Andy Yeoman: When AI Rewrites Risk Transfer | TRP #162

[Music] Hey there, welcome to the Reinsurance Podcast, the place where we dive into all things reinsurance, the coolest part of insurance. We're your hosts, Jared and Ben, a couple of ex-practitioners who loved the industry so much we found it supersede to tackle some of its biggest headaches and we're here to share our insights and stories with expert guests as we uncover what's really going on in the industry. Welcome to the Reinsurance Podcast everyone, we're here with a special guest as well as Jared, of course. Andy Yoman from Consurrices here to take us through InsureTech, Past, Present and Future especially to me. Welcome Andy, how are you? I'm good, I am the Machina survive my line bike ride, the Mall out which is a very great challenge. And Andy, as a bit of context I guess for the audience here, we've known each other and met a very long time ago. So before we start Jared and I started supersede when I was at A on, you were already conserious so you're very much sort of the originals I suppose of the InsureTech movement. And I wondered if you could humor us and our audience with giving us a bit of your backstory, conserious then and maybe up till now. Sure, it's a long backstory, we've been at it for a while so in my mind I divided it into Conserious version 1 and Conserious version 2 which is where we are. So Conserious version 1 was never actually meant to be an InsureTech, I should drift it into it, there's a phrase in the UK and you can take the girl out of Essex because Essex out of the girl and I've been running companies for quite a while. So I just thought Conserious, the name connected cloud as she is consulting for her family self, starting her software company and then family self raising money and going into marine analytics. And I think version 1 of the company which is all about marine analytics was tremendous fun. We had done something that nobody else had done. We introduced a new category of data and analytics and decisioning and it was fantastic and people would look at it and go that's really, really interesting. And we would come back and present again and again and to be a world to realise that really interesting just means that you can come back and present to me as many times as you wish but we're not necessarily going to buy it. Although we did sell it dozens of times. It went wrong for us in lockdown. We got a strategic decision wrong, the markets didn't really spend any money for two years on technology such as ours. And so where we went into lockdown full of cash and hope we came out with neither of those. And that resulted in 2022 and us having to downsize the company in 2023 we recapitalised the company and from there Conserious version 2 was born. Conserious version 2 will come to what we do in a moment but in essence what we saw was let's try and create something which is compelling not interesting. We actually did what is called I2C, interesting to compelling. And our sort of governing thought on what we were going to create was actually quite simple is when we present this I want people to point at the screen and go if we don't have that we just don't going to be able to compete. And that's sort of governed our thinking and so that's where the latest decision of the company was formed. So we've been on this really be roller coaster. Of loads of money coming into the organisation more going out and then running out. So it's been a fascinating time. When you benefit massively from the learnings of that first pass, right, you have it where like there's areas of the market that you saw opportunities and challenges and friction points and all these sorts of things. And when you sort of go back and you say okay where is the if you're in the I2C journey, you have an idea where that sits. Right, you're kind of like direction from all the conversations you've been having and the clients you've been having over the last several years. You know broadly where you're trying to point at. It's not like a blank piece of paper as such. It's more of a heavily guided direction of travel there. It is a direct travel. And strange enough, I met with Andrew Johnson yesterday and seen him for a while and I gave him a hug and said I feel like across to the dark side, Andrew. And so why is that? Because I keep getting pitches for new ideas for companies in the insurance market. And I'm like well that's never going to work. Stop telling me insurance is broken because it's non-broken. What you're doing is never going to work. And I said I feel like I've reached a happy level of optimism and cynicism to understand how to operate it in the market. And that I think really comes from that deep sort of like I mean 10 years now so I wouldn't say domain x but I clearly understand a bit about insurance. And that the insight you get from having both technical expertise and domain expertise I think is fascinating. I could have done without some of the hard lessons. We are going. And I guess the context changed a lot for all of us during that, that windy. I can describe, I remember the heyday of inshortek when there was a lot of education actually finding out what is inshortek. What do you mean that we could do things differently? And then as you said pandemic came along. And since the pandemic, we thought we'd all got out and that it was business as usual and we could go out and do things in a similar way to before the pandemic but then crash landing of AI. How's that changed things? I think it's changed everything. I, my catch phrase for once of a better way of describing it is I just think we're in a goal in age now from a technology and insurance perspective. There are many people describe AI in many different ways but from my perspective it allows you to do things that you couldn't either previously do or couldn't do in the time that you can now do them. And I think we're going to see