Is AI Existential for Banking? The Evolution of Fintech Innovation with Ron Shevlin

0:00 I think a lot of banks and credit unions are just not thinking about correctly when they think about AI as think, oh gosh, you know, we're just going to plug this stuff in and it's going to create huge capabilities for us. Well, and it doesn't. And the reason it doesn't is because quite simply, their data sucks.

0:16 They can't access the data that they want. It's sitting all over the place. There's no common definition. The systems and applications don't agree on the actual balances. Working with a bank at a mode that has three systems that think you're at any point in time your account balance is at 3 different levels because of the way they're updated and because of the definitions and things like that. So there's huge issues around that. Open banking in and of itself does not solve any of those problems.

0:45 It just forces people, forces the institutions to move data around, does not take care of the problem. You don't take care of the problem, you don't reap the benefits of the techno, the AI technologies, unless you are improving the quality of the data. Hi there, are you a professional or business owner teaching investors for sharing big ideas? What your presentations just are not keeping up with your vision? That's exactly why I use Gamma app are proud partners. Gamma is an AI powered platform that helps you build beautiful decks, documents and web pages in just minutes. No complicated formatting, no design headaches, just clean, modern presentations that communicate your ideas clearly and professionally.

1:36 If you want to present like a pro, Sign up today using the tool stack link in the description below. Esther creators smarter gamma app real live. Ladies and gentlemen, once in Welcome to the Kingdoms podcast, in the Diary of a CFO, and today I have to be a very, very special guest fighting out of Boston, MA.

1:58 Big welcome, Rod Shamlin. How are you today, Rod? Great, Adam. Oh, thanks a lot for having me on. OK. Thank you very much for sharing your time today with us on the podcast. Today, we welcome one of the most influential thinkers shape in the global Fintech conversation. Ron Scheblin is Chief Research Officer at Cornerstone Advisors and recently named 1 of Fintech Magazine's Top 10 Global Fintech Influencers.

2:28 For more than two decades, he has helped financial institutions and fintech innovators separate not signal from noise using disciplined research and strategic clarity. On this show last year, we hosted Brett King where we explored AI era leadership, digital scale and the structural sheets away from branch economics.

2:49 The conversation examined the vision to dig go deeper into the data, the execution at the economic consequences. Ron Shevlin welcome once again to the Kingdoms Podcast. Great to be here. Let's start with the vision, you know, breakings episode actually exposed AI era leadership and digital skill. And you know, today we're going into deep dive into the data that is the execution and the economic consequences. So when it comes to branch economics and banking, right, Brett argued that branch economics are structurally broken because digital skill is fundamentally more efficient. From your research, is that economic conclusion inevitable across all markets?

3:32 Yeah, I wouldn't argue with Brad on this one, Adabola, but I I would take it a little bit different tact in when I try to convince banks that they need to stop investing so much money in the physical branches. And let's take a step back and ask the question, why do we even have branches? And the reason is because if you go back 60-70 years, at least in the United States, people moved there was a a population shift back in Emo even more than 70 years ago at this point.

4:00 Got to go back 80 years or so, post World War 2, when people started moving out of the cities and into the suburbs. And when a lot of the population was congregated in the cities, you could reasonably expect them to come down to the brand, to the bank headquarters to do their banking. But as people moved out to the suburbs, somebody in some bank somewhere had this great idea, hey, let's put a building near where the people actually live and make it more convenient for them. And so it was really all about convenience, the idea of putting physical locations out near where consumers lived. It isn't simply because people want to talk to people at a moment.

4:42 We're having a great conversation here. You're thousands of miles away, and yet I can see you and you can see me. We don't need to be in the same room together. And the problem with the not just the branch economics, but the problem with the branch concept is that, yeah, people do like to talk to people, but we don't have to physically be in the the same space to do that.

5:06 And so look, I don't, I advocate tearing down or or closing branches. What I advocate is that if you've got a dollar to invest in capabilities, that dollar is far better spent creating some digital capability, which is going to create better convenience, better capabilities, better delivery, better products and services then you're going to offer than by building out your branches.

5:32 What has amazed me for 10/15 years now is this branch of the future concept that some banks and and credit unions like to put out that there's all this technology in the branch Edible. This is ridiculous thinking. Why would you build out technologies in the branch for technologies that people can use sitting at their desk or in their living room or at their kitchen table? So I think Brett has right where he talks about the branch economics, but it isn't about closing down the branch. It is about where do you spend spend the next dollar of investment and I find it really hard to believe that for the vast majority of banks that spending that money in the branch is a good use of money.

6:16 You know, overtime, the cost benefit analysis for ranch economics has been swinging in a negative direction. As you know, digital scaling across the industry is happening. But you know, Bread also highlighted that top fintech platforms now serve more customers collectively than top incumbent banks.

6:38 So does the data support this claim that fintech has already won or is that a narrative that is premature? Well. I think we make a little too much of the fintech economics. If you take a a company of fintech like Chime for example, yeah, it's got 1015 million, 20 million customers, whatever it's up to these days.

6:59 But I think it's important to recognise that while their delivery costs are lower because they don't have branches, their marketing costs are astronomical. The cost it it that they the investment they have to make in creating a brand. A brand name is brand name driving awareness, driving affinity.

7:18 They are spending an incredible amount of money compared to what the average bank has to spend. Now what's interesting is that from a local banking perspective, simply having a branch in town actually creates some awareness and creates some affinity because it makes people think, oh, here's a bank that's committed to our town, our community, They're here and they're advertising. Tell you a quick story.

7:42 A couple years ago, I was at ACEO Roundtable for a community banks and the CEO's were talking about their branch strategies. And one of the CEO's, it was from a bank in a relatively small bank in Texas. And he says, yeah, I'll tell you what my branch strategy is. He's got that Southern accent. He goes, my branch strategy is Basil.

8:02 And everybody's looking at him like Basil AM. I What is Basil? What that means they go, it's an acronym BASTL, Big ass sign, tiny lobby. And his point was with a big sign and a tiny lobby, he didn't need a lot of physical space. He needed to say, here we're in town and that might be perhaps one of the better benefits of having a local presence is that it is a marketing.

