47,400 --> 00:40:52,360 service to not fixing their underlying problems with data 671 00:40:52,520 --> 00:40:56,600 technology and legacy. And all our PA was really doing 672 00:40:57,080 --> 00:41:01,040 was it was like a Band-Aid tech that stitched together old 673 00:41:01,040 --> 00:41:03,600 systems to get them functioning more effectively. 674 00:41:03,600 --> 00:41:06,920 It was very useful. But in terms of could you use 675 00:41:06,920 --> 00:41:10,800 RPA to replace thousands of people unless it was a very high 676 00:41:10,800 --> 00:41:13,840 throughput process, very repeatable, very predictable. 677 00:41:13,840 --> 00:41:16,400 Of course you couldn't. No, this was like a patchwork 678 00:41:16,400 --> 00:41:18,640 technology. You, you know, if you want to, 679 00:41:18,640 --> 00:41:24,000 if you want to say, hey, we have 1000 people answering calls in 680 00:41:24,000 --> 00:41:27,560 the Philippines for our consumer products that we're selling, 681 00:41:27,760 --> 00:41:31,400 right? That's people on mass at scale 682 00:41:31,480 --> 00:41:35,400 where you need technology that can actually have some empathy 683 00:41:35,400 --> 00:41:40,240 with clients, that can replicate CX behaviour, that can actually 684 00:41:40,240 --> 00:41:42,040 do the job. And I think that's the big 685 00:41:42,040 --> 00:41:45,720 difference right now is agentic is much, much closer to doing 686 00:41:45,720 --> 00:41:49,200 the job of human beings than RPA was, which it really wasn't. 687 00:41:49,200 --> 00:41:53,480 It was a patchwork back office break fix technology that was 688 00:41:53,480 --> 00:41:57,440 great if you wanted to keep, you know, your old kicks mainframe 689 00:41:57,440 --> 00:42:01,880 working with a cobalt system, working with ASAP system, for 690 00:42:01,880 --> 00:42:04,800 example. But now it's much more, you 691 00:42:04,800 --> 00:42:07,160 know, you can see where this is all shifting. 692 00:42:07,160 --> 00:42:10,560 And I think there's a, there's a real exhaustion with companies 693 00:42:10,560 --> 00:42:15,640 having to keep maintaining real, you know, creaking old systems 694 00:42:16,360 --> 00:42:18,920 in a world where if, you know, competition is much more 695 00:42:18,920 --> 00:42:22,720 cutthroat and you got to be really slick and on the ball if 696 00:42:22,720 --> 00:42:24,640 you're going to be effective in this economy. 697 00:42:25,000 --> 00:42:30,240 Phil, I get the sense that what you're really also saying is 698 00:42:30,240 --> 00:42:37,720 that the the difference between RPA and agentic AI is that that 699 00:42:37,840 --> 00:42:44,800 15% of early adopters of agentic AI have demonstrated that in 700 00:42:44,800 --> 00:42:48,720 fact it really does work. It really can bring these kinds 701 00:42:48,800 --> 00:42:52,160 of benefits and savings that you've been describing. 702 00:42:52,560 --> 00:42:59,440 Yeah, you can actually create a virtual Co worker to complete 703 00:42:59,440 --> 00:43:02,200 end to end processes. It's proven it works. 704 00:43:02,240 --> 00:43:05,720 We've all seen the demos, we've worked with companies who are 705 00:43:05,720 --> 00:43:07,800 piloting it, we have done it on ourselves. 706 00:43:07,800 --> 00:43:11,360 And a lot of enterprises, more advanced ones in particular are 707 00:43:11,360 --> 00:43:14,440 at least working with single agents and some move into multi 708 00:43:14,440 --> 00:43:17,120 agent models. So they're on the path. 709 00:43:17,240 --> 00:43:22,240 And it, it's a different type of technology that removes the need 710 00:43:22,240 --> 00:43:24,880 for constant human oversight of complex processes. 711 00:43:25,040 --> 00:43:29,200 It's a transformational tool rather than a task automation 712 00:43:29,200 --> 00:43:31,520 tool, which RPA was. That was about tasks. 713 00:43:31,520 --> 00:43:37,600 This is about human oversight, support and real process 714 00:43:37,600 --> 00:43:42,400 capability and the fact that you can build these Co workers. 