1 00:00:10,000 --> 00:00:12,400 Hello, and welcome to the Retail Podcast. 2 00:00:12,680 --> 00:00:15,320 Now today we're going to do a bit of a deep dive in an area 3 00:00:15,320 --> 00:00:18,040 that has had a lot of press coverage. 4 00:00:18,480 --> 00:00:21,840 But translating press coverage into the reality for retailers 5 00:00:21,840 --> 00:00:25,640 sometimes is half the trick. I'm joined by Rob Mckendrick 6 00:00:25,640 --> 00:00:28,720 from REO Blue. Rob, why don't you tell us a 7 00:00:28,720 --> 00:00:32,000 little bit about yourself, your background, a little bit about 8 00:00:32,000 --> 00:00:34,680 what Oreo Blue is, and then we'll come back to the 9 00:00:34,680 --> 00:00:37,480 conversation around AI. Sure. 10 00:00:37,480 --> 00:00:39,200 Great. And thanks for having me on. 11 00:00:39,640 --> 00:00:42,520 Yeah, So I'm Robin Kendrick. I work for Harry Blue. 12 00:00:42,520 --> 00:00:46,400 We're a consultancy in data and AI and my role is field Chief 13 00:00:46,400 --> 00:00:50,520 data officer. So I help our customers in terms 14 00:00:50,520 --> 00:00:54,320 of understanding data strategy, how they're using data and how 15 00:00:54,320 --> 00:00:57,240 to, you know, do things safely and in a positive way. 16 00:00:57,240 --> 00:00:59,400 A lot of that comes from a background of working at the 17 00:00:59,400 --> 00:01:02,920 Co-op for six years. We're at a variety of of data 18 00:01:02,920 --> 00:01:06,160 roles in ethics, data governance and engineering. 19 00:01:06,440 --> 00:01:08,880 So yeah, so really excited to be doing this with you. 20 00:01:09,400 --> 00:01:12,240 Oh fantastic. I know Aria Blue released a 21 00:01:12,240 --> 00:01:15,760 report on Responsible AI. For those who've not had a 22 00:01:15,760 --> 00:01:19,360 chance to read it or unaware about, you know, what it means, 23 00:01:19,680 --> 00:01:24,480 What is responsible AI and why? Why are you guys focusing on it? 24 00:01:24,480 --> 00:01:26,800 Yeah, thanks. So I mean, put really simply, it 25 00:01:26,800 --> 00:01:30,680 is what it what it says, it's about using AI in a responsible 26 00:01:30,680 --> 00:01:35,280 way, which doesn't just mean from an ethical point of view. 27 00:01:35,280 --> 00:01:38,400 It can also mean from a business point of view, environmental 28 00:01:38,400 --> 00:01:40,840 point of view. Obviously it's become really 29 00:01:41,560 --> 00:01:45,000 talked about as you said in the press because of generative AI 30 00:01:45,000 --> 00:01:48,280 and chat bots. But it also applies to things 31 00:01:48,280 --> 00:01:50,840 such as machine learning algorithms that you might use 32 00:01:50,840 --> 00:01:55,080 for optimization or pricing. And I think that, I mean, you've 33 00:01:55,080 --> 00:01:58,240 only got to look recently right at dynamic pricing, which is a 34 00:01:58,240 --> 00:02:01,760 great algorithm for Uber, you know, where you've got an 35 00:02:01,760 --> 00:02:04,640 increase in demand means an increase in price, which means 36 00:02:04,640 --> 00:02:07,560 an increase in supply because more drivers comes that area. 37 00:02:07,560 --> 00:02:11,240 But obviously say, let's say concert ticketing, it's not 38 00:02:11,600 --> 00:02:14,040 doesn't work so well. So it looks a bit like you're 39 00:02:14,040 --> 00:02:17,280 using, well, it looks like a bit like being opportunistic and 40 00:02:17,280 --> 00:02:19,760 greedy. Everybody's blaming is on an 41 00:02:19,760 --> 00:02:22,360 algorithm and saying, well, it was the algorithm that said, 42 00:02:22,640 --> 00:02:24,960 let's let's double or triple the price of these seats. 