'We can't do that': the phrase that's finally disappearing from insurance technology, with James Benham

James Benham, CEO of JBK and co-founder of Terra, has spent 25 years helping insurance organisations modernise their core systems, and he has some strong views on why so many are still stuck. In this episode, he and host Georgie Simister dig into the real barriers to technology adoption, from shadow AI use and legacy system debt to the role of company culture and leadership in getting innovation off the ground.

Creators and Guests

Host

Georgie Simister

Georgie is the Commercial Partnerships Lead at Artificial, a London-based insurance technology partner to global insurers and brokers.

What is Insurance Technology: Fact or Fiction??

Host Georgie Simister speaks with key insurance industry figures to separate fact from fiction in a world of insurance, debunk industry myths and explore the game-changing innovations shaping its future.

Get in touch to suggest our next guest: https://artificial.io/contact/

Georgie: 00:01

We've never had better technology in insurance. AI, automation and tools that promise to transform everything. And yet, most of them still aren't being used. So here's the real question: Is technology the solution or is adoption the problem we're still not solving? Today I'm joined by James Benham.

Georgie: 00:21

He is the CEO of JBK, a technology consulting company he co founded from his dorm room at Texas A and M that help insurance organizations transform legacy systems into modern, revenue generating platforms. He is also the co founder of Terra, an AI enabled core insurance platform, and the host of the InsureTech Geek podcast. James has previously built and exited SmartBid, one of the world's leading construction bid management platforms. James, welcome to the podcast.

James: 00:53

Thanks, Georgie. Good to see you.

Georgie: 00:54

I'm very happy that you're here. But first of all, need to ask you, when do you find the time to sleep?

James: 00:59

I made the mistake of buying a whoop strap.

Georgie: 01:02

Yeah.

James: 01:02

And now it tells me how little I sleep. I got four hours last night.

Georgie: 01:06

You got so much going on.

James: 01:08

Yeah. It's you know, they say the idle hands are the devil's work. I need to be busy. It's just not good for me. So and I enjoy work.

James: 01:16

I love insurance. I love technology. Last night, I was spinning up some cloud agents to to build a code at

Georgie: 01:23

As you do.

James: 01:23

At midnight Yeah. Before I went to sleep. And I woke up and checked the results of the regression test that I built on a claims and policy system and then checked in with my team. And then I have meetings right after this. You know, it's it's busy, but I'm here to work.

James: 01:38

You know, that's my my role, and it's a very exciting time to be alive.

Georgie: 01:43

It definitely is. I'm slightly intimidated having a fellow podcast host, so be kind to me.

James: 01:49

Yeah. Ten I've been podcasting for ten years now between my two podcasts. Mhmm. I love it. I love the community that it builds.

James: 01:56

I think podcasting is brilliant. I love that it's international. I love I've got listeners all over the world. And in my previous podcast, I really had listeners all over the world. And on this one, we've built a a strong European following too and an interesting group in The Netherlands that that have been following us.

James: 02:12

But it's really fun to be on on other people's shows and not be the host.

Georgie: 02:16

Well, I'm honored. So thank you for being here. I'd love to start by exploring some of the differences between The UK and US insurance market. So from your kind of perspective, how would you describe the level of technology adoption that's happening in The US right now? Are there any areas that you think that The US is moving faster or perhaps slightly different to what we're seeing here in London?

James: 02:38

Well, you know, I I worked primarily in North America. My previous company that I sold in 2018, I had some huge clients in The UK. And so I can speak to my personal experience there that I I did find that The US market moved a little quicker on adopting new technology, both from a governmental perspective, just a cultural in fact, we had an InsurTech breakfast this morning here as part of InsurTech Week in London, which is really exciting, by way, that you have a whole week for it. And we were we were talking about this exact topic. And so I was the only American there.

Georgie: 03:09

Mhmm.

James: 03:09

And then there was a a large group of of Britons and a couple of Israelis there. And we were talking about this, and I don't think there was a definitive conclusion as to why. Maybe cultural, maybe political and bureaucratic and regulatory reasons why there's different rates of adoption between the two countries. But I've definitely seen on the startup scene, there's just a lot more capital and funding available in The United States. And so that definitely encourages more innovation, and there's a lot more new companies that are constantly pitching new ideas to to insurance companies.

James: 03:43

I think that impacts but at the end of the day, it comes down to to culture. Right? Countries have culture. Industries have culture. But there's quite a bit of innovation going on in The UK.

James: 03:53

I mean, London is the home of modern insurance. Right? The I would say the Babylonians invented insurance three thousand five hundred years ago with their bottom re contracts. But but the coffee shop, Lloyd's, was was just down from the street from here, and that's pretty exciting. And I think it still plays a pivotal role in in insurance.

James: 04:11

And there's more innovation going on than I think Britons give themselves credit for.

Georgie: 04:15

Well, I think one of the terms and one of the phrases that are used in the London market is traditional. And the insurance market in London is is built on these traditions, which is incredibly relationship driven. And so understandably, that means that people are probably a bit more hesitant with technology because it can disrupt the trust. So I kinda see from that lens why, yeah, there's the hesitation of the adoption because it is so built on trust. But from from The US side, what are some of the kind of the adoption challenges that you see with people getting more comfortable with tech?

James: 04:47

Well, you know, the funny thing is that even in The US, the key to being a successful InsurTech MGA Mhmm. InsurTech startup MGA is capacity, and the only way to get capacity is with personal relationships. Yeah. So, I mean, I as good as the tech is, if you don't have the relationships to get capacity and get guess what capacity requires? Someone trusting you individually.

James: 05:10

Exactly. And so look. We go back to Lloyd's, what's an underwriter? It was someone who went up to the chalkboard and and wrote their name under the risk. Mhmm.