8:26 And so without physical presence, the fintechs have to spend a lot to generate name awareness and affinity. And so well their delivery economics and operational economics might be superior because they don't have the branch they are have to outspend on marketing. The other thing they have to do Adamola, which I think sometimes gets overlooked, is while the era customer support costs have to be really, really good. I mean, they're operational capabilities have to be really good because fintech customers can't just take a drive or walk down the street and go to the branch and talk to somebody. And so when you've got phone trees and that prevent people from talking to people and you try to keep them and say, oh, you can do this on the website.

9:11 Hey, fintech customers know you can do it on the website. They signed up and have been interacting on the website or the mobile app since the day they started. When they've got a problem, they need to talk to some embody and so the the call centre contact centre costs are are I think higher for the Fintech than it is for a branch, but for a bank with branches, but again that's a departmental thing.

9:34 So I think we make a little too much of the Fintech operational cost benefits because they do have to spend a lot more on marketing to to create the branding. It's quite a delicate situation, handling the customer satisfaction. It is very frustrating and it's interesting. It's something that came up in the last conversation on this podcast that sometimes you call an institution and an AI part is responding to you and you're like, can I just speak to a human, please?

10:04 It's a delicate situation. So I think that makes some some people very sceptical about doing big banking with the fintech giants. But I think how do we achieve a position of trust in society for the fintech operators? Well, here's another one where I think people are looking at it the wrong way. Trust isn't developed just simply because you can talk to somebody or you trust somebody, a organisation, and I don't know, it doesn't even have to be a fintech or infinite financial services.

10:40 Any provider that delivers and provides a good product or service that's reliable and it works and it gets the job done, people will start to trust that provider of the pricing aspect becomes an important aspect as well. There's a value equation and that's a very subjective thing in people's heads. Am I paying the right amount?

11:03 Am I getting the value for what I'm I'm getting? And what's interesting in banking is that thanks to the old days of free checking accounts, you know, a lot of people said, well, this is a free product or service. It, it allows me to do things and OK for free, it's great. But once banks started charging 5-6 dollars a month, people started thinking, you know what, this value equation doesn't work for me. And as a result, a lot of fintechs swooped in and said, well, we have lower fees or no fees and we won't charge you $35 for an overdraft.

11:39 In fact, we'll see you $200 and basically, basically they were giving away money. Now that might not be profitable in the long run, but as a strategy for attracting customers, it worked. And so the idea that well, the only way to develop trust is people talking to people is absolutely wrong. It's if you are delivering and providing a reliable product or service for a price that consumers the OR the user sees as value added, that's driving a lot of trust. And I think that's an advantage that fintechs have had is that for the segment of consumers they're going after, they're creating a lot of trust because there's that value equation and the product or service delivers.

12:20 Now, if there's a service issue and I can't talk to somebody and I can't get the product, service or problem resolved, that's when the trust breaks down. Interesting. So let's talk about the AI ambition gap. The AI era leadership was a core theme in that discussion. And in your view, what is the biggest gap between executive AI ambition and operational AI capability?

12:48 Well, this is all over the map animal and we've got a, we've got a terminology problem here. I've said this many times in the last year and a half or so, and my colleagues are actually getting really sick and tired of me saying this, but I'm going to say it again one more time here. We got to stop using the term AI. It's an umbrella term that relates to a lot of different technologies under the manner of AI. There's machine learning, there's conversational AI, there's generative AI, there's agentic AI, there's computer visioning, there's all these other things. And it's really the the 1st 4:00 so that I met machine learning, conversational AI, generative AI and agentic AI that are really having the biggest impact on

13:28 banking right now. It's important to recognise that machine learning's been around for a long time, 40-50 years we've had machine learning. The problem of the challenge we've had is that we really haven't had the technology infrastructure to really support the scaling of machine learning, meaning it wasn't until we had PCs sitting on people's desks.

13:50 And by the way, I'm sure that you and most of your listeners don't even remember back when PCs, personal computer that is, started hitting people's desks. You know, there was no Internet back then. In fact, they weren't even tied together. They weren't even local area networks when PCs started hitting people's desks.

14:08 So it wasn't that ability to kind of collect a mass and share data. It wasn't until the Internet came that help, but then we needed mobile banking and then we needed the the cloud computing. So we've gone through this evolution of capabilities that have finally now allowed us to really take advantage of machine learning. But then along comes conversation away, which helps us understand computer human language and to understand it and reply to it whether it's typed or or verbal. And then that got integrated with generative AI capabilities that were powered by large language models that never existed before because we didn't have that infrastructure. And now the agentic AI sits on top of those large language models as well.

14:51 So this idea of a gap is totally understandable. Listen, when personal computers first hit people's desks, there were no PC based enterprise applications. It took 10 to 1520 years for those applications to move off of mainframes onto the PCs. And it's the same thing with AI. We have these. The mainframes.

15:13 Now, but the, the, the initial uses for these are going to be personal AI uses, just like it was with PCs. When PCs hit people's desks, what, what did it give us? It gave us word processing, spreadsheets, PowerPoint, e-mail, calendaring. They were personal Productivity Tools. And that's what's happening now with all these artificial intelligence based whether it's, and it isn't really machine learning because that's more for modelling and prediction. But the conversational AI, the generative AI now the emerging agentic AI, the early applications for this are all going to be personal productivity enhancers, not enterprise type things.

15:55 That's going to take a few years to get there. So I don't know why people are so hung up on this gap. It's totally predictable. It takes a few years. It will be happen a lot faster now that we have all this infrastructure built and what we didn't have back in the mid to early 80s, nineteen 80s. So that gaps, if that really is a gap, is closing very rapidly. The banks that you observe today, are they transforming strategically, right And what is the strategy about having modernising that legacy infrastructure and just calling it transformation?