43 00:02:24,960 --> 00:02:27,560 So I think I think it's really important from that point of 44 00:02:27,560 --> 00:02:31,560 view of understanding how you're using some of these techniques 45 00:02:31,560 --> 00:02:33,240 and being mindful in the way you do that. 46 00:02:33,560 --> 00:02:36,800 It's interesting, I mean, because I've seen grocers do 47 00:02:36,800 --> 00:02:40,320 this with users that are on applications and even at 48 00:02:40,320 --> 00:02:43,880 conferences, I've heard people say, you know, if you're 49 00:02:43,880 --> 00:02:47,480 offering everyone 10% discount, why you're offering your loyal 50 00:02:47,480 --> 00:02:52,440 customers who would spend the money anyway, another 10%, 51 00:02:52,440 --> 00:02:54,800 right? And so that was like, OK, I 52 00:02:54,800 --> 00:02:58,760 don't know how it's business. And so there's this sort of line 53 00:02:59,120 --> 00:03:05,040 between business operations and, I don't know, profit margins and 54 00:03:05,040 --> 00:03:08,120 then what's the right thing to do or you know, what your 55 00:03:08,120 --> 00:03:10,080 customers would actually want you to do. 56 00:03:10,320 --> 00:03:11,760 I think that's the way to look at it. 57 00:03:11,960 --> 00:03:13,640 What surprised when you looked at the report? 58 00:03:13,720 --> 00:03:16,960 I, I think firstly, a lot of, a lot of people are waiting for 59 00:03:16,960 --> 00:03:21,400 the government to say something about AI specifically and 60 00:03:21,400 --> 00:03:24,400 feeling like until that happens, they don't need to worry about 61 00:03:24,400 --> 00:03:26,800 it. But you know, CMA have announced 62 00:03:26,800 --> 00:03:29,880 a, a review of what happened with Ticketmaster. 63 00:03:30,040 --> 00:03:33,680 If you, if you use a fuse data badly, use personal data badly, 64 00:03:33,680 --> 00:03:35,920 you can fall, fall foul of the ICO. 65 00:03:36,040 --> 00:03:40,040 So, so you know you can fall foul of GDPR, right? 66 00:03:40,200 --> 00:03:42,520 So that that's a really easy thing to do. 67 00:03:43,000 --> 00:03:46,760 As you say, if you use pricing, you can get into a situation of 68 00:03:46,760 --> 00:03:50,920 accidentally introducing bias where you might give discounts 69 00:03:50,920 --> 00:03:55,120 to one particular set of people versus another based upon 70 00:03:55,360 --> 00:03:59,360 something like gender or race that that might be a byproduct 71 00:03:59,360 --> 00:04:02,880 of your products or the way you intend to do the pricing. 72 00:04:02,880 --> 00:04:07,920 But you know, it's all those kind of things that people can 73 00:04:07,920 --> 00:04:12,880 get in trouble for, can get prosecuted for, but it doesn't 74 00:04:12,880 --> 00:04:15,880 rely on a new law or regulation coming out. 75 00:04:15,880 --> 00:04:18,240 It's just part of the existing regulations. 76 00:04:18,240 --> 00:04:21,480 And I think so the first thing is people are kind of waiting 77 00:04:21,480 --> 00:04:25,080 for the government to say this is what you can do and before 78 00:04:25,080 --> 00:04:26,600 they make a change. And I think you really need to 79 00:04:26,600 --> 00:04:30,280 make that change now, I think. I think secondly, a lot of 80 00:04:30,280 --> 00:04:35,360 people think about the bad consequences of using AI as 81 00:04:35,360 --> 00:04:39,560 being something just in terms of brand reputation or say 82 00:04:39,560 --> 00:04:42,800 regulation. But a badly trained model might 83 00:04:43,360 --> 00:04:47,000 introduce pricing reductions which are unsustainable for your 84 00:04:47,000 --> 00:04:50,120 business. It could predict, you know, you 85 00:04:50,120 --> 00:04:54,720 could have an AI algorithm in in supply chain, which might over 86 00:04:54,720 --> 00:04:58,240 purchase something because it's got some bad data or the models 87 00:04:58,240 --> 00:05:02,480 badly trained. So there's potential downside 88 00:05:02,480 --> 00:05:05,400 financial there. And also, you know, isn't it's 89 00:05:05,400 --> 00:05:10,720 just not cheap, right? So a large language model, which 90 00:05:10,720 --> 00:05:15,760 might give you a few percent uplift on a chat bot might cost 91 00:05:15,760 --> 00:05:18,080 you five times as much as the alternative approach. 