Bank 5.0 and The Return of The King with Brett King

Speaker A: Welcome to Fintech Daydreaming, the podcast that dives into the world of banking technologies and the ever changing landscape of fintech companies. We bring you real life examples from global and local thought leaders as well as experts working within the financial industry and seek out the best stories from the front lines of financial services innovation. Where dreams of industry pioneers meet reality. Hosted by Paul Krogdahl and Villa Songtu, this is Fintech Daydreaming.

Speaker B: Welcome back everybody to another episode of Fintech Daydreaming, where Villa, uh, and I get the opportunity to talk to some of the most interesting people in the universe when it comes to banking and fintech and everything else. I'm Paul Crowdel. I'm going to be your host for this episode. Special episode. We've got a very good friend of the podcast and me personally back on to, uh, talk to us. But as always, I never do this alone. I have to bring, uh, my friend, my wingman, my partner in Fintech, Villa. So, Villa, how are you?

Speaker C: I'm doing good. This season is actually going quite, quite great because I think we have said on every episode that this is a special episode. So looking forward to the first non special episode, I suppose, for the, for this year. But, uh, that's a good thing. So, so anyway, doing good.

Speaker B: Non special episode. Would that be you and me at, uh, Christmas drunk, just talking about how good life is?

Speaker C: No, that's actually a very special episode.

Speaker B: You know, I'm actually, I'm having an interesting thought here, Vila. I have decided that I'm going to categorize myself as a trendsetter. Not a sort of first of a kind trendsetter, but a trendsetter. Because very recently I published a book and then every other fintech and banking extraordinaire and their dog decided to publish a book as well, which means that my ranking went very rapidly down on, um, Amazon. But I'm very, very proud of all of my fellow, uh, authors out there. I can call myself an author now, having written a book, uh, publishing books as well. And I am so stupidly excited to have, um, the, the king, see the pun here? The king of the banking 2.0, uh, 3.0, 4.0, all the way up to. I'm sure we're going to end up with banking 20.0 at some point in time, but I am so excited to have our friend Brett with us. Brett, how are you?

Speaker D: I'm well, man, you know, it's uh, uh, I want to take exception at your trend setting, um, characterization, though given that I've been at this for 15 years. I would say I just started little bit earlier than you. But I am very proud of you. Um, I'm very proud to have had a minor involvement in the project and to welcome you to the very exclusive club of best selling authors in fintech and banking. Welcome.

Speaker C: So in short, Paul, take a seat.

Speaker B: Yeah, yeah, sit down, shut up and see how the.

Speaker D: I tried to do it in the nicest way possible.

Speaker B: Oh, fantastic. Wonderful. But I mean Brett, you've, you've sort of gone off in multiple different directions. You've written an awful lot of books around banking. Then you've sort of stepped out of. Hang

Speaker D: on, did you say I've written awful books on banking? Come on. That

Speaker B: as well. I mean I have to be honest sometimes. No, awful lot of books on banking. I think that's what I was trying to say. Anyway, then you've sort of stepped out of the focus on banking and moved into futurism. You're doing a podcast on futurism. Now you're back again and you've written uh, banking 5.0, uh, which is. Let's be careful here, we're recording now. Hopefully this will come out just before the book is published. But what to you has been the major shift from the last one. 4.0 to this one. 5.0?

Speaker D: It's obviously AI, and in this instance, um, it's my understanding of how, um, completely and how extensively agentic AI is going to reshape not only banking but industry, uh, in general. Um, I, uh, think as a futurist, I wrote about AI, uh, in my book Life in the Smart Lane In 2015, I wrote about what life would be like with a personal AI agent, for example, and that's just now starting to come to fruition. But I didn't really realize the extent to which agentic AI was going to change the way we think about the operation of the world, uh, particularly from a commercial perspective. Um, and uh, that's coming and it's coming fast.

Speaker B: You're saying that the fundamental shift here is AI? What about everything else, like quantum and other things that are happening at the same time? Do you see all of this playing into each other or is it really just AI, agentic AI, it's going to change everything?

Speaker D: No, no, no. I think, um, we are, I guess this is the new renaissance, right? And uh, we have a series of technological breakthroughs, mathematical breakthroughs, uh, that we're going to see unfolding a whole lot of new knowledge for uh, mankind, uh, material science, we're going to eliminate disease, most likely within the most chronic diseases within the next 10 years or so uh, as a result of this, and that's even before into quantum, which is going to help us unlock uh, the mysteries of the universe in ways that we couldn't imagine. So the amount of knowledge that humanity is going to accrue and unfold in the next, uh, decade or two is the equivalent of probably a thousand years of scientific uh, progress in the old days. Right. So um, it's enormous to think about the change that that's going to thrust on human society. And that's of course my biggest concern is that um, it's not just about how banking is going to change, but it's really about what it means to be human in this sort of technologically advanced, uh, era of the singularity.