this new generation of risk entrepreneurs come through. So I think it's absolutely fantastic. I do think however that as vendors, we are clearly both vendors but as another vendors, we have a responsibility to think through the so what of our technology. If I have a conversation with you and you say, what do you do? I'm an AI vendor in insurance. You go, okay, that's good. That's interesting. That just means that neither of you know what you do. We need to do something very specific and get it out there in two or three sentences. I think what's the outcome of what you're doing? Don't tell me the input. We're an AI vendor. That's an input. Don't tell me the outputs. We organize your submissions for you. Okay, tell me the outcome. We write more business. We don't be writing more business, more profitable in less time. That's a good outcome. Yeah. You have to have those things inextricably linked to a problem statement of some sort. When you do fundraising or early founding of businesses, you always start with problem statement then solution statement that your product allows to happen. But so oftentimes, especially with a technological wave, people get so excited about it can do this. That might be really interesting and really cool. But if that thing isn't attached to a problem that needs solving, it is to your other point. It is interesting. It's not compelling because I don't know if I need it. But if you begin to then use that to say, you have this challenge, your businesses face this specific challenge and the challenge exists for these reasons and now we're able to solve that challenge in this way. Therefore, giving you these benefits, now that is a far more compelling and thing rather than just an interesting piece of kid. It absolutely is. I think, no, the simple way is that if you're a hammer, the world looks like nails. You want to go and hit them all. That's not not the answer. I was working on this morning some copy for our website. I've just been thinking through what does, what actually is it that we're doing? And I think this is all of us that we're doing. And I realized that we're solving a problem which actually is bigger than insurance. It actually sits to the fundamental of how we're educated. Which is we're brought up with phrases such as look before you leap. Can you think of one that's opposing that? He who hesitates is lost. Can we help? Yeah. We've got those phrases. We've got, you know, hurry up and wait. We have, we have more speed less taste and this, that, and the other. And underlying all of those is actually a dynamic that exists in insurance, which is I can make an underwriting either insurance or re-insurance decision quickly. But I give up. If I make it really quickly, I give up certainty. I need to take a bit more time. We need to do some analysis. We need to pull our data together. I have a process that needs to go through. I need to sanction this check. I need to go through clearance compliance, et cetera. I need to get certainty. So I need to take a bit of time. And so you've got this scalar of what I want to do. Am I far, do I act with speed? Do I Do I have to sit? I know you've recently got married. I've got married many years ago. The questions I got asked, "Is your marriage in your marriage "would you like to be right? "Or would you like to be happy?" I've been happy to marry for 30 years. So this speed and certainty, can under them, is an interesting one. So if you look at this technology, what is actually doing? It's actually bringing both of those together. So the technology has actually said, "Well, I tell you what, "you tell me everything you need to check." I'll do it. But instead of it taking 48 hours, I'll do it in 90 seconds. So now you've got speed and certainty. That's really interesting. Because that opens up all sorts of opportunities. And I think at the end of it, if I sit across an under, I'd say, you know, you can double your book, but I've got to keep it profitable. Well, we can do, you've got to have to stay in compliance. And over the years, we've seen more and more stuff that you have to do, that's what you can do the stuff you want to do. And I think that this is what this technology trend is about, is it's about, this is doing things really quickly. Yeah. Yeah. Let's do them with much greater certainty. They're not either all, so you can have them both. It's an interesting, I see, and I suppose, that comes to mind from the Betty Pross. You've come across that in any channel that there's, the character is told, "Oh, we found a way to make you get from A to B." Much, much faster. And the character responds well. What would I do, you know, with the time? When I got there, earlier than planned. And it's sort of suggesting, like, I guess I just walk very, very slowly for the last five steps. And enjoy the extra time. But it does raise the question around talent, right, if we can get things, both speed and clarity, satisfied so quickly. What happens to talent? I'd love to hear your thought about how that's changing around the market, with sudden speed and clarity. Sadly, I run a lot of run the cycle. And I like it because I have a busy brain. I know I'm slightly on the spectrum. And it just allows me to think these things through. So let's start with what we don't do. If you buy a load of technology that makes the process quicker, and then you don't do anything with your time, all you've done has increased the cost base of your organisation. So all the ROI's only work. If you take that time and invest it, you get a return that's at least as good as my salary. Otherwise, it doesn't work. So you then start to think about it. One of our case studies we've got is a company in the States. They've grown in the last two and a half years from 25 million GWP to 100 million GWP. They didn't with no proportionate headcount increase. Nice. That's a nice ROI. But if you're not growing, what are you