92 00:05:18,080 --> 00:05:20,920 So all of those things aren't just about, you know, 93 00:05:20,920 --> 00:05:22,960 regulation. They're also about the bottom 94 00:05:22,960 --> 00:05:25,200 line to, to grocers and retailers. 95 00:05:25,400 --> 00:05:29,520 And the third thing I think is that people assume that it might 96 00:05:29,520 --> 00:05:32,920 slow them down. And I, I always use the analogy 97 00:05:32,920 --> 00:05:35,480 of if you've got a, you know, racing car with a really good 98 00:05:35,480 --> 00:05:38,240 brakes and steering and go a lot faster because you've got the 99 00:05:38,240 --> 00:05:40,920 safety built in. So I, I think a lot of people 100 00:05:40,920 --> 00:05:44,920 worry that if they implement something around being mindful 101 00:05:44,920 --> 00:05:47,800 and responsible with AI that, that that might slow them down, 102 00:05:47,800 --> 00:05:50,520 slow their innovation. But I think it really opens up 103 00:05:50,520 --> 00:05:54,440 the ability to do things and do things purposefully and, and do 104 00:05:54,440 --> 00:05:56,640 them quickly and well. Who would you normally be 105 00:05:56,640 --> 00:05:59,800 presenting this to? Who in the business should care 106 00:05:59,800 --> 00:06:00,920 about this? Yeah. 107 00:06:00,920 --> 00:06:03,120 So I think that's a really interesting thing. 108 00:06:03,120 --> 00:06:05,680 So I think there are a lot of people that care a little bit 109 00:06:05,680 --> 00:06:07,720 about it. If you talk to say head of 110 00:06:07,720 --> 00:06:10,600 compliance, they're they've got to worry about AI. 111 00:06:11,080 --> 00:06:13,880 Head of data protection would be worried about it. 112 00:06:13,880 --> 00:06:16,040 Data scientists, they're kind of worried about it. 113 00:06:16,040 --> 00:06:19,280 So you've got quite a lot of people who have a concern but 114 00:06:19,280 --> 00:06:21,520 don't feel they're empowered to do something about it or don't 115 00:06:21,520 --> 00:06:23,400 feel it's necessarily their scope. 116 00:06:23,840 --> 00:06:27,560 So when we talk to people about implementing something around 117 00:06:27,760 --> 00:06:31,040 responsible AI, we say what you've got to do is you're going 118 00:06:31,040 --> 00:06:33,280 to take this top down. You've got to talk to the 119 00:06:33,280 --> 00:06:36,720 executive when you've got to set some AI principles that really 120 00:06:36,720 --> 00:06:40,320 set the tone of of how you're going to use AI for an 121 00:06:40,320 --> 00:06:43,120 organisation. So, you know, go into the 122 00:06:43,120 --> 00:06:47,960 C-Suite, talk about, talk about some of these opportunities, 123 00:06:47,960 --> 00:06:53,120 potential downsides and, you know, make sure it's up to them 124 00:06:53,120 --> 00:06:55,640 to show the responsibility and accountability within their 125 00:06:55,640 --> 00:06:58,720 organisation. So put in place some principles, 126 00:06:59,520 --> 00:07:03,680 put in place some cheques and measures and and then maybe 127 00:07:03,680 --> 00:07:06,520 enable those heads of compliance, heads of data 128 00:07:06,520 --> 00:07:09,920 science to do the right thing in their organisation. 