Speaker C: Since we're talking about the upside and the future that you see, obviously as a futurist, you see it as a very optimistic future. But let me try to turn that around a little bit. So what could derail this progress? I mean we've seen this progress now with AI agentic happening really. We're seeing even flying cars from Elon Musk now. Um, it's all crazy stuff. But what could happen that would derail this progress and you know, take us to a bad direction instead of the future that we are seeing in front of us?

Speaker D: Uh, this is an interesting question. I get asked this a lot in terms of, you know, can we stop it, um, and what could we do? What could prevent AI from um, these changes? The problem is that the advances we're seeing in AI are tied to the market. So absent the complete uh, dismantling of capitalism, I don't see AI being prevented or stopping in terms of these trends. Which means that most likely we're going to see a period of chaos. And um, my fellow futurists like Gerd Leonhard or Mo Gadat talk about a 10 to 15 year period of chaos as we adjust and adapt. Um, and that is, um, Bill Gates, uh, talked about this on Ezra Klein's show recently as well. The risk is not AI like Terminator robots coming and killing humans. This is not the greatest risk to us right now. The risk is the dismantling, the rapid dismantling of human systems, the way we govern, the way economic run, um, the collapse of um, um, human capital in terms of labor, and all of these things happening very rapidly in a short period of time because of the level of automation that we're pursuing and we're pursuing that because the market wants these capital returns, because we've entered a period of slowing growth because of declining birth rates. And um, we know um, that most of the growth in the 20th century happened from uh, increased consumption, that increased GDP. So as birth rates are declining around the world, growth is harder to achieve. So investors look for capital growth through uh, productivity and M. The best productivity tool the planet's ever seen is AI. So you apply AI from a productivity term means you don't employ humans anymore, you don't have wage distribution of wealth anymore. The whole economic system potentially collapses. And we're seeing evidence of the frustration of this system in places like France and elsewhere right now. Um, and uh, I can sort of talk about some of those structural elements as well if you like. But um, the chaos of the end of these human dominant systems, um, and how large scale autonomy is going to change is largely inevitable unless we decide, uh, that we don't want growth as a primary measure of economics and that we want to find a different economic system, system of operation, which is unlikely before this collapse happens, if you like.

Speaker B: So if we do

Speaker D: so,

Speaker B: uh,

Speaker D: it's not all optimism,

Speaker B: but if we do have a fundamental collapse of the financial infrastructure, what's the role of the banks?

Speaker D: Well this is really the point of Bank 5.0 from an architectural perspective is that if you look at the way we're moving, we're very clearly making a shift from a posture perspective to large scale autonomy. Um, and China probably leads the way in respect to this with their automation of supply chain, their dark factories, their autonomous factories, autonomous vehicles, autonomous ports and shipping capabilities, et cetera. Um, and in that environment your primary customer is no longer a human or a corporation, it's AI, it's agentic. Um, and so we're really seeing the emergence of a new financial services system, one that's built for machines, not for humans. And the more autonomy we put in the system, the more that the banking system will have to cater for this autonomous, uh, layer of infrastructure and the less important humans become in that loop. Which, that's a big fundamental shift. Because if you think about the way we have regulated banking, the compliance aspect of banking, we always assume that there's a human at the end of the chain with a wet signature on a piece of paper. That's legally how we've thought about it. And that itself is, uh, you know, fundamentally broken. In the future world of banking, um, the products disappear, um, you know, sort of turns the bank into a pure form of utility rather than the product, uh, structures that we have today. So uh, you know it is a pretty fundamental change in banking, the biggest change in what we think of as banking in the last 500 years for sure.

Speaker C: How much do you think banks have actually say in that and how that will play out? Because the uh, we look at for example when Meta launched Muse, uh, then a lot of people started experimenting using uh, giving access to their E banking and screen scraping all the things out of there, executing payments on behalf. And the reaction from the banks was uh, at least JP Morgan, Citi, few others I believe blocked these AI,

Speaker D: blocked the agents. Yeah,

Speaker C: so yeah. So I mean is other banks able to dictate all uh, this in the future or is it just like trying to put a stop sign in front of a moving freight train? Or how would you describe it?