715 00:43:42,400 --> 00:43:44,720 You can, you can engage with these things, you can talk to 716 00:43:44,720 --> 00:43:47,720 them, right? I don't even want to get into 717 00:43:47,920 --> 00:43:51,720 some disturbing things about teenage boys building 718 00:43:51,720 --> 00:43:54,880 relationships with AI girlfriends and things like 719 00:43:54,880 --> 00:43:55,680 that. I don't know if you've been 720 00:43:55,680 --> 00:43:58,840 reading about some of these things, but this, this stuff is 721 00:43:58,840 --> 00:44:00,040 real. People are building 722 00:44:00,040 --> 00:44:01,680 relationships with their software. 723 00:44:02,880 --> 00:44:05,760 The software, you know, you can ask the question, if you ring up 724 00:44:05,760 --> 00:44:08,840 customer service today, do you care that you're dealing with a 725 00:44:08,840 --> 00:44:10,440 computer or dealing with a human being? 726 00:44:10,840 --> 00:44:14,920 When you're go checking into your airline, do you really want 727 00:44:14,920 --> 00:44:17,320 to talk to the gate agent? No, of course you don't. 728 00:44:17,320 --> 00:44:19,600 You just want to use your app and get on the plane, you know 729 00:44:19,600 --> 00:44:22,080 what I mean? So we're getting to this whole 730 00:44:22,080 --> 00:44:26,640 next layer of, you know, technology becoming part of our 731 00:44:26,640 --> 00:44:29,960 daily lives much more than ever before, to the point where we're 732 00:44:29,960 --> 00:44:32,200 actually engaging with technology in a much more 733 00:44:32,520 --> 00:44:36,840 humanistic real way. And I don't know if you saw the 734 00:44:38,120 --> 00:44:42,120 CEO of Google Deepmine the other day, He was, I dig this out, I'd 735 00:44:42,120 --> 00:44:46,840 saw it on X this morning. But he was saying how the new 736 00:44:46,840 --> 00:44:51,760 way to develop code is inviting creators and people with 737 00:44:52,200 --> 00:44:54,760 heuristic creative skills to develop the code. 738 00:44:54,760 --> 00:44:58,120 Rather than in the old days you were going to like a computer 739 00:44:58,120 --> 00:45:00,960 science engineer and having to kind of explain in a very clunky 740 00:45:00,960 --> 00:45:03,800 way, this is what we need. We're getting to the point where 741 00:45:03,800 --> 00:45:07,600 we can create code and we can create applications without 742 00:45:07,600 --> 00:45:10,680 being technically proficient. And one of the things where you 743 00:45:10,680 --> 00:45:12,800 still want to talk about the difference between agentic and 744 00:45:12,800 --> 00:45:18,000 RPA is agentic is the first time ever really that we can take 745 00:45:19,040 --> 00:45:23,520 non-technical C-Suite or leaders within enterprises and have them 746 00:45:23,520 --> 00:45:25,440 dictate what they want from their technology. 747 00:45:25,880 --> 00:45:29,240 But we are seeing technology that is pivoted towards the 748 00:45:29,240 --> 00:45:32,680 business professional and we're already in a situation where I 749 00:45:32,680 --> 00:45:37,440 think 46% of IT decisions are made outside of the CI OS walls 750 00:45:37,720 --> 00:45:42,520 of their offices. This is the age where the CFO, 751 00:45:42,520 --> 00:45:45,560 the head of supply chains, the head of marketing, these people 752 00:45:45,560 --> 00:45:48,400 are making their own technology choices because they can start 753 00:45:48,400 --> 00:45:53,840 to build technology that is very much answering the needs of the 754 00:45:53,840 --> 00:45:59,320 business versus something that you're having to be dictated to 755 00:45:59,320 --> 00:46:01,120 by engineers, that sort of thing. 756 00:46:01,520 --> 00:46:07,920 It is extraordinary the level of research support that, for 757 00:46:07,920 --> 00:46:11,200 example, that these tools can provide. 758 00:46:11,640 --> 00:46:18,280 I had a networking issue of my own here in our studio and doing 759 00:46:18,280 --> 00:46:21,440 a little bit of research I was able to figure out a fairly 760 00:46:21,440 --> 00:46:24,160 complex question having to do with routing. 