James: 05:19

And we're still functionally there. You know? Who's gonna give you power of the pen as a tech enabled MGA? Mhmm. Someone who knows you individually.

James: 05:27

So I think it's important to understand that that that's still really important that personal relationships in insurance are really important because at the end of the day, everybody who wants to really innovate needs capacity. Now if you're talking about, though, taking existing insurance companies, which is what I spend a lot of my day doing Mhmm. Is taking existing carriers, third party administrators, brokers, MGAs, and then trying to help them revolutionise themselves with both tools and services.

Georgie: 05:54

Yep.

James: 05:56

There is across the world a lot of of hesitance. Not enough people have personal experience working with automation tools and AI. They're getting there. They they're maybe a year into their AI journey. Yeah.

James: 06:08

And maybe some of their experiences early on a year and a half ago with ChatGPT four point o or or an early version of Claude, Bioanthropic, or an early version of Gemini, which really wasn't that good until three point o. They're judging their experience with AI through the lens of working with early stage large language models rather than looking at the breadth and depth of AI. And so they kinda dig their heels in a little bit and say, hey. Look. You know, it hallucinates, and it has issues.

James: 06:35

It's not accurate. I'm like, hey. Look. A lot of that's been trained out of it. You need to you know, this space moves daily Yeah.

James: 06:41

Not weekly, not even monthly. Like, the AI space is moving on a daily pace right now with releases because AI is writing AI now.

Georgie: 06:49

Yeah.

James: 06:49

And it's getting way more accurate and way better at predicting outcomes and way better at automating. And so there's definitely reluctance there. There's also reluctance because there's really two big ways that my service company JBK and that Terra, my my plat my core platform, that we really help insurance companies innovate, and that is through workflow automation and decision support. Right now, a core system, we process all the transactions. And in our services shop, we work on their bespoke internal proprietary core systems.

James: 07:20

And so you're definitely getting a lot of buy in on using AI to help with tech development and retirement of technology debt. But when you're looking at using AI for automating like, core workflow items, like underwriting a policy or adjudicating a claim or summarizing a medical record on a claim or summarizing a claim summary or or even potentially communicating with a claimant or communicating with a policyholder, you've gotta tread carefully with that area. There's definitely people that feel very threatened

Georgie: 07:51

Yeah.

James: 07:51

By this, and that is absolutely a 100% happening. There are absolutely a lot of insurance professionals. And I I would say probably a third of people that work at insurance companies are clerical in nature. They sit at a computer for a living. They input, process, and output data into Excel spreadsheets and websites for a living.

James: 08:12

And so those folks need to train up. Like, we need to get them into legal license adjusters or certified underwriters. And and so because those jobs are now remarkably easy to automate. And so but we have a shortage now of qualified licensed adjusters and a shortage of qualified certified underwriters. And so we we need people to retrain.

James: 08:40

And so I say, I'm even seeing a reluctance on some people to retrain. They just don't want to. Yeah. They're far enough in their career. They wanna keep doing their job.

James: 08:48

You're seeing the same thing. I think you call them broking firms. We call we call them brokers. You call them broking. But there's there's probably a third of staff at broking firms that are clerical in nature that pull submissions together and normalise submissions or or, you know, key data into carrier systems or and all of that can be automated now and really streamline the broking process.

Georgie: 09:10

I completely agree. And I think what what one of the the trends and what I like to speak about is that change is coming, and that's okay, but you need to learn how to evolve with it and adopt AI. Yeah. Like, the but I think there's also what it brings is opportunities, and I see those as being opportunities rather than threats. And it's, you know, chances to evolve and develop with the technology, new yeah.

Georgie: 09:33

New job opportunities. You're gonna have new people now managing AI agents. That's a new role that we've not seen before.

James: 09:39

It is.

Georgie: 09:39

So it's a And

James: 09:40

doing QA on AI? Yeah. Like, I'm in the process of automating, you know, some some pretty difficult tasks. Mhmm. And I I just think that's something where there is a huge opportunity for us to upsize people.

James: 09:53

Yeah. The surprising thing for me because I love learning new things. Mhmm. I watch YouTube every morning, and I have a few AI channels I follow and a few technology channels, few insurance channels. I listen to insurance podcasts because I'm a podcaster.

James: 10:06

I listen to podcasts. I always am trying to learn new things, listening to audiobooks. But there are quite a few people that really I mean, a material percentage of people in the workforce that really are interested in just doing what they've always done. And so that that kind of organizational traction creates some challenges. But I agree with you.

James: 10:26

There's there's opportunity to take people into higher paying positions, higher skilled positions. You know? And you saw this happen with the industrial revolution, which really kinda started here in England. You know? And and, you know, that that phrase of being a Luddite is from England.

James: 10:42

It was the English textile workers, if I remember correctly, two hundred years ago. You know, they said, well, all this machinery is gonna automate all of our jobs out of existence. And it did. But they said, we're all gonna be permanently unemployed, and that was wrong. They weren't.

James: 10:56

Unemployment went up by 6% for nine years, and then it went down below the pre industrial revolution levels. And then because of the automation, wages went up, GDP went up, productivity went up. And then we started having the time to now it took a labor movement as well, by the way. So I don't wanna leave out the fact that there was a really important labor movement that led to a five day work week instead of a six to seven day work week that led to child labor laws being reformed. But it was the technology that enabled the productivity that took us there.

James: 11:25

I do fundamentally believe that good things will outcome at the end of this, but this is going to be a much faster pace of fundamental structural change than most people are comfortable with.

Georgie: 11:39

How do we get people comfortable then, like, with the adoptions?