Speaker D: That's a great illustration. I actually should have used that for the book because that is a perfect illustration. Look, the motivating factor behind the growth of AI is the market. And so it's not a stop sign in front of a freight train, it's a stop sign in front of the market. And the market wants this to happen. So um, I'm sorry, Jamie Dimon, I'm sorry. You know, you don't really have a choice in this. The problem that um, Chase has identified and that other banks are identifying right now is we don't have um, identity schemas to identify agents coming in that they are authorized. So this is a failure of the regulatory infrastructure around identity. Um, but it's not going to be the first regulatory point of failure. Um, we are not going to be able to supervise agentic AI operation in banking unless we have agentic AI at the regulatory level. So we need regulatory technology infrastructure to come in to safely regulate this sort of emerging financial services system as well, neither of which we have right now. So this was a big part of the focus of uh, the book as well. Um, so the blocking of the AI that's happening right now is a short term strategy. What I would rather see banks like Chase and Citi do is say let's invest in a digital identity authority or digital identity infrastructure that can make this safe. Because we're not going to have really any choice of participating in this system at some point in time. That could be done fairly simply because you could um, uh, you know, tokenize identity for your agent. You could link uh, transactions specifically, um, you know, with, with the idea your identity information parsed in the agent's identifier. You know, there are ways that we could sort of hack A, uh, path around this. But ultimately we need a digital identity authority for machines.

Speaker B: I think we're sort of sitting at the moment in a bit of a transition where as, as, as you've already said, to a certain degree, we're on this journey. Your book is looking at a horizon which is not next year. It's looking a little bit further, although it's taking us on a journey. I've had the pleasure of reading your book already, so, uh, I'm not gonna, I'm not gonna divulge too much. I want the people out there to read it. But at the moment, we're sort of stuck in a, Almost a dilemma where the agents, the technology is advancing quite rapidly. The banks aren't quite sure how to deal with this. The humans, the users, are struggling with how do they put the boundaries around this, how do they get the intent correct. And at the same time, the agents are hallucinating, they're drifting, they're playing, they're. I mean, very recently we read an article. They're

Speaker D: swarming.

Speaker B: They're swarming, yes. They're stealing people's cryptocurrencies so that they can fund their own things and everything else.

Speaker D: CPU time. Yep.

Speaker B: Yeah. So I think there's an issue here of not just trying to get the regulation in place and getting the banks aligned to how they're going to deal with this. It's also about us getting to a point where we can, not only as financial institutions, but also as individuals, actually trust these agents. And I'm wondering, what are the banks going to do in the meantime? Because I hate to say this, not that I want to say that we're all stupid, but humans sometimes tend to be stupid, and therefore they rely a little bit too much on these agents. I mean, there was a. Was it Elon Musk said, you can take a photograph of your credit card? Yes. You know, wow, let's slow down. That's a great

Speaker D: idea.

Speaker B: Yeah, let's slow down a little.

Speaker D: Hey, Grok, show me all the credit cards you've already got from other users. Yeah, exactly. Um, yeah. Um. Well, I will say, um, you know, I look at trust as very different in the AI world. Um, where trust, you know, when we talk about banking or we, you know, it's a better illustration is actually healthcare is the reason I go to a doctor and ask a doctor's opinion on my symptoms when I'm not feeling well. If I've got an illness, is I trust that the doctor has more knowledge in this space than I do. Um, and the technical term or the scientific, uh, term we give to this is information asymmetry. He has more information about this topic than I do. So therefore he's in a more commanding position to offer advice on this. And this is the foundation of every advisory role that we have in our economy today, whether it's an accountant, a, uh, tax estate advisor, whether it's a healthcare provider, uh, uh, whether it's a bank or a financial planner. And um, um, you know, when we look at AI, AI is much better able to absorb information that can be turned into advice today because it can process so much more information. So we're losing information asymmetry to the machines. That's definitely going to happen. There's going to be a certain point and for many people, uh, like myself now, I'm already at that point where I'm not going to trust an individual doctor anymore to give me advice. I'm going to be in the doctor's surgery using ChatGPT and using Claude and using these other models to um, test the assumptions here in real time. Uh, I just did an interview, uh, for the Pod for the Futurist podcast, interviewed Jean Baptiste Isnard, who's working on sovereign, ah, uh, AI systems for government, for automating government, showing a really interesting example of this where he had a friend testing one of his models and fed in the information, um, the blood work and recent tests and things like that. And the AI said you need to go to the hospital right now. You need to go to the hospital right now and you need to ask for this type of doctor and you need to give them this information. And the doctor actually was debating with the AI, uh model in, you know, at the clinic, at the emergency room. And it's like, uh, this is unbelievable that you. There's maybe five or six people in the world that really understand this, you know, and that's, I think, a good illustration of what we're going to see in banking that the trust is going to shift to, you know, this, uh, um, to AI because of the information capability, but also because we're just going to have a lot more faith in utility. And that's already started to happen in financial service with wallet providers and other things. Now you can see it in China. Alipay and WeChat Pay are much more trusted in their environments than the traditional banks because that utility infrastructure we've built, the next generation of financial services infrastructure is just much more reliable and so much easier to use because it's been designed, uh, experientially Rather than from a product perspective.