16:29 Yeah, these are just labels that managers and executives love to just slap on these things, transformation this and automation that. You know, you go back to when computers and even mainframes, nobody talked about Cobalt transformation or Fortran transformation. They weren't these buzzwords that that existed back then.

16:50 It's the same thing now that the tools and the AI technologies are simply technologies that enable something to get done and without a vision for how are you competing and on what are you, what are you selling and to who, and how do you ensure that what you're selling is the right product or service to meet those customers needs and making it more efficient, effective, convenient, value added. The technology is just an enabler and putting the the technology ahead of the business strategy is a huge mistake. The idea that, yeah, we're, we're going to be an AI native bank, who cares?

17:29 Customer doesn't care. The product might suck. If the product sucks, what good is being AI native and an animal? This is, this is too much. There are too many, too many damn influencers out there trying to go, oh, we need to transform around AI and there's going to be an AI transformation. It's like it's all nonsense buzzwords.

17:48 They're just, we're just looking for attention on LinkedIn. Too many buzzwords out there actually. But you know, let's talk about the behavioural banking and financial Wellness. So Brett King on his own episode, he talked about moving, how it was designed to help customers save through behavioural nudges and real time spending feedback. Why have so few banks fully embraced behavioural financial design despite clear evidence it drives better outcomes? Yeah, well a lot of that's cultural and and history. The idea if you've spent twenty, 30-40 years in in any job, it doesn't even have to be necessarily within a bank. But if you've worked in a bank twenty, 30-40 years, you start to look at the world from a very

18:30 internal perspective. How are we organised? What are we good at building and delivering and not taking necessarily an outside in view. And if you're a big bank, it's really hard to historically been very hard to serve niches of the market because you're a huge organisation, you're spending a lot of money on technology and people and delivery capabilities. And to customise and personalise every product or service for every niche segment in the market has historically become an incredibly expensive proposition. And so the idea of how do you develop products or how do you design products from a big bank perspective has always been. So what's the common denominator of features and functionality that we could package and put

19:16 together that will reach the mass market? The behavioural approach that Brett has always advocated for is driven more by a customer segmentation type view that says, OK, let's identify the niches in the market and understand what their individual needs and wants are. And, and from a behaviour, your perspective to how do they like to interact? What channels do they like to use? How do they prefer to use technology or not? And so the behavioural aspect is really a segmentation type approach.

19:49 But you know, I, I don't think that's the end all be all either. Sorry, Brett. And I know he'll agree with me on this actually that the behavioural piece is a is a great view to take, but you've really got to look at it not just from the behavioural perspective, but from the needs slash wants perspective, which is. You know, how does a particular segment of the population need a differentiated type of product?

20:12 Let me give you an example of that. In the US, we've got 10s of millions of gig workers, folks who really don't have what the old traditional full time nine to five job is. They don't have a single employer. They work for multiple employers or more on projects or gigs, and they have very unique financial needs because they don't have a regular paycheck coming in.

20:35 They have very unpredictable cash flow. 2nd, and very importantly, is they're really kind of like small businesses in that they have to invoice for their services. They have to account for these things. They're working for multiple employers even though they're not really traditional employees.

20:53 And so they have a lot of needs. And the traditional bank account offered by most banks, at least in the US, don't serve the needs of those gig workers. So it's great if you understand what they want from a behavioural perspective. But banks need to start looking at these segments from a product need perspective, not just a behavioural perspective.

21:16 All this AI stuff, Adam, all, it may help in actually creating and delivering the product, but it's got to start with more of a vision and understanding of the segmentation in the market and understanding how these things, how they're, those needs differ. And that's not something that AI is necessarily going to just do by itself. It could help from a product design perspective, but you've got to start with some human intelligence first. I appreciate that and I think that, you know, AI can actually help accelerate a product market feet for these financial institutions, you know on talking about AI agents as money coaches.

21:54 If AI agents become integrated money coaches embedded in daily life, what happens to traditional revenue models built on overdrafts and credit margins? Well, so a couple of things. First of all, I think we overestimate AI agents ability to just kind of take over the management of our financial lives.

22:13 And the reason is, is that sure agents can do things, but they do what what we tell them to do. And so in their be some some capability that says, OK, agent go off and open an account with the best provider in the market. Well, now that begs the question out of Mola, who's determining who the best provider in the market is? Now, we've gone through a period here, 1520 years where companies like Google have tried to create product selector tools that say, OK, we're going to capture all of the attributes of a particular product and list them. And you know, what are they?

22:50 Prices, fees, interest rates? You might be looking at some digital banking capabilities to make that decision. But the way we actually decide incorporating things like, well, where are the nearest branches? Have I had an interaction with them? What do my friends think about this provider? We have family members who've worked with this provider before.

23:11 The way we actually make decisions is pretty complex, and I'm not convinced that just because you have an AI tool that it is necessarily helping you to make a better decision. You can go on to bankrate.com and find plenty of providers, at least here in the USI don't know where what which you've got out in South Africa. There are other parts of the world, the US, we've got bankrate.com, Tom, we can go on there and find the provider with the lowest fees or the highest interest rates on a checking account or the lowest rates for a loan or a mortgage. But it's that's never, almost never the only criteria that somebody uses to make a decision. Now, some AI tool or agent might be able to go out and get more information, but MO, what do you

23:57 do when when the the agent says, OK, we're going to move 100,000 of your dollars from bank A to bank B? Oh, and by the way, it ignored the fact that there's a pricing stipulation that says if your balance goes below $100,000, you're going to get hit with a $5 a month fee. Now, maybe the AI tools can, can capture that, but the idea that there's going to be all this kind of money moving around and, and the idea, So what happens next week when somebody else comes out with a better rate or a better fee?

24:27 Does your AI tools move the money again? So I think we're kind of over thinking the real impact of a lot of this stuff because it's just not the way people really want their how to manage, manage their money. They want to feel like, OK, I got my money in the right place and it's doing the right things for me.