doing? Well, then I'm doing better. I'm doing a risk selection. Am I thinking about my portfolio a lot more? And then there's the question the hard question to ask is, well, why don't you get rid of people then? If you can do it in half the time, it's a question that you're going to ask them, why don't you have half as many people? It's about its source. Not a big proponent of getting rid of good, good, well-skilled people. But you have to, if you write the check, you need to bang the check. And so I think that understanding it, but when I spoke to that customer about what they did and he described the benefits, he said clearly we've grown, that's been fantastic. Two things which kind of caught me a little bit by surprise. One was, he said, our profitability's improved. Why is that? Well, whilst we're really, we're really for the, to quote, in 90 seconds. We're not quoting 90 seconds. He said, we're quoting it about 30 minutes. He said, but what we then do, answer the quotes really, we then think about, do I want that business? How does this fit in my portfolio? How shall I structure deductibles? How do I think about that? And then we're not in the apologetic corner with the broke going back in the last minutes. He said, so we can actually think about. So that one, I'll go, well, that makes sense. I could have probably got there in my head. The other one which I didn't get to in my head, as a benefit, he said, he's underwriter satisfaction. He said, the job satisfaction has gone through the roof. He said, it's because they're doing the things that they want to do. Because our tradition, we were hiring underwriters and then turning them into data entry clocks. And the putting them under pressure. Now they're, they're now deal makers. They're doing what they're writing that sort of, it's nexus, that the right word, or where they want to apply their skills. That's really exciting. I think when I was, when I was, first listening to you talk about this sort of, do we do it fast or do we do it right? The thought that came to mind is, is kind of exists in this order of operations requirement. And then the thread around people, I think, plays into that really nicely because you get a piece of business and you have two options. You can first validate whether you can completely write that business and it doesn't meet sanctions and all these things. And so the individual, the underwriter, undertakes that work. And it turns out they can't underwrite it. So they didn't review the risk, but they spent, they spent more and more their time first looking at compliance, which is tedious and boring and miserable. Or they spent a bunch of time reviewing that risk, getting excited about that risk to then find out they can't write it, which feels like a waste of their energy and a waste of their effort. But if both of those things are able to happen immediately, and you can go not compliant, ignore it, or immediately compliant, and I'm sorry to look at it, you're right. Both of those things start to go away very quickly, or combine very quickly, by which the thing you're left with is the deep thought around, is this a risk that fits evaluably into our portfolio? So you can see that, the satisfaction, like skyrockets, because you're pulling out the sort of risk of doing either of those two things first, and either having your work feel like it was a waste of time or your work being only compliance data loading, tedium. And so you get to a point where actually the work between those, that nexus between those two things is the most interesting thing. You're absolutely right. And of course the irony is, you get to the end of the year. I think the other benefit of some of these technology gets you into the year, and you go, "Hey, listen, I've had a great year. I grew my book from whatever, 50 millions or 100 million. I maintained my loss ratios, what combined, everything under control." They can go with fantastic. Okay, next year, can you grow by 10%? Make sure we don't lose our loss ratio control. So where's that growth going to come from? It's likely to come from your declarations. Now, in prior days, your declarations are, I don't know, they're probably in your deleted folder and your email. Now, they're in the same system for the things you quoted and won are on. So you're going to get the technology to now say, "Okay, ask it. "If I want to grow my book by 10%, which bits should I look at?" "Am I stressing my sanctions? Am I stressing my compliance? "What is it that calls that declaration and start to do some more things?" You've got the data there. It's great, isn't it? It's almost like saying, "Great, we've got AI to read all of the books on Planet Earth. What do we do now? We've run out of content." Go and read them again, but this time you're looking at it and slightly differently. And it's one of the things which I thought would be interesting for this conversation is just to sort of be a bit edgy in terms of where the market could go with the sort of technology. And the impacts both of us, the market as a whole. And this is only, I've been thinking about, four, probably five years plus. And this week, it finally, the penny dropped. And that is that if you look into another market, which looks awfully like insurance, you end up gambling. You're gaming the gambling market and the insurance market. You're taking a fee for an uncertain outcome. Clearly, one of them is an industry where people hang around the pub and dark and corners smoking and talking all day. And the other one's gambling. You see with that one, you know, head of the Oppen Pass. But actually, in the gambling and gaming industry, about 10, 15 years, going a little bit more, we saw this concept of in-play betting. It's Alvaro's playing tennis. I could always bet that it was going to win the match. Now I can bet that he's going to win the point. So how did that happen? What happened? Nobody invented gambling. They