Speaker B: I have to admit I'm very positive about AI. I think that, uh, what's happening in the developments around agentic AI is fantastic and I can see agentic capabilities growing. Absolutely. But I also have a little bit of this notion that an idiot with an advanced tool is still an idiot with an advanced tool. And a little bit back to what you were saying, that if you, if you rely too much on the tool, uh, you have potentially the risk that you will trust more the false information you get from the tool, even though the tool might be correct, but not 100%. So at what point do we get to where the banks lose that trust? Or even more importantly, would a bank's agent be happy to tell a customer, yes, you're right, I suggest you move to another bank?

Speaker D: Yeah, that's an interesting one. Um, uh, I mean, uh, I don't know whether the bank's agent is going to need to be involved in that conversation because, you know, the more I think about agentic in terms of its operation, um, and let me take you through a train here in terms of how this transitions likely to occur and why it's a problem for banks in terms of this concept of a primary financial institution and all of those things that we've had, uh, in the last one, uh, hundred years. So today if I need to send you Paul, I need to send you €100, for example. What's the process? Well, I'll go into my Revolut account if I can get into it, um, and I go into my Revolut account where I've got some euro stored and I have your phone number so I can send you via Revolut easy. But, um, if I didn't have that information, I would then say to my personal AI, um, I need to send Paul €100, here's his information. What's the best way to do it? And AI might come back and say, all right, here are three paths. Uh, you can send it as a wire transfer, you can send it, um, uh, using wise, um, and you can send it using stablecoins. And here's the pricing and uh, here's the time. What's your objective? You want to get there as fast as possible or do you want to get there as cheap as possible? So over time I develop this process. If I have to do this type of transaction repeatedly, where I'm starting to choose, I'll choose this path. This looks like the best path. Over time my personal AI is going to learn which path I like and uh, at some point, every time it recommends, I'm going to say, yes, that's the right recommendation. You got it right again. Um, and at a certain point I'm going to say to my personal AI, just send €100 to pull. Right? And at that point, um, as a bank, I've sort of lost control over that process. Because if you extend that analogy, then to credit or other aspects of our banking relationship, once you have that level of trust in your personal AI, the AI or the agent is going to choose. And the way the AI agent chooses is it spawns all of these sub agents. They go out and do the investigation. Those agents come back and report back to your agent. It makes a decision based on your preferences and executes. At no point in that process is the agent saying, oh, but hang on, what about airline miles or cashback rewards or, um, these sorts of things that we have traditionally used to attract human customers. And uh, all bets are off in terms of whether you're the right bank in that scenario at that point.

Speaker C: So you're describing a world where everybody has their own personal AI agent, effectively in one way or another provided by one of the big ones, or even having a real personal one. But nonetheless, the agent is personal, it's yours. So you're saying that this, uh, kind of competing vision where banks provide agents to their customers in order to do their banking through the agent of the bank provided agent is not a scenario that you believe in, is that right?

Speaker D: So what I see is that, um, you know, you're going to, and we already see this developing. Right now a lot of banks are going to be putting a semantic interface in their app, right? So this will be like a chat, um, window, uh, you'll be able to talk to it, you'll be able to use text, uh, you know, text instead of voice if you wish, and you'll tell your, the bank's agent, you know what you want, or the bank semantic interface and it'll go off. And then, um, within that banking environment, give you some options and, but fairly quickly we're going to leapfrog that where your personal AI is then talking to the bank. It'll be a discussion between your agent and the bank's agent. This is the engagement layer that we talk about in orchestration in Bank 5.0. That's the engagement orchestration layer. Um, there'll be a negotiation back and forth. But remember that this is not happening in isolation, that if I've asked my personal AI agent to do something unless I've directed it Specifically to go to my bank. It's going to be having that conversation a dozen times with a dozen different um, financial services platforms to find out the best approach it should use. Because your personal AI agent is very outcome based. It's not brand based or loyalty based in that respect. It's trying to achieve the primary outcome that you've tasked it with. So that bank agent um, modality that you talked about I think is sort of short lived. Having said that, some people will still find it a hard habit to break using a banking app. Um, but even then, um, if you think about the banking apps of the next few years that are developing, we describe a world in the book as very different from the banking apps today. Whereas you will tell your banking app what you need, I need access to some credit to do this purchase, um, and, and the agent will work away and it will generate an interface with the options for you in real time. So it's not really like we think of a banking app today. It's very liquid, generative, uh, um, experiences based specifically on what you're trying to achieve. It sort of becomes really outcome focused.