24:46 And if something comes up and I got a problem, I might not even know who to call because I can't keep track of how my AI agent is moving the money around. And I don't think a lot of the folks who are pushing a lot of this stuff are really thinking through the details on this. Oh, look at all these great capabilities and we're going to do this and AI agents are going to do all of this. Now, here's the other question that that's really important here.

25:11 Adam Mullen. OK, let's say there is this amazing AI agent capability. Where are you going to get that AI agent from? Who's going to provide that? Google, Microsoft, Amazon, Do you trust them enough to do that? They're going. To move to. AI, whoever's paying them the most to do it. So it's more likely in my eyes that you're going to get that AI agent from the financial institution or fintech that you trust the most and like the most. This is why I had a conversation with Klarna CEO about two years ago at this point when they started announcing all their AI customer service capabilities.

25:49 And I was like, OK, this is this is not the story. They were touting that AI was going to do the work of 700 customer service agents. And I looked at that and said, something's wrong here because I know Clara doesn't even have 700 customer service agents because it outsources its customer service to a third party. And Clara wouldn't have the slightest idea how many customer service agents that company has.

26:13 So I had a chance to talk to the CEO and said, OK, this is not the story. What's really behind the story of your AI use? And he said, we envision becoming the the provider of services to directly to the consumer to help them make the best choice of product and provider, which I thought was very interesting.

26:32 And that's where AI agents could come in. So you see even companies like Klarna want to be the provider of those agents. So I think that at least in the short run at meaning five years, maybe even 10 years out, the idea that AI agents are going to come in and run everybody's financial life, shake things up and put financial institutions out of business is way oversold. You know, go back to the late 90s when Wingspan Bank came out and the online banks started appearing. Everyone's like, oh, these are going to put the banks in out of business.

27:04 And look who's left standing out of all. It's JP Morgan Chase, Bank of America, Wells Fargo all still there, little tiny digital Internet banks not still there. So, I mean, look, five years from now, I could be really wrong on this one, but history says the incumbents don't get tossed out as easily as as you think.

27:26 Absolutely, and delving into financial Wellness, return on investment from your research, does financial Wellness generally increase long term profitability for institutions or is that still more theory than measurable practise? So we've got a terminology technology problem again, because what is financial Wellness here?

27:48 A lot of that in terms of financial health, financial Wellness, financial education, financial literacy, those are all very worthy goals. But I would argue, Adam, all that what most people want is better financial performance, Meaning I want my money to return more. I want it there when I need it. I want the interest rates that I'm earning to be as high as possible.

28:10 I want the return on my investments to be as high as possible, I want to pay the least amount of fees for the services I get, and I honestly don't care if I'm educated or literacy. It's like driving a car. I have a pretty good driving record. Knock on wood. I'm a fairly safe driver. I don't have. I haven't had a ticket in a long time, haven't had an accident in a long time, knock on wood.

28:34 And I don't know the first thing about the insides of the car. I know too get an oil change, I know to keep gassing it, I know to do a couple of things. But I don't have to be car literate in order to be a good driver. And I think it's the same thing with managing your money. You don't need to have the highest level of literacy as long as you have some good behaviours and and good tools and technologies to help you do that.

29:00 And I think what's happened in the industry is that the terms Wellness, literacy, financial health tend to apply to the lower to middle income type of consumer. But it's really more of a spectrum. Listen, but even even the person making a quarter of $1,000,000 a year could have be in a worse financial literacy or Wellness or health space or level than somebody making a fifth of that if there's good money management capabilities and tools that apply to them. So I try to get financial institutions and even the fintechs to think more in terms of financial performance. What's going to help people to get better value, better, higher performance on their money, not just Wellness or health or literacy.

29:50 Amazing. So in terms of metrics for measuring or tracking financial indices for CFOs, right, So what metrics should CFOs track if they want to quantify the return on behavioural banking initiatives that is CFOCFOS of small businesses, especially these platforms? Yeah. So from a small business perspective, you know, they're looking at cash flow, do they have the cash they need? They're looking at, you know, a lot more operational type things.

30:21 And I think it's important for a lot of financial institutions to not just applying the financial Wellness, health literacy to a from a retail perspective, but the small businesses are, are in dire need of better money management type capabilities as well. It's funny how a lot of banks will go out there and talk about their treasury management capabilities where a lot of small businesses and startups don't have the slightest clue what treasury management is.

30:49 They're just thinking about it. As you know, I need cash when I need it. I want to make sure that my cash flow is good and I've I'm able to pay my bills and pay my people and have enough money coming in. So there's a lot of opportunity from a banking perspective to, to better support some of the startups and small businesses.

31:07 And this is where I'll go back to my example of the gig workers. I mean, these are, they don't think of themselves as small businesses, but they they operate and act behaviorally a lot more like a small business than they do with a typical retail consumer. OK. So in terms of the infrastructure, the structural infrastructure of banks, it was stated in our conversation with Bret King that EI based banking cannot function without open banking connectivity.

31:35 Do you agree that Open Data Infrastructure is a direct site for AIDTCP Financial Services? My gosh, there's a huge can of worms on that one. I could come back right now and go, what do you mean open banking? What is that? And once again, we've got a problem around terminology because if you ask 10 different people what open banking is, you're you're bound to get at least six or seven different definitions.

31:59 The idea that some people have around open banking is that it forces the financial institution or the provider to have to share all of the data around a customer because it's they because the customer quote owns the data. That's too simplistic of you. If a bank could just simply say, OK, you know, here's here's your OK, Adam all I will share your banking information with the the other provider you want me to and I tell the other provider, OK, here's Adam, always account balances. What good does that do?

32:31 That's that, that count balance might be different an hour from now after you've made some big deposits or do withdrawals. There's no transaction history involved with that. There's no context around that. So I get where Brett's coming from pushing the whole open banking thing. But what's really more important is better definition around what the data means. I mean, even internally, and this is the, you know, you may, we may get to this in this conversation, or maybe we'll get to it sooner rather than later.