James: 11:42

Education. I mean, there's just no that and using it themselves. Now I'm seeing a lot of shadow AI adoption inside of insurance companies where people are pretending to be reticent to use it, and then quietly they're using it behind the scenes. Mhmm. And they recognise that the productivity gains mean that the job the way it is right now is gonna be a lot less hours in a day to get the same thing done, but they're recognizing the productivity gain.

James: 12:05

They're using it on their own. They're just not telling anybody they're using it. And that's a really interesting phenomenon where people are like, oh, no. No. They they wanna capture the productivity gain for themselves.

Georgie: 12:15

I don't I don't think it's anything to be shy of. Like, I think you should be you know, if you're not using AI, then I don't think and nowadays, you're you're not you're not working to the highest productivity level that you can be, surely, because

James: 12:29

Yeah. You're not. There's there's a lot of regulatory concerns in some markets around how accurate the AI is and and how but look, I mean, I'll be honest, we do a lot of QA work. It's it's good.

Georgie: 12:40

But even as a day to day a day to day user, you know, having an AI meeting taker. Yeah. Some of them aren't perfect.

James: 12:46

But It's better than anybody anybody oh, okay. Those AI meeting notes, Microsoft Teams, they're built in Copilot for meetings, is the best notetaker I've ever used. It takes better notes than any human I've had take notes. So, I mean, like, I'm I'm like so sometimes I say, like, are you saying that humans do this perfectly? Because show me a perfect claims adjuster.

James: 13:05

Show me a perfect underwriter. Show me a perfect notetaker. Show me a pro I mean, I think that we're at the point where we're we're getting remarkable productivity gains Mhmm. Out of the simple things like meeting note takers, but we're also getting remarkable productivity gains out of summarization of large documents out of and, of course, really in the machine learning spaces where I'm probably even more interested than in large language models because I find large language models assisting with some decision support. Largely, we're using it in workflow automation.

Georgie: 13:35

Yeah.

James: 13:35

But machine learning models is where we really get some juice on decision support on predicting outcomes. Mhmm. And the entire insurance market is predicated on predicting outcomes. Yeah. It's one giant betters book.

James: 13:47

Mhmm. Right? I mean, it is. Yeah. And so I'm very, very excited about the using the power of machine learning to predict better outcomes and be better at claims adjudication, be better at underwriting, be better at renewals, be better at decision making.

James: 14:01

Mhmm. Those are the areas where I think we're gonna add a lot more value. But workflow automation is really real. And allocated loss adjustment expense, all this stuff that goes on top of the claim cost, it's real. We can reduce we can reduce it pretty significantly.

James: 14:16

And so I think we've gotta take it seriously, but also say, okay. What's the plan for these people? What can they apply for? Where where can they go next? How can they fill the unfilled roles that we're having a hard time filling?

Georgie: 14:27

And and

James: 14:28

all of that, by the way, comes back to education. Training, training, training, and more training. Mhmm. But if they're not willing to train, which I have run into probably half the time, then I don't have an answer. Yeah.

James: 14:40

I don't have an answer for someone who doesn't wanna enter training again.

Georgie: 14:43

Another reality that we face across the organization is people are still operating on legacy systems that they've had for decades. What risk do you see for companies that are delaying modernization because of those systems?

James: 14:56

I mean, let's be honest. One of the most important core values in insurance is stability. Mhmm. So I think it's a bigger risk if an if a carrier whose neck is on the line is willing to be haphazard with decision making around the core thing that runs their company. So I think there's a good reason and justification to have stability as a core value to focus on making sure that you're making good underwriting decisions, you're making good claim decisions, you're making good renewal decisions, And that the systems that have supported those forever, that the knowledge that's trapped in those systems Mhmm.

James: 15:38

Doesn't get lost. Yeah. What's changing right now is that it's way easier to convert those systems into modern tech stacks.

Georgie: 15:48

Okay.

James: 15:49

Because the cost of writing individual lines of code is going down. Now we're able to go through backlog, and what you're really talking about is technical debt. Technical debt is that old legacy COBOL system or, you know, that runs on a s 400, which, by the way, I run into all the time.

Georgie: 16:07

Yeah.

James: 16:07

Or this old legacy Java application that's running on an old server. These are really significant problems because it's hard to update them. It's hard to expand. It's hard to expand lines and states Mhmm. In The United States.

James: 16:19

That's what we care about is line line and state expansion because we have 51 regulated markets, the 50 states plus Washington DC. Yeah. And so I think that's a a really important thing that that AI is unlocking for us Mhmm. Is the opportunity to take really old COBOL systems, really old Java systems, and have a cloud agent Mhmm. Spun up for a period of time, rewriting that into a modern tech stack that we can use to move forward on.

James: 16:44

Mhmm. So I think one of the really brilliant things and as a you know, my custom development business, JBK, is very busy right now. You have to wonder a lot of things around the technology industry because the first thing that coders that were building Anthropic and ChatGPT did was teach it how to write code. Yeah. So that's a big disruptor for the technology industry, but it actually is enabling us to retire technical debt, which are these old systems much quicker than we could in the past.

James: 17:14

Mhmm. In the past, you just looked at the backlog and said, well, we're never gonna get to 90% of those things. Now I'm going, okay. Maybe we can get to half or more, maybe all of the backlog. Maybe we'll just rewrite the whole thing and retire all the technical debt all at once.

James: 17:29

And that was not an option before. There just wasn't enough time in the day to go back. Some of these old carriers have 32 to 35 systems that are involved in quote, buying, pay, rate, underwrite, issue, renew. 32 different, 35, 38 different systems involved.

Georgie: 17:45

It can completely depend by different team, business. You know? Correct.