129 00:07:09,920 --> 00:07:11,920 But you know, it really needs to come down from the top. 130 00:07:12,000 --> 00:07:15,760 Have you seen that in action like in do retailers do that or 131 00:07:15,760 --> 00:07:18,960 grocers? Are they effectively doing that? 132 00:07:18,960 --> 00:07:21,680 Are they trying to to do? That the the research we 133 00:07:21,680 --> 00:07:24,040 conducted said there were certainly very few are talking 134 00:07:24,040 --> 00:07:26,920 about having done that, right. Yeah. 135 00:07:26,920 --> 00:07:30,960 So it, it was literally sort of, you know, one or two out of 40 136 00:07:30,960 --> 00:07:34,080 retailers that we looked at in the UK have said something about 137 00:07:34,080 --> 00:07:38,080 that as a Co-op where I used to work. 138 00:07:38,080 --> 00:07:41,240 We, we worked quite a lot on that, but it was part of our 139 00:07:41,240 --> 00:07:45,640 whole environmental social purpose of the Co-op to do 140 00:07:45,640 --> 00:07:48,880 things responsibly. So we consider it partly a good 141 00:07:48,880 --> 00:07:51,440 thing to do. And the, the retailers we're 142 00:07:51,440 --> 00:07:56,040 working with at the moment, it's putting this together as a bit 143 00:07:56,040 --> 00:08:00,520 of a new thing for them. So it's something that they 144 00:08:00,520 --> 00:08:03,640 might have been doing, but it's something that they kind of do 145 00:08:03,640 --> 00:08:05,400 it in pocket. So they, they haven't really got 146 00:08:05,400 --> 00:08:09,040 their hands around it. Last year, Global Data Science 147 00:08:09,040 --> 00:08:11,720 from Unilever talked about the fact they've been looking at at 148 00:08:11,720 --> 00:08:15,920 this worldwide and they they'd assessed 40 of their AI 149 00:08:16,120 --> 00:08:19,040 algorithms for being responsible huge amount. 150 00:08:19,200 --> 00:08:22,600 They had 400 to go. That AI was so prevalent in 151 00:08:22,600 --> 00:08:24,960 their organisation that, you know, it was literally the tip 152 00:08:24,960 --> 00:08:26,200 of the iceberg that they'd looked at. 153 00:08:26,760 --> 00:08:31,240 But if I'm not a top three retailer in the UK, right, I 154 00:08:31,240 --> 00:08:36,760 probably won't have that depth of field or that I'm relying on 155 00:08:36,760 --> 00:08:39,159 other people to help me. So where do I start? 156 00:08:39,159 --> 00:08:43,799 Maybe is the better question. Yeah, I, I think so. 157 00:08:43,919 --> 00:08:47,840 So when you start from, from that, you know, let's assume 158 00:08:47,840 --> 00:08:49,960 you've done the C-Suite conversation and you said, OK, 159 00:08:49,960 --> 00:08:53,320 we've got, we've got in place our, we're going to place our 160 00:08:53,320 --> 00:08:54,800 principles. We want to do something about 161 00:08:54,800 --> 00:09:00,560 it. I think creating a creating your 162 00:09:00,560 --> 00:09:04,120 understanding, your list of all the applications you might 163 00:09:04,120 --> 00:09:07,480 already have or systems, you know, because you might be, you 164 00:09:07,480 --> 00:09:10,920 might have, you know, be using Microsoft and you go, oh, 165 00:09:10,920 --> 00:09:13,000 switched on Co pilot. That's cool. 166 00:09:13,160 --> 00:09:18,120 You know, you've got AI there. Most of the ERP systems and 167 00:09:18,120 --> 00:09:20,560 finance systems are claiming they've got some kind of AI 168 00:09:20,560 --> 00:09:22,120 capability. So you've got something there 169 00:09:22,120 --> 00:09:23,800 that's, you know, already in place. 170 00:09:23,800 --> 00:09:26,840 So understanding that to begin with and saying, what are we 171 00:09:26,840 --> 00:09:30,600 using? And then looking at them and 172 00:09:31,760 --> 00:09:34,440 from a risk perspective and say, what are the potential things 173 00:09:34,440 --> 00:09:36,800 that could go wrong with some of these things? 