761 00:46:24,440 --> 00:46:27,520 Rather than need to call an IT person and bring a consultant 762 00:46:27,520 --> 00:46:29,560 in. It is amazing, but we have 763 00:46:29,560 --> 00:46:32,320 we're, we're almost out of time and we have a number of 764 00:46:32,320 --> 00:46:35,360 questions that are left. So I'm going to ask you, Phil, 765 00:46:35,360 --> 00:46:39,760 to answer these questions pretty quickly, pretty concisely. 766 00:46:40,240 --> 00:46:46,120 First one is from Prem Kumar Aparanji and he says when LLMS 767 00:46:46,120 --> 00:46:51,080 powering the AI agents aren't reliable or predictable, how do 768 00:46:51,080 --> 00:46:56,760 we rely on them to automate unpredictable scenarios that 769 00:46:56,760 --> 00:47:00,120 need to work? You have to train the model to 770 00:47:00,120 --> 00:47:03,040 work is is my answer. So if there's something wrong 771 00:47:03,040 --> 00:47:09,120 with the LLMS, then you need to really have a look at the 772 00:47:09,120 --> 00:47:12,440 underlying technology that you're using and find the right 773 00:47:12,440 --> 00:47:16,160 LLMS that can deliver the scenarios you want. 774 00:47:16,160 --> 00:47:21,240 So I think there's a lot more technology based conversations 775 00:47:21,240 --> 00:47:25,200 we got to have to get this, you know, and really enterprise 776 00:47:25,200 --> 00:47:28,160 grade ready. And what I would say is, you 777 00:47:28,160 --> 00:47:30,560 know, I know from a lot of friends in the industry that 778 00:47:30,560 --> 00:47:34,200 like open AI, for example, is very, very obsessed with 779 00:47:34,200 --> 00:47:38,800 becoming enterprise ready. Like, you know, you know, the 780 00:47:38,800 --> 00:47:42,400 leadership within that company are spending all their time with 781 00:47:42,520 --> 00:47:45,400 the C-Suite within Fortune to 500. 782 00:47:45,480 --> 00:47:48,240 They're trying to figure it out. So I would say it's great 783 00:47:48,240 --> 00:47:52,480 question and there are a lot of faults in the system right now. 784 00:47:52,840 --> 00:47:56,880 And a lot of this is honing the models and training the models 785 00:47:56,880 --> 00:47:59,560 until you get them working. I mean, as I said, we're doing 786 00:47:59,560 --> 00:48:02,080 our own model, we're putting our whole business into into an 787 00:48:02,080 --> 00:48:05,760 agentic solution. And it's it's taken us three 788 00:48:05,760 --> 00:48:07,960 years to get to a point where we still haven't gone live with the 789 00:48:07,960 --> 00:48:10,080 new system yet. But you've got to learn 790 00:48:10,080 --> 00:48:11,600 yourself, you've got to learn your business. 791 00:48:11,600 --> 00:48:12,920 You've got to learn these models. 792 00:48:12,920 --> 00:48:15,520 You've got to try it and try it and try until you know what 793 00:48:15,520 --> 00:48:19,040 works and what doesn't. And I remember when shortly 794 00:48:19,040 --> 00:48:24,160 after ChatGPT came out, I remember that your company HFS 795 00:48:24,160 --> 00:48:27,920 Research was one of the first analyst firms that I was aware 796 00:48:27,920 --> 00:48:32,800 of that was making that attempt to put your research online into 797 00:48:32,800 --> 00:48:36,520 an LLM. So you, you truly are an early 798 00:48:36,520 --> 00:48:39,560 adopter at this. We have a question very quickly 799 00:48:39,560 --> 00:48:41,960 now because we're just going to run out of time from Elizabeth 800 00:48:41,960 --> 00:48:47,560 Shaw, who says CE OS and boards are driving the use of AI 801 00:48:47,560 --> 00:48:50,720 agentic and beyond. There's serious implications for 802 00:48:50,720 --> 00:48:54,440 worker and social society impacts. 803 00:48:54,680 --> 00:49:00,080 What concerns do CE OS and other senior business leaders have 804 00:49:00,680 --> 00:49:03,200 with these concerns? And very quickly, please, 805 00:49:03,200 --> 00:49:04,400 please. I know it's a complicated 806 00:49:04,400 --> 00:49:07,640 question. Data privacy and cyber are the 807 00:49:07,640 --> 00:49:12,360 number one problems and biggest concerns by country mile, to be 808 00:49:12,360 --> 00:49:16,200 honest with you. And after that, it's, you know, 809 00:49:16,200 --> 00:49:19,240 it, it's, it's other areas around transformation and 810 00:49:19,240 --> 00:49:21,240 replacing process and compliance. 