Speaker C: Yeah. So once upon a time, I mean you are well known for being one of the co founders of MovenBank, one of the leading first neobanks. Kind of almost started the era of neobanks that we're seeing coming to fruition today with likes of Revolut and others. But take that thinking. So if you would start a bank today that would be optimized for this new era of AI agents, uh, what would be the kind of first principles that we would do differently perhaps than traditional banks are doing in order to make that bank the best possible option in this new agentic world.

Speaker D: I think this is a good question. Um, and I've thought about this and actually I'm working with a couple of friends on toying with this right now. So uh, it's a timely question. Um, the key understanding is that by 2030 your main customer is going to be agents, not humans. Um, so we're not that far away. Paul was talking about not in the next year or so, but certainly with the next two to three years this trend is going to be I think apparent. Uh, um, and then what happens to the way you deliver banking in that environment? You've really got to have it optimized for very rapid decision making. Um, the options that you present have to be presented in agentic friendly format. So it looks more like uh, um, when you ask ChatGPT to design a prompt for you or Claude to design a prompt for you. What you get back in that text is sort of what I would expect from a bank to send to an AI agent rather than, uh, what we think of today. Um, and then really just sort of trying to optimize around outcome to get to you, to the outcome as quickly as possible. But that needs that context also of, um, how do I make your money safe, how do I make your money work for you, and what are your objectives and goals as a human, as an individual? And how does that frame the way I, uh, give you a recommendation? So every time that I'm recommending that utility of the bank, how am I framing that utility based on what I already know about you, and how can I improve that knowledge to provide better context for you so that I become that sort of primary form of utility that your personal AI agent is using?

Speaker B: Let me break all of this apart a little bit and maybe this will take us off on a tangent. And then I've been a bit of a rabbit hole. But if we look at uh, the advice that industry leaders have been giving to the banks, uh, for the last five years, we've talked very much about the fact that you need to move from being output focus to become outcome focused. You need to focus on the customer journeys. You need to understand where it is that your, your touch points are, uh, with the customer as they're going through their flows in their life, etcetera, Very much like you just said. But at the same time, what we're saying now is that we move to an agent based economy. The customer will have their own agents that will be firing off through swarms, whatever else, to get, you know, touch points to one or multiple financial institutions where it's asking for a slice a bit. It's no longer an outcome, it's I need this and I want it now and I want it as quickly as possible. And if we're then separating the consumer away from the financial institution where the interactions become purely agent based, the outcome, the journey ends up being actually embedded into the customer's agent, not into the bank's agent. Which means that you were actually now suddenly potentially telling banks to pivot completely and say, actually, whoa, whoa, whoa, stop. Yes, we did say you need to be focused on the out the customer journeys, but actually that's wrong. You now need to focus very much on output, very thin, focused output in your agents and in your agent layer so that the agents out there can come and nitpick from a plethora of financial institutions to satisfy the Outcome and the journeys for the consumer, the customer, which no longer is your primary customer, they're just a touch point in a small dot for you. Is this where we're heading to or have I just completely blown things apart? Brad? It,

Speaker D: uh, it's definitely one scenario that's possible. But I do think that just like let's again use the healthcare analogy, um, I could go out to multiple doctors but the doctors who have the best data history on you as an individual, on your overall health, um, the latest test, the latest blood work and so forth is your genome for example, is going to have a better shot at answering that than every other doctor that you ask the question of. And I think the same is true in financial services. I still think there is personalization to be had here which is based on your financial history. This is why banks building out these models of transactional and behavioral financial services become so foundational in terms of what comes next. Um, you can't have siloed data. You need to have your data lake. You need to have a foundation model which understands financial behavior and that needs to, that is your secret sauce in being able to stay in the loop of a personal AI agent's choice of utility when it comes to decision making. Because again remember your AI agent, your personal AI agent is very outcome focused but that outcome has been been molded by your behavior, your desires, your wants that it trains on over time. So the more closely a bank can align with that uh, the more chance of success they have. So it's not a zero sum game but it's also one where you do have some control over this as a bank if you really understand your customers needs, desires, wants and behaviors more fully. And this is a challenge for banks today because banks aren't asking this question. We don't know what the intent is. The simplest example of this is a mortgage buying a property. Um, and uh, I'm going through that process in Dubai right now and I can tell you um, I'm not receiving calls from my banks of record asking me if I want a mortgage because they don't know that I intend to tend to buy a home here. Um, but so many other organizations do. Um, you know Facebook does because I'm using WhatsApp, uh, for this and I'm interacting with ads on Instagram. Google knows because I've done um, property searches there. Um, Apple knows because of the apps I've got on uh, my um, uh, iPhone. There's so much uh, data out there that shows intent and this is really, I think the Question banks have to ask themselves is what data do you need in this environment to really be a competent partner for your customers in the future? And it's not product data and it's not optimizing product. It's really optimizing for the outcome for customers. It's really understanding the customers. And AI gives us an enormous tool set to be able to do that today if banks are willing to invest in building out their own foundation models and capabilities. In that respect