174 00:09:37,080 --> 00:09:41,760 What would we, what do we need to put in place to investigate 175 00:09:41,760 --> 00:09:44,640 them more fully? And then you can start to work 176 00:09:44,640 --> 00:09:47,400 back if you've got, if you've got a system that's, you know, 177 00:09:47,400 --> 00:09:52,920 where we've built a, we've built a brand new system to give 178 00:09:52,920 --> 00:09:54,880 offers to customers and it's based on AI. 179 00:09:56,120 --> 00:09:59,720 Do you know if the data going into that model is correct? 180 00:10:00,600 --> 00:10:03,440 And what kind of ways can the data be incorrect that would 181 00:10:03,440 --> 00:10:05,320 that would cause that model to go badly wrong? 182 00:10:06,080 --> 00:10:08,080 And then you can sort of kind of trace that back. 183 00:10:08,240 --> 00:10:13,360 And and as you say, not everybody has the money of a 184 00:10:13,360 --> 00:10:15,400 Unilever. So what things can you do? 185 00:10:15,840 --> 00:10:18,920 You can do things like use the algorithm, but make sure you've 186 00:10:18,920 --> 00:10:21,480 got some testing around it as the algorithm rolls out. 187 00:10:21,480 --> 00:10:26,000 You know, we used AI solutions to work out that the Co-op what, 188 00:10:26,000 --> 00:10:28,480 what bread to bake and what cookies to break bake. 189 00:10:28,760 --> 00:10:31,560 But there was a Baker who looked at that and said, well, that's, 190 00:10:31,560 --> 00:10:34,480 that's clearly on because I'm not going to do exactly that, 191 00:10:34,480 --> 00:10:37,680 but I'm, I'm mindful. So you call that human in the 192 00:10:37,680 --> 00:10:41,640 loop in, you know, in the same way that, you know, in, in 193 00:10:41,640 --> 00:10:46,440 banking, you would hope that your application for a credit 194 00:10:46,440 --> 00:10:48,760 card might have somebody looking at it at some point. 195 00:10:48,840 --> 00:10:52,160 You know, if it's on the margin, you might go, OK, is there 196 00:10:52,160 --> 00:10:54,800 somebody looking at that? And in the same way you you want 197 00:10:54,800 --> 00:10:57,200 to see if there are any high risk algorithms, whether you've 198 00:10:57,200 --> 00:11:00,600 got a human in the loop who can, you can look at those things. 199 00:11:00,600 --> 00:11:02,600 Is that what term? Is that a Rob term? 200 00:11:02,600 --> 00:11:03,640 I love that term. That's a. 201 00:11:03,680 --> 00:11:06,680 That's a. That's AAI term. 202 00:11:07,160 --> 00:11:09,800 In the, in the, in the, in the know they know, right? 203 00:11:10,000 --> 00:11:12,680 Yeah, yeah, yeah, exactly. So it's like, yeah, but it but 204 00:11:13,040 --> 00:11:16,040 it's a good one because it's like unlike a lot of the other 205 00:11:16,040 --> 00:11:18,840 techno jargon that that actually tells you what it is right there 206 00:11:19,040 --> 00:11:22,760 in that making such a. Futuristic term. 207 00:11:22,840 --> 00:11:25,360 I can imagine that term. It's the first time I'm hearing 208 00:11:25,360 --> 00:11:27,960 it, so I can imagine that term, you know, was there a human in 209 00:11:27,960 --> 00:11:29,480 the loop? I was like, Oh yeah, right. 210 00:11:29,480 --> 00:11:31,720 We, we didn't have a human in that loop because it was just 211 00:11:31,720 --> 00:11:35,360 all process, which I guess then leads into I, I think what 212 00:11:35,360 --> 00:11:37,960 you're talking about and what the report was, is about your AI 213 00:11:37,960 --> 00:11:40,480 framework. So what, what are the pillars to 214 00:11:40,480 --> 00:11:42,360 the framework? Was that what you were just 215 00:11:42,360 --> 00:11:44,720 covering there? So yeah, that that was that was 216 00:11:44,720 --> 00:11:48,080 around the whole AI safety pillar. 