811 00:49:21,520 --> 00:49:26,200 But cyber is by far and away I think the biggest headache as 812 00:49:26,200 --> 00:49:29,280 companies look at shifting to these models is, is maintaining 813 00:49:29,280 --> 00:49:32,920 a secure infrastructure. Arslan Com comes back and says 814 00:49:34,040 --> 00:49:37,880 who gains the most value from magentic AI, small companies or 815 00:49:37,880 --> 00:49:40,720 large companies? I would say at the moment small 816 00:49:40,720 --> 00:49:46,160 companies, it allows I, I hate using my own example, but it 817 00:49:46,160 --> 00:49:49,760 allows mid sized businesses to really punch above their weight 818 00:49:49,760 --> 00:49:54,080 because you can scale fast, you can act nimbly and you often 819 00:49:54,080 --> 00:49:57,280 don't have as much legacy within the business to change. 820 00:49:57,280 --> 00:49:59,280 You don't have as many people resisting change. 821 00:49:59,280 --> 00:50:03,640 And I think large companies can also be truly beneficial in 822 00:50:03,640 --> 00:50:05,600 terms of how they can leverage this. 823 00:50:05,600 --> 00:50:09,360 But I just found with a lot of large businesses, it's harder 824 00:50:09,360 --> 00:50:12,200 for them to beset their legacy. You know, look at the technical 825 00:50:12,200 --> 00:50:16,280 debt they have the the lock in they've got with legacy software 826 00:50:16,280 --> 00:50:17,800 providers, you know, that sort of thing. 827 00:50:17,800 --> 00:50:20,040 So I think it's harder for large companies to change because 828 00:50:20,360 --> 00:50:25,600 there's a huge amount of training and, and, and cultural 829 00:50:25,600 --> 00:50:27,480 change and shifting that needs to happen. 830 00:50:27,480 --> 00:50:32,200 And I think SM ES a better place to pivot and and I see a lot of 831 00:50:32,200 --> 00:50:34,600 people I know wanting to go and work for smaller businesses 832 00:50:34,600 --> 00:50:37,560 'cause they are more nimble and you got to be nimble in this 833 00:50:37,560 --> 00:50:40,440 market. What advice do you have for 834 00:50:40,440 --> 00:50:45,720 enterprise, technology and business leaders when it comes 835 00:50:45,720 --> 00:50:52,120 to how they should be relating to this agentic AI world today 836 00:50:52,120 --> 00:50:54,880 and very quickly, please? Get on top of it, learn it, 837 00:50:54,880 --> 00:50:58,520 understand it, experiment with it, do boot camps with it. 838 00:50:59,440 --> 00:51:03,480 You've got to educate yourself. The days of BS ING around 839 00:51:03,480 --> 00:51:06,320 technology are over. You've got to be much more 840 00:51:06,320 --> 00:51:11,480 proficient at knowing what is possible and engaging and 841 00:51:11,480 --> 00:51:14,960 building relationships with the whole emerging AI ecosystem 842 00:51:14,960 --> 00:51:18,720 around you, you know? And with that, a huge thank you 843 00:51:18,720 --> 00:51:21,960 to Phil. First, he's the CEO of HFS 844 00:51:22,000 --> 00:51:24,320 Research. Phil, thank you so much for 845 00:51:24,320 --> 00:51:26,000 being here. I'm just so grateful to you. 846 00:51:26,400 --> 00:51:28,000 Yeah, a pleasure, Pleasure, Michael. 847 00:51:29,000 --> 00:51:32,720 Now I went quickly, enjoyed it very much and I look forward to 848 00:51:33,120 --> 00:51:35,400 more interactions. And a huge thank you to 849 00:51:35,400 --> 00:51:39,440 everybody who is watching today. You guys are an awesome 850 00:51:39,440 --> 00:51:41,840 audience. You're so intelligent, so smart. 851 00:51:42,000 --> 00:51:46,200 Go to cxotalk.com, subscribe to our newsletter and join us. 852 00:51:46,200 --> 00:51:50,120 Join our community and join our live shows. 853 00:51:50,120 --> 00:51:53,720 We have one next week and the week after that, so check it 854 00:51:53,720 --> 00:51:55,040 out. Thanks so much everybody, and I 855 00:51:55,040 --> 00:51:56,960 hope you have a great day and we'll see you next time.