33:03 But you know, one of the things that I think a lot of banks and credit unions are just not thinking about correctly when they think about AI is think, gosh, we're just going to plug this stuff in and it's going to create huge capabilities for us. Well, and it doesn't. And the reason it doesn't is because quite simply, their data sucks.

33:20 They can't access the data that they want. There's different, it's sitting all over the place. There's no common definition. The systems and applications don't agree on the on on the tool balances. You may have to be working with a bank at a mode that has three systems that think you're at any point in time your account balance is at 3 different levels because of the way they're updated and because of the definitions and things like that. So there's huge issues around that. Open banking in and of itself does not solve any of those problems.

33:53 It just forces people, forces the institutions to move data around, does not take care of the problem. You don't take care of the problem. You don't reap the benefits of the technology, the AI technologies, unless you are improving the quality of the data. I wrote a blog post a couple of, I guess a couple of months ago at this point now and it was titled if your data quality is low, your AI strategy will blow.

34:19 And I use that in every speech. I hope it doesn't offend anybody, the terminology, but it's really is the point is that having just the technology itself does not solve any problems because of some of the data issues. And I don't think open banking solves anything either because it's just forcing institutions to move data around and that it doesn't doesn't, it doesn't fix the problem or or accomplish what anybody thinks it's accomplishing.

34:46 Garbage in, garbage out to see, you know, Speaking of our security, quantum and systemic risk. So quantum resilience, the possibility of quantum competing breaking current and creation standards was raised as a serious threat. Our financial institutions, they can quantum proceed and CSB now oh.

35:06 Gosh, so again, you won't remember. You're not old enough. That's not a bad thing. This could be, and I would imagine a lot of folks listening will not remember. Let's go back to history. Personal computers. Oh my gosh, personal computers are going to change. Everything is going to put, it's going to, you know, 'cause all these problems and issues for financial institutions, they adapted that the Internet, oh, this new technology is going to be the death of financial institutions. It wasn't mobile.

35:34 No, it's the death of financial institutions. It isn't. And now, you know, the next set of technology is AI and the one after that quantum. Everybody adapts. There will always be new providers that will come in and and capture share. But you know, if you look at the history of failed banks over the last 30 years, almost if not every one of them, 99.99% failed because of poor risk related decisions, not because of they did not deploy or employ technology.

36:08 So there's going to be a huge shift in, in, in changes and adapt, adoption and adapting to quantum computing, quantum technology. But it's hard for me to think that today in 2026, that a lot of banks, CEO's and management teams need to be that worried about quantum computing right now. And I know Brett, you know, he's a futurist. He his mind, his head's 10 years from now, you know, quick story. I went to bred a number of years ago. I've been friends with him for 1516, maybe even a little more years now.

36:43 And you know, he speaks all over the world. I went and I go, you know, you ought to really bring me along on all your speeches. And he kind of laughed. And in that great Australian accent, he goes, and why would I do that, mate? I said because Brett, you're a futurist. You're telling people what's going to happen in 10 to 20 years from now.

37:00 I'm a nowist. I tell people what's happening now and what they need to do now. So he laughed and I'm still waiting for that invite to go with him on all the cool places he goes to. The Futurist and the Now is so Speaking of now, how should boards think about AI risk, biometric authentication, and cryptographic transformation over the next five years?

37:25 The boards should be hammering the management teams for what problem are you solving or what opportunity are you taking advantage of that AI is going to be involved and they should also be hitting them with why do we need AI? Could we do this with some other technology? And that would make the management teams really think through what is the, the business opportunity or challenge that they are addressing with the technology and making them be able to make forcing them to explain well, what is it about the quote AI technology that that's going to help solve the problem. I think it gets too much buzzword and gets too buzzwordy. And you know, well, we need to do something. And I've said into way too many

38:11 board meetings and I see this over time. If you go back 15 years or so, maybe not quite 20 years at this point, Adam, all, you know, I would hear this in the board meetings all the time. I'd hear my granddaughters on Facebook. What are we doing about social media? And then the management teams would all just sort of run and go and you know, go well, we now have a Facebook account and what did it do for them?

38:35 Nothing. The boards need to be asking and I think many of them do ask smart questions around what's the business opportunity? What's the the OR business challenge that we are addressing with the technology and why AI technologies? Can't we get this stuff done with the technology we already have announced or why can't we all get it done with the technology we have in house?

38:58 And that forces the management team to answer the question of why AI for some of the other stuff. And then the question also that the boards need to be asking is should we have be developing these capabilities internally or can we rely on 3rd party sources to do it? Because I think it's going to get to a point where a lot of banks are going to say, you know what, we're especially smaller banks, mid sized banks in this country, call them 100 billion in assets and below. I'm going to get to the point where they say, gosh, our, our vendors are not delivering this technology fast enough for us. And should we need to be building some better capabilities around technology internally and not just be vendor managers.

39:44 We're also getting to a point at a MO where the whole model around this is potentially getting shook up. And what I mean by that is look when when 12 year olds sitting in their bedrooms can use clawed code to develop applications in minutes, a lot of software vendors need to be rethinking their existence. Why do they even exist when 212 year olds can develop an application and get it out to the market? And so what's happening is the whole model of generating and building competitive advantage on top of technology is really getting challenged.

40:22 It's something that I have embraced for nearly 40 years, the idea that technology can help create a competitive advantage. And I think it's getting to a point where that logic is doesn't take hold anymore. The idea, you know, for the past 40 years is that you had a vision for how technology could create a better product, deliver something faster, create more convenience, and then you went out and built it. And that was a serious commitment of time and resources to develop that capability.

40:54 And then you have to sell it and support it. And so it wasn't easily copied or as easily copied as you think that in terms of really making those advantages, of creating those advantages. But now the time to develop that technology, get it out to market and support it and sell it is minutes and days, not years and years.

41:16 And so it's really time for the boards and and management teams to be thinking what generates competitive advantage in the market today and creates A sustainable competitive advantage because it's no longer just technology. And I think it actually, interestingly enough, shifts back to the human. It still needs to be the human that has a vision for what is the problem that we're solving or the opportunity we're taking advantage of.