didn't invent the calculating of odds. They didn't inventions. They're taking them a bet. What happened was we've got more data, much more data coming in in real time. And now I can assess that risk in real time. I can place the bet. And actually, the bookies can actually lay the bet off in a fraction of a second. Yeah. And there are two interesting aspects to that. It's number one. The market's gone from zero to trillions of dollars in that time period. Number two, the regulators of that market are not the same regulators of the terrestrial markets. So the Maltz, GDP, something like 10, 12, 15,000, I'm not sure if it's recently, comes from gaming companies that are domiciled there. So I've been thinking, what's the equivalent in insurance? Is it, was cars going down the road? Is it going to hit the car in front? Yes or no? Yes or no? That's not, I don't think that that's the equivalent. But I do think the equivalent could be this week, and it has both an insurance and a re-insurance capability. Is this? When an insurance perspective, you get these very short tail risks, and it's embedded insurance. So in India, when you get an Uber and an Ola, you can pay another 10 rupee and you get insured. And of course, that policy is a very short cycle, very short risk, absolutely. And you know, it's like, now I've assessed that risk, I've placed it, priced it, it's all done in that time. So I think that there's under of very noses. There is this entire market for new insurance products that it's been there all the time we've definitely never had the data. - I think you're right. I think interestingly, it's starting to appear very nascently already. So as a super commuter myself, and who buys regular long-distance train tickets, I'm now offered, but only quite recently, through my train booking provider, do I want insurance on this specific train journey that I'm about to take? And as you see, historically, that would have been, do you want an annual travel policy or some kind or somehow accommodate all the train journeys you're going to take, which is very hard to price, is very hard to actually understand what risk is going to be taken in the year ahead, whereas they can quite specifically look at, okay, how likely is it based on all the data we have that that specific journey between station X and station Y is actually going to be disrupted? And to your point, then there's a reinsurance angle as well. Can they get cover on top of that for the possibility of a catch train and then where all the trains get there? - They could get there. But I don't know that you'd need, that wouldn't be a cat event in the same way, but there's definitely a reinsurance angle. So the same, take the theory to another level. So one of the things that challenges me about the insurance market is risk is only traded once. All right, policy, that's it. You can write some factory insurance, but that's it. What would be fascinating is if I could say, if we do multi-line now, but in the marine example, I might write, let's go with aviation example, I might write your 50 aircraft, under those 50 aircraft, they might be five I don't like. Well, I would like to be able to place that out. Yeah, I'd like to be able to lay that risk off. We could, you could fact it off. But it's a very manual process doing that. But if that was algorithmic, I could do that. Yeah, almost, I could do that actually, co-term, so at the same time, as I bind the policy and I place that, and I bind that bit out the back. And of course, what you've then got is if people are doing that all the time, there's a secondary market where you can scoop those up, and then start doing this. You get that, so portfolio management, which is today about risk selection. So it's what I'm getting in the door could also be, what I'm pushing out of the back. Yeah. And now I can start, I can actually dynamically trade my risk. And you add a bit of blockchain in there, so you've got like an immutable later in there, you suddenly got like a stock market style dynamic going on. Yeah, it's super interesting. I was in a conversation in Monte Carlo a couple years back now with James Slaughter. And we were talking about this, we were sort of doing this, XXX says we're doing now, where you kind of go, okay, what that happens, you could then do this thing, and you could therefore then do that thing, and you sort of progress backwards. And we got to this idea of, could there be, or would there emerge a world where treaty re-insurance did not exist? The idea being every individual policy that got insured could go get facultative re-insurance, but it because you'd be able to sink these things up at such scale, and you'd be able to even at like the point of underwriting the original policy, have a lens as to the cost to re-insure that policy should you want, and like this, the model that says annualized portfolio, purchasing of re-insurance, is the mechanism that has been put in place because the only way to do it was to say, how many car insurance policies did you write last year? What do you think you'll write this year, and therefore you buy that in advance? But could you move the world towards a place where, I'm literally writing a singular car insurance policy that will then be automatically appended to this facultative thing where someone might offload 10% of that thing they write to their own. And again, I don't know how far we're away from those things being actually possible and being actually a thing that's happening, but you begin to keep pulling this thread in some really interesting alternative ways this market works, begin to emerge, and new capital sources as a result of that as well. And essentially, because you then challenge the nature of the head just because you need a certain level of uncertainty to allow people to use it to use it as a risk. - So the question then becomes, how does the arrest go turn into a market dynamic? And