Speaker B: it does still have a dilemma of the fact of whose data, what data, where the data is, whether it's with the bank, whether it's with the consumers, uh, agents. We're moving towards that to a certain degree with open data, the evolution zone from open banking. But it's still a journey to go there. Are we stuck in a paradigm now where there is too much happening at the same time to enable banks to be able to transition, to modernize not just their IT but their governance, um, their ways of working, their culture to be able to actually grasp this in time? Or is the answer going to be just what JP Morgan has done and just say close all the doors, let's uh, stop and hope that we can weather out this storm.

Speaker D: Well JP Morgan will not remain the largest bank in the world if it continues down that path, that's for certain. Clearly they are working. They've led on tokenization with deposit tokens and things like that. So I think this is a short term strategy maybe to push regulation to some extent or maybe just to give them some time time. Um, but in 10 years time let's have a think about what the world of banking looks like. Paul, to answer your question, I think the largest uh, emerging banks in the world will be AI native banks. So think about the next wave of fintech. So we have Nubank and Revolut getting a lot of uh, traction right now in terms of awareness. Um, and their valuations are picking up and their spreading out and they're growing very rapidly. In 10 years we're going to have the AI native equivalence of this uh, um, developed in various markets to capture agentic behavior side by side with stablecoin agentic accounts versus human fiat currency accounts and so forth. Right then you'll have the fintechs that have very agile technology stacks that have incorporated AI into their capabilities capabilities and have adapted. Um, but let's keep in mind that they too have legacy infrastructure now, particularly fintechs that are 10 years plus sold. So they have to get over that legacy and start thinking about their new technology. Layer but they're more culturally and technically agile so they've got a much better chance of making this transition than incumbents. And then the third layer of players 10 years out is, is the incumbent banks who've been able to adapt to this. But this requires a level of cultural and technical agility we haven't yet seen from most banks. There are a few isolated examples and I'm seeing some of these banks emerge in terms of uh, sort of cultural shift. Uh, that's pretty aggressive. So there's going to be some there. But that's how I sort of see the tiers of leading uh, banks in 10 years time.

Speaker C: So we mentioned a couple of banks and fintechs already in this episode. But I mean can you mention any that you see today that are doing an exceptional job of becoming that agentic AI bank or a fintech, uh, any shout outs effectively?

Speaker D: Well, you know um, I think what uh, both Revolut and nubank are doing from the interface perspective and they've got some different approaches, is positive and they're definitely dipping their toe in the water. Um, but I would probably point to Stripe because I think Stripe is a great example of what next generation financial services infrastructure looks like. They have the smart M contract capabilities, they offered stablecoin accounts uh more than a year ago now in 100 countries. Um, they've been able to use artificial intelligence to reduce fraud at a significant rate, 40% decline in fraud rate, uh, using AI. So you know I would say from an infrastructure perspective Stripe is probably the best example that we have seen. Um, and you know, I think to be clear, uh, um, you know I was reading a post with Richard Turin today, um, you know, on LinkedIn and he was saying that you know um, we had this uh, expectation, you know, probably a year or two ago that all banks would be operating stablecoin companies, custodies and things like this, um, and uh, adopting blockchain uh, for transmission and things like that. And it hasn't quite happened like that but certainly we have seen banks having to respond to the types of moves that players like Stripe and others have made. Even the recent swift announcements about how they're changing their network has been forced on by what's happened in uh, the defi world and the increasing risk that stablecoins and stablecoin transmission capability represent to organizations like Swift, um, MasterCard and Visa's posture on this is I think a clear indication of that as well. So um, we are seeing the uh, infrastructure layer definitely uh, morphing to sort of cope with this,

Speaker C: you mentioned MasterCard of Visa, so I have to ask. We've seen both MasterCard, Visa doing a lot of, of stuff in agenda Commerce, so enabling AI agents to use Visa and MasterCard cards effectively to make purchases on behalf of the user. Uh, how do you see the card network role, uh, in this, let's say five, ten years out? Uh, do you see them as uh, dominant and relevant as they are today, or would they be less of a factor in the future?