James: 17:49

Each line of business could have its own policy administration system. Each line of business could have its own billing system. Like, you can have some really and it's because they just don't have the capital

Georgie: 17:58

to

James: 17:58

hire enough engineers to rebuild that code. Well, now maybe we do. And so that's like the exciting part for me is that we can actually tackle backlog. I don't I don't see less of a need for engineers.

Georgie: 18:09

Yeah.

James: 18:10

But I do see us tackling a lot more problems with the same number of engineers.

Georgie: 18:14

So for some leaders that that are listening that might be, I don't know, a bit stuck and they don't know where to start, where would you recommend that when they're kind of fearful of transformation or hesitant to transform, where do you recommend that they do start?

James: 18:26

You know, where to get started is a really hard one. If you don't control your core system Mhmm. If you have an off the shelf core system, you've gotta have a core system provider that's willing to partner with you. Even then, I think one of the best places to start is just having a an AI policy in place Mhmm. That actually allows people to use AI Yeah.

James: 18:45

That lets them use it on the desktop. The easiest way to do that is if you have a Microsoft Office license, which 98% of insurance companies use Office three sixty five Mhmm. To go get a Copilot license added on to that to your Office three sixty five and release a secure version because, you know, Copilot's ChatGPT. Right? Microsoft owns 49% of OpenAI.

James: 19:11

They bundle ChatGPT inside of Copilot. But you can opt out of training and opt out of abuse monitoring, which means that you can comply with a lot of the requirements for AI usage that the data can't be used for training, can't be used for abuse monitoring. And then you can you can unleash and I've actually done this in a few of my insurance company clients is we we've first, we unleashed Copilot. Yeah. Because what what's happened with at most of them until now is their corporate general counsel came down and said, absolutely no AI usage.

James: 19:41

Do not go on the Internet and sign up for Claude, ChatGBT, etcetera. They're using our data. And by the way, they're right.

Georgie: 19:46

Okay.

James: 19:47

If you look at those license and those terms, like, you gotta really make sure you're using the enterprise edition, that you're opting out of abuse monitoring, you're opting out of training.

Georgie: 19:54

Yeah.

James: 19:54

Well, the easiest way to start is to start with Copilot because most of these guys have Office. And Copilot's embedded in Teams, and Copilot's embedded in Word and PowerPoint. Now personally, not a big fan. Like, the very fact that I went into Copilot last night, my Office three sixty five Copilot, and I said to summarise them with the emails I've gotten for the last two days, and it's already in my okay. Microsoft hosts my data, owns Outlook, owns Azure, and owns ChatGPT, and it couldn't do it.

James: 20:24

Yeah. Because instead of I'm not connected, I don't have access, I don't have permissions that you have to do this. I mean but you can if you have ChatGPT with an Office three sixty five connector, by the way. It's they've still got a little bit of plumbing to hook up. Mhmm.

James: 20:38

But Copilot will enable people to start understanding the potential that a large language at least a large language model can provide. If you really wanna get into some juicier tidbits, you go into Power BI.

Georgie: 20:53

Yep.

James: 20:54

And you start using their smart insights feature, which is some machine learning models that help you identify the two most important things when you're doing business intelligence, and that's correlation and causality.

Georgie: 21:03

Mhmm.

James: 21:03

So I think it starts with desktop tools. Like, Copilot that's a long answer. I'm sorry. It's like the longest

Georgie: 21:09

answer ever. Answer, though, because it can be overwhelming.

James: 21:12

It can be overwhelming. It's overwhelming for everybody. So start in Copilot. Yeah. Start in Power BI.

James: 21:17

Start using Smart Insights and Power BI. Start using Copilot, and then move into your first actual pilot of a machine learning model or an AI model. Mhmm. And if you really wanna get snazzy, go get in an enterprise edition of Claude and start using Claude Cowork and Claude Code Mhmm. Will change your life.

James: 21:34

And opt out of abuse monitoring, opt out of training, and get an enterprise license, and then start playing with that. Mhmm. And then if you the next step is put together an automation team Mhmm. And have that automation team go to one team in your company and literally document every manual tax they do Mhmm. And then automate all of them.

James: 21:53

And then take those people and retrain them on something where you're short in.

Georgie: 21:59

Yeah.

James: 22:00

And I'm sure there's a backlog of things you need resources for, And you just literally created bandwidth and capacity at your company. So, I mean, that's that's really kind of the path I would take. Mhmm. So maybe a three or four step path.

Georgie: 22:13

It brings me on nicely to my next question. So you said on LinkedIn recently, something that stood out to me. So 80% of insurance leaders say AI is critical and only 22% have it in full production. The gap isn't technology, it's about decisions.

James: 22:27

It is.

Georgie: 22:27

Tell me a bit more about that. Why do you think so many organizations get stuck in the pilot phase? And more importantly, what helps companies move successfully out of the experimental phase into the real production deployment?

James: 22:41

I feel like and I I had breakfast at a law firm this morning, so I do feel like part of the answer lies in the general counsel's office.

Georgie: 22:47

Mhmm.

James: 22:48

So really good general counsels of insurance companies work to enable innovation, and really bad ones work to stifle innovation. You know, the head lawyer is really important. And one of the most important things you can do in any organization is engage your general counsel early. Engage them in the development meetings. Engage them in the testing of of AI.

James: 23:11

Engage them in the testing of technology and the rollout. Engage them all along the way. Don't come to them at the end and say, give me permission to do this. Yeah. Because it's a great way to waste a lot of effort and to get it spiked at the end.

James: 23:24

The general counsel has to do their job. Now there are some general counsels that are just doctor nos. Mhmm. They specialise. They got a PhD in saying no.

Georgie: 23:32

Yeah.