so our consumers version one, we realized that we ran up against that subscription market. If one client has superior information to the rest, that actually doesn't work because they might price it lower, and everyone else, not quite sure I want to be on that slip, 'cause it doesn't meet so the asymmetry of information doesn't quite work. But this is not that, 'cause the current reinsurance market moves the process annual Monte Carlo, et cetera, moves at the speed the people move at. When you take the people out of it, and I'm not just saying this again, but if there's an electronic exchange, things can speed up and be slightly different. And it doesn't need the entire market to do this. It only needs one player. So a good friend of mine is a gentleman called Peter Graham, a direct line, he was the founder of E-Shore, et cetera. And he told me when back in the day, when they were emoting insurance, I think it was one of the big rental car companies that said, "Why don't we come and do credit hire?" So if you involved an accident, it's not your fault, I'll rent your car, and I'll get it back off the other insurer. And he said, "We resisted doing that, because we'll have to pay." It's not like the others are gonna pay for that. We'll have to pay for that. It's just gonna increase the expense in the industry. So they resisted doing it, and kind of like, everyone else would resist do it as well, but doing it as well. But then one insurer cracked, said, "No, I think we'll do that." And of course, that everybody has to do it. And I think that's the way that these market changes are going to play out. It's one person will do it. One company will do it. There's a bit of an arbitrage opportunity, yeah, before everyone else does it. And then once the majority of the market's doing it, it's no longer an advantage. It's just the way the business works. - Yeah, it's interesting. So I guess where would you see things, possibly going from here then, 'cause earlier you mentioned, do you, now at the stage of your career, where you're both the enthusiast, but also the cynic. Where are you cynical, where are you enthusiastic and optimistic? What parts of the golden age do you think we can actually get the fruits from and which parts, maybe too far a pipe dream? - No, I think that. It's a big answer. I'll just go back maybe 25 years to answer it. So one of the few benefits of getting older is you get to see trends. And one of the trends that I've lived through is, one of the things I've seen, sorry, is the evolution of new technologies. So I've lived through the introduction of AI and blockchain and mobile phones and the internet and email and all that stuff. I'll embarrass myself and go back to it for a while. And I've seen three organizational responses to each of those technologies. People, some people just ignore it. Some people adopt it. And then there's the last one, which I've called the "but four". I'll come to define them. So let's take an example in banking. So the internet comes along. People look at who's going to put banking across the internet. Why would anyone want to do banking across the internet? It's fundamentally insecure. Let's just ignore that technology. And then you get the adopters who say, "This could be quite good, because instead of me standing you a statement, posting it out to you, what I could do is, I could put on my website, you could download it, and say we're fortunate in postage. It's a lot of the top to that technology. The last category are the "but fours", which is "but four" this technology. My core business model doesn't work. It's not a feature of function. It's the actual business model. So there's going banking and look at a neo bank. The monzoes, the revolutes. You can name lots of them. If the internet, you unplug the internet. There's no going. There's no plan B. There's no backup system. You know, their business model fails. And we can all name them for the internet. Amazon, you know, if the internet goes down, what are we going to do? Facts the ordering. We'll run the call and pick it up. That's not going to work. And you can clearly, though, you can go on and on. So if you look at where we are now with artificial intelligence, I think we can all probably come up with a list of who's ignoring it. That's never going to work. It's untrusted. Let's block these chat to you, BTN, or organization, because we're not quite sure what that means to us. We've got, we're going to name those companies. You've got the adopters. In many respects, what we do today, as a company, is we provide technology to the adopters. We're making processes quicker. But we haven't yet seen the butt-force. In the same way as we've got the butt-force in banking and the internet. And then, clearly, those mobile phones, those sorts of stuff, the butt-force with AI haven't come out. when they do. Yes. stand back. Because when I go back to my-- if I go back to my-- do you want to be happy? Or do you want to be right? Do you want to make the right decision? Do you want to make a quick decision? We're using AI and technology, which is would you like a system quickly or would you like something that works for your business? Would you want a custom made or do you want it quick? AI is breaking that one, because now I can give your custom made decision really quickly. So where does it go? And I think where it goes is some of these-- you're building-- and I think what you're being said-- you're building the information super high on the risk super high way that can really process things quickly. What you need is that AI-powered risk transfer model, which I think will start to fuel those. And it's a chicken and egg. We talked the other day. Our students are already about actual real models. So now, with the ingestion capabilities, I can ingest hundreds of parameters. Fantastic. I told you an act, and I said, how constrained are you by how many parameters you