Speaker D: Well, full disclosure, I advise MasterCard as a resident futurist in their uh, um, operation out in this part of the world in the EMEA region. Um, but I've had had this conversation with Michael Maybach and Ken Moore and others and I um, think Ken and I share this view that MasterCard M just becomes a data network in the future. The card is not important. The pan, the number, the 16 digit number that we've had on the card number is certainly not long for this world. We've already tokenized cards uh, for um, wallet payments and so forth. I think the plastic and the card number disappears fairly quickly. Some people will still keep vestiges of their plastic card. But ultimately um, we're really looking at um, a data transmission network that's optimized for these types of uh, transactions. The really interesting thing to think about is um, twofold. Um, one is why do we only do the authorization on a network basis? You know, as the last step of the credit card transaction. The final oath does we go to the bank, we ask the bank to um, um, approve or deny the request based on the data we're sending them. We get that authorization from the bank and we proceed with the transaction. Um, I don't see that the network needs to work like that anymore. We can have identity um, as part of uh, sort of instantiation of the packets that flow across the network. It doesn't have to be a single identity uh, point that requires pre identification of a customer at a bank. This could be done in real time with the identity infrastructure sort of captured in the packet. Um, and then I also think we need to think very differently about um, how we manage hallucinations and fraudulent transactions. Because right now all we do is we have chargebacks when a fraud has occurred and then either the bank carries the weight of that or the customer carries the weight of that. But I think in the future we're probably going to develop network capabilities to do rollbacks where we say that was a bad transaction, the agent made a mistake, or um, it was a criminal entity, uh, doing that fraud and we able to reverse that transaction. That's the sort of capabilities we could build into blockchain. So I see that um, we could have a bit more flexibility on that front in terms of transaction rounds as well. I do see sort of the network developing into more of a data network than um, just purely a transactional payments network for players like MasterCard.

Speaker B: We, uh, as always tend to run out of time rather rapidly on these uh, podcasts. And sitting down with you Brett, having discussions, we tend to uh, to get very excited. Uh, but I'd like to just come back to the book before we round off here. Um, on your book, one of the things that you talk about in your book obviously is you describe AI making a financial expertise, you know, more widely accessible to everybody. And um, but also at the same time in your book you do you warn that it could displace the income people depend on. Right. You talked about this right at the beginning, the fact that we might see new economic models, um, and new ways of recognizing wealth and money. Supposedly automation actually arrives faster than the mechanism for sharing its wealth. So from your book's perspective, what does banking 5.0 look, uh, like if the technology arrives before the new economic model?

Speaker D: Well, I think it already has to some extent, Paul. I think this is the point of the conversation we're having. And there's a lot of debate over whether AI is going to take jobs from humans. And the purists, the religious capitalists, I call them, the fundamentalist capitalists, will always say yes, yes, but we constantly destroy jobs. Jamie Dimon's quote on this is, yeah, tractors destroy jobs. In fact, farms, um, but you know, then we had to have people in factories building tractors and we have to have people driving tractors. So you know, it created, it created new jobs. And this is the analogy we like to hear all the time. But we've never had a technology like AI that's designed to eliminate human labor from the workforce. That's its defining characteristic. So we have to create new jobs that are currently not in the workforce. I mean we're talking about an entire set of jobs. And this is um, blue collar, white collar, you know, um, disruption across at a scale and speed with that we're just not really familiar with. So I think this is highly likely, we're highly likely to see large scale technology and employment in the2030s. And uh, the proposed solution to this has been universal basic income. Anthropic is sort of trying to create a foundation to sort of create a sort of UBI Right now. Um, but, you know, we would need to tax agents and tax robots at scale to replace lost income, uh, income tax anyway, so we definitely need a new system there. Um, but then the question ultimately becomes, well, what do humans do if AI and robots take all these jobs? Um, and that's really the real philosophical question. And this is where we need to make a shift and agentic and AI is going to force us to really examine what is it we're here for as humans, what is our life purpose for? It's no longer going to be as an economic unit of wealth creation or capital growth for the economy. That was sort of a hallucination over the last 300 years that we created through capitalism. It's going to become something more meaningful. Um, and so this, this comes out of this 15 or 20 years of chaos that we have, um, with this displacement of labor and so forth, where we come into a period of time where humans are able to do much more meaningful things as a collective species that's not just oriented towards economic growth. Um, and this, this is where it gets really interesting, the science. We're going to be living longer. We're going to have the capabilities to explore the solar system and invent, um, new materials, uh, transform the damage we've done to the climate, um, examine consciousness, uh, um, the hard problem of consciousness and all these things. I mean, this really is the most exciting time to be alive, but it is going to be a period of disruption first, and then there's something new. And when I always have this conversation about capitalism, the first thing that happens is people attack me for being a socialist. And I'm saying, no, no, this is not about capitalism versus socialism. This is about a net new system of economic thinking, something we've never done before. And we need that sort of new thinking to take us into this age of AI at the foothills of the Singularity, as Demis Hassabas calls it.