41:43 The AI technologies and tools can help you visually that solution and actually build and support that solution, but it's got to come from some human creativity to really define and understand what is the right problem to solve or what is the right opportunity to take advantage of. That reminds me of an episode we've had on the podcast before.

42:03 Title Humans fest AI Next. And with the proliferation of generative AI and agentsic AI these days, the obsolecence, the speed of obsolecence of new technology is very, very alarming. It's amazing. And with agentsic AI of autonomous liability, so if AI agents transact autonomously on behalf of customers, who ultimately carries the liability?

42:33 Look, we love to to pick on the AI technologies where all it could be making mistakes and doing these things and what are the liabilities. But a human is just as likely if not more likely to make a major mistake as well. I mean, the idea of going back to last fall when I'm blanking out on who the company was minted $300 trillion in stable coins they make and everyone made a big deal.

42:57 See, see what the problems are with AI. It's going to create these problems. But reality was is that the tools caught the mistake in incredibly speedy time and fixed things before it became a major mistake. If you go back a couple years, there was a big UK investment provider where some guy fat fingered, you know, a $3 billion trade instead of a $3,000,000 trade. And that rippled through major implications of of issues and problems.

43:26 So I think it's funny as a lot of the naysayers, you know, everybody, I think both sides of the coin have of the AI coin sort of have their weaknesses. The proponents love to say look how this is going to transform everything and the life's going to be great. And then on the other side of the coin are the people say, yeah, but look at the liabilities and look at the problems, overlooking the fact that, hey, people make mistakes too.

43:49 So you know, the answer I think comes somewhere in the middle, but I do think it's inexorable that we are moving towards a very heavy automation computing delivery of of everything and the AI tools and technologies are going to be key pieces around this. Complaining about the liability, complaining about this or that may slow things down a little bit, but I don't think it changes the the overall trajectory of the changes. Interest in.

44:19 So let's talk about the global markets and Greenfield opportunities. So the Greenfield advantage, emerging markets, we're framed as having potential Greenfield opportunity because they are not burdened by legacy infrastructure. Do you see data supporting that advantage? Oh yeah, absolutely. Just look at South America and the growth of new bank.

44:41 But once again, well, they want to put this in a historical context. You know, one reason why mobile phones were adopted so much more rapidly in emerging countries, especially around in South Africa, South America, and in certain Asian countries, is because they didn't have the telco infrastructure that developed companies have developed countries have. So this isn't just a banking specific thing, nor is it specific to just AI.

45:10 It's a history of lesser developed countries from a technology perspective, leapfrogging certain building of infrastructure with newer technologies. But even even those things change. And what's what should be scary for a lot of countries that are more on the developing side is that the change of of pace in the from a technology perspective means that, yeah, you might be not burdened by legacy technologies that a country like the United States has, but in five years, you could be burdened by the legacy of the today's technologies.

45:45 If you're not keeping up on that might go back to your quantum computing type example where that even not just certain countries, but companies that are more AI native in five or ten years might find that to be the new legacy technology. Everything becomes legacy after some point in time. And what's what's kind of scary is that the time from deployment to legacy status seems to be declining very rapidly.

46:11 It's typically a very short period of time nowadays and you know, talking from a geographical innovation standpoint standpoint. So from a research standpoint, where do you see the most aggressive AI driven banking innovation happening? Right now from a functional perspective, probably the the risk areas part of that's because there's been a history of using machine learning from a risk perspective. But now with broader large, large language models incorporating a lot more data points, I think the risk world. I would also say that from a compliance perspective, huge opportunities to to kind of leverage and change the way compliance gets done in a lot of financial institutions to build on mostly generative AI

46:57 capabilities, but to a certain extent agentic as well. It's interesting that I've done a study recently, I do it every year actually, called What's going on in banking. I generally survey between 3 and 400 senior executives at mid sized U.S. financial institutions. Past couple years I've asked them about where they see applications in the use of generative AI and agentic AI from a functional perspective in their organisations.

47:24 And it's kind of scary to me that there are some really low numbers for areas like finance and compliance and even IT where it is going to be totally transformed by a lot of these technologies. And yet there seems to be a a real lack of education around what these tools and technologies can do and will do. So from an innovation perspective, I think what happens is especially in a lot of mid sized institutions, they lock in on a particular function. I was actually recently on a podcast we surveyed the there was about 350 folks.

48:01 And again, could not have no idea if it was representative of any population or anything like that. But we asked them, where are you with your AI in your AI journey? And Adam Mullet was ridiculous. What a high percentage of them said that yeah, we're doing in AI pilot or an AI experiment, really just one. I mean, this impacts every function in the organisation. So, you know, I think, I think part of the problem or challenge is that it starts with even the, the use of the term innovation. Well, we're going to do a test or a pilot. It's like, no, just throw these tools at people, get them to use generative AI to review their documents, start using them to, to build capabilities to answer

48:46 questions about policies and procedures in the organisation. The contact centre has long been a use of, of AI or AI ish type tools around chat bots and things like that. But I think it's also important to remember that a lot of those chat bots were not necessarily AI technologies, they were just rules based engines. But some of the, the leaders are really looking at how to employ, first of all, are truly employing AI driven conversational AI capabilities now starting to use RAG techniques to to improve that and to get more of the human aspect of intelligence into the equation. So you know, this is happening in spots in certain areas. So I don't know.

49:28 I'm, I think I'm at this point rambling and not answering your questions. So I'll go back to focus more on this one. OK. So in terms of tech versus banks, right, is the future of financial services more likely to be dominated by technology companies that do banking rather than banks that become technology companies? I struggle with that because I'm not even sure you know the distinction between a technology company and a bank. I don't think makes any sense to me. A bank is sampling, at least in the US, is simply an organisation that is chartered or licenced to provide financial services.

50:07 What is chime? Is chime a bank or a technology company? At a MOA they don't have a financial licence or charter so therefore they are not allowed to call themselves a bank. But to their 10/15/20 million customers, if you say who do you bank with, they'll say Chime. The idea that a lot of bankers have that you know, well, we're really a technology company that just happens to have a banking licence. No, that's not correct in it because there's an implication of what a technology company is.