James: 23:33

And that's not particularly helpful. It's easy, by the way, and I'm a other life. I'm a politician too. I didn't we didn't talk about that. I I am in politics.

Georgie: 23:44

So zero sleep is what we're down to I

James: 23:46

am in politics. And I was an elected official for two terms. I've been an appointed official for two terms.

Georgie: 23:53

Wow.

James: 23:53

You know? It's easy to look smart saying no. Mhmm. Everybody wants to be a food critic. No one wants to be a chef.

James: 24:02

Yeah. Because being a chef is hard. Mhmm. It's one of the reasons I love Gordon Ramsay. I don't like Gordon Ramsay.

James: 24:08

I love Gordon Ramsay. First, because he has a proclivity for language that is entertaining. Yeah. Secondly, because he he he says it like it is. Thirdly, because he simplifies every organization.

James: 24:21

Kitchen Nightmares was a brilliant show. Fantastic. Walks into everyone, and he says, we're gonna chop the menu by 80%. We're gonna go down. We're gonna get a few things people love to eat.

James: 24:30

Yeah. We're gonna simplify your kitchen. We're gonna clean everything up. We're gonna refresh, do a new menu, new design. I mean, the playbook is the same every time.

James: 24:38

Mhmm. You know? But you look at that same the the same principles apply. Mhmm. It would be very easy for him to just go into those restaurants, tell them their food is a a certain word that he likes to use.

James: 24:53

I'm not gonna say in your podcast. I don't want you to to get an explicit tag, but he's gonna he's gonna go in, and he's gonna he's gonna say your food is beep Yeah. And then leave. But, see, he's not a critic. He's a chef.

James: 25:04

Mhmm. So he actually gets involved and fixes it. And I think that's, like, what we need. We need more we need more chefs, less critics. Mhmm.

James: 25:11

And so it's easy to be a critic. It's easy to say no. It's easy to look smart saying no because you can say, I'm protecting the organization. We're protecting the organization. Okay.

James: 25:19

Are you? Are you protecting the organization? Are you holding it back from being disintermediated? Maybe a lot of these guys won't ever be disintermediated, but maybe they'll get blown by by somebody else. And so I I do work with some amazing insurance companies who have really proactive general counsels and risk management departments that are interested in enabling innovation, and they behave very differently as an organization.

Georgie: 25:41

I wanna bring us on to investment in the InsurTech space. So we are here for InsurTech Week, which is group of well, group, masses of insurance companies and insurtechs that kind of come to London to, yeah, share their experiences, see what what the new trends are. I attended an event yesterday that was talking about investment into the insurtech space and what is right for new startups. And a shift that we're seeing in the space is the rise of founder operators that are leading and building companies that have deep insurance expertise. Do you think that model is gaining traction right now?

Georgie: 26:17

Is that something that you're also seeing?

James: 26:18

I mean, I'm a founder operator. So, you know, I I'm a bootstrap founder operator, though. Mhmm. You know? Thus, the the book.

James: 26:25

Mhmm. Right? I'm militant about using self funding, but it's not always possible to do that. I do think that some of the best companies I'm seeing are owner operator founder led Mhmm. Because they have the vision and the fire, and they carry the they carry the torch every day.

Georgie: 26:43

Bonobar, he's now our chief product officer. He used this phrase in a meeting the other day that we are frustrated practitioners. And I think that really comes through because you feel passionate about the product product, and it's something that you've experienced and frustration that you have, and you're trying to fix it. And I think, yeah, that passion comes across really well in meetings when you are someone that is an insurance background.

James: 27:04

Well, something I say to my actually, in my speeches, you have to be an archaeologist of pain. Have to dig for pain. And if you've experienced the pain that you're solving, you're gonna be way more passionate about it. You're gonna be way more intelligent speaking about it. You're gonna build a better solution.

James: 27:20

Mhmm. You're gonna prompt your LLMs better because you understand the pain points that they and it's hard it's hard to solve problems that you don't understand. That sounds logical, but there's there's a lot of folks in the last twenty years. And I've been now I've got a good bit of gray hair, so I'm in my twenty fifth year of business.

Georgie: 27:38

Wow.

James: 27:39

I started this company in my dorm room in uni, as you would say. We call it college. We call it uni. And I've watched a lot of generations of outside disruptors come into an industry that they don't understand and pain points they don't understand and Mhmm. And spend a lot of investor capital and not make a lot of headway because they're not solving the most important problem.

James: 28:00

Mhmm. To really get traction, you've gotta solve their top one or two problems Yeah. Or they're not gonna buy from you. And you you and you gotta solve it directly. It has to pass the obviousness test.

James: 28:12

They have to look at your software and go, yes. That solves my biggest one or two problems. When you see people take a long time to buy or they this you know, you get a lot of demos, but not a lot of sales Mhmm. You're not hitting a big enough nerve. And so I think that's something that's really nice.

James: 28:29

If you're a frustrated practitioner, then you've really got an axe to grind on the problem you're trying to solve. Mhmm. And it's it's it's lovely. I mean, that's why I personally get embedded so deep with my clients. Mhmm.

James: 28:42

I get to know their staff. I go visit them. I spend time. You know, I've I've got a few clients that I've embedded super deep with because I've got to understand the pain. Otherwise, I'm gonna be building a solution in search of a problem.

James: 28:57

Yeah. And and, you know, if you look at the technology industry and the InsurTech space, most of the capital slash money that's been wasted was not wasted on bugs or bad code. It was wasted on building the wrong thing.

Georgie: 29:10

Mhmm.

James: 29:11

They built a mousetrap nobody wanted.

Georgie: 29:14

Talk to me a bit more about bootstrapping then. So that's a topic on the podcast we haven't actually covered. So, yeah, you you bootstrapped, and is that part of your book that you've kind of given me? So do you think that can lead to stronger, more sustainable companies in this space? Okay.