217 00:11:48,080 --> 00:11:51,760 So, So what we, what we've done is we've built it up from the 218 00:11:51,760 --> 00:11:54,040 core of data governance and architecture. 219 00:11:54,040 --> 00:11:59,720 So do you understand your data? Do you understand how it got 220 00:11:59,720 --> 00:12:02,280 into the system? Do you understand how it was 221 00:12:02,280 --> 00:12:04,440 collected? So that's the first, first item. 222 00:12:04,600 --> 00:12:06,240 You think of it like expanding out, right? 223 00:12:06,440 --> 00:12:08,240 So that's, that's the first, that's the first part. 224 00:12:08,240 --> 00:12:11,640 The second part is looking at safety, which is where the human 225 00:12:11,640 --> 00:12:16,120 loop comes in and understanding how the system is built, what 226 00:12:16,120 --> 00:12:19,160 tools you're using and having those cheques and balances. 227 00:12:19,160 --> 00:12:21,520 You've got AI safety as the next thing. 228 00:12:22,360 --> 00:12:25,560 Building out from that, you can start to look at biosynthetics. 229 00:12:25,640 --> 00:12:29,680 So, you know, just because something is doable, should we 230 00:12:29,680 --> 00:12:33,760 do it which can which, you know, as I mentioned at the beginning, 231 00:12:33,760 --> 00:12:37,280 that could be something around it's going to cost quite a lot 232 00:12:37,280 --> 00:12:41,480 to build this and maintain it. Is is that cost worth the worth 233 00:12:41,480 --> 00:12:43,200 the benefit at the end of the day, just because it's like 234 00:12:43,200 --> 00:12:46,720 cool, funky to have that AI, should we do it? 235 00:12:46,960 --> 00:12:50,000 But also from a, you know, customer experience, as you 236 00:12:50,000 --> 00:12:53,840 said, if everybody gets, if everybody starts getting dynamic 237 00:12:53,840 --> 00:12:56,800 pricing and you know, some of the customers feel like they're 238 00:12:56,800 --> 00:13:02,080 being, you know, being miss sold or something that that's, 239 00:13:02,080 --> 00:13:04,760 that's, that's a bad thing. You can do that, but you know, 240 00:13:04,760 --> 00:13:06,080 it might be legal to do it as well. 241 00:13:06,080 --> 00:13:08,760 It's just that case of is it good from a customer safety and 242 00:13:08,760 --> 00:13:12,720 ethical point of view? And then then the last pillar 243 00:13:12,720 --> 00:13:16,720 around that, the top thing is, is there somebody that's at the 244 00:13:16,720 --> 00:13:19,480 top of the organisation from a governance point of view saying, 245 00:13:20,040 --> 00:13:24,080 do we have a strategy? Do we have overall control of 246 00:13:24,080 --> 00:13:26,760 this? Have we enabled people with the 247 00:13:26,760 --> 00:13:29,000 right roles and responsibilities to look at this? 248 00:13:30,280 --> 00:13:34,000 And it's those four areas, the data, the safety, the ethics and 249 00:13:34,000 --> 00:13:38,120 the governance and the components we've put within 250 00:13:38,120 --> 00:13:41,440 those that really says OK, and now I've got everything I need 251 00:13:41,440 --> 00:13:44,080 to be confident I'm being responsible with AI. 252 00:13:44,200 --> 00:13:46,680 Rob, thank you so much for taking us through. 253 00:13:47,600 --> 00:13:52,320 You know what is responsible AI, some of the outcomes good and 254 00:13:52,320 --> 00:13:55,560 bad, where in the business we need to to look at it. 255 00:13:55,760 --> 00:13:59,560 I guess I'm curious in terms of, you know, where can people get 256 00:13:59,560 --> 00:14:03,240 hold of the report and what next, what should they be doing? 