can reasonably expect someone to key in? So you've got a fun-- you've got a risk model. I can give them 10 parameters after that. They're not going to key stuff in. But of course, if the technology is keyed in, you can have 100 parameters. So you then get hyper personalized risk models, which drives a different risk application process. That's a chicken and egg. So I think that you've got one piece part. And call it to me, one piece part, which is we've now got loads of data that we can transfer between it really quickly, that an insurance product could be built on. And once that's been built, it'll need more data. And that will spin up really quickly. I don't know if insurance to conceive of the product. But I do know that we're built in collectively. That framework that sits behind it, so it's a really exciting time. Yeah. Do you think the reinsurance market, the insurance market, are as enthusiastic as the technologists at the moment? Oh, they still stuck in their sort of hard soft market cycles, focus, business as usual. How much of the AI excitement is making its way into the actual traditional conversations that these insurers are still thinking about? I think it's both. I think there's pockets. All of its education. We have to face the fact that 20%, 30% of the population, underwriting population in Rotara in the next 10 years, would I be investing in time if I was going to do it probably not? I think we did a survey. I was going to go back to my Consurers version one, by board said, you keep putting this product in front of them. Yeah, just do a quick survey. Ask me a question. If you have anything you wanted, what would you have? And the answer, predominant answer, came back as a variation of, I'd like it to be, how it was 15 years ago, just make a bit more money. I don't want the pressure and the intensity of this. It was nothing to do with technology. It was to do with that ease of life and simplicity of life. Yeah. I think it's very easy to look at some of these future states. I think Amazon's a great example. Amazon is a butt-for because without the internet existing, Amazon's entire model collapses. However, there are still brick and mortar shops that exist. I think the same analogy can be used in this industry. Because if you think about our industry's willingness to deploy diversified streams of capital, you had the various sort of rated agency, rated carriers, and collateralized carriers, then you saw new money come in in the form of like ILS, and that sort of asset class. It didn't mean that the whole market flipped to just having that. So what you could begin to see is not to say all undreaders will fundamentally change how they operate. We might see one or two that say, we have this model, and it's designed to work like this. And we have two people who work here. And we deploy $500 million worth of capital into this market. It's not to say that they're going to be the lead underwriter on major worldwide programs. But could you see a scenario whereby the buyer, though, that's the seed into our primary insured or someone, would be happy to give a 5% or 10% stake to this purely automated player alongside AIG, or a municrier, or whomever. Feasibly yes. And then if those firms begin to become successful just like ILS, you might have it where they might represent 5% or 10% of the market. This not going to be the case where every single risk is done purely without people involved. But there's a new pool of capital that is very, very efficient, and therefore can write meaningful lines alongside the other infrastructure. And that could be an interesting way where this industry sort of gently sees this evolution become the past. I actually realize, as I've explained to you that we've created a but for-- so my customer I gave who's grown from $25 to $100 million. But for the technology, their business model doesn't work. So I'm just-- they would have had twice as many-- At least twice as many people. --25% more people. It's not the good ones, but the early ones. They haven't fired hard to it. They should never digit entire them. So they effectively become a but for operating model. That is interesting. So it's fascinating. And it is a big grandiose statement. I don't want to go there. But I think that it's a-- my kids are in the 20s. If I've got-- you look back on photographs. And I can see who they are today. I can see who they are. They're clearly different, bigger, hairier, all those things. And-- but day by day you can't see the change. But year on year you can. And I think that's what's going to happen. We're doing these day by day changes that will actually be unrecognizable in five years. Yeah. Exciting times, look forward to it. It is exciting times. And it's been a pleasure having you on the show. I really enjoyed your thoughts. And any final things you'd like to leave the audience with before we hang up for today? No, I think when I talk about these sort of things, I think the big thing in this sort of technology age is actually-- it's just starting-- is that there's many things that are-- you can agree to like this and you can agree to not like it. You can agree to do it, not do it. But you can't agree that it's not going to happen. You can't decide that it's not going to happen. So this is going to happen. So my big encouragement to people is just try. We run AI education days. We just come try some of the tools. Some of them are really simple. Some of them are really complex. I think you can only lose by not trying. Yeah. Brilliant. Thank you so much. Nice pleasure. [MUSIC PLAYING] Thanks for tuning in to the Reinsurance Podcast. If you enjoyed our show, don't forget to subscribe wherever you get your podcasts. And leave us a review to let us know how we're doing-- preferably five stars. For more insights and updates, follow us on LinkedIn and visit our website at supersede.com/podcasts. You'll find the links in the show notes.