James: 29:28

Absolutely. Yeah. Bootstrapping forces efficiency. Mhmm. Effectiveness, it's a forcing function.

James: 29:36

Mhmm. Because you don't have vast amounts of capital available, it forces you to make leaner, meaner decisions. Mhmm. It is more possible. There's much less of a technology moat around a lot of these companies now.

James: 29:50

So, you know, AI has enabled a lot more startups. But bootstrapping is is lovely. I mean, what is the definition of bootstrapping? Because most people don't really understand what it means. Yeah.

James: 30:02

It means doing something that generates revenue and profit and then taking that profit and then plowing it into the thing you really wanna do. I what I say in the book is build what you have to build so you can build what you want to build. Mhmm. So I built hundreds of websites. I built a lot of custom software.

James: 30:21

I used profits that I generated to start other companies Mhmm. Under the same umbrella of family of companies. And it forced us to be very cost efficient, forced us to look for solving the biggest problem and most obvious problem first. It forced us to be very innovative in how we delivered solutions. And we've managed to survive as a bootstrap company, now almost 300 employees for twenty twenty five years, you know, twenty five profitable years.

James: 30:50

Mhmm. It's possible. And so I think, you know, ironically, when I sold one of my companies in 2018, I became a limited partner in four VC funds. So I'm actually on the other side of the table Mhmm. On the VC table.

James: 31:04

But I I think that if you can Mhmm. Bootstrap as long as humanly possible, if not forever, but there's some opportunities that are time boxed so severely and are so capital intensive that you know, it's like it's hard to bootstrap a carrier. Mhmm. I mean, you need a lot of capital.

Georgie: 31:23

Yeah.

James: 31:24

You can bootstrap an MGA. And I have a few friends who went from being MGAs being carriers, A rated carriers. Mhmm. But it took them decades. Yeah.

Georgie: 31:31

Of course.

James: 31:32

I I think I heard a really funny phrase. I think it was last week. Like, insurance is is like a a lesson in how to make money slowly. You know? It's like it's it's like it does take forever.

James: 31:42

Mhmm. It takes a long time.

Georgie: 31:43

I need to get it right that it's profitable and a healthy book that you're not just gonna burn through a hemorrhage of cash.

James: 31:48

Correct. Correct. Yeah. Yeah. I mean, it's and we've seen this over and over again in, like, the property markets and insurance.

James: 31:55

Like, you want premium? You can buy premium.

Georgie: 31:57

Lower

James: 31:58

your rates enough, and the brokers will pile in. Mhmm. And then God help you when you have a bad your first bad year.

Georgie: 32:06

Yeah.

James: 32:07

You know? I mean I mean, you saw in the Florida property market three years ago

Georgie: 32:11

Mhmm.

James: 32:12

We had three or four major carriers go belly up. I mean, that's a big deal. It still happens today even with as tight as the regulations are. And, of course, Florida passed a new raft of legislations afterwards to to try and prevent, you know, carriers from going belly up. But, I mean, the reality is it's it's very tempting to chase premium by lowering your rate.

Georgie: 32:35

Definitely. Just circling back then to some of the clients that you're working or some of the organizations that you're working with, the ones that are truly succeeding in driving forward innovation, are you finding that that they are tending to focus more culturally or in a strategic way when delivering the technology?

James: 32:54

Well, I do find that it has to start at the top. If the CEO is not preaching to everybody that they're going to be innovative, that they're going to try new things, that they're going to disrupt the way they're doing things, it is really hard to get it done at a lower level. It's really hard to innovate at a grassroots level. If you're a senior management, your CEO, COO, CTO, CFO

Georgie: 33:20

Mhmm.

James: 33:20

Chief underwriting officer, if these folks aren't in lockstep about the need to change the way that they perform many of their tasks, then everybody will use that as a crutch and an excuse to not change.

Georgie: 33:35

So it's more of a mind mindset shift that

James: 33:37

needs to happen. At the top. They've gotta believe it. Yeah. They have to, like, they have to believe that they need to change or there's going to be severe consequences for the business.

James: 33:47

There are, by the way, gonna be severe consequences for companies who are not interested in changing. Mhmm. Because, ultimately, new entrants will come into the market Mhmm. And we'll do it more efficiently and more effectively.

Georgie: 34:05

They'll get left behind.

James: 34:06

Eventually. Yeah. And you look at the the S and P 500 when the, you know, the Standard Report is the top. You know? Almost that whole list has changed in a hundred years.

James: 34:16

Right? I mean, industries persist, but companies don't. I think we kinda forget, like, just because a company is huge doesn't mean it's gonna be here forever. Mhmm. I mean, Kodak was one of the world's largest corporations, and they they had the upper leg Mhmm.

James: 34:32

On digital photography. And they chose not to do it because it would have cannibalised their film business.

Georgie: 34:37

I think look. Another great example of that is Nokia. I think Nokia didn't want to move into doing touch screens.

James: 34:42

They didn't.

Georgie: 34:43

And then look whether has a Nokia now?

James: 34:45

I know. Nobody. Same with BlackBerry.

Georgie: 34:47

Yes. Yeah.

James: 34:48

Watch any of the documentaries about BlackBerry, and it's a case study lesson on how ignoring customer preferences Yeah. Ignoring new trends ultimately leads to the demise of an organization. Even if they're the greatest in the world, BlackBerry had the entire enterprise market corner.

Georgie: 35:06

Beat the BlackBerry Messenger. Yeah. You would have it.

James: 35:09

Everyone had it. Yeah. I had it. Yeah. You probably did.