257 00:14:03,560 --> 00:14:10,160 Reports on our website, arioblue.com names on that 258 00:14:10,280 --> 00:14:12,840 report contains and some of the other things on the website 259 00:14:12,840 --> 00:14:16,160 contain the framework. So people can look at that and 260 00:14:16,160 --> 00:14:19,160 understand a little bit more about the framework and, and 261 00:14:19,360 --> 00:14:23,080 they can make a, an assessment for themselves on how, how well 262 00:14:23,080 --> 00:14:25,560 they've considered some of those things in the framework around 263 00:14:26,000 --> 00:14:29,280 around data around safety, ethics and governance. 264 00:14:29,360 --> 00:14:32,120 We also have some tools that help us so we can quite quickly 265 00:14:32,120 --> 00:14:35,360 help people with looking at, at that maturity and, and then help 266 00:14:35,360 --> 00:14:37,800 them plan out what to do next. So start with the report. 267 00:14:37,800 --> 00:14:40,320 There's some really great, great ideas in there about how to, how 268 00:14:40,320 --> 00:14:42,960 to get going. And as I say, the the framework 269 00:14:42,960 --> 00:14:45,360 which people can use and and have a think about themselves. 270 00:14:45,880 --> 00:14:47,920 I get I get this is just a personal curiosity. 271 00:14:48,080 --> 00:14:50,720 How long do these projects take? How long would you be? 272 00:14:50,760 --> 00:14:53,560 Like is this an expensive thing to be doing? 273 00:14:53,880 --> 00:14:56,280 I I think that that there are two things. 274 00:14:56,280 --> 00:14:59,880 One is the way we run that kind of maturity settlement, it's 275 00:14:59,880 --> 00:15:02,400 really quick. It's a few days of interviews in 276 00:15:02,400 --> 00:15:06,640 somebody's office or you know, virtually more and more so, so 277 00:15:06,640 --> 00:15:11,040 that that's not a long process. I think that if you are an 278 00:15:11,040 --> 00:15:15,840 organisation who has started to think about these things and 279 00:15:15,840 --> 00:15:18,800 have made a mistake, then the regulator's going to be a lot 280 00:15:18,800 --> 00:15:22,480 more lenient on you than if you haven't started at all. 281 00:15:22,480 --> 00:15:26,680 So the process of getting going is really important and then you 282 00:15:26,680 --> 00:15:32,480 can take it at your own speed. As you know time funding allows. 283 00:15:33,280 --> 00:15:35,520 You'll probably find that a lot of a lot of places in your 284 00:15:35,520 --> 00:15:38,800 organisation have started thinking about these things and 285 00:15:38,800 --> 00:15:40,640 maybe just needs writing down the things they're already 286 00:15:40,640 --> 00:15:42,440 doing. So you might not be in a bad 287 00:15:42,440 --> 00:15:45,080 situation as you as you sort of to begin with. 288 00:15:45,080 --> 00:15:48,280 So having a look at that, understanding the framework and 289 00:15:48,280 --> 00:15:52,040 then applying it and thinking about what gaps you might have 290 00:15:52,040 --> 00:15:55,520 would be, you know, really first, great first step and 291 00:15:55,520 --> 00:15:58,880 something that will help you on the journey and help you keep 292 00:15:58,880 --> 00:16:00,760 you on the right side of the regulators as well. 293 00:16:01,440 --> 00:16:04,480 That's brilliant and definitely regulation is coming. 294 00:16:04,960 --> 00:16:07,360 And so I guess it's better to be on the right side of it before 295 00:16:07,360 --> 00:16:11,880 it's even there exactly because it's a lot harder to unpick some 296 00:16:11,880 --> 00:16:14,000 of these spaghetti systems that you see. 297 00:16:14,560 --> 00:16:17,920 Rob, thank you so much for taking the time to to talk to us 298 00:16:18,160 --> 00:16:20,800 and we'll speak to you soon. Brilliant. 299 00:16:20,800 --> 00:16:21,720 Thanks a lot. Take care.