50:39 The implication of or the the insinuation when you say a technology company is somebody who sells technology to somebody else, what is the bank? The bank is someone who provides financial services. Now if you look at especially in the US and you look at most mid sized banks and credit unions, the digital banking platforms that they have, vast majority of them are not in house developed platforms.

51:05 They come from companies like FIS and Fiserv and Jack Henry and Q2 and Alchemy and Candescent and other providers as well. Now here's the here's the problem or not the problem, but the challenge once again around terminology. I would argue that to the typical 30 year old consumer, the banking product that they buy is the mobile, is the mobile or digital banking app.

51:32 That's what they use predominantly. Yeah. Every once in a while they have a problem and they might go and have to call somebody, but the product they're buying is a product that is almost 100% delivered and used through that mobile banking or digital banking app. And that app is being provided by some third party technology company.

51:55 What's the difference? It is the bank a technology company? Is the technology company a bank? This doesn't make any sense. The terminology doesn't make any sense, and it doesn't matter whether you're fintech, a bank, a bank, tech company. The terminology doesn't make any sense to me at all. Regardless of where you stand, you have responsibilities to the customers and speaking from a global landscape perspective as well, we have the ongoing prices in what I say the Middle East, right. So what implication would this have on the global financial services landscape?

52:36 Oh, you got to you're every time you use the word global, you scare me because I am so narrowly focused on the US banking market. And I think you're asking a question that you know, really goes more to politics and society and not banking. So I am going to respectfully ask you to move on to the next question and because I'm not sure I have great answers for that.

53:02 One, OK, OK, it's just we'll pass on that one. So Speaking of AI, Aaron, leadership for boards and CFOs. So we have a platform called the Diary of a CFO and our listeners are very much finance leaders and would appreciate advice on how to evaluate AI investment differently from traditional IT spend. So I am fond of saying that a lot of AI tools and technologies are infrastructure.

53:36 We wrestled with this once again historically with every new technology investment like personal computers and Internet and things like that and wanting the CFO's always want to know what's the ROI, what's the ROI on this. But at some point they learned that putting APC on everybody's desk did not have a direct measurable return on investment.

53:57 It's infrastructure. There are certain capabilities you have to have around hardware, software, communications capabilities, things like that, data collection, data movement. That in and of itself, the investment in those capabilities in and of themselves does not produce a direct ROI. It enables something else to get done. It enables those things to that get done maybe faster, maybe with better quality, maybe with less people at a lower cost, or it may enable some something to be sold that actually generates a revenue or it might actually replace people. That's the getting things done with less people. And so there's a cost reduction.

54:40 I think we got to remember that technically speaking, turn on investment strictly the way it's defined means you're either generating revenue or you're reducing costs. That's the only way to truly get return on investment from a return on investment perspective. But that's not to say that there are other and other benefits.

55:00 And so it's important to look at the the process that's being impacted by these technologies and is it enabling something to get done faster? If it enables something to get done faster, there's a ripple effect that might be measurable from a dollar perspective, or maybe it isn't and it just creates more productivity, something some of these tools might require less people to do the same amount of work.

55:25 That's a benefit. Now, the ability to actually generate the return on investment by firing people might not be there, but if it creates capacity, meaning, well, we could take on twice the volume that we're doing today with the same number of people. There isn't a hard return on investment. But that is certainly a benefit.

55:45 And I do think that a lot more CFOs are, are really grasp this now and understand that it isn't simply a hard dollar cost, whether it's a there, it's a reduction in cost of people or a increase in revenue where the benefit is important to look at this from a process and productivity perspective. But some even still, sometimes it's, it's hard.

56:06 I mean, the ability to just use a tool to to have a bank employee query a large language model and say, hey, I've got this customer who's asking about ABC. What do we do when that type of problem comes up? You know, today there are some tools, but they're not very good at providing the answer to that. And generally what happens is somebody has to ask somebody else who then has to go back and respond to that original person. That could take 24 to 48 hours or more to do. But imagine a technology tool that actually answers the question in real time and saves that time. Well, it's only saving somebody 5 minutes of work, but over housing interactions over the course of a year, that starts to add up.

56:54 Are you going to fire that person who's answering answering the questions? Because they no longer have to spend 5 minutes 100,000 times during the year. No, they're going to be doing more higher value added work. In fact, they're already doing that work. They just have to spend some of their time at this lower app value work. So I think it's really important for the CFOs who are kind of looking at the impact of this to not be totally focused on just the hard dollar cost and the ROI, but to look at it from a process impact perspective and just really make some logical, rational decisions and recognise that this is infrastructure. It is going to enable other, It's going to enable either people to do things or it's

57:38 going to enable other systems to do things. And that's where the whole agentic AI type thing comes in. You can't just slap an agentic AI tool into your tech stack and think it's going to have an impact unless you've got a large language model already in place that drives that. And going back from that model isn't going to be doing much good unless there is good data quality, good data definition, and good data flow moving from wherever it is into the large language model. So, you know, the CF OS really need to start taking up more of an infrastructure type view of AI and stop asking what's the ROI?

58:13 What's the ROI? Interest intake? So at the board level, what are the board level questions every financial institution should be asking about AI right now? They need to start by asking how are we really going to be competing in the market over the next couple years? What's going to differentiate us?

58:33 What are the threats to our existing positioning? And that's not even a technology question until they get answers to those questions. And when they get, I should say, when they get answers to those questions, then they can say, OK, how is AI going to create opportunities for us? How is AI going to be a threat to us?

58:51 I think there are sometimes the board's a little too focused on the functional level when they should be looking at this more broadly from a broader strategic and positioning perspective, looking at the opportunities and threats. Anxious. And lastly, the uncomfortable truth, right? So what uncomfortable truth about AI in banking do executives need to hear?