Georgie: 35:11

Absolutely. It was it was everything. Need any of well, the AGI was. We didn't need any other feature on a phone. No.

Georgie: 35:17

We just needed BBM.

James: 35:18

Yeah. You needed BBM. You needed the messenger and you their their email client, their BBM. Yeah. And by the way, it was better than the stuff that exists today at delivering enterprise secure messaging.

James: 35:27

It was super efficient, super low, super lightweight, and they ignored. And you have to watch them if you don't watch the movie. The documentary is so so fascinating. And I actually got to I I have a friend in insurance who was in the rooms when that happened.

Georgie: 35:43

Mhmm.

James: 35:43

She worked in the leadership team at BlackBerry and got to watch the whole rise and fall.

Georgie: 35:49

Oh god.

James: 35:50

And oh, yeah. And it was it's really interesting talking to her about it. She was in the room when they watched Steve Jobs' keynote in 2007 where he unveiled the iPhone, and they dismissed it. And, of course, it was the BlackBerry killer

Georgie: 36:04

Yeah.

James: 36:05

And the Nokia killer and the Sam almost the Samsung. And then Samsung obviously partnered with Google and then you know? But I guess the point is that all of us have to be cognizant that there's very few Lloyds out there. Lloyds just passed the test of time. How many years?

James: 36:20

Six hundred?

Georgie: 36:21

Something like that. Billions of billions and billions of pounds worth of premium and a long standing legacy.

James: 36:27

Six hundred years. Yeah. But they're quite innovative. And that, you know, they have Lloyd's lab, and they I mean, they're they're doing some really innovative things. Right?

James: 36:35

But there's very few of those.

Georgie: 36:36

So looking to the future, and as we said, we're here for InsureTech Week. And every time we go to a new conference, there is a new trend. We are back here in a year. What do you think the trend will be in 2027?

James: 36:51

You know, artificial intelligence, the very definition of it, is a field of computer science dedicated to the development of tools typically reserved for humans. Mhmm. I'm not sure there's any tools that are being reserved. I I how how can I put this? The cognitive capabilities of what I'm seeing AI be able to do.

James: 37:11

As someone who's I'm 46. I've been writing code since I was 11.

Georgie: 37:14

Wow.

James: 37:15

This is the first year in my professional life, the last twelve months, where I don't say the words we can't do that.

Georgie: 37:23

That's interesting.

James: 37:24

That's staggering.

Georgie: 37:24

Yeah.

James: 37:25

I look at analyzing unstructured data, text, photos, documents, videos, anything. I mean, I drive a Tesla Model y, a new one.

Georgie: 37:35

Mhmm.

James: 37:35

I don't drive anymore. I don't touch my steering wheel. It opens my garage, pulls my car down the driveway, closes the garage, takes me to the office, finds a parking spot, parks me. And then when I come out of the office, I hit someone, and it comes around and picks me up, takes me home. Mean, we live in the the future's now.

James: 37:50

Yeah. We're there. So in a year, fully unsupervised self driving

Georgie: 37:55

Okay.

James: 37:56

In The US. Yeah. Not in The UK. You guys have some regulatory hurdles that we don't. In a year, that's gonna continue to transform the auto insurance market, I believe.

James: 38:07

I think we're I I I believe this is a a product. I think we're gonna continue to see like, in workers' comp, I think we're gonna continue to see the market soften Mhmm. Because severity and frequency are going down. We have some things happening in workers' comp. But in the technology space and the Insurtech space, I think it's gonna continue to be hard to get funding.

James: 38:25

Mhmm. Like, we're we're seeing some challenges around funding right now. I'm not sure what's gonna be out of bounds for AI in a year. And I mean I mean I mean that seriously. Like, its capabilities, what I'm seeing in robotics because I'm an investor in some early stage robotics companies that are now doing crazy things Mhmm.

James: 38:45

With humanoid robots.

Georgie: 38:46

Wow.

James: 38:47

Like and that are deploying into actual facilities. A year from now, we're gonna be in full scale, full deflection deployment across enterprise on large language models, machine learning models, prediction models, autonomous vehicles, robotics. Like, it's happening now. Mhmm. Starlink's gonna IPO.

James: 39:08

That's gonna change a lot. I mean, Starlink is connecting every device on the planet together in a satellite Internet. I've got two Starlink dishes that I use all the time. So I'm not trying to be like a a, you know, Ray Kurzweil or or Peter Diamandis talking about like but, like, we're at this precipice of a point where we're gonna have the most connected planet we've ever had Mhmm. With the cheapest chip cost we've ever had with the, you know, small chips, not the GPUs.

James: 39:37

Those are gonna continue to be expensive. And I believe there will be another revolution in the next twelve months on models, on AI models Mhmm. That make them consume less power and less GPU. Because we we almost saw that happen, and that's when the the NVIDIA stock tanked was when we saw the Chinese models come on Yeah. Come online, DeepSeek

Georgie: 40:01

Mhmm.

James: 40:01

Come online. There's gonna be another DeepSeek moment in the next twelve months.

Georgie: 40:06

Well, then if we're not careful, what traps do you think that we could fall into?

James: 40:10

Well, I think we can we can fall into the trap of assuming that this is all gonna get adopted at a desk level quickly. Mhmm. And forget that at the end of the day, people have to change. And half the battle in in technology and in InsurTech is getting getting people to come along with us on the journey Mhmm. And training them and teaching them and providing tools new tools for them and helping them in decision support, getting them comfortable with the recommendations that their that their machine learning model makes, getting them comfortable with the summaries that their LLMs are providing, getting them comfortable with I think that's the mistake we can make is we can get lulled into a bit of an echo chamber in InsurTech where we view it as inevitable and forget that we've gotta bring millions of people in the insurance industry along with us for the ride.

James: 41:03

Involve them, train them, retrain them, you know, get them excited and motivated about this, explain how they fit into this new world. So I think that I'm not seeing enough effort go into training, rollout, and deployment. That's the trap that we can fall in is forgetting to to train and retrain all the people.

Georgie: 41:23

And finally, I always like to end on a positive note. What gives you the most optimism for the future of the insurance industry?

James: 41:29

Well, it's the same optimism I have for the future of humanity. Mhmm. I just think everything's getting better all the time.

Georgie: 41:34

Great.

James: 41:35

When you look at the data, it supports it. Mhmm. I think I think we're functionally in our third or you could argue third or fourth industrial revolution right now.

Georgie: 41:43

Mhmm.

James: 41:44

You know, the first one started here in England two hundred years ago, I believe. That's my facts are right. Early eighteen hundreds, textile workers. But it started all over the planet. Right?

Georgie: 41:53

Yeah.

James: 41:53

And if you look at the last two hundred years, average lifespans have doubled. Mhmm. I believe we're approaching longevity escape velocity. Yeah. Do you know what that is?

James: 42:03

No. That's where we live long enough to live to the next invention, to live long enough to live to the next invention, to live long enough to live to the next invention.

Georgie: 42:10

Do you want to live that long?

James: 42:12

Well, I love life.

Georgie: 42:13

Okay. Fine.

James: 42:14

I'd like to live to at least 130. God, no. I wanna see what happens. I wanna go to space. I am hopeful.

James: 42:20

I am an optimist. I do believe that things are getting better. We work on average 50% fewer hours per week than our grandparents did. Mhmm. Childhood mortality is down, like, ninety five percent.

James: 42:33

Mhmm. It's actually safe to give birth Yeah. And have children. We need to have more children because we have that's another problem we have right now is we have a population cliff we're falling off of. Mhmm.

James: 42:42

But I think that's solvable too. I think we we can get cost in line, cost of living. And I think AI helps with that. Robotics helps with that. Mhmm.

James: 42:49

I I know this is getting a little heady.

Georgie: 42:51

Yeah.

James: 42:52

But all of this impacts the insurance industry.

Georgie: 42:54

Mhmm.

James: 42:55

And so why do I think it's getting better? Because I think everything's getting better. I think we're gonna have higher productivity, better work life, better health care, better quality of life, longer lives. And we've gotta work hard, I think, at teaching people how to learn and making sure that they don't stop learning. That is the big, big, big danger that we have.

James: 43:16

And it's been proven that if you have children who only use large language models in school, that they deactivate, like, half their brain. It was terrible. In Texas, they banned all electronics in schools. Did you know that?

Georgie: 43:28

I didn't, but I can believe why they would. Yeah.

James: 43:31

You're only allowed to use paper and pen and pencil in Texas schools now. Except for computer science classes, everything else is paper only because they want them to learn and activate their brain before they start using the tooling. Mhmm. And so I think that we're gonna come up with creative solutions like that to get people to continue to learn and then give them the most powerful tools ever created in the history of humanity. Mhmm.

James: 43:52

And I think that our quality of life is gonna be better. Mine already is. A robot drives me to work every day. A robot vacuums the floors in my house. A robot mops the floors in my house.

James: 44:03

I've got 64 connected devices in my house. I can talk to at any room in my house, control everything in my house, in my car, And it wasn't that expensive. I got it all at Best Buy

Georgie: 44:11

Yeah.

James: 44:11

On sale. So it's like not like it's that big of a leap Mhmm. For people to have a very comfortable life. I do think the future looks a little more like WALL E than Terminator. You know?

Georgie: 44:21

Okay. Good. That's that's I'm glad that you said that because people have the Terminator.

James: 44:25

Yeah. Well, they do. And it it is there is a world there is a possibility that we result either in Matrix or Terminator. In Matrix, we're batteries. In Terminator, our skulls get crushed by the robots.

James: 44:35

I'd say there's a single digit chance Mhmm. Percent that that happens to us. There's a single digit chance that Starlink is Skynet

Georgie: 44:42

Mhmm.

James: 44:43

And that Sarah Connor is having her baby somewhere. You know? I mean, like, I mean, that could be happening.

Georgie: 44:50

Yeah.

James: 44:50

Or we could just all be headed for a

Georgie: 44:54

Utopia.

James: 44:55

We could be headed for Utopia. I choose to think Utopia. I choose to think that we're gonna have better quality lives because it's possible, but it does require people to learn.

Georgie: 45:06

Well, that is all we have time for today. That has been a really engaging and fantastic conversation. So just a shout out to your podcast. So anyone that's listening, also, you should listen to the InsureTech Geek podcast. And then also, do you wanna quickly mention your book?

James: 45:20

Yeah. The book's at jamesbenham.com, benham, jamesbenham.com. It's called Be Your Own VC. It's on Amazon. I did my own Audible on it, you can listen to that.

James: 45:29

I do obviously a lot of podcasting at insure insuretechgeek.com. Companies are JBK. So jbknowledge.com. And then Tara is at tara.insure, I n s u r e. But all of that's available from jamesbenham.com if you wanna just go there.

James: 45:44

And then the only social I'm super active on is LinkedIn. So I'd love for people to hit me up on LinkedIn and hang out with me there.

Georgie: 45:49

I'll put all your details in the show notes so people know where to find you. And thank you once again for coming.

James: 45:54

Thanks, Georgie.

Georgie: 45:55

Thank you so much for listening to Fact or Fiction. If you like what you hear, then please subscribe to our channels. We'd also love to hear from you about future guests. So please get in touch via the link in the show notes or reach out via LinkedIn. Thanks so much again and we'll see you next time.