Speaker B: And with that and some very deep thoughts for everybody to go away with. Uh, Brett, we have a tradition on this, uh, podcast. Vi normally does that at the beginning, I do it at the end. But we like to lighten the mood a little bit with a joke from our guests. So I throw it upon you, sir. Do you happen to have a good but bad joke you would like to share with us and our listeners?

Speaker D: Well, you know, I could joke about putting the king in banking or something like that, but no, um, um. Well, you've heard the expression, give, uh, a hungry man a fish and you feed him for a day, teach him to fish and he'll never go hungry. Right. Well, do you know what the banking equivalent of that is, Paul?

Speaker B: Uh, no, but I'm sure you're going to share it with us.

Speaker D: Well, if you give a man a gun, he can rob a bank, right? Um, especially if you live in the United States, where guns are easily accessible. But if you give a man a bank, he can rob everyone else. That's my joke.

Speaker C: Yeah. Give a man a stable coin and he will rob the entire planet. But anyway, that's.

Speaker D: Yes,

Speaker B: that was very good. That was very good. Thank you. I'm

Speaker D: glad you liked it.

Speaker B: Yes, no, that's good. That goes up there with some of the best ones we've had. Brett, if anyone wants to get in touch with you, talk a bit more about your book, um, have a general discussion with you about the future of agentic banking and the future world we're heading towards. How can a DCS reach you?

Speaker D: So, um, for the book, go to bank5book.com, that's the number five bank5book.com to find more about the book. And you can also follow me@brett king.com or of course go to LinkedIn, um, where you just search on Brett King. And, uh, you should be able to find me there. And also check out, of course, uh, our part, the partner podcast for fintech Daydreaming, which is Breaking Banks, where we've done a mashup, mashup, uh, podcast series together before. And also the Futurist podcast, which is where I get really sci fi about the future and innovation. You, the world's top futurists.

Speaker B: Oh, wonderful. Thank you very much. Uh, Villa, any last two words from you?

Speaker C: Well, only one thought. Uh, I've heard that in the Silicon Valley, many of the AI optimists have started smoking because they believe that, uh, the AI will cure cancer. So it doesn't matter. So it's the ultimate show of faith that the AI future is a good one. And that level of optimism. Well, I still need to reach. Reach it. Uh, I think we have heard in this episode that maybe it's something that we can wait for.

Speaker B: Absolutely.

Speaker D: I do think that, uh, I do think we are on the cusp of curing cancer. And there's a number of, uh, uh, therapeutics on that. That's a whole new podcast. But, uh, definitely check out the work that Demis Hassabas and, uh, David Sinclair, um, are, uh, doing in this field. David Sinclair, who heads up Harvard Medical. Cool.

Speaker C: Yeah. And to be clear, don't start smoking.

Speaker B: Not

Speaker D: unless it's plant medicine for therapeutic means. Yes,

Speaker B: exactly. And with that, I will just summarize to say, we're going to cure cancer. We're all going to live longer. We don't need to work. The agents will do it all for us. We will have universal income. So we'll live like rich people with nothing to do, sitting on the beach, drinking margaritas. It sounds like the perfect life to have for the future. With that, we are grateful that you have stuck with us until the end of this episode. Really hope you've enjoyed the discussions we've had with, uh, Brett. Please go and buy his book. I have enjoyed all of his other bank 2.0, 3.0, 4.0, and now 5.0 banks. So please read them. Go and have a look. Get back to him and let him know what you thought of the book. After you have read it, there'll be a lot more. Um, I am sure that he will share in the future around banking. But if you're still with us, give us a thumbs up on the channel that you're listening to us on. Or if you're on YouTube. Do that. Give us a comment. We love the comments. Good comments, bad comments. Any comment is better than the silence that, uh, sometimes us podcasters have to endure. But we know you're out there. We can see our listeners growing slowly. Share the podcast with some friends. But until next time, this has been fintech. Dave.

Speaker A: This is fintech Daydreaming.