59:14 I mentioned this before, I'm going to reiterate it now. If your data quality is low, your AI strategy will blow. This all starts. It all starts from having great data quality in terms of definitions, appropriateness, accuracy, timeliness, integrated capabilities. It all starts from data at Ebola, and without that, the ripple effects of poor data through the AI capabilities and infrastructure is going to to be a real problem for a lot of financial institutions over the next couple years.

59:49 Amazing take so I, we typically end with the quick fire round or before I put you through the quick fire round, I'm going to put you on the hot seat. So it's actually been a really, really amazing session with you today and it's amazing having. The what I say the elite level thought leaders of my own time right have conversations that share really bad and highly resourceful insights. I grew up reading the likes of the great philosophers like the Aristotle's, the Socrates, the Plateaus, and I wish I could delve into a conversation with them. But now Many thanks to the technology in our agent time, we're able to connect across several kilometres to make these important, you know, conversations happen. So it's a very, very big thank

1:00:40 you from me to you. I say gratitude is good attitude and when a person shares the time with you to share the life with you. So which you are the hearted, who would you like to see on this podcast, right? So she named one name you can invite yourself and one name that I should invite myself. So leave it to names you would like to see on this forecast. Yeah, I'd love to see you get Jack Dorsey the podcast. You know, he has been the creator of a couple of amazing companies, Twitter and Square now called Block. He also just fired 40% of his staff and said that AI is and will be creating the capabilities to replace them. So I would love for you to to put him through the ringer and defend a lot of his actions and

1:01:24 beliefs. So would love to see you get Jack Dorsey on the. Jack Dorsey at ease, right? So I would invite Jack Dorsey myself. So who are you going to invite on the forecast? Oh, I have to invite somebody. Okay, but I got to think about that a little bit. So somebody that I know that that you could put on the podcast.

1:01:44 So I think you should get Wade Arnold runs a company, founded a company called Move MOO. He also was the founder of a company called Bano that was acquired by Jack Henry. And when it comes to really understanding technology and how it applies, especially in the payments world these days, but just in general into the financial services world, he's super, super knowledgeable. He can go into depths around the technology that I couldn't, excuse me, ever come close to.

1:02:14 And if you'd like, I'll make an introduction to Wade for you. Absolutely, absolutely amazing. And of course, shout out to him right now. So we are officially inviting you on the podcast. So we'll go with a quick fire. Round five quick questions and quick answers. So #1 is what is 1 Fintech narrative that the data simply does not support. The data does not support the idea that the impact of AI technologies will have such an immediate impact as some people think it will.

1:02:47 Interesting. So #2 what is 1 AI capability in banking that is under height? 1 AI capability in banking that is under height. OK, ready for this one? The answer is spell check. I cannot tell you how many times I write stuff and Microsoft does not catch the spell check or the grammatical errors. I upload my document into either chat sheet PT or Claude and it comes back and tells me all the mistakes that Microsoft Word didn't capture.

1:03:15 Spell check. Amazing capability, totally under hype. Amazing, amazing, amazing. Just like what I tell people that my favourite productivity tool is the calendar. Calendar. And you're surprised because the simpler the solution, the more effective and more efficientities for me. I, I stand to be corrected #3 what is your three-year prediction that you're willing to stand behind?

1:03:41 Well, first of all, I'm always willing to stand behind every one of my predictions, and I don't even really make that many predictions. Remember, Brett's the Futurist. I'm the nowest I'm looking. At The Futurist. Right now, so Brett's the the guy and he'll even say that he doesn't look out three years from now.

1:03:59 But here's my prediction is that again, I go back to the the research I did that found that, you know, a relatively low, and I'm talking single digit percentages or even low double digit percentages of bankers think that tools like generative AI and agentic AI will not be impacting many of their functions. Three years from now.

1:04:19 Generative AI will become a core P of every function within the bank. Interesting. So what is 1 Leadership behaviour every CEO must develop in the AI era? They must develop not a behaviour, but an understanding that technology can be used to develop a relationship with a customer. Relationships are not just people to people.

1:04:45 The idea that people do business with people, yeah, that's nonsense. I can't even begin to tell you, Adam MOA, how many Apple products I have in my house and I don't know anybody who works at Apple, but yet I got a million Apple products that the attitude or behaviour, it's not really a behaviour, but the attitude, the technology can be critical in developing a customer relationship has to be embraced by every financial services leader.

1:05:11 Interest in. So the last one is a challenge to Brett King. So Brett positioned AI as existential for banking. We have. Would you respectfully challenge him on this? Every new technology development is gets threatened to be existential. And it is. I think we're making too big an idea to make a deal of.

1:05:34 It is every bank is on the Internet, Every. Every bank eventually had PCs on people's desks, every bank eventually had Internet access. Every bank has some form of cloud computing in it. Every bank is utilising mobile technologies now. Isn't it existential? Yes, of course it's existential. Is that a bad thing, A threat?

1:05:55 No. Everybody will adapt to this. Some will do it sooner rather than later. They will have temporary advantages that their lives will be a little bit easier because they will have spent more money, but ultimately have made their lives a little more easier because they were early adapters or adopters.

1:06:14 But guess it's existential. But it's that's a good thing, not a bad thing, so. Finally, Ron, if you had one word on one piece of advice for the next generation of financial services leaders listening today, what would that piece of advice be? Get your data in order. Amazing. I know it does. It is not very sexy, but you got to get your data in order.

1:06:42 Wow. Thank you very much for your time, Ron. So I like to tell the audience, guys, I'll just go for something in your lives, conceive beliefs, tractor pride and achieve. But do you want to leave a charter to anyone before you take your? Hey, shout out to my buddy Brett. Thanks for leading the way on this.

1:07:01 He's been a visionary in this space for a long time now. And I I walk in his shadow, on his coattails to to translate all his great visions into like, What do banks and credit unions do now? Amazing, amazing. Thank you very much once more. Like I always love to see when a person he shares a town with you this year like it because the SI unit for measuring time for measuring life is time.

1:07:25 So thank